Close
About
FAQ
Home
Collections
Login
USC Login
Register
0
Selected
Invert selection
Deselect all
Deselect all
Click here to refresh results
Click here to refresh results
USC
/
Digital Library
/
University of Southern California Dissertations and Theses
/
Gender differences in symptom presentation of depression in primary care settings
(USC Thesis Other)
Gender differences in symptom presentation of depression in primary care settings
PDF
Download
Share
Open document
Flip pages
Contact Us
Contact Us
Copy asset link
Request this asset
Transcript (if available)
Content
GENDER DIFFERENCES IN SYMPTOM
PRESENTATION OF DEPRESSION IN
PRIMARY CARE SETTINGS
by
Afsaneh Nasserbakht
A Dissertation Presented to the
FACULTY OF THE GRADUATE SCHOOL
UNIVERSITY OF SOUTHERN CALIFORNIA
In Partial Fulfillment of the
Requirements for the Degree
DOCTORAL OF PHILOSOPHY
(COUNSELING PSYCHOLOGY)
May 2001
Copyright 2001 Afsaneh Nasserbakht
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
UM I Number: 3027757
Copyright 2001 by
Nasserbakht, Afsaneh
All rights reserved.
___ ®
UMI
UMI Microform 3027757
Copyright 2001 by Bell & Howell Information and Learning Company.
All rights reserved. This microform edition is protected against
unauthorized copying under Title 17, United States Code.
Bell & Howell Information and Learning Company
300 North Zeeb Road
P.O. Box 1346
Ann Arbor, Ml 48106-1346
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
UNIVERSITY OF SOUTHERN CALIFORNIA
THE GRADUATE SCHOOL
UNIVERSITY PA R K
LOS ANGELES, CALIFORNIA 9 0 0 0 ?
This dissertation, written by
.
under the direction of h Dissertation
Committee, and approved by all its members ,
has been presented to and accepted by The
Graduate School, in partial fulfillment of re
quirements for the degree of
DOCTOR OF PHILOSOPHY
Dean of Graduate Studies
Date ...Jfey...U A ..2QQL
TATION COMMITTEE
LCm df,
Chairperson
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
1
Afsaneh Nasserbakht Rodney Goodyear
ABSTRACT
GENDER DIFFERENCES IN SYMPTOM PRESENTATION
OF DEPRESSION IN PRIMARY CARE SETTINGS
This study investigated gender differences in symptom presentation of
depression. Data were obtained through seven managed care organizations in
different parts of the U.S. Patients visiting their primary care clinicians in five
urban areas (Los Angeles; San Antonio; Twin cities; San Luis Valley; and,
Columbia, MO) were screened for depression. Participants completed demographic
questionnaires as well as the Center for Epidemiological Studies - Depression (CES-
D) Scale, the Composite Interview diagnostic Interview (CIDI), and the Daily
Functioning and Well-being using Short Form Health Survey (SF-36). The 1,187
(867 women and 320 men) participants ranged from 18 to 90 years old (M = 43.50 ).
Although many studies have examined differential presentation of
depression symptoms in men and women, findings have been inconsistent. This
study examined three research questions: Do men and women present different
symptoms of depression? Do depressed men and women differ in their daily
functioning? and, Does severity of depression interact with gender in predicting
symptom presentation?
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
2
Women were found to score lower on symptoms of lack of positive affect,
and higher on guilt and somatization factors. They also reported a greater number of
symptoms and were more likely than men to report specific symptoms, especially
crying spells and eating too much. Women also reported greater impairment in
functioning and well-being (especially in vitality and overall mental health).
Depressive symptoms also varied according to level of depression. But unlike
Newmann (1984), we did not find that depression severity and gender interacted in
predicting symptomatology. Implications of these findings are discussed.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Acknowledgments
I would like to express my appreciation to my dissertation committee at the
University of Southern California. First and foremost, I would like to thank my
advisor Dr. Rodney Goodyear, for his invaluable guidance and support throughout
my doctoral studies. He was always available when I needed his input and he
continuously supported and encouraged me to pursue my interest. I have been
fortunate to have him as my advisor. I would also like to thank the other members of
my committee, Dr. Merril Silverstein and Dr. Kaaren Hoffman. I have benefited
tremendously from their support and advice during the course of my dissertation. I
am particularly grateful to Dr. Hoffman for giving me feedback and advice on the
analysis of the data on numerous occasions, which proved to be invaluable.
I am extremely grateful to Dr. Ken Wells of UCLA and RAND Corporation
and Dr. Cathy Sherboume at RAND Corporation who were there from the
beginning of this research, offering ideas, suggestions and support. The
encouragement of Dr. Wells and the insightful comments and suggestions of Dr.
Scherboume throughout the writing of my dissertation have been invaluable. This
study would not have been possible without their support and I am forever grateful.
The data of my dissertation was collected as part of the Partners In Care (PIC)
project. I have been fortunate to work on a project with such an extraordinary group
of talented and committed individuals. I would like to thank the project staff, in
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
iii
particular Maureen Carney, Bernadette Benjamin and Gail Yeaple who have been
wonderfully helpful in providing details of the study.
I also would like to thank the University of Pittsburgh, Counseling Center’s
staff, particularly Dr. Mary Jo Loughran and Dr. Penny Crary for providing
incredible support and encouragement during my internship year and allowing me
the flexibility to work on my dissertation. A special thank you to my fellow interns,
Victoria Creighton, Marla Somova and Keith Beard for making our internship and
dissertation writing time a supportive and enjoyable experience.
I also would like to express my gratitude to other faculty, staff and
colleagues at USC who have truly enriched my graduate experience. A special
thanks to Dr. Jose Abreu, Dr. Shing-Shiong Chang, ILeslie Kay, Yvette Barraza-
Reyes, Liz Gross, David Ettleson, Joy Davis, Hsing-fang Chang and Linda
Matthews.
I also like to thank my good friends Corinne, Natalie, Alex, and Mildred.
Thanks for their support and friendship. Last but not least, my special thanks to
Chuck who has been tremendously helpful throughout the course of my dissertation.
I am grateful for his encouragement and support throughout the years.
As I reflect back to acknowledge and express my gratitude to those who
have supported me over the years, I realize that my journey started long ago.
Reaching my goals would not have been possible without the help of my family. I
would like to thank my parents who with their love and support provided an
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
iv
environment where I could pursue my career and also to thank my sisters who have
enriched my journey every step of the way.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Table of Contents
Acknowledgments................................ ii-iv
List of Contents................................ v
List of Tables ................................ vi
Chapter I
Review of Literature ............................ 1
Chapter II
Method ................................ 18
Chapter III
Results ................................ 35
Chapter IV
Discussion ................................ 80
References ................................ 91
Appendix
Study Design .................................. 100
CES-D & DSM Questions............................. 101
SF-36 Questionnaire ............................. 102
SF-36 Measurement Model ........................... 107
SF-36 Scoring ...................................108
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
vi
List of Tables
Table 1 23
Table 2 37
Table 3a .............. 39
Table 3b ........................................ 40
Table 3c ........................................ 41
Table 4a ........................................ 42
Table 4b ........................................ 43
Table 4c ........................................ 44
Table 5a ........................................ 45
Table 5b ........................................ 48
Table 5c ........................................ 51
Table 6 56
Table 7 59
Table 8 61
Table 9 64
Table 9a ............................ 65
Table 9b ........................................ 66
Table 10 71
Table 1 0 a ........................................ 75
Table 1 0 b ........................................ 76
Graph 1 79
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
1
Chapter I
Introduction
Many compare depression to the “common cold” of mental illnesses. It is
one of the most common problems that primary care and mental health clinicians
face today. It has been estimated that about 10% of the adult population suffer from
some sort of depressive disorders (Kessler, McGonagle & Zhao, 1994).
Depression has an incapacitating effect on people’s day to day functioning.
It has been reported that patients suffering from depression have greater difficulty in
functioning than most patients with other chronic medical illnesses (Wells, Strum,
Sherboume & Meredith, 1996).
Society pays a high cost not only for the treatment of depression, but also for the
loss of productivity that depressed individuals experience (Wilson & Drury, 1984).
Interestingly, the rate of depression is not the same for everyone in the
society. It appears that the prevalence of depression is much higher in women.
Many studies have repeatedly shown a difference in the prevalence rate of
depression between men and women. In 1987, NIMH published the results of
epidemiological studies of depression in the US and Western countries. These
results indicated that women are twice as likely as men to have some sort of
depression. (Amenson & Lewisohn, 1981; Cleary, 1987; Kessler, McGonagle &
Zhoa, 1994; Weissman & Klerman, 1977).
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
2
Most investigators using DSM categories consider reports of sadness,
helplessness, eating disturbances, loss of ability to concentrate, feelings of
worthlessness, lack of energy, and tension to be symptoms of depression. In spite of
the extensive research in this area, one major question remains unanswered. Are the
manifestations of symptoms the same in men as in women?
The purpose of this study is to investigate the presentation of symptoms in
depressed men and women. In this chapter, first an overview of the diagnostic
criteria, the various ways of assessing depression, as well as the impact of
depression on functioning and well being will be discussed. Second, the literature
related to the prevalence of depression and various ways of presenting depressive
symptoms for both genders will be reviewed. In the last section theories regarding
gender differences and the study purpose will be presented.
Detection of depression in primary care settings
The appropriate treatment for any illness depends in a vital way on its
correct diagnosis. Practitioners typically employ two primary ways of assessing
depression. One is to look at depressive symptoms; another is to consider specific
criteria. Both rely on signs and symptoms as defined by Diagnostic Statistical
Manual (DSM) or International Classification of Disorder (ICD). Significantly,
there seem to be differences in which diagnostic practice is used between
professions. Whereas most mental health clinicians use the Diagnostic Statistical
Manual (DSM) or International Classification of Disorder (ICD) criteria, most
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
primary care clinicians consider depressive symptomatology prior to making actual
diagnoses. The different approaches used related to professional orientation seems
to be important, since it has been reported that in the U.S. about half of patients who
receive any mental health care receive that service from their primary care
clinicians. (Wells, Manning, Duan et al. 1986). Wells, Strum, Sherboume and
Meredith (1996) found that the rate of underdetection of depression by primary care
clinicians was high. Also previous findings by Wells, Hays, et al. (1989) showed
that this degree of underdetection was high for both men as well as women.
Some studies report that primary care practitioners frequently do not
consider the underlying affective disorders, and consider only the somatic
symptoms (Goldberg & Bridges, 1988; Katon, Kleinman & Rosen, 1982; Potts,
Buman & Wells, 1991).
Assessing depression
In this section, the definition of "depression" as used in this research will be
discussed. An overview of the diagnostic criteria as well as various ways of
assessing depression will be reviewed. The two ways of assessing depression, using
the clinicians criteria (DSM or ICD) and using self-rating instruments (BDI,
MMPI,...) will be discussed later.
An important technical issue in research regarding depression is the actual
definition of it. Depression is defined in the dictionary as a mental disorder, which is
marked by symptoms of sadness, hopelessness, difficulty concentrating, change in
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
sleep and/or change in appetite and inactivity. The symptoms can range from mild
to severe, and can have a profound impact on individual’s functioning.
In the U.S., the most accepted criteria for many types of mental disorders are
the criteria provided by the Diagnostic and Statistical Manual (DSM). The DSM
also differentiates among different types of unipolar depression: Major Depressive
Disorder (single, recurrent), Dysthymic Disorder and Depressive Disorder Not
Otherwise Specified.
According to the DSM, a diagnosis of Major Depression is given only when
five out of nine symptoms are present during a two-week period. These symptoms
are: depressed mood, loss of interest, sleep problems, significant gain or loss of
weight or appetite, loss of energy, feelings of guilt or worthlessness, reduced ability
to concentrate, psychomotor agitation or retardation and suicidal ideation. In
addition to the DSM criteria for a depressive disorder, ICD-10 requires both feelings
of sadness and loss of interest or one of these symptoms plus the presence of
chronic fatigue to fulfill the criteria. A depressive episode is not considered present
if there has ever been an episode of mania or hypomania.
According to the DSM, the episode is characterized as having Melancholic
features if the depression has a distinct pattern of being worse in the morning, and
symptoms of early awakening, retardation or agitation, significant loss of appetite or
excessive guilt. ICD-10 would characterize the same episode being of the somatic
type (the depression is worse in the morning) based on an additional set of
symptoms, one of which is a marked loss of libido, which is not included in the
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
5
DSM. But it excludes excessive guilt as a criteria, which is included in the DSM.
(CEDI manual, WHO 1997)..
The criteria for Dysthymic Disorder is having a depressed mood for more of
the time than not for at least 2 years with the presence of 2 or more of the following
symptoms: appetite problems, sleep problems, low energy, low self-esteem, poor
concentration or difficulty making decisions and feelings of hopelessness.
According to the DSM, there must not have been any episodes of mania or
hypomania and major depressive episodes must not have occurred in the first two
years, whereas in ICD-10 it must be rare or mild. According to the ICD, Dysthymia
is characterized by withdrawal and feelings of being unable to cope with everyday
life.
Self-rating Instruments
As mentioned previously, another way of assessing depression is through the
use of self-rating instruments. Several instruments have been developed to assess
depression by measuring symptoms. The most common ones are mentioned here
such as MMPI, BDI, and CES-D.
MMPI (Minnesota Multiphasic Personality Inventory) has several scales,
one of which is Scale 2, the “Depression scale”, which measures depression. This
scale has 33 items and includes symptoms of depression such as fatigue, anhedonia,
dysphoria, pessimism, hopelessness, crying easily, feelings of guilt or emptiness,
loneliness or suicidal ideation. Harris and Lingoes (1968), who studied MMPI
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
scales, report the following subscales for the “Depression Scale” such as:
‘subjective depression’, ‘psychomotor retardation’, ‘physical malfunction’, ‘mental
dullness’, and ‘ brooding’.
Reliability studies on the MMPI indicate that it has moderate levels of
internal consistency. For example, Hunsley, Hanson and Parker (1988) reviewed
studies performed on the MMPI and reported that all scales are reliable with values
ranging from .71 to .84 (Hunsley, Hanson and Parker, 1988; Parker, Hanson,
Hunsley, 1988).
Other studies provide extensive support for MMPI’s construct validity and
studies investigating MMPI validity are also increasing (Dahlstrom, Welsh &
Dahlstrom, 1972; Sheppard, Smith & Rosenbaum, 1998). However, it has been
reported that the MMPI-D Scale has more somatic items, which reduces its
suitability for assessing depression in those who suffer from chronic conditions or in
older adults. Butcher et al. (1991) found that older men had higher scores on the
Depression scale than younger men. In addition to the need to adjust to this result,
many researchers prefer not to use MMPI-2, because of its length and time it takes
for individuals to fill out the questionnaire.
Another useful instrument to measure depression is the “Beck Depression
Inventory” (BDI). This instrument was initially developed to assess depression in
clinical settings. However, these days BDI is used in both clinical as well as
community settings (Beck et al., 1961). The BDI has 21 items. Each item has four
possible answer choices, which are given a value of 0-3. The total score is
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
calculated by adding up the score of each item. A total score of 16 to 19 indicates
mild to moderate depression, 20-29 moderate to severe depression and scores
greater than 30 severe depression.
BDI has been extensively studied over the years. Beck, Steer and Garbin
(1988) reported the test-retest reliability of the BDI in psychiatric settings ranged
from .48 to .86, whereas for non-psychiatric settings, it ranged form .60 to .90.
(Beck, Steer and Garbin, 1988). A number of studies have indicated that the BDI
can differentiate psychiatric patients from non-psychiatric patients as well as
differentiating between those who have severe depression from dysthymia. It has
also been reported that this instrument has construct validity with other measures:
MMPI (.76), Zung (.76) (Steer, Beck, Brown & Berchick, 1987; Steer, Beck,
Riskind & Brown, 1986).
Some studies report a two sub-scale structure for BDI: the “psychological”
and the “somatic” item portions. Rapp, Parisi, Walsh and Wallace (1988) found that
the psychological subscale demonstrated a better overall performance in assessing
depression than the somatic items. Although this instrument has been widely used, it
has been reported that some of the BDI items, which measure somatic difficulties,
may not be as good in differentiating between those who are depressed versus not
depressed. (Rapp, Parisi, Walsh and Wallace, 1988)
Another instrument, which measures depressive symptomatology, is CES-D,
(Center for Epidemiological Studies Depression Scale). Unlike BDI, this instrument
was first developed to assess depression in the community sample (Radloff, 1977).
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
However, CES-D, too, has also been extensively used in both clinical and
community settings. (Santor, Zunoff, Cerrantes, Palacias and Ramsay, 1995). Many
studies have shown this instrument’s reliability and validity, with a score greater
than 16 being the indicating criterion for depression. In order for someone to obtain
a score of 16 or greater, a person must report at least six of the symptoms for most
of the time or a majority of the symptoms during shorter periods (Boyd, Weissman,
Thompson & Meyers, 1982; Weissman, Scholomskas, Pottenger, Prusoff & Locke,
1977). Santor et al. (1995) reported that although this instrument was initially
developed for community settings, it is more effective than BDI in assessing
individual differences in severe depression in outpatient settings. Therefore, many
investigators prefer to use this instrument over BDI.
One of the reasons that various studies report different findings in terms of
symptom presentation in men and women may be due to the use of different
instruments. The lack of consistency in terminology and use of specific instruments
raises the question of whether certain criteria were considered. For instance, MMPI,
CES-D and BDI have items asking about low appetite (in MMPI, CES-D and BDI)
and weight loss (in BDI). However, they do not ask about an increase in appetite or
weight. On the other hand, when a clinician wonders if someone has depression,
they often consider different criteria including an increase as well as a decrease in
weight or appetite. This raises the question of whether specific criteria were
considered in assessing depression in clinician’s ratings vs. self-rating instruments.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
9
Functioning and Well-being
Recently more attention has been paid to the individual’s functioning and
well-being. The impact of mental health disorders is not only assessed in terms of
their prevalence, but also in terms of individuals’ functioning. (Johnson, Weissman,
& Klerman, 1992; Lehman, Ward & Linn, 1982; Wells et al., 1989; Wells, Golding
& Brum, 1988). Some have reported that depressed patients experience more
limitations in terms of their functioning compared to chronically ill patients
(Broadhead, Blazer, Gorge & Tse, 1990; Klerman & Weissman, 1992, Wells, et al.,
1989). Other researchers have shown that people with major depression as well as
people with subthreshold depression (those who have some symptoms of
depression, however, do not meet all the criteria for major depression or dysthymia)
experience limitations in their functioning such as being bed ridden, missing work
and a loss of wages (Broadhead et al., 1990; Johnson et al., 1992).
Wells et al. (1992) report that individuals with dysthymia show low
functional status. They are also more likely to have recurrence of their symptoms in
a follow-up compared to those with major depression, in spite of the fact that
dysthymia is a less severe form. Other investigators have also reported more
unfavorable outcomes for those who experience some of the symptoms instead of
the actual major depressive disorders (Wells, et al., 1992). Others have also found
unfavorable outcomes for those who have depressive symptoms and not the actual
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
10
disorder (Coulehan, Schulberg, Block, Janosky, & Arena, 1990; Johnson,
Weissman, Klerman, 1992).
As mentioned previously, a higher prevalence rate of depression in women
has been documented in a number of studies. Robb, Young, Cooke & Joffes (1998)
report that in their bipolar sample, the patients report greater impairment in
functioning and well-being, particularly in the area of physical health and pain.
Moreover a high rate of morbidity in women has also been reported (Chou, 1994).
Hammen and Peters (1977) investigated attitudes towards depression in men and
women. They report that when the ability of depressed persons to function in
various roles in relation to their gender is considered, depression in men is perceived
to have a much more debilitating effect than in women. Wells, Strum, Sherboume
and Meredith (1996) report that the degree of limitation may vary among subgroups
of patients, including gender as a subgroup. They report that the rate of functional
limitation remained highest among the oldest, the women, and the unmarried. These
studies raise the issue as to “what are the gender differences in symptom
presentation and are they associated with differences in functioning and well
being?”
Depression and gender differences
Many studies have reported different symptom presentation in men and
women. However, the results have not been consistent. The question is “Are women
more likely to show endogenous types of symptoms such as appetite change, diurnal
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
mood variations and psychomotor retardation or experience more role limitations in
terms of physical health?” Angst and Dobler-Mikola (1984) used the interview
ratings in a community sample to assess depression. They reported that even among
those who were severely depressed, women report more symptoms in general. They
also reported more sleep disturbances, appetite or weight changes, and feelings of
worthlessness or guilt. Katz et al. (1993) who studied a sample of patients with
major depressive disorder using structured interviews to assess depression reported
that women showed more distress, in the areas of somatic complaints, psychomotor
retardation and anxiety.
Unlike the previous studies, which found that women experience more
somatic symptoms, Hammen and Padesky reported more somatic disturbances in
their male sample. Hammen and Padesky (1977) showed that a gender difference
was evident in the college sample using BDI, where depressed males were more
likely to report somatic complaints, loss of social interest, sense of failure, and
inability to cry. Whereas, depressed women were more characterized by
indecisiveness and self-dislike.
Oliver and Toner (1990), using BDI in a college sample, reported that men
demonstrated more somatically oriented problems such as problems sleeping,
decreased libido, social withdrawal, and more dissatisfaction than women. Women,
on the other hand, were characterized by more emotive symptoms such as crying
and sense of punishment than were men. Unlike Hammen and Padesky (1977), in
this study, women were more likely to show somatic symptoms.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
While results regarding gender differences seem contradictory, Radloff
(1977), who performed factor analysis on CES-D items, found that the factor
structure was similar across a variety of demographic characteristics including
gender. However, Ross and Mirowsky (1984) who studied depressive symptoms in
married men and women reported a slightly different factor structure. They reported
that women scored higher on symptoms of depressed affect, enervation, and lack of
lack of positive affect compared to men. Studies, which used instruments besides
BDI and CES-D also report, gender differences. In another study, Craig and Van
Natta used a number of instruments including BDI, Zung, MMPI-D, Raskin, and a
scale developed by Gardner which was used in both a clinical and a community
sample. In their study, men had a higher persistence rate for one symptom, ‘all an
effort’. Women showed significantly higher prevalence rates for ten symptoms and
significantly higher persistence rates for the four symptoms of ‘poor appetite’,
‘couldn't get going’, ‘fearful’, and ‘people unfriendly’. One symptom, ‘crying
spells’, was almost exclusively a symptom reported by women (Craig & Van Natta,
1979).
In a community sample study, Chino and Funabiki (1984) used the Inventory
of Depressive Behaviors and reported that females scored higher on items
measuring seeking personal support, and writing to express feelings as well as
eating more, along with more physiological disturbances, and gastrointestinal
complaints (e.g. T feel sick inside’). Whereas males scored higher on items showing
smoking and a desire to be with other depressed people.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
13
Young, Scheftner, Fawcett and Klerman (1990) in a clinical sample of major
depression based on clinician ratings found that only two clinical features occurred
significantly more frequently in one gender, with more women reporting appetite
and weight disturbances. While, they identified gender differences in appetite, they
suggest that this may be mostly due to an increase rather than decrease appetite.
Frank, Carpenter and Kupfer (1988) who used a clinician sample and used
both clinician rated and self-report instruments reported that, taking all modes of
assessment into consideration, it appears highly likely that increases in appetite and
weight are more common features of depression in women than in men.
The major factors, which may have contributed to the different findings of
these studies, are the different types of setting in which each study was conducted
(clinical vs. community) and the different types of measure used. In summary, most
of the studies mentioned above report a difference in the presentation of symptoms
in men and women. Unfortunately, the findings have not been consistent and no
stable cluster of symptoms was reported that is specific to one gender. However,
regardless of the different settings, certain themes are evident:
1. The majority of these studies (those which have adequate sample size)
report a different presentation of depressive symptoms among men and
women.
2. 2. In both clinical and community samples, where the concept of
appetite change / weight change was considered, whether in the
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
14
clinician’s interview or the self-rating instrument, it has been
consistently reported that depressed women report greater weight or
appetite disturbances compared to depressed men.
In assessing depressive symptoms, what was done differently in the current
study was to add ten other items to approximate the criteria closer to the DSM
criteria. Although the higher prevalence of depression among women has been
documented by a number of studies, only a few studies have investigated the
manifestation of symptoms and limitations in daily functioning in clinical settings
and no study has investigated this difference in primary care settings. Gender
differences in these settings may be a crucial factor in the diagnosis and treatment
planning of depression.
Theories on gender differences
Many studies have reported the higher prevalence of depression in women.
This higher prevalence has been documented over many years and across many
countries. It was not until very recently that investigators started gathering data to
test several hypotheses which attempt to explain this phenomenon. (Amenson &
Lewisohn, 1981; Verbrugge, 1979; Weissman & Klerman, 1977).
Some studies have tried to relate these gender differences to the willingness
of women to report their concerns (Verbrugge, 1986), or the reluctance of men to
discuss their personal concerns (Morgan, 1976). Different theories have offered
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
1 5
alternative explanations for these observed differences. In summary, they suggest
the following causes:
1. different socialization for men and women,
2. biological differences, (Klerman & Weissman, 1980),
3. Women’s disadvantaged social status and learned helplessness,
(Hammen, 1982; Repetti & Crosby, 1984; Weissman & Klerman, 1977),
4. Women’s worse health status and higher morbidity (Marcus & Seeman,
1981; Verbrugge, 1976),
5. The instruments’ psychometric properties and potential biased items
(Barsky, 1979; Clauser & Mazor, 1998),
6. Severity of depression (Newmann, 1984; Young, 1990).
Weissman and Klerman (1977), who have extensively studied different
theories regarding gender differences, report that the gender difference is real and is
not due to reporting or care-seeking behavior. They encourage the other
investigators to pay more attention to sociocultural factors affecting women’s
mental health. Other reports have strengthened Weissman and Klerman’s
perspective with findings of the interaction between sex and marital status in
relationship to depression. For instance, married women tend to be more depressed
than married men, whereas single men tend to be more depressed than single
women (Ahnlund, K., Frodi, A., 1996; Van Grootheest, D.S., Beekman, A.T.F., Van
Groenou, M.I.B., Deeg, D.J.H., 1999; Weissman & Klerman, 1977).
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
16
Some investigators have discussed the inadequacy of certain instruments in
assessing depressive symptoms in men. They report that most studies use BDI for
measuring depression. According to Bradsky (1979) this instrument contains several
items which measure somatic symptoms such as sleep disturbance, weight loss, and
loss of libido. However, it does not ask about other somatic symptoms, such as
headaches, back pain, gastro-intestinal distress, etc. Therefore, BDI does not fully
assess whether the participants are somatizing or not.
Other investigators have tried to explain the gender differences in relation to
the severity of depression. For example, Kessler and her colleagues report that there
is a gender difference in help-seeking behavior which only manifests itself at the
first stage of problem recognition (Kessler, Brown & Broman, 1981). However,
once men admit they had problems with their mood, they were just as likely as
women to seek professional help. Young and his colleagues (1990) suggest that the
relationship may not be the same over the range of depressive disorders. Newmann
(1984) reported that women showed more depressive symptomatology than men.
But the differences did not stay constant over the range of depressive disorders.
There were small differences for the most severe symptoms, mixed patterns for
moderately severe symptoms and bigger differences for the less severe symptoms.
Hypotheses and Research Questions
The purpose of this study is to look at the clinical presentation of symptoms
in depressed men and women. Specifically, the hypotheses is:
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Women will report more endogenous types of symptoms such
as weight or appetite change, sleep disturbance and experience
more irritability and anxiety; whereas, men will report more
interference with social functioning, withdrawal and substance
abuse. (Angst and Dobler-Mikola, 1984; Chino and Funabiki,
1984; Hammen and Padesky, 1977; Katz et al., 1993).
Three major questions will be addressed:
Research question 1: Do men and women present differently in
terms of the symptoms of depression?
Research question 2: Do depressed men and women differ in their
daily functioning?
Research question 3: If we do find different profiles of symptom
presentation, does the pattern vary by severity of depression
and the range of depressive disorders?
Comparing the presentation of symptoms and limitations in daily
functioning allows us to refine the investigative approach and thus develop
improved indicators of depression in men and women.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
18
Chapter II
Methods
In this section, the design of the study will be discussed. First, the
characteristics of the participants are described, and then the measures used in the
study as well as the background information regarding the measures will be
presented.
Participants
There were 1,187 (867 women and 320 men) study participants who were
physicians’ (internal medicine and family practice) patients in five cities, including:
Los Angeles, California; San Antonio, Texas; Twin Cities, Minnesota; San Luis
Valley, Colorado; and Columbia, Maryland.
Demographics included: gender: females (N = 867), males (N = 320);
married (N = 630), non-married or divorced (N = 522); working part-time or full
time (N = 705), not working (N = 403); ethnicity: Native American (N = 86),
Asian/Pacific Islander (N = 24), African American (N = 74), Hispanic (N = 273),
Caucasian (N = 706); education (mean = 13.19, range: 0-17); age (mean = 43.48,
range: 18-90).
Measures
Participants all completed demographic questionnaires as well as five
measures. These were: a screener, the Center for Epidemiological Studies (CES-D),
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
19
the Composite Interview diagnostic Interview (CIDI), and the SF-36 (See Appendix
C).
Screener
The screener was a thirty-item measure used to determine eligibility for the
study. In addition to all of the demographic information, the screener asked patients
whether they had any of the following chronic conditions: asthma, high blood
pressure, hypertension, arthritis, physical disability, trouble seeing, trouble
breathing, cancer diagnosed within the last 3 years, neurological condition, stroke,
heart failure, angina, other heart disease, back problem, stomach ulcer, chronic
inflamed bowel, thyroid disease, and migraine headache.
In order to assess depression at this stage, five items from the major
depression and dysthymia sections of the 12-month World Health Organization’s
Composite International Diagnostic Instrument, Version 2.0 (GDI 2.0), were also
used. Two additional items regarding the presence of depressed mood and/or loss of
interest during one week or more in the past month were also added. (See Appendix
A.)
Center for Epidemiological Studies-Depression (CESD) Scale
CES-D was first developed by Radloff (1977) to study depression in
epidemiological settings. (See Appendix B.) Twenty items were chosen to measure
certain aspects of depression such as depressed mood, psychomotor retardation,
feelings of helplessness and hopelessness, loss of appetite, feelings of guilt and
worthlessness, and sleep disturbance (Radloff, 1977).
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
20
This instrument measures depressive symptoms in the past week, as opposed
to the past month. The questions ask “how often have you had depressive symptoms
in the past week?” Items are labeled from 1 to 4 on a four-point scale: 1. Rarely, 2.
Some of time, 3. Occasionally, or 4. Most of the time). These options are scored
from Oto 3. Some of the items, which measure lack of positive affect such as “I feel
happy”, etc., are reverse coded. (Specifically, items number of 4, 8, 12, and 16.)
After these items are reverse coded, all the responses are added up and a total score
is calculated which can range from 0-60 with a higher score indicating more severe
symptoms.
Many studies have shown the reliability and validity of the CES-D to assess
depressive symptomatology (Radloff, 1977; Weissman et. al., 1977, Craig Van
Natta, 1979.)
The CES-D has been shown to have high consistency and adequate test-
retest reliability. Radloff (1977) reported internal consistency of .84 to .90. High
reliability is also reported in several other studies (Ross & Mirowsky, 1984;
Roberts, Andrews, Lewinsohn & Hops, 1990).
This instrument has been reported to have good discriminative validity.
Several studies have reported that CES-D scores are capable of differentiating
between depression and other psychological problems. According to Radloffs
study, schizophrenics and drug addicts had lower scores than depressed patients.
(Myers et al., 1984; Radloff, 1977; Weissman et al., 1977). Other studies have
reported that scores of this instrument correlate highly with other instruments
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
21
assessing depression such as symptom Checklist-90, BDI and Zung (Bemdt, 1990;
Weissman et al., 1977).
The research on the most appropriate cut-off points does not seem to be in
agreement in terms of what specific cut-off point is the most appropriate one to use.
There have been mixed reports in terms of what the cut-off scores should be used.
Radloff (1977) himself suggested a cut-off score of 16 indicating severe depression,
whereas Schulberg (1985) and his colleagues report a cut-off score of 27. Weissman
and her colleagues agreed with Radloff and reported that the cut-off score of 16 or
above correctly diagnosed about 99% of their depressed sample (Hankin & Locke,
1983; Weissman et al., 1977). Whereas Schulberg and his colleagues (1985)
reported that the cut-off of 16 or above diagnosed only 84% of their depressed
sample. (Schulberg et al., 1985). While Zimmerman and Coryell (1994) reported
that 88% of their sample were identified as depressed using the same cut-off score
of 16.
In this study, to make CES-D more comparable to DSM, 7 items were
deleted from the original scale (See Appendix B.) Several other items were added to
measure weight/appetite, sleep, anhedonia, psychomotor agitation, guilt,
concentration, and suicidal ideation. After several items were reverse scored, scores
of each item were added. Then, the scores sum was transformed into a 0-100 scale.
A higher score indicated greater symptomatology of depression. The procedure
described below was used to calculate the CES-D cut-off score.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
22
CES-D Cut-off Score
As previously mentioned, 13 original CESD items were retained and 7 items
of the original CES-D were dropped since they were not directly related to DSM-IV
symptoms. Then, ten DSM items were added bringing the number of items to 23.
Since the number of original items was different in this study, Item
Response Theory methodology was used to equate the observed scores of the 23-
item version to the original 20-item scale. In the combined 43-item test (20+23), 13
items were similar in both forms (as 13 items were taken from the original CES-D),
and there were 30 unique items that were different. An equating procedure on a
sample of responses to those 30 items was performed as follows: Samejima’s graded
model was used to calibrate these 30 items together. (Samejima, F., 1969 & Zeng &
Kolen, 1955.) The likelihood of each summed score on the 23-item scale and on the
20-item scale was separately calculated. Then, the theta (EAP) of each summed
score likelihood was assessed. (Thissen et al., 1995.) Table 1, shows the two
summed-score EAP (theta) for both the 23-item scale and the 20-item scale. Using
this table any summed score on one scale can be matched with (EAP or) theta in the
other scale. For instance, summed scores of 1 on the 20-item scale correspond to
EAP (Theta) o f-1.7. Looking at the 23-item scale and finding the same or closest
EAP (Theta) of -1.7 and looking at the summed score of this scale, we find the
summed score of 2. (See Table 1.)
As already noted, previous research has supported a cut-off score of 16 on
the 20-item scale. The current analysis shows that score of 16 is equal to a
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Table 1. IRT for CES-D items
20-Item Scale
Summed Score EAP
23-item Scale
Summed Score EAP
0 -2 .10 0 -2.20
1 -1.70 1 -1.90
2 -1.50 2 -1.70
3 -1.30 3 -1.50
4 -1.20 4 -1.40
5 -1.00 5 -1.30
6 -0 . 93 6 -1.10
7 -0.82 7 -1.00
8 -0.73 8 -0.93
9 -0.65 9
1
o
CO
10 -0.56 10 -0.76
11 -0 .48 11 -0.68
12 -0.40 12 -0 . 61
13 -0.33 13 -0.54
14 -0 .27 14
1
o
15 -0.20 15 -0.41
16 -0.14 16 -0.35
17
00
o
o
1
17 -0.29
18 0.00 18 -0.23
19 0.03 19 -0.17
20 0.09 20
CN
\— 1
O
1
21 0.15 21 -0.06
2 2 0.20 22 -0 .10
23 0.26 23 0.04
24 0.31 24 0 . 09
25 0.36 25 0.14
26 0 .41 26 0.19
27 0.46 27 0.24
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Table 1 Cont. IRT for CES-D items
20-Item Scale
Summed Score EAP
23-item Scale
Slimmed Score EAP
28 0.51 28 0.29
29 0.56 29 0 .34
30 0 . 62 30 0.38
31 0.67 31 0.43
32 0 .72 32 0 .48
33 0.77 33 0.52
34 0.82 34 0.57
35 0.87 35 0 . 62
36 0 . 92 36 0 . 66
37 0.97 37 0.71
38 1.00 38 0.76
39 1.10 39 0.80
40 1.15 40 0.85
41 1.20 41 0.90
42 1.25 42 0 .95
43 1.30 43 0.99
44 1.40 44 1.00
45 1.45 45 1.10
46 1.50 46 1.15
47 1.55 47 1.20
48 1.6 48 1.25
49 1.7 49 1.30
50 1.80 50 1.40
51 1.85 51 1.45
52 1.90 52 1.50
53 2.00 53 1.55
54 2.10 54 1. 60
55 2.20 55 1.70
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
25
Table 1 Cont. IRT for CES-D items
20-Item Scale
Summed Score EAP
23-item Scale
Summed Score EAP
56 2.30 56 1.75
57 2 .50 57 1.80
58 2 . 60 58 1.90
59 2 . 80 59 2.00
60 3 .00 60 2.05
61 2 .10
62 2.20
63 2.30
64 2 .40
65 2.50
66 2 .70
67
2.80
68 3.00
69 3.20
theta of -0.14. Now, the closest value to this theta (-0.14) on the 23-item scale is -
0.12, which corresponds to the summed score of 20 on the 23-item scale. Therefore,
a summed score of 16 on the 20-item scale is equivalent to a summed score of 20 on
the 23-item scale. Consequently, participants scaling 20 or higher on the 23-item
scale were classified as depressed in this study. (Orlando, Sherboume & Thissen,
2000.)
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
26
Composite Interview Diagnostic Interview
The CIDI 2.1 is a comprehensive structured interview, which assesses
various psychiatric disorders based on the criteria of International Classification of
Diseases (ICD) developed by World Health Organization in 1992 and the
Diagnostic and Statistical Manual of Mental Disorders (DSM) developed by the
American Psychiatric Association in 1994.
This interview takes about 75 minutes to administer. It can be used in large
epidemiological studies and can be administered by trained lay individuals. It is
designed for adults 18 years of age and above. In order to administer this interview,
the questions need to be understood by persons with different levels of education,
cultural backgrounds and intellectual capacities. Initially, individuals with dementia,
mental retardation or psychosis were screened out.
CIDI covers a wide variety of disorders of DSM-IV and ICD-10 diagnoses
(See Appendix B.) In this study, only modules D,E, and F covering mood disorders,
bipolar disorders and anxiety disorders were used. Later on, those who fulfilled the
criteria for mania were excluded.
In order to score the interview, the values of each variable are assessed first.
If all the criteria of a diagnosis are positive; then the diagnosis is considered present
at some time in the individual’s life, and in those cases information will be gathered
regarding when their symptoms were last experienced. Following that, the
assessment is made as to whether the diagnosis is current or not.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
27
These steps are followed in the CIDI scoring program for the PC (R.DAT
files). This scoring system gives both the ICD-10 and DSM-IV diagnoses. This
program, first covers the criteria of each diagnosis and then combines the criteria to
develop the diagnoses. (CIDI Manual.)
The SF-36
The patient’s functioning and well-being was assessed using SF-36. (See
Appendix B.) This instrument was first developed in the Medical Outcomes Study
to measure an individual’s health (Hays, Sherboume & Mazel 1993; Ware &
Sherboume, 1992; Ware, Kosinski & Keller, 1995). The items of this instrument
were based on studies measuring health over the last 20 years (Stewart, Ware &
Brook, 1977; Ware, Davies-Avery & Donald-Sherboume, 1978b; Hays & Dimatteo,
1987; Hays & Stewart, 1990; Stewart et al., 1992). Currently, there are two forms
available, one, which has 36 items and the other, which is a shorter version and has
only 12 items. Ware and his colleagues reported that the same norm that has been
used for SF-36 can be used for this version as well. (Ware, Kosinski & Keller,
1994).
Several studies provide support for this instrument’s constructive validity.
Internal reliability estimates are .78 or greater for each of the eight scales
(McHomey, Ware, Rogers, Raczek, & Lu, 1992).
Factor analysis on this instrument was performed and the analysis suggested
a two-factor solution of “Physical Health Composite Scale” and “Mental Health
Composite Scale” as well as eight sub-scales. (Hays et al., 1993).
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
28
As can be seen in Appendix D, four of these sub-scales together construct
the “Physical Health Composite Scale” and the other four sub-scales construct the
“Mental Health Composite Scale”. The “Physical Health Composite Scale”
consisted of items related to perception of health, limitations in moderate activities,
ability to climb several flights of stairs, experiencing bodily pain, accomplishing
less than one would like to, and being limited in the kind of work which can be
performed due to physical health in general. The range of the scores is 10.80 to
71.80 with a higher score indicating more difficulty in the area of physical health.
The “Mental Health Composite Scale” consisted of items related to feeling
calm and peaceful, feeling downhearted and blue, amount of energy, accomplishing
less than one would like or doing work less carefully than usual due to emotional
problems, and frequency and degree of interference of health with social activities.
The range of the scores on this sub-scale is from 4.30 to 69.0 with higher score
indicating more difficulty in the area of mental health.
First, the four sub-scales, which construct the “Physical Health Composite
Scale”, will be discussed. Second, the other Sub-scales representing the “Mental
Health Composite Scale” will be discussed.
Physical Functioning (PF)
This sub-scale consists of two items assessing the individual’s ability to
perform day-to-day activities such as bathing, climbing the stairs, etc. The goal of
this sub-scale is to not only assess the differences between healthy and chronically
ill individuals, but also to differentiate between those who have mild versus
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
29
moderate limitations in their physical functioning. Therefore, a variety of activities,
ranging from the ability to self-care to more vigorous activities, were included in
this sub-scale. Scores on this sub-scale range from 0 to 100 with a higher score
indicating more difficulty in functioning physically. (For more information, see
Appendix D and E.)
Role-Physical (RP)
This sub-scale measures the extent to which physical health affects doing
work or other activities, or engaging less in things they used to do or accomplish
less than before. Including a variety of activities, makes this sub-scale more reactive
to different limitations that people may experience. Scores on this sub-scale range
from 0 to 100 with a higher score indicating more role limitation due to physical
health problems.
Bodily Pain (BP)
Bodily Pain is defined as the subjective discomfort of the body. Measuring
pain is somewhat challenging since people may have different pain thresholds, or
people may experience pain in different ways depending on their ethnicity or gender
(DiMatteo, Friedman, 1982). Not all instruments measure the quality as well as the
duration of pain. In this sub-scale, not only bodily discomfort is measured, but also
the extent to which it affects normal activities is assessed. Previous studies have
shown that the extent to which pain interferes with activities seems to be the best
predictor of health. Again, scores on this sub-scale range from 0 to 100 with a
higher score indicating more bodily pain.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
30
General Health Perceptions (GHP)
General health perception is defined as the personal evaluations of health
status in a more general sense. This sub-scale is strongly correlated with more
specific measures of physical, psychological and social health. (Tessler &
Mechanic, 1978; Ware, et al., 1978a; Ware, Davies-Avery, Donald, 1978b; Wan &
Livierators, 1978). Again, scores on this sub-scale range from 0 to 100 with a higher
score indicating worse perceptions of overall health.
VitalitvfV)
Vitality is defined in terms of a general perceived level of energy. Lack of
energy seems to be a common symptom of many mental and physical disorders.
This scale is adapted from the National Center for Health Statistics General
Well Being Schedule. Several studies have documented this sub-scale’s
effectiveness in measuring the impact of illness on energy level. The range of scores
on this sub-scale is from 0 to 100 with a higher score indicating feeling more tired.
Mental Health Index (MHD
This sub-scale not only measures the participants’ general mood or affect
(e.g. feeling down or happy) but also assesses the presence of certain mental
disorders (e.g. depression or anxiety) (Ware et. al, 1979).
This sub-scale was developed based on the Mental Health Inventory, which
was designed to measure mental health. This sub-scale showed high reliability and
validity in a number of settings (Veit & Ware, 1983). Again, scores on this sub-
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
31
scale range from 0 to 100 with a higher score indicating more difficulty with general
mood and affect.
Social Functioning (SF)
Social functioning is defined as the individual’ s social interactions with
friends and relatives and how it affects their physical or emotional health. (Donald
& Ware, 1984). Again, the range of scores on this sub-scale is from 0 to 100 with
higher score indicating more difficulty in social functioning.
Role-emotional (RE)
This sub-scale measures the extent to which emotional health affects doing
work or other activities (that are typical of their age and social role), or if they
accomplish less than they used to do or are less careful than before.
Most instruments do not consider role limitation due to emotional problems,
in spite of the fact that this factor seems especially relevant in studying mental
health and particularly depression. Most people do not mention limitations unless
they are specifically asked.
Like Role-Physical, this way of questioning, makes this instrument more
effective in assessing even small differences in limitations. Again, scores on this
sub-scale range from 0 to 100 with a higher score indicating more role limitations
due to emotional problems.
Procedures
Seven managed care organizations (MCOS) from different states were
selected (one company had two organizations). The locations of the study were:
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
32
Los Angeles, California; San Antonio, Texas; Twin Cities, Minnesota; San Luis
Valley, Colorado; and Columbia, Maryland.
All primary care clinicians (Internal Medicine specialties and Family
Practice physicians) at these organizations were asked to participate; from which
(97%) agreed. Informed consent, at the screening process was not obtained from
the subjects since no identification information was gathered.
Screener Questionnaire
The interviewers screened patients to see if they were eligible based on their
language comfort, and cognitive impairment prior to giving them the screening
questionnaire. (See Appendix F.) To assess depression, a two-stage process was
used to recognize depressed patients. In the first stage, all participants completed the
screener. In each clinic, patients then were asked to complete a Screening
Questionnaire, which collected information on eligibility, demographic
characteristics, as well as health status. The process of collecting this information
took several months in each clinic. The exclusion criteria were: if the participants
had a medical emergency, were under 18 years old, or did not have an eligible
insurance.
In the second stage, patients who scored equal to or higher than the cutoff
score on the screener were administered the CIDI 2.0 -12 month version to assess
the patient’s current depressive disorder. Patients were categorized into two groups.
One group, those with “Current depressive disorder”, who had an episode of DSM
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
33
major depression or a period of DSM dysthymia during the prior 12 months with no
remission, and a second group, those with “Subthreshold depression” who had
depressive symptoms as reported in the screener (positive score for depression) but
did not have any current depressive disorders.
After patients with depression were identified, they were informed about
the study and were asked to participate. If they agreed and signed the consent
forms, they were administered the affective disorders section of CIDI 2.0. by staff
in person or by phone using the computer administered personal interview (CAPI).
Study participants were also asked to complete a self-administered questionnaire,
which included the measures mentioned previously (CES-D, 7-DSM items, SF-36).
At the beginning of the study, a 2-4 week pilot phase was initiated. During
this time, enrollment was offered only to patients with 12-month depressive
disorder. Depressed patients who were Pregnant or had lifetime mania or recent
alcohol abuse were excluded. After the pilot phase was over, these exclusions were
dropped and patients who scored positive on the screener were offered enrollment.
Those who were excluded during the pilot phase (N=77) were added and continued
in the main study.
Study Participants
Of the many patients approached during the screening phase, 23% were ineligible
and 15% refused (they felt too ill or were not interested). From the 27,332 patients who
completed the screener, 3,918 were potentially eligible. From the remaining 3,918, some
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
scored negative, some refused or walked out. Of those who were 2,417 eligible, 10% had
ineligible insurance, and 13% walked out. Of the 1,854 confirmed eligible cases for whom
informed consent was sought, 20% refused and 1,485 signed consent and took the full
CIDI. From those who took the CIDI, 74 disenrolled and 55 were excluded because of
pilot phase exclusion criteria. From those who were potential eligible and enrolled (N=
1,356), 169 did not return their questionnaires. In the end, (N= 1,187) patients were
entered the study. Therefore, the number of participants was 1,187, of which 867 were
women and 320 were men.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
35
Chapter III
Introduction
This chapter is divided into two sections. In the first section, I discuss data
preparation. I describe the preliminary analyses as well as sample
representativeness, data distribution, missing data and potentially confounding
variables. In the second section of this chapter, I describe the data analyses that
focus on the three research questions.
Data Preparation
Preliminary Analyses
Of the 27,332 patients who completed the screener, 2,417 were eligible
participants. Of these 10% (241) had ineligible insurance, and 13% (322) walked
out. Of the 1,854 confirmed eligible cases, 369 (20%) refused and 1,485 signed
consent forms and took the full CIDI. From those who took the CIDI, 74
disenrolled and 55 were excluded because of pilot phase exclusion criteria. From
those who were potentially eligible and enrolled (N= 1,356), 169 did not return
their questionnaires. In the end, (N= 1,187) patients were entered the study.
Therefore, the number of participants was 1,187, of which 867 were women and
320 were men.
Representativeness of Sample
Several analyses were performed to ascertain if the 1,187 enrolled
participants were representative of the general eligible population. Specifically,
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
36
participants were compared with non-participant counterparts with respect to
demographic variables that might influence the study results. These variables were
gender, ethnicity, age, education, number of chronic conditions, work status and
marital status.
No statistical differences were observed with respect to gender, number of
chronic conditions, age and marital status. Those who participated in the study
seemed to be on the average, working and slightly more educated. Caucasians were
also more likely to participate. These results are summarized in Table 2.
Missing Data
In studies such as these, there always are some missing data. For this data set
only 1 - 2.4 percent of the data were missing across each item. The exceptions were
for two of the SF-36 scales (Physical Health and Mental Health composite scales)
for which 3.1% of the data were missing. The distribution of missing data for each
item was analyzed for the differences in gender. No striking pattern related to
specific items was observed for the data as a whole, or for each gender. Because the
missing data were so few, data imputation was not employed in this study. Instead,
observations with partial data were eliminated from the particular analyses.
Check on Data Distribution and Accuracy
Item variability distribution, was examined for all variables. The results
showed that the response distribution of CES-D and SF-36 items were very similar
across gender. The distributions of CES-D items were normally distributed with the
exception of one item, the suicide question, which had a narrow variance (indicating
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
37
Table 2. Demographic characteristics of non-participants vs.
participants
Demographic
Characteristics
Eligible,
refused
N = 860
Eligible,
enrolled
N = 1,187
X2 or t-value
(p value)
Gender
Female
Male
619 (42.0%)
241 (43.0%)
867 (58.3%)
320 (57.0%)
X2 =.28, p<.59
DF = 1
Ethnicity-
Native Am.
Asian/Pac IS
African Am.
Hispanic
Caucasian
47 (35.30%)
20 (45.40%)
73 (49.71%)
272 (49.90%)
403 (36.31%)
86 (64.70%)
24 (54.50%)
74 (50.30%)
273 (50.10%)
706 (63.70%)
X2 =34.43,
pc.Ol*, DF=4
Age
838(X= 42.44;
SD= 15.20)
1,187(X= 43.48;
SD= 14.3) T =-1.57,
p< . 11
Education
821 (X= 12.59;
SD= 2.94)
1,169(X= 13.19;
SD= 2.71) T =-4.63,
p<.01*
No. of chronic
conditions
0
1-2
3 +
191 (45.10%)
347 (41.30%)
274 (41.20%)
233 (12.1%)
494 (58.7%)
391 (58.8%)
X2 = 1-97,
p<.37, DF =2
Work status
Not working
Working
256 (38.9%)
551 (43.9%)
403 (61.2%)
705 (56.1%)
X2 = 4.47,
p< . 03 *, DF =1
Marital status
Unmarried
Married
403 (43.6%)
450 (42.0%)
522 (56.4%)
630 (58.3%)
X2 =-74, p<.39
DF = 1
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
38
that most people overall, 85.0%, 84.0% men and 85.3% women, responded that they
had not considered suicide). The unadjusted means, standard deviation for all
symptoms, and functioning and well-being items were measured for the entire
sample. (See Table 3a-c and 4a-c.)
As one check accuracy for our data, item-correlations were performed for
the CES-D. These correlations followed the expected pattern. For instance, the
negative items were positively correlated with negative items and they were
negatively correlated with positive items. For instance, ‘I felt depressed’ (item 2)
and ‘I felt sad’ (item 7) were positively correlated (r = .68), whereas T felt happy’
(item 5) and ‘I felt depressed’ (item 2) were highly negatively correlated (r = -.55).
This Correlation matrix is summarized in Table 5.
Potentially Confounding Variables
Prior to assessing symptomatological differences among gender and severity
of depression categories, it was necessary to ascertain if any demographic variables
might interact with these variables and therefore, confound the results. The
following variables were considered as potentially confounding : age, presence of
chronic disease, level of education, work status, marital status and ethnicity.
Analyses indicated that the various samples of depression type were similar with
respect to these variables but some differences in these variables were found
between male and female samples.
For gender, two variables were significant: age (T=-5.49, pc.00) and marital status
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
39
Table 3a. Descriptive information for the CES-D and D S M
variables for the entire sample (N= 1187)
Items Kf N Miss Mean SD
#
Scored
Bottom
Distr.
#
Scored
Top
Distr.
P00CES1 1171 16 1.86 1.00 576 97
P00CES2 1171 16 2 .75 1.02 154 339
P00CES3 1167 20 2.65 1.05 196 310
P00CES4 1161 26 2 .83 1.10 185 431
P00CES5 1172 15 2.38 0.97 232 186
P00CES6 1167 20 2 .49 1.02 223 236
P00CES7 1170 17 2.66 0.99 170 273
P00CES8 1166 21 2 . 65 1.02 191 278
P00CES9 1166 21 2 .32 1.09 351 218
P00CES10 1153 34 1.24 0.63 980 022
P00CES11 1164 23 2 .23 1.00 328 155
P00CES12 1160 27 2 .49 1.02 225 228
P00CES13 1167 20 2.18 1.19 490 254
P00CES14 1163 24 1.99 1.10 550 157
P00CES15 1163 24 2.45 1.06 278 235
P00CES16 1167 20 2.34 1.04 314 187
P00CES17 1165 22 2.66 1.13 252 363
P00CES18 1164 23 2.52 1.13 290 313
P00CES19 1165 22 2.57 1.04 229 259
P00CES20 1159 28 2.52 1.04 240 245
P00CES21 1159 28 1.74 1.00 662 099
P00CES22 1165 22 2.41 1.06 272 240
P00CES23 1169 18 1.94 1.05 555 126
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
40
Table 3b. Descriptive information for the CES-D, D S M
variables for the female sample (N= 867)
Items N N Miss Mean SD Effect
Size
#
Scored
Bottom
Distr.
#
Scored
Top
Distr.
P00CES1 855 12 1.85 0.99
O
O
!
423 070
P00CES2 854 13 2 .77 1.01 0.07 109 252
P00CES3 8 52 15 2.67 1.04 0 . 07 137 232
P00CES4 845 22 2.84 1.10 0 . 04 137 315
P00CES5 854 13 2.35 0.96 -0.11 168 124
P00CES6 849 18 2 .45 1.00 -0.14 163 154
P00CES7 854 13 2.71 0.98 0.17 114 210
P00CES8 851 16 2 . 69 1.02 0.17 135 220
P00CES9 849 18 2 . 32 1.10 0 . 00 257 159
P00CES10 846 21 1.24 0.63 0.00 722 017
P00CES11 847 20 2 .22 1.00 -0.04 243 106
P00CES12 845 22 2 . 47 1.00 -0.07 163 157
P00CES13 851 16 2 . 25 1.22 0.23 346 210
P00CES14 847 20 2.03 1.11 0.14 387 119
P00CES15 847 20 2 .45 1.05 -0 . 01 200 164
P00CES16 849 18 2 .34 1.04 0 . 01 232 134
P00CES17 849 18 2.61 1.13 -0.15 191 245
P00CES18 852 15 2 .52 1.13 0.01 211 227
P00CES19 847 20 2.56 1.04 -0.04 171 186
P00CES20 842 25 2 . 53 1.04 0 . 03 174 179
P00CES21 843 24 1.73 1.01
in
o
o
i
495 074
P00CES22 849 18 2 .40 1.05 -0.04 196 172
P00CES23 851 16 2 .11 1.07 0.60 329 114
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
41
Table 3c. Descriptive information for the CES-D, DSM
variables for the male sample (N= 320)
Items N N Miss Mean SD
#
Scored
Bottom
Distr.
#
Scored
Top
Distr.
P00CES1 316 04 1.89 1. 01 153 027
P00CES2 317 03 2 .70 1.02 045 087
P00CES3 315 05 2 .60 1.06 049 078
P00CES4 316 04 2 .80 1.10 048 116
P00CES5 318 02 2 .46 1.02 064 062
P00CES6 318 02 2.59 1.07 060 082
P00CES7 316 04 2.54 1.00 056 063
P00CES8 315 05 2.52 0 . 99 056 058
P00CES9 317 03 2.32 1.09 094 059
P00CES10 307 13 1.24 0.62 254 005
P00CES11 . 317 ■ 03 2.26 1.02 085 049
P00CES12 315 05 2.54 1. 05 062 071
P00CES13 316 04 1.98 1.09 144 044
P00CES14 316 04 1.88 1.07 163 038
P00CES15 316 04 2 .46 1.09 078 071
P00CES16 318 02 2 .33 1.04 082 053
P00CES17 316 04 2 .78 1.14 061 118
P00CES18 312 08 2 .51 1.15 079 086
P00CES19 318 02 2 . 60 1.03 058 073
P00CES20 317 03 2.50 1. 04 066 066
P00CES21 316 04 1.78 0.97 167 025
P00CES22 316 04 2 .44 1.08 076 068
P00CES23 318 02 1.48 0.85 226 012
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
42
Table 4a. Descriptive information for the SF-36
variables for the entire sample (N= 1187)
Items N NMiss Mean SD Quantile
1
Quantile
3
Pain 1163 24 52 . 84 26.62 32 74
Physical
Function
1186 01 69 .93 28.04 50 95
Role
Physical
1161 26 49.10 42 .10 00 100
Role
Emotion
1161 26 35.99 38.91 00 67
Vitality 1164 23 34.47 19 .59 20 50
Mental
Health
1165 22 48 .37 19.69 36 64
Social
Function
1164 23 52 .48 26.13 37 75
General
Health
1187 00 51.82 23 .56 35 72
Physical
Comp. Score
1151 36 44.07 12 .42 35 54
Mental
Comp. Score
1151 36 33 .94 11.64 25 42
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
43
Table 4b. Descriptive information for the SF-36
variables for the female sample (N= 867)
Items N NMiss Mean SD Effect
Size
Quant.
1
Quant.
3
Pain 851 16 52 . 82 26.50 0.00 32 74
Physical
Function
866 01 70.32 27 .30 0.05 50 95
Role
Physical
850 17 50.05 41.84 0.08 00 100
Role
Emotion
852 15 36.21 39 .16 0.02 00 67
Vitality 854 13 33 .13 19 .43 -0.26 20 45
Mental
Health
854 13 47 .77 19 .48 -0.11 36 60
Social
Function
852 15 52.10 25.66 -0.05 37 75
General
Health
867 00 51.85 23 .64 0.00 35 72
Physical
Comp.
Score
844 23 44.28 12.12 0 . 06 36 54
Mental
Comp.
Score
844 23 33 .55 11.60 -0 .13 25 42
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
44
Table 4c. Descriptive information for the SF-36
variables for the male sample (N — 320)
Items N N Miss Mean SD Quantile
1
Quantile
3
Pain 312 08 52 .90 26.97 31 74
Physical
Function
320 00 68.88 29 .98 45 95
Role
Physical
311 09 46 .49 42 .77 00 100
Role
Emotion
309 11 35.38 38.26 00 67
Vitality 310 10 38.16 19.59 25 50
Mental
Health
311 09 50.02 20.18 36 64
Social
Function
312 08 53 .53 27.39 31 75
General
Health
320 00 51.75 23 .39 30 72
Physical
Comp. Score
307 13 43 .48 13 .19 33 55
Mental
Comp. Score
307 13 35.04 11.70 26 44
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Reproduced w ith permission o f th e copyright owner. Further reproduction prohibited without permission.
Table 5a. Correlation matrix of CES-D and DSM items
Items Item 1 Item 2 Item 3 Item 4 Item 5 Item 6 Item 7 Item 8 M SD
1. Poor appetite 1.87 1
2 . Felt depressed 0.28 2 .74 1.01
3 . Effort 0.25 0.6 2.64 1.04
4. Restless sleep 0.27 0.43 0.42 2 . 83 1.09
5 . Felt happy -0.2 -0.55 -0.43 -0.27 2.38 0.96
6. Enjoyed life -0.17 -0.5 -0.44 -0.22 0.75 2.49 1.02
7 . Felt sad 0.24 0.68 0.48 0.33 -0.51 -0.47 2.65 0.99
8. Couldn't get going 0.24 0.5 0.57 0.38 -0.39 -0.37 0.5 2.64 1.01
9. Shake off blues 0.24 0.69 0.56 0.36 -0.49 -0.46 0.64 0.56 2 .32 1.09
10.Considered suicide 0.19 0.3 0.18 0.14 -0.22 -0.22 0.23 0.16 1.23 0.62
11.Nothing enjoyable 0.23 0.57 0.47 0.29 -0.45 -0.43 0.52 0.44 2.22 1
12.Less interest 0.29 0.59 0.55 0.37 -0.45 -0.43 0.54 0.52 2.49 1.01
13.Eating too much -0.33 0.17 0.19 0.08 -0.08 -0.07 0.14 0.17 2.19 1.19
14.Slept too much 0 0 .22 0.2 0.03 -0.13 -0.13 0.19 0.31 1.98 1.09
15.Moved slower 0.26 0.44 0.53 0.35 -0.35 -0.34 0.41 0.52 2.45 1.06
16.Felt restless 0.26 0.43 0.39 0.37 -0.29 -0.26 0.41 0.31 2 .34 1.03
17.As good as -0.09 -0.34 -0 .27 -0.17 0.39 0.43 -0.34 -0.22 2.65 1.13
18.Blamed self 0.13 0.37 0.3 0.24 -0.25 -0.22 0.34 0.25 2 .52 1.13
19.Keeping mind 0.26 0.52 0.46 0.34 -0.36 -0.32 0.43 0.5 2.57 1.04
20.Couldn't concent. 0.29 0.54 0.5 0.38 -0.37 -0.34 0.44 0.51 2.52 1.04
21.Thought of death 0.2 0.36 0.25 0.18 -0.23 -0.22 0.32 0.18 1.74 0 . 99
22.Hopeful -0.09 -0.32 -0.24 -0.17 0.39 0.41 -0.31 -0.22 2 .41 1.05
23.Crying spells 0.23 0 .42 0.27 0.22 -0.29 -0.24 0.44 0.27 1.93 1.05
Reproduced w ith permission o f th e copyright owner. Further reproduction prohibited without permission.
Table 5a (continued). Correlation matrix of CES-D and DSM items
Items Item 9 Item 10 Item 11 Item 12 Item 13 Item 14 Item 15 Item 16 M SD
1 . Poor appetite 1.87 1
2 . Felt depressed 2.74 1.01
3, Effort
2.64 1.04
4. Restless sleep
2.83 1.09
5. Felt happy 2.38 0.96
6. Enjoyed life 2.49 1.02
7 . Felt sad 2.65 0.99
8. Couldn't get going 2.64 1.01
9. Shake off blues 2 .32 1.09
10.Considered suicide 0.3 1.23 0.62
11.Nothing enjoyable 0.61 0.29 2 .22 1
12.Less interest 0.6 0.25 0.65 2 .49 1.01
13.Eating too much 0.17 0.05 0.14 0.18 2.19 1.19
14.Slept too much 0.23 0.09 0.2 0.25 0.27 1.98 1.09
15.Moved slower 0.48 0.16 0.41 0.52 0.2 0.34 2 .45 1.06
16.Felt restless 0.44 0.18 0.4 0.4 0.13 0.14 0.37 2.34 1.03
17..As good as -0.33 -0.18 0.28 -0.3 -0.14 -0.13 -0.18 -0.21 2.65 1.13
18.Blamed self 0.36 0.2 0.29 0 .32 0.16 0.11 0.22 0.31 2.52 1.13
19.Keeping mind 0.49 0.18 0.41 0.51 0.12 0.24 0.44 0.39 2 .57 1.04
20..Couldn't concent. 0.51 0.17 0.43 0.51 0.16 0.23 0.46 0.45 2.52 1.04
21.Thought of death 0.3 0.47 0.3 0.28 0.08 0.11 0.25 0.32 1.74 0.99
22.Hopeful -0.32 -0.17 -0.3 -0.26 -0.05 -0.07 -0.19 -0.16 2 .41 1.05
23.Crying spells 0 .43 0.3 0.32 0.33 0.14 0.16 0.23 0.33 1.93 1.05
Reproduced w ith permission o f th e copyright owner. Further reproduction prohibited without permission.
Table 5a (continued)« Correlation matrix of CES-D and DSM items
Items Item 17 Item IS Item 19 Item 20 Item 21 Item 22 Item 23 M SD
1. Poor appetite 1.87 1
2. Felt depressed 2 .74 1.01
3. Effort 2.64 1.04
4. Restless sleep 2.83 1.09
5. Felt happy 2.38 0.96
6. Enjoyed life 2.49 1.02
7. Felt sad 2.65 0.99
8. Couldn't get going 2.64 1.01
9. Shake off blues
2.32 1.09
10.Considered suicide 1.23 0.62
11.Nothing enjoyable
2.22 1
12.Less interest
2 .49 1.01
13.Eating too much
2.19 1.19
14.Slept too much
1.98 1.09
15.Moved s1ower
2 .45 1.06
16.Felt restless
2.34 1.03
17.As good as
2.65 1.13
18.Blamed self -0.21
2 .52 1.13
19.Keeping mind -0 .23 0.34
2.57 1.04
20.Couldn't concent.
in
CM
o
i
0.31 0.84
2 .52 1.04
21.Thought of death -0.2 0.28 0.32 0.32 1.74 0.99
22.Hopeful 0.38 -0.15 -0.16 -0.16 -0.18 2 .41 1.05
23.Crying spells -0.27 0.3 0.34 0.35 0.29 -0.13 1.93 1.05
<1
Reproduced w ith permission o f th e copyright owner. Further reproduction prohibited without permission.
Table 5b. Correlation matrix of CES-D and DSM items for female sample
Items Item 1 Item 2 Item 3 Item 4 Item 5 Item 6 Item 7 Item 8 M SD
1. Poor appetite 1.85 0.99
2 . Felt depressed 0.28 2 .77 1.01
3 . Effort 0.26 0.63 2.67 1.04
4. Restless sleep 0.29 0.42 0.46 2.84 1.1
5. Felt happy 0.2 0.53 0.42 0.27 2.35 0.96
6. Enjoyed life 0.19 0.49 0.44 0.22 0.74 2.45 1
7. Felt sad 0.24 0.69 0.5 0.34 0.53 0.49 2.71 0.98
8. Couldn't get going 0 .21 0.51 0.57 0.39 0.39 0.37 0.52 2.69 1.02
9. Shake off blues 0 .24 0.7 0.57 0.37 0.49 0.49 0.67 0.57 2.32 1.1
10.Considered suicide 0.21 0.31 0.19 0.15 0.21 0.2 0.24 0.18 1.24 0.63
11.Nothing enjoyable 0.24 0.56 0.47 0.31 0.44 0.42 0.52 0.46 2.22 1
12.Less interest 0.27 0.59 0.56 0.39 0.44 0.44 0.57 0.55 2.47 1
13..Eating too much -0.39 0.19 0.19 0.07 0.1 0.08 0,14 0.19 2.25 1.22
14..Slept too much 0 0.25 0.22 0.04 0.13 0.14 0.17 0.3 2.03 1.11
15..Moved slower 0.21 0.45 0.54 0.37 0.35 0.35 0.42 0.52 2.45 1.05
16.Felt restless 0.27 0 .45 0.41 0.37 0.31 0.28 0.42 0.32 2.34 1.04
17.As good as 0.12 0.34 0.26 0.19 0.4 0.42 0.33 0.23 2 .61 1.13
18..Blamed self 0.14 0.39 0.31 0.25 0.25 0.24 0.38 0.26 2.52 1.13
19..Keeping mind 0.25 0.55 0.51 0.34 0.38 0.36 0.46 0.52 2 .56 1.04
20..Couldn't concent. 0.27 0.57 0.54 0.39 0.4 0.39 0.45 0.51 2.53 1.04
21..Thought of death 0.23 0.37 0.27 0.18 0.25 0.24 0.32 0.17 1.73 1.01
22 , .Hopeful 0.09 0.33 0.24 0.19 0.41 0.4 0.33 0.24 2.4 1.05
23..Crying spells 0.25 0.5 0.3 0.26 0.33 0.28 0.46 0.28 2.11 1.07
Reproduced w ith permission o f th e copyright owner. Further reproduction prohibited without permission.
Table 5b (continued). Correlation matrix of CES-D and DSM items for female sample
Items Item 9 Item 10 Item 11 Item 12 Item 13 Item 14 Item 15 Item 16 M SD
1. Poor appetite 1.85 0. 99
2 . Felt depressed 2 .77 1.01
3 . Effort 2.67 1.04
4. Restless sleep 2 . 84 1.1
5 . Felt happy 2 .35 0.96
6. Enjoyed life 2 .45 1
7. Felt sad 2.71 0.98
8 . Couldn't get going 2.69 1.02
9. Shake off blues 2.32 1.1
10.Considered suicide 0.31 1.24 0.63
11.Nothing enjoyable 0.62 0.28 2 .22 1
12.Less interest 0.61 0.26 0.65 2 .47 1
13.Eating too much 0.17 0.04 0.13 0.19 2.25 1.22
14.Slept too much 0.23 0.11 0.18 0.26 0.27 2.03 1.11
15.Moved slower 0.47 0.16 0.41 0.53 0.21 0.36 2.45 1.05
16.Felt restless 0.43 0.18 0.39 0.4 0.13 0.15 0.36 2.34 1.04
17.As good as 0.32 0.17 0.27 0.32 0.12 0.13 0.22 0.22 2 . 61 1.13
18.Blamed self 0.38 0.2 0.28 0.34 0.17 0.14 0.23 0.3 2.52 1.13
19..Keeping mind 0.49 0.21 0.42 0.53 0.14 0.25 0.45 0.39 2.56 1.04
20..Couldn't concent. 0.51 0.2 0.44 0.53 0.17 0.23 0.47 0.43 2 .53 1.04
21..Thought of death 0.32 0.49 0.3 0.29 0.07 0.11 0 .25 0.3 1.73 1.01
22 . .Hopeful 0.32 0.17 0.29 0.26 0.07 0.08 0.2 0.17 2.4 1.05
23 . .Crying spells 0.49 0.31 0.38 0.37 0.14 0.16 0.26 0.35 2.11 1.07
Reproduced w ith permission o f th e copyright owner. Further reproduction prohibited without permission.
Table 5b (continued). Correlation matrix of CES-D and DSM items for female sample
Items Item 17 Item 18 Item 19
1. Poor appetite
2. Felt depressed
3. Effort
4. Restless sleep
5. Felt happy
6. Enjoyed life
7. Felt sad
8. Couldn't get going
9. Shake off blues
10.Considered suicide
11.Nothing enj oyable
12.Less interest
13.Eating too much
14.Slept too much
15.Moved slower
16.Felt restless
17.As good as
18.Blamed self 0.22
19.Keeping mind 0.27 0.36
20.Couldn't concent. 0.29 0.33 0.87
21.Thought of death 0.2 0.29 0.35
22.Hopeful 0.39 0.18 0.16
23.Crying spells 0.29 0.37 0.4
Item 20 Item 21 Item 22 Item 23 M SD
1.85 0.99
2.77 1.01
2.67 1.04
2.84 1.1
2.35 0.96
2.45 1
2.71 0.98
2.69 1.02
2.32 1.1
1.24 0.63
2.22 1
2.47 1
2.25 1.22
2.03 1.11
2.45 1.05
2.34 1.04
2.61 1.13
2.52 1.13
2.56 1.04
2.53 1.04
0.34 1.73 1.01
0.17 0.2 2.41.05 °
0.4 0.32 0.15 2.11 1.07
Reproduced w ith permission o f th e copyright owner. Further reproduction prohibited without permission.
Table 5c. Correlation matrix of CES-D and
Items Item 1 Item 2
1. Poor appetite
2. Felt depressed 0.3
3. Effort 0.23 0.55
Item 3
4. Restless sleep 0.25 0.46 0.34
5. Felt happy 0.2 0.56 0.46
6. Enjoyed life 0.14 0.5 0.44
7. Felt sad 0.22 0.65 0 .43
8. Couldn't get going 0.33 0.49 0.56
9. Shake off blues 0.25 0.71 0.54
10.Considered suicide 0.2 0.28 0.18
11.Nothing enjoyable 0.2 0.6 0.46
12.Less interest 0.33 0.56 0.51
13.Eating too much -0.17 0.15 0.16
14.Slept too much 0.03 0.16 0.15
15.Moved slower 0.35 0.42 0.47
16.Felt restless 0.24 0.41 0.33
17.As good as 0.03 0.34 0.29
18.Blamed self 0.09 0.31 0.24
19.Keeping mind 0.3 0.45 0.33
20.Couldn't concent. 0.34 0.46 0.4
21.Thought of death 0.14 0.34 0.2
22.Hopeful 0.11 0.31 0.26
23.Crying spells 0.21 0.32 0.2
DSM items for male sample
am 4 Item 5 Item 6 Item 7 Item 8 M
1.89
2.7
2.6
2.8
SD
1.01
1.02
1.06
1.1
0.24 2 .46 1.02
0.18 0.77 2.59 1.07
0.32 0.46 0.46 2.54 1
0.36 0.4 0.39 0.46 2.52 0.99
0.37 0.49 0.43 0.59 0.53 2.32 1.09
0.15 0.24 0.26 0.23 0.14 1.24 0.62
0.28 0.46 0.46 0.54 0 .41 2.26 1.02
0.34 0.47 0.43 0.51 0.49 2.54 1.05
0.12 0.03 0.04 0.16 0.11 1.98 1.09
0.03 0.13 0.12 0.22 0.31 1.88 1.07
0.31 0.35 0.34 0.39 0.5 2 .46 1.09
0.39 0.26 0.25 0.4 0.29 2.33 1. 04
0.1 0 .37 0.42 0.37 0.19 2.78 1.14
0.24 0.24 0.2 0.23 0.21 2.51 1.15
0.34 0.3 0.26 0.35 0.48 2.6 1.03
0.36 0.27 0.21 0.4 0.5 2.5 1.04
0.18 0.19 0.2 0.32 0.23 1.78 0.97
0.11 0.38 0.43 0.29 0.21 2 .44 1.08
0.12 0.16 0.14 0.37 0.17 1.48 0.85
Reproduced w ith permission o f th e copyright owner. Further reproduction prohibited without permission.
Table 5c (continued). Correlation matrix of CES-D and DSM items for male sample
Items Item 9 Item 10 Item 11 Item 12 Item 13 Item 14 item 15 Item 16 M SD
1. Poor appetite 1.89 1.01
2. Felt depressed 2.7 1.02
3. Effort 2.6 1.06
4. Restless sleep 2.8 1.1
5. Felt happy 2.46 1.02
6. Enjoyed life 2.59 1.07
7. Felt sad 2.54 1
8. Couldn't get going 2.52 0.99
9. Shake off blues 2.32 1.09
10.Considered suicide 0.28 1.24 0.62
11.Nothing enjoyable 0.59 0.36 2.26 1.02
12.Less interest 0.57 0.24 0.66 2 .54 1.05
13.Eating too much 0.21 0.11 0.19 0.2 1.98 1.09
14.Slept too much 0.19 0.06 0.24 0.22 0.23 1.88 1.07
15.Moved s1ower 0.47 0.19 0.39 0 .52 0.15 0.28 2 .46 1.09
16.Felt restless 0.44 0.22 0.44 0.42 0.12 0.1- 0.39 2.33 1.04
17.As good as 0.35 0.23 0.29 0.24 0.18 0.11 0.07 0.21 2 .78 1.14
18.Blamed self 0.3 0.21 0.29 0.25 0.15 0.02 0.19 0.34 2.51 1.15
19.Keeping mind 0.47 0.12 0.38 0.44 0.07 0.22 0.44 0.39 2.6 1.03
20.Couldn't concent. 0.5 0.15 0.4 0.46 0.14 0.21 0.44 0.49 2.5 1.04
21.Thought of death 0.25 0.45 0.26 0.25 0.12 0.12 0.24 0.35 1.78 0.97
22.Hopeful 0.34 0.22 0.34 0.27 0.01 0.05 0.17 0.14 2 .44 1.08
23.Crying spells 0.33 0.33 0.25 0.29 0.07 0.11 0.21 0.31 1.48 0.85
Reproduced w ith permission o f th e copyright owner. Further reproduction prohibited without permission.
Table 5c (continued). Correlation matrix of CES-D and DSM items for male sample
Items Item 17 Item 18 Item 19 Item 20 Item 21 Item 22 Item 23 M SD
1. Poor appetite 1.89 1.01
2. Felt depressed 2.7 1.02
3. Effort 2.6 1.06
4. Restless sleep 2.8 1.1
5. Felt happy 2 .46 1.02
6. Enjoyed life 2.59 1.07
7. Felt sad 2 .54 1
8. Couldn't get going 2 .52 0.99
9. Shake off blues 2 .32 1.09
10.Considered suicide
1.24 0.62
11.Nothing enjoyable
2.26 1.02
12.Less interest
2.54 1.05
13.Eating too much
1.98 1.09
14.Slept too much
1.88 1.07
15.Moved slower
2 .46 1.09
16.Felt restless
2 .33 1.04
17.As good as
2 .78 1.14
18.Blamed self 0.2
2.51 1.15
19.Keeping mind 0.12 0.3
2.6 1.03
20.Couldn't concent. 0.1 0.24 0.78
2.5 1.04
21.Thought of death 0.22 0.28 0.22 0.25
1.78 0.97
22.Hopeful 0.35 0.11 0.14 0.11 0.14 2.44 1.08
23.Crying spells 0.22 0.15 0.25 0.26 0.31 0.1 1.48 0.85
54
(X2 =11.55, Pc.0007, DF = 1). In addition, severity of depression (consisting of
Major Depression and Dysthymia, Major Depression or Dysthymia, and Sub
threshold Depression) (%2 = 8.30, p<.016, DF = 2) and gender showed a significant
association. In general, the female sample was younger, less likely to be married and
had a greater severity of depression. Therefore, where possible, these factors were
controlled or used as covariates in analyses investigating gender differences.
Examining the Research Questions
Research Question One
To answer the first research question “Do men and women present
differently in terms of the symptoms of depression?”, gender differences in
depressive symptoms were analyzed at both the item and the factor levels.
Examining Research Question one at the Item Level. Table 3a-c shows the
means, standard deviations and effect size (the group mean difference divided by
the pooled SD) for males and females for each of the 13 CES items and the 10 DSM
items, which made up the study instrument regarding the symptoms. ( See Appendix
B. for item names.) Preliminary Chi-square analyses were performed to examine
gender differences across depressive symptoms (CES-D & DSM items). Because of
the multiple t-tests and the increased risk of a type I error, the alpha level was
adjusted to 0.002 (.05/23) using a Bonferonni correction. Gender differences were
observed on the following items: item 23 (CES-D), ‘crying spells’ % 2 (3, N = 1169)
= 99.90, p <.00*, effect size of 0.60; items 13 (DSM item), ‘eating too much’, % 2 (3,
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
N = 1167) = 18.17, p <.00, effect size of 0.23*. Men scored nearly one quarter
(23%) SD lower than females on item 13, ‘eating too much’ and women scored 60%
of a SD higher than men on item 23, ‘I had crying spells’. (See Table 6.)
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
5 6
Table 6. Gender differences on CES-D and DSM items
Items (symptoms) Men
(N)
Women
(N)
x2
Value
P Value,
DF= 3
1 . Poor appetite 326 887 0.66 0.89
2 . Felt depressed 327 886 1.14 0 .70
3. Effort 325 885 1.44 0.69
4. Restless sleep 326 877 5.12 0.16
5. Felt happy 328 887 6.04 0.11
6. Enjoyed life 328 882 9 .50 0.02
7 . Felt sad 326 887 6.61 0.08
8. Couldn't get going 325 884 8 . 64 0.03
9 . Shake off blues 327 882 0 .15 0.98
10. Considered suicide* 317 877 0 .74 0 . 86
11. Nothing enjoyable* 327 880 4.19 0 .24
12 . Less interest* 325 877 2 .53 0.47
13 . Eating too much* 326 884 18.41 0 . 00*
14. Slept too much* 325 880 4.59 0 .21
15 . Moved slower* 325 880 3 .06 0.38
16. Felt restless* 328 882 4.24 0 .24
17. As good as 326 880 7.44 0.06
18. Blamed self* 322 885 0.39 0.94
19 . Keeping mind 328 879 0 . 64 0 . 89
20 . Couldn't concent.* 327 874 0 .56 0 .91
21. Thought of death* 326 876 7.30 0 .06
22 . Hopeful 326 882 1.98 0.58
23 . Crying spells 328 884 100.69 0 .00*
* Added DSM Items
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
57
Examining Research Question One at the Scale Level.
Results of Radloff s (1977) factor analysis demonstrated a four-factor solution:
“depressive affect”, “somatic symptoms”, “positive affect and “interpersonal
factor.” In his study, the depressive affect factor included the following items:
‘crying’, ‘sad’, ‘depressed’ and ‘blues.’ The “somatic factor” included ‘appetite,’
‘getting going,’ ‘effort,’ and ‘insomnia.’ The “positive affect” factor included
‘hopeful,’ ‘enjoyed,’ ‘happy,’ and ‘as good.’ And the “Interpersonal factor”
included ‘friendly’, and ‘dislike.’
Based on their subsequent research, however, Ross and Mirowsky (1984),
suggested eliminating the seven items that contained the interpersonal factor
because these items were confounded by factors other than depressed mood (Ross,
Mirowsky, 1984; Schoenbeck, et al. 1982; Boyed, et al., 1982). In addition to
previous findings, for the purpose of approximating the criteria closer to the DSM-
IV, we decided to add 10 additional DSM-IV items, making this a 23 item scale.
Three of these four factors were represented by the 13 original CES-D items
in the present study. The items representing the interpersonal factor were not
included in the present study for reasons discussed above. The 13 items were used
to create scales based on the first three-factor structure (i.e., “positive affect,”
“depressive affect,” “somatic factors”) Radloff (1977) reported. Scores on these
scales were obtained by summing the scores on the items previously shown to load
on these factors and then divided by the number of items. Tables 7 reports ANOVA
results for the three scales using gender as the independent variable. Age and
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
58
severity of depression were entered as co-variates in the analysis. Although
significant results also were obtained for different age groups and severity of
depression, entering these co-variates did not change the result for gender. Gender
differences were noted for scale I “depressive affect” and scale HI “positive affect”,
(see Table 7). Specifically: for scale I, women scored higher (M = 2.48) than men
(M = 2.25) and for scale HI, men scored higher than women (men M =2.57; women
M = 2.45).
Certain items seemed to contribute disproportionately to the significant
gender differences. For instance, it became apparent that the ‘crying item’ from
factor I and ‘enjoyed life’ and ‘as good as others’ from factor HI showed large
gender differences. Did these differences unduly affect the differences noted on the
scales? Would scale differences remain without these three items? To answer these
questions, significance tests between males and females were run on each factor
with and without each individual item on the scale. After dropping those items, the
results became non-significant for factor I “depressive affect” when the ‘crying
item’ was removed, while still significant for factor HI “positive affect” when
‘enjoyed life’ was removed. The result became non-significant for factor HI when
‘as good as others’ was removed. The t value in factor I changed drastically, when
the ‘crying’ item was dropped (from t =-4.23 to t =-1.53), whereas the t value
changed less, when item 17 ‘as good as others’ was removed. (From t =2.22 to t
=1.71.)
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
59
To summarize, whereas significant results were found for the “depressive”
and “positive” affect scales, these differences seemed to be dependent on only two
items. To ensure that this result was not by chance, each item in each scale was
dropped and the t-test was preformed. However, only the two items as mentioned
above, ‘crying’ and “as good as others’, caused a change in the analysis result when
dropped.
Table 7. Gender differences on reported three Scales
Scales Variables DF F-value P Value
Scale I
Gender 1 17.93 <.0001*
Age level 3 2 .46 <.06
Severity of dep 2 63.46 <.0001*
Scale II
Gender 1 1.66 <.19
Age level 3 3.02 <.03*
Severity of dep 2 51.93 <.0001*
Scale III
Gender 1 5.74 <.01*
Age level 3 1.80 <.14
Severity of dep 2 57 .27 <.0001*
Note: Scale I is "depressive affect"; Scale II is "somatic
symptoms"; and Scale III is "positive affect".
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
60
Research Question One: Factor Analysis of CES-D and DSM
Items. For the next analyses, to see if there were gender differences on CES-D and
DSM items, the 13 CES-D items were combined with 10 DSM-TV items to create a
total pool of 23 items. The added DSM questions were interesting, because
clinicians use these criteria in their assessment and some of these items may reflect
gender differences. To investigate how the gender differences were grouped across
these 23 items, a principle components exploratory factor analysis was performed
with a varimax rotation. Four factors were retained by the Kaiser-Guttman rule,
which indicated a four factor solution with eigenvalues of greater than one. (see
Table 8.) The scree plot also suggested the same number of factors. Items with
factor loading of higher than 0.45 were kept in the analysis.
Factor I was named enervation factor (factor loading: .77-.51) and contained
the following items: ‘concentration’, ‘keep my mind’, ‘get going’, ‘moved slower’,
‘everything effort’, ‘less interest’, ‘blues’, ‘depressed’, ‘restless sleep’, ‘ jittery’,
‘nothing was fun’, ‘sad’. Factor II was named lack of positive affect (factor loading:
.78-.61) and contained the following items: ‘enjoyed life’, ‘happy’, ‘hopeful’, ‘and
as good as others’. Factor HI was named guilt factor (factor loading: .75-.47) and
contained the following items: ‘death’, ‘suicide’, ‘crying’, ‘and blaming self’. Factor
IV was named somatic factor (factor loading: .82-.52) and contained the following
items: ‘eating too much’, ‘appetite poor’, and ‘sleep too much’ (see Table 8).
Reliability was measured using Cronbach’s Coefficient Alpha for all the
scales: Scale 10.92, Scale II 0.76, Scale HI 0.62, Scale IV 0.21. Reliability was also
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
6 1
Table 8. Factor pattern with varimax rotation on all the CESD
& DSM items for the entire sample
Items (CES-D and DSM)
Enerv.
factor
( .77-
. 51)
Lack of
pos.affect
( .78-
.61)
Guilt
factor
( -75-
.47)
Somatiz.
factor
( . 82-
.52)
1. Poor appetite 0.44 0 .03 0.18 -0.66
2 . Felt depressed 0 .59 0.47 0 .33 0.06
3 . Effort 0.68 0.35 0.09 0.09
4. Restless sleep 0.58 0.12 0.11 -0 .14
5. Felt happy 0.35 0 .74 0.08 -0.03
6. Enjoyed life 0.31 0.78 0.06 -0.02
7. Felt sad 0.51 0.50 0.27 0.06
8. Couldn't get going 0 .72 0.25 0.00 0.14
9 . Shake off blues 0. 60 0 .46 0.28 0 .08
10. Considered suicide 0.02 0.18 0.72 -0.07
11. Nothing enjoyable 0.52 0.45 0.24 0 .04
12 . Less interest 0.66 0.38 0.17 0.07
13 . Eating too much 0.11 0.07 0.13 0 .82
14. Slept too much 0 .33 0.01 0.03 0 . 52
15 . Moved slower 0.69 0.16 0.03 0 .16
16. Felt restless 0.53 0 .12 0.32 -0.05
17. As good as 0.08 0.61 0.23 0.19
18 . Blamed self 0.27 0 .12 0.47 0 .14
19 . Keeping mind 0.75 0.05 0.28 0.04
20 . Couldn't concentrate 0 . 80 0.05 0.26 0.04
21. Thought of death 0.17 0.09 0.75 -0 . 02
22 . Hopeful 0.04 0.68 0.12 -0.02
23 . Crying spells 0.28 0.21 0.56 0 . 07
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
62
measured for both genders, for females: Scale 10.92, Scale II 0.77, Scale HI 0.64,
Scale IV 0.19 and for males: Scale 10.91, Scale H 0.76, Scale HI 0.57, and Scale IV
0.25.
Again the variable, gender, was entered into the analysis to explore its association
with the discovered factors. Gender loaded only on factor IV the “somatic” factor.
In addition, the crying item, which initially loaded on factor M, moved to factor IV
(suggesting its association with gender).
The result of factor analysis was used to create scales. Scores on these scales
were obtained by summing the scores on the items shown to load on these factors
and then dividing them by the number of items. Items were exclusively incorporated
into the factor for which they had the highest loading. Therefore, the scales did not
share any common items. Another method of computing scales was considered by
utilizing the weight of all 23 items and including all of items in the summation,
which constitutes each scale. Because no differences were found in the resulting
scales related to the method of summation, the first method was used in the
analyses.
MANOVA was run to assess the overall difference. ANOVAs were also run
to assess gender differences on each of the four scales, since the effect of gender
might not be the same for each scale. MANOVA was found not to be significant.
This may not be surprising considering that more items have loaded on scale I,
which made the overall effect non-significant. However, significant gender
differences were noted for Scale H, HI and IV. The result indicated an effect size of
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
63
-.15 for scale n, .24 for scale HI and .20 for scale IV. Women scored lower than
men on Scale II (lack of positive affect): M for men = 2.57, women = 2.45. Women
scored higher on Scale HI and higher on scale IV. Scale ID (guilt factor): M for men
= 1.73, women = 1.90, scale IV (somatic factor): M for men = 0.64, women = 0.82.
ANOVAS also showed a significant result for the severity of depression on
all of the scales. We also entered marital status and age as covariates. Entering these
variables did not change the result on gender differences. However, it showed a
significant result for marital status and age, which will be discussed later (see Table
9).
Again, it was of interest to see if the noted gender differences were unduly
influenced by specific items as was reported earlier for the CES-D scale.
Accordingly each of the four scales was recreated without successive items and the
significance tests were repeated.
For scale II, the result became non-significant only when item 17 ‘as good as
others’ was removed (the T-value changed from T = 2.24 to 1.71). For scale M, the
result became non-significant when item 23 ‘crying’ was removed (the T-value
changed from T=3.77 to 0.15) and for scale IV, it became non-significant when item
13 ‘eating too much’ was removed (the T-value changed from T = -3.12 to -1.82).
Please note that ‘eating too much’ is a DSM item.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
64
Table 9. Multivariate and Univariate analyses for combined
CES-D & DSM scales
Source Multiv. Univariate
Enerv. Positive Guilt Somatiz
Factor Affect Factor Factor
DF
p !a>
DF F DF F DF F DF F
Gender 4 0.91 1 0.37 1 5.11* 1 13.94*** 1 8.91**
Sev. 8 4.29*** 2 62.7*** 2 53.30*** 2 29.28*** 2 2 .22
Age 12 2.85** 3 4.79** 3 1.71 3 1.36 3 0.96
Mar. St. 4 2. 01 1 3 . 82 1 3 .14 1 4.15* 1 0.69
Gender*Sev 8 1.58 2 1.13 2 2.12 2 1.81 2 0.56
Gender*age 12 2.08** 3 2.43 3 1.56 3 1.23 3 1.33
Gender*mar 4 1.45 1 1.00 1 0.02 1 1.16 1 0 . 09
Sev.*age 24 1.27 6 0.72 6 0.36 6 1.46 6 1.74
Sev.*mar 8 0.85 2 0.57 2 0 . 80 2 0 . 07 2 0 . 06
Age*mar 12 1.45 3 0.67 3 1.18 3 3.36* 3 0.78
Note. (a > . F ratios are Wilks's approximation of Fs .
Scale I is "enervation"; Scale II is "lack of positive affect";
Scale III is "guilt factor" and Scale IV is "somatization factor".
* P <.05, ** P <.01, *** P <.001
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
65
Table 9a. Our combined CES-D & DSM scales based on severity
of depression
Scales Depressive
Symptoms
only
Major dep.
or
Dysthymia
Major dep.
&
Dysthymia
Post hoc
Results
M SD M SD M SD
Enervation
Factor
2.25 0.71 2.65 0.71 2.96 0.72 1< 2,3; 2<3
Positive
Affect
2.24 0.77 2.63 0.78 2.94 0.66 1< 2,3; 2<3
Guilt
factor
1.69 0.62 1.91 0.66 2.14 0.69 1< 2,3; 2<3
Somatiz.
Factor
0.73 0.67 0.76 0.79 0.91 0.83
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
66
Table 9b. Our combined CES-D & DSM scales based on different
age groups
Scales Age group1
18-39
Age group2
40-60
Age group3
61-74
Age group4
75 & over
Post
hoc
M SD M SD M SD M SD
Enervation
Factor
2.44 0.77 2.62 0.75 2.47 0.69 2.58 0.59 1 < 2
Positive
Affect
2.52 0.82 2.55 0.80 2.39 0.76 2.34 0.69
Guilt
Factor
1.89 0.70 1.85 0.64 1.71 0.57 1.88 0.60
Somatiz.
Factor
0.76 0.75 0.80 0.77 0.67 0.62 0.61 0.80
Research Question Two
To answer the second question, “are depressed men and women different in
their functioning and well-being”, MANOVA was performed two times using the
following subscales and scales of SF-36 separately as dependent variables:
“Physical Health Composite Scale”,
“Mental Health Composite Scale”,
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
6 7
Subscales:
“Physical Functioning”,
“Bodily Pain”,
“Role Limitation” (due to physical health),
“Role Limitation” (due to emotional health),
“Vitality”,
“General Mood and Affect”,
“Social Functioning” (degree to which health interferes with social
functioning),
“General Health Index” (overall current health; Hays, Sherboume,
and Mazel, 1993).
It has been reported that women have a higher rate of depression and a
higher rate of morbidity in general. It seemed reasonable, therefore, to predict a
lower rate of functioning. Specifically, we expected to find gender differences in
somatic subscales of SF-36 including: “Physical Functioning,” “Vitality,” “Role
Limitations due to Physical Health,” “General Mood and Affect.”
Scales were formed by summing the SF-36 items previously shown to load
on the reported factors. (For a complete description of how this was done please see
Appendix C & D.) For SF-36 scales and subscales, a 2 (gender) by 3 (depression
severity) MANOVA was run. Age and marital status were also added as co-variates
(See Table 10).
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
68
The results indicate that there is a multivariate effect attributed to gender F
(8, 1117) = 2.44, Pc.01; severity of depression F (16, 1117) = 3.39, Pc.001; age F
(24, 1117) = 6.09, Pc.001; and marital status F (8, 1117) = 1.99, Pc.05, as defined
by the Wilks’ Lambda criterion.
No interactions were significant: gender X severity of depression F (16,
1117)= 1.61, P>.05; gender X age F (14, 1117) = 1.37, P>.05; gender X marital
status F (8, 1117) = 1.16, P>.05; severity X age F (48, 1117) = 1.08, P>.05; severity
X marital status F (16, 1117) = 0.95, P>.05; age X marital status F (24, 1117) =
1.00, P>.05.
Univariate analyses for each of the separate variables also were performed
following the multivariate analysis. These analyses indicated a significant result for
the following scales: for the “Physical Health Composite Scale”, age F (3, 1117) =
42.83, Pc.001 was significant; for “Mental Health Composite Scale” gender,
severity of depression and age were significant, (gender F (1, 1117) = 4.85, Pc.05;
severity of depression F(2, 1117) = 70.85, Pc.001; age F (3, 1117) = 5.44, Pc.001).
The analyses also showed a significant result for the following subscales:
Severity of depression was significant for all the subscales. Gender was significant
for only the “Vitality” subscale F (1, 1117) = 16.43, Pc.001; age was also
significant for “Pain” F (3, 1117) = 21.82, Pc.001; “Physical Functioning” F (3,
1117) = 58.46, Pc.001; “Role Limitation due to physical health” F (3,1117) =
24.17, Pc.001; “Social Functioning” F (3, 1117) = 3.69, Pc.01; and “General Health
Index” F (1, 1117) = 11.97, Pc.001. Again there were no interaction effects.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Because the result indicated that there are significant differences in the
severity of depression and for different age groups with several of the SF-36 scales,
a post hoc test (Scheffe) was used to determine where these differences lie. This test
is used to determine which of the differences between group means are significant.
An examination of the means in Table 10b indicates that there are no differences
between severity of depression for the Physical Health Composite Scale. However,
there are differences between group 1 (depressive symptoms only) and group 3
(double depression) for “pain” (group 1: M = 54.99, SD = 26.85; group 3: M =
48.11, SD = 25.52); “Physical Functioning” (group 1: M = 72.08, SD = 27.89;
group 3: M = 64.42, SD = 28.47); and “Role Limitation due to physical health”
(group 1: M = 53.31, SD = 41.93; group 3:M = 40.10, SD = 41.39).
There are also significant differences between all groups: group 1
(depressive symptoms only) and group 2 (Major Depression or Dysthymia), and 3
(double depression) for the following subscales: “Role Limitation due to emotional
health”, “Vitality”, “Mental Health Index”, “Social Functioning”, “General Health
Index”, and “Mental Health Composite Scale”. (Please see Table 10b for
information regarding the M and SD of each group).
Table 10c shows the same analysis for different age groups, except for the
“Role Limitation due to emotional health”, “Mental Health Index” and “Vitality”
subscales which did not show significant differences between their means, the
following two subscales show significant differences between age group 1 (18-39)
and 2 (40-60) for “Social Functioning” (group 1: M = 55.12, SD = 25.16; group 2:
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
M = 49.91, SD = 26.23), “General Health Index” (group 1: M = 54.92, SD = 23.00;
group 3: M = 49.72, SD = 23.80). For the “Pain” subscale, the first age group is
different from the rest. For the following subscales not only was age group 1(18-
39) different from the rest, but age group 2 (40-60) was also different from the rest.
The “Physical Functioning”, “Role Limitation due to physical health”, Physical
Health Composite Scale”. For the “Mental Health Composite Scale”, age group 3
was different from age group 1 and age group 2. (Please see Table 10c for
information regarding the M and SD of each group).
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
71
Table 10a. Multivariate and Univariate analyses for SF-36
scales
Source Multiv. Univariate
DF
p (a|
Pain
DF F
Physical
function
DF F
Role
physical
DF F
---
—
Gender 8 2.44** 1 0.02 1 0.86 1 2.38
Sev. of Dep. 16 3.39*** 2 4.23** 2 4.95** 2 5 .66**
Age 24 6.09*** 3 2 1. 82*** 3 58.46*** 3 24.17***
Marital St. 8 1.99* 1 0.03 1 0.26 1 0.75
Gender*Sev. 16 1.61 2 0.04 2 0.27 2 2.92
Gender*age 24 1.37 3 0.57 3 0.08 3 1.66
Gender*mar. 8 1.16 1 1.94 1 1.00 1 1.03
Sev.*age 48 1.08 6 1.09 6 0.46 6 1.46
Sev.*mar. 16 0.95 2 0.52 2 3.37 2 0.31
Age*mar. 24 1.00 3 0.93 3 0.90 3 0.69
Note. < a> F ratios are Wilks's approximation of Fs .
* P <.05, ** P <.01, *** P <.001
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
72
Table 10a Cont. Multivariate and Univariate analyses for SF-
36 scales
Source Multiv. Univariate
DF
p ( a )
Role
emotional
DF F
Vitality
DF F
Mental
health
index
DF F
Gender 8 2.44** 1 0 . 04 1 16.43*** 1 3 . 62
Sev. of Dep. 16 3.39*** 2 30.12*** 2 49.68*** 2 69.71***
Age 24 6.09*** 3 0.15 3 2.55 3 2.22
Marital St. 8 1.99* 1 0 .37 1 0.38 1 0.97
Gender*Sev. 16 1.61 2 1.57 2 0 .27 2 0 . 90
Gender*age 24 1.37 3 0.39 3 1.81 3 1 . 88
Gender*mar 8 1.16 1 1.77 1 3 .16 1 0.05
Sev.*age 48 1.08 6 0 .74 6 0 .40 6 0.45
Sev.*mar 16 0.95 2 0.18 2 0.15 2 0.92
Age*mar 24 1.00 3 0.38 3 1.13 3 0.30
Note. < a) F ratios are Wilks's approximation of Fs.
* P <.05, ** P <.01, *** P <.001
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
73
T a b le 1 0 a Cont. Multivariate and Univariate analyses for SF-
36 scales
Source Multiv. Univariate
Social
function
General
health
Physical
Composite
Scale
DF
p ( a )
DF F DF F DF F
Gender 8 2.44** 1 0.68 1 0 .13 1 1.11
Sev. of Dep. 16 3.39*** 2 39.67*** 2 9 .09*** 2 1.24
Age 24 6.09*** 3 3.69** 3 3.82** 3 42.83***
Marital St. 8 1.99* 1 11.97*** 1 0.03 1 0.03
Gender*Sev. 16 1.61 2 1.99 2 0.65 2 0.69
Gender*age 24 1.37 3 0.90 3 2.30 3 0.81
Gender*mar 8 1.16 1 1.06 1 2 .77 1 3.38
Sev.*age 48 1 . 08 6 1 . 81 6 0.87 6 1. 07
Sev.*mar 16 0.95 2 0.41 2 2 . 67 2 2.36
Age*mar 24 1.00 3 1.61 3 0.98 3 1.41
Note. ( a ) F ratios are Wilks's approximation of Fs.
* P <.05, ** P <.01, *** P <.001
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
7 4
Table 10a Coxxt. Multivariate and Univariate analyses for SF-
36 scales
Source Multiv.
Mental
Composite
Scale
DF
p < a>
DF F
Gender 8 2.44** 1 4.85*
Sev. of Dep. 16 3.39*** 2 70.85***
Age 24 6 . 09*** 3
5.44***
Marital St. 8 1.99* 1 1.71
Gender*Sev. 16 1.61 2 1.35
Gender*age 24 1.37 3 0.66
Gender*mar 8 1.16 1 0.18
Sev.*age 48 1.08 6 0.15
Sev.*mar 16 0.95 2 1.29
Age*mar 24 1.00 3 0.85
Note. < a) F ratios are Wilks's approximation of Fs.
* P <.05, ** P <.01, *** P <.001
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
75
Table 10b. SF-36 scales based on severity of depression
Variables Depressive
Symp. only
Major dep.
or
Dysthymia
Major dep.
&
Dysthymia
Post hoc
Results
M SD M SD M SD
Pain 54.99 26 . 85 52.31 26.57 48 .11 25 .52 1 < 3
Physical
Function
72 . 08 27 . 89 69.61 27 .88 64.42 28.47 1 < 3
Role Phys. 53 .31 41.93 48.00 42 .06 40.10 41.39 1 < 3
Role Emo. 46.02 41.01 31.34 37 .02 22.15 30.90 1 <
2 <
2,3;
3
Vitality 41.09 20.01 31.53 18.00 25.01 17.37 1 <
2 <
2,3;
3
Mental
H. Index
56 .00 18 .74 44.85 18.70 37 .91 17 . 61 1 <
2 <
2,3;
3
Social
Function
60.20 25.86 48.98 24.98 41.47 24.49 1 <
2 <
2,3;
3
General
H. Index
54.87 24 . 01 50.92 22.74 45.66 23 . 80 1 <
2 <
2,3;
3
Physical
C. Scale
44.15 12 .67 44.36 12 .30 42 .74 12 . 03
Mental
C. Scale
38.58 1 1. 60 31.66 10 .77 27.99 09 .52 1 <
2 <
2,3;
3
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
7 6
Table 10c. SF-36 scales based on different age groups
Variables Group 1
18-39
Group 2
40-60
Group 3
61-74
Group 4
75 & over
Post hoc
M ? SD M ; SD M s SD M ; SD
Pain 60.28;
25.72
48.17;
25.76
47 .20;
27.67
3 9.97;
24.82
1 < 2,3,4
Physical
Function
80.97;
22 . 61
65.14;
28.56
53.12;
27.76
42.50;
25 .28
1 < 2,3,4
2 < 3,4
Role Phys. 60.31;
40.69
44.40;
41.58
3 0.36;
39.37
18.68;
28.32
1 < 2,3,4
2 < 3,4
Role Emo. 3 8.06;
38.86
34.80;
39.31
34.85;
38.98
28.74;
30.50
-----
Vitality 3 5.81;
19 .20
32 .78;
19 .73
37.42;
19.95
34.94;
19 . 90
Mental
H. Index
48.66;
19 .38
46.99;
19 . 90
54.42;
20 .14
49.69;
15.13
Social
Function
55.12;
25.16
49.91;
26.23
55.99;
28.16
46.98;
27 . 88
1 < 2
General
H. Index
54.92;
23.00
49 .72;
23 . 80
50.90;
24.48
45.42;
19.48
1 < 2
Physical
C. Scale
48.46;
11.11
42.05;
12 .21
37.34;
12 .37
33 .33;
10 .45
1 < 2,3,4
2 < 3,4
Mental
C. Scale
32.85;
11.66
33 . 82;
11.52
39.01;
11.66
37.25;
08 .72
3 < 1,2
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
77
Research Question Three
The third question was “Does severity of depression symptoms vary by
gender?” To answer this question, a 2 (gender) by 3(depression severity)
MANOVA was performed with the four scales derived from the factor analysis of
the CES-D and DSM items. Age and marital status were also added as co-variates in
the analyses.
The result indicates that there is a multivariate effect attributable to the severity of
depression F (8, 1006) = 4.29, Pc.001; age F (12, 1006) = 2.85, Pc.01; and
gender*age F (12, 1006) = 2.08, Pc.01, as defined by the Wilks’ Lambda criterion
(See Table 9).
Univariate analyses for each of the separate variables were also performed
following the multivariate analysis. The following results were significant: Scale I.
severity of depression F (2, 1006) = 62.7, P c.001 and age F (3, 1006) = 4.79, P
<■01: Scale n , gender F (1, 1006) = 5.11, P c.001, severity of depression F (2, 1006)
= 53.30 P <001; Scale m gender F (3, 1006) = 13.94, P c.001, severity of
depression F (2, 1006) = 29.28, P c.001, marital status (1, 1006)= 4.15, P c.05,
Scale IV gender F(l, 1006) = 8.91, P <01.
The only interaction that the result of the univariate analysis showed to be
significant was between age and marital status for Scale IH (guilt factor) F (3, 1006)
= 3.36, Pc.05. It seems that for age group 1 (18-39), people who were not married
experienced more guilt compared to their married counterparts. For the second age
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
78
group (40-60), the score for both married and non-married individuals was the same.
The score for unmarried people decreases until the third age group (61-74), and then
the score increases for the last age group of 75 and over.
For married individuals, the score for the first age group (18-39) is less than
their not married counterparts, and it slightly increases for the second age group (40-
60). For married individuals when we move from age group 2 (40-60) to 4 (75 and
over); their score steadily decrease. (See Graph 1).
Since the result indicated that there are significant differences in the severity
of depression and different age groups for our created scale, a post hoc test
(Scheffe) was used to determine where these differences lie. Looking at Table 9a
indicates that there are significant differences between all severity groups, group 1
(depressive symptoms only), group 2 (Major Depression or Dysthymia) and group 3
(double depression) for all the Scales except Scale IV. (Scale I (group 1: M = 2.25,
SD = 0.71; group 2: M = 2.65, SD = 0.71; group 3: M = 2.96, SD = 0.72); Scale E
(group 1: M = 2.24, SD = 0.77; group 2: M = 2.63, SD = 0.78; group 3: M = 2.94,
SD = 0.66); Scale IE (group 1: M = 1-69, SD = 0.62; group 2: M = 1-91, SD = 0.66;
group 3: M = 2.14, SD = 0.69)). We also observed that the mean score increases as
the severity of depression increases. (See Table 9a.)
Table 9b shows that the only significant difference for all age groups and our
created scales is between age group 1 (18-39) and age group 2 (40-60) on Scale I
(enervation) (group 1: M = 2.44, SD = 0.77; group 2: M = 2.62, SD = 0.75). The
mean score for group 1 is less than group 2.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
79
Graph 1. Marital status by a g e for Scale ill
2.5
2.25 -
‘Married
‘Not Married
— 1.75
1.25
3 1 2 4
Age groups
Going back to the third question, the univariate analysis results showed a
significant result on all of the created scales for severity of depression and a
significant result based on gender for only scales II (positive factor), IH (guilt
factor), and IV (somatization factor). However, there was no interaction effect,
suggesting that the depressive symptomatology varies based on gender and across
the range of depressive disorders. But, depressive symptoms over the severity of
depression do not vary based on gender. (See Table 9).
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
80
Chapter IV
Discussion
Introduction
The cost of treating people with depression is rapidly increasing. These
days, based on the need for cost-containment strategies, many managed care
organizations are sending clients with depression to primary care clinicians rather
than to specialists.
Therefore, the detection of depression in its earlier stages by primary care
clinicians seems to be very crucial. Unfortunately, many primary care clinicians,
when dealing with depressed patients, often miss the underlying factors related to
their patients’ affective problems, and pay more attention only to their physical
symptoms.
Previous research reports that men seen by the primary care clinicians are at
a higher risk than women for having their depression undiagnosed. Perhaps this is
because women are twice as likely to be diagnosed as being depressed and therefore
clinicians are more attuned to their likelihood of depression. (Potts, Bumam, Wells,
1991.)
It is possible that the higher proportion of diagnosis of depression in women
is related to the way they present symptoms. Awareness of the factors that influence
symptom presentation can be of great help in diagnosis, especially when the
diagnosis depends on what patients think is important enough to mention. Being
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
8 1
unaware of those crucial factors and their impact on the patients’ responses may
lead to missing important diagnostic cues. (Zola, 1966). Therefore, in this study, we
first examined symptom presentation in men and women, asking “Is there a gender
difference in symptom presentation of depression?”
To answer this question, several analyses were performed based on both the
individual items related to symptoms and on scales created from those items.
Examining the mean differences, the result indicated that women reported more
symptoms and specifically reported “crying spells” and “eating too much” more
frequently.
To investigate the differences at a broader construct level, three scales were
created based on a factor analysis by Radloff s study (1977). These were
“depressive affect”, “somatic symptoms”, and “positive affect”. Age and severity of
depression were entered as co-variates in the analysis. Although significant results
were also obtained for different age groups and severity of depression, entering
these co-variates did not change the result for gender. Significant gender differences
were noted on scale I “depressive affect” and scale III “positive affect”. Women
scored higher on “depressive affect” and lower on “positive affect”. Contrary to our
expectation, no gender differences were found in the somatic factor.
Meaning of depression and gender differences
Some authors have reported that women characteristically present with a
more emotional way of expressing depression (Hammen & Peters, 1977). We
became interested as to whether men and women interpret the symptoms differently
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
82
and whether certain items function differently for men and women. Hammen and
Peters (1977) found that society evaluates depressed males and females differently.
Specifically, men with psychological problems are viewed more negatively than
their female counterparts; this effect is much more pronounced when patients
present with depressive symptoms (Hammen & Peters, 1977).
As we consider depressive symptoms, we need to be mindful that expressing
and reporting a symptom in and of itself, is part of a social process and ignoring the
meaning it has to each gender may lead us to miss important aspects of the
diagnosis (Zola, 1966). Results from this study suggest that crying has a different
meaning for men and women. We need to be aware of the implications and the
impact of the reaction, which is expected in society for men and women, to be
different. Therefore, more caution needs to be paid in interpreting this item in the
context of gender.
Factor analysis on the original CES-D items & DSM items
As mentioned before, a factor analysis was performed on the CES-D and
DSM items together. The result suggested a four-factor solution. The following
items loaded on factor I, which we called “enervation factor”: ‘concentration’, ‘keep
my mind’, ‘get going’, ‘moved slower’, ‘everything effort’, ‘less interest’, ‘blues’,
‘depressed’, ‘restless sleep’, ‘ jittery’, ‘nothing was fun’, and ‘sad’. Items on factor
II, which we called “lack of positive affect” were: ‘enjoyed life’, ‘happy’, ‘hopeful’,
and ‘as good as others’. Items on factor HI, which we called “guilt factor” were:
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
83
‘death’, ‘suicide’, ‘crying’, and ‘ blaming self. Items on factor IV, which we called
“somatic factor” were : ‘eating too much’, ‘appetite poor’, and ‘sleep too much’.
Consistent with Radloff s findings, we were also able to produce a four
factor structure. We also reproduced a somatization factor. But, at the same time,
this factor was different from the one reported in Radloff s study. In addition to
appetite/sleep disturbance, Radloff s scale included ‘get going’ and ‘effort’. These
two items factored out in our first factor.
Adding gender as a variable, changed the factor structure. The crying item,
which initially loaded on factor HI, moved to factor T V with gender. This indicates
the close association of this item with gender.
Effect of gender on created scales based on our findings
Using the four factor (analytically derived) scales created from our analysis
of the CES-D and DSM items, ANOVA showed a significant gender difference on
scale II (lack of positive affect), HI (guilt) and IV (somatization), but non-significant
results on scale I (enervation factor). Women scored lower on scale II (lack of
positive affect) and higher on scale HI (guilt) and scale IV (somatization). The
analysis also showed a significant result for the severity of depression on all of the
scales. We also entered marital status and age as covariates. Entering these variables
did not change the results related to gender differences. However, it showed a
significant result for marital status and age, which will be discussed later (see Table
9a).
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
84
The differential impact of items on each gender was considered. Factor
structures (scales) were recreated without each item and t-tests were repeated. The
result became non-significant only when the following items were removed: for
scale HI, 17 ‘as good as others’ and item 23 ‘crying’; for scale IV, item 13 ‘eating
too much’. This result is the same as the previous findings reaffirming the high
impact of items 23 and 17. Note that item 13 (eating too much) was a DSM item and
therefore was not included in the previous analysis. Therefore, it appears that there
is a possibility that this item is biased as well.
Gender differences on functioning and well-being
We were also interested in learning whether one reason for the higher
proportion of women being diagnosed with depression could be related to their
physical health being generally worse than men. The high prevalence of depression
in women has been documented by previous research as well as the higher rate of
morbidity in women (Chou, 1994). One reason this is a matter of concern is that,
according to previous findings, depression has a more debilitating effect on overall
functioning and well being than many chronic illnesses. Wells, Strum, Sherboume
and Meredith (1996) report that the rate of functional limitation remained highest
among the oldest, women, and the unmarried.
Women, in fact, may have more limitations in terms of their functioning and
sense of their well being. The second question that this study attempts to answer is
“Do depressed men and women differ in their functioning and well-being?”
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Using a sample of bipolar patients, Robb, Young, Cooke & Joffes (1998)
found that women experienced greater impairment in functioning and well-being on
subscale scores of SF-36, with significant impact on physical health and pain. We
did not, however, replicate this finding. We expected to find gender differences in
somatic subscales of SF-36 including: “Physical Functioning”, “Vitality”, “Role
Limitations due to Physical Health”, “General Mood and Affect”. Unexpectedly, in
our results, the only variables that showed significant gender differences were the
“Vitality” subscale (amount of time spent on feeling energetic vs. being tired) and
the “Mental Health Composite Scale.” Where women scored lower than men in both
cases. For severity of depression, the result indicated a significant difference on all
the scales and subscales of SF-36 except “Physical Health Composite Scale.” The
results also indicated that there were no interaction effects between gender and
severity of depression.
Age and marital status were entered as co-variates in the analysis, but with
no change in the results for gender. However, the analysis did show a significant
result for marital status and age. For marital status, the analysis showed that married
individuals scored higher on “Social Functioning” than un-married individuals. For
age, the analysis showed that individuals between ages 18-39 scored higher on
“Social Functioning Scale,” and “General Health Index” than individuals between
40-60. No interaction effect was found to be significant.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
86
Gender differences on depressive symptoms across the severity of depression
Young et al. (1990) suggested that the relationship between gender and
depression may not be the same across the range of all depressive disorders.
Newman (1984) reported that, whereas women showed higher levels of depressive
symptomatology than men, the differences between men and women were not
constant over the range of the severity of depression. They found small differences
for the most severe forms, a mixed pattern for moderately severe forms and
substantial differences only for the less severe forms. Therefore, the third question
we attempted to answer is, “Does depressive symptomatology over the severity of
depression vary by gender?”
To answer this question, a 2 (gender) by 3 (severity of depression)
MANOVA on the scales formed by the CES-D and DSM items was performed. This
analysis showed significant results on all of the created scales for severity of
depression except scale IV (somatization) and a significant result based on gender
for scale II (lack of positive affect), HI (guilt factor), and IV (somatization factor).
Again, age and marital status were entered as co-variates in the analysis,
which is worth mentioning. The analysis indicated a significant result for age on
scale I (enervation factor) and for marital status on scale HI (guilt factor).
Returning to our third question, Newman’s (1984) and Young (1990)
findings that tried to explain gender differences in relation to the severity of
depression by suggesting that the gender differences are not constant over the
severity of depression. In contrast, we did not find any interaction to be significant
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
87
between severity and gender. This suggests that the depressive symptomatology
varies according to gender and also that it varies across the range of depressive
disorders. However, depressive symptoms over the severity of depression do not
vary based on gender and vice versa. The only interaction was between age and
marital status for Scale III (guilt factor). Those in the 18-39 year age group who
were not married experienced more guilt compared to their married counterparts.
Guilt scores for un-married people decreased until the third age group (61-74), then
increased for the last age group of 75 and over. These results are depicted in
Graph 1.
Study Limitations
In most studies conducted in clinical settings, patients are identified as
depressed by their clinicians. In contrast, patients in this study were evaluated for
depression based on their self-reports. We realize that this type of approach, may
create a less severely depressed sample.
Other potential limitations of this study, were: inclusion of only specific
managed care organizations that may not be representative of all managed care
groups in the U.S., and exclusions of patients who were too physically sick to
participate in the study. However, the main purpose of this study was to examine
gender differences, and both genders reasonably would be similarly affected by this
self-report feature of our design. Therefore, they do not seem to be functioning as
limiting factors in this particular study.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Summary
Many people compare depression to the common cold of mental health. Also
an increased number of patients are being sent to primary care clinicians, who
therefore are put in the position of diagnosing this highly prevalent disorder. For the
primary care clinicians to accurately interpret the patient’s symptoms and detecting
depression early seems to be very crucial.
In general, women are twice as likely to be diagnosed as depressed when
compared to men. But at the same time primary care clinicians tend to underdetect
depression in men (Potts, Bumam, Wells, 1991). One possible reason for this gender
difference may be due to the manner in which patients present their depressive
symptoms. Many studies have reported gender differences in symptom presentation.
In considering gender differences, it is important to pay attention to influencing
factors which impact men and women differently. It is especially when the actual
diagnosis very much depends on what they report (Zola, 1966).
In answer to our first question, “Is there a gender difference in symptom
presentation of depression?”, we found several gender differences. In almost all of
the analyses, the result indicated that reports of frequent crying may be a good
indicator of depressed mood in women, but not in men. The same is true for reports
of eating too much. The assumption that a symptom has the same meaning in men
and women may not be true in all cases.
If a symptom is to be generalized across two groups, the meaning that
symptom has to both groups’ members should be the same. Society in general is
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
89
more accepting of the emotional expression of depression by women, especially
through crying. This likely accounts for our finding that men report crying less
often. There is also more stress and pressure concerning women’s appearance and
body weight. Therefore, it is not surprising that many women report ‘eating too
much’ symptoms with higher frequency. Perhaps ‘Crying’ and the other item
mentioned here (‘eating too much’), may not be good indicators of depressed mood
in men.
In response to our second question, “Do depressed men and women differ in
their functioning and well-being?”, we found that women also reported greater
impairment in functioning and well-being on subscale scores on “vitality” and
“Mental Health Composite Scale”. It appears that women’s depression tends to limit
their functioning mostly in the vitality and mental health area.
In regards to some other mental health disorders, the diagnostic criteria in
DSM dictates that clinicians evaluate the extent to which the problem interferes with
one’s functioning. However, the clinicians do not need to consider the extent of
limitation in functioning when evaluating someone for depression. This result has
been brought up by several investigators and may be worthy of further investigation.
In answer to the third question, “Does depressive symptomatology over the
severity of depression vary by gender?”, unlike some previous studies which
reported that the relationship between gender and depression may not be the same
across the range of the severity of depression, we did not find that to be true.
Depressive symptomatology varied across the range of depressive disorders. They
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
90
also vary by gender. However, depressive symptoms over the severity of depression
did not vary by gender. Therefore, it is not likely to have substantial gender
differences in the less severe forms of depression and smaller gender differences
when the severity of depression is high.
In summary, open and full reporting of gender comparisons should provide a
basis for gender-sensitive awareness in order to detect depression early. As we gain
an understanding of the presenting symptoms of depression, primary care clinicians
may benefit from broadened training in the diagnosis of this illness in men and
women. These types of comparisons may provide information for primary care
clinicians, which not only allow for better diagnosis of their patients, but also allow
for better patient management.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
91
References
1. Ahnlund, K., Frodi, A. (1996). Gender differences in the development of
depression. Scandinavian Journal of Psychology. 37, 229-237.
2. Amenson, C.S., & Lewinsohn, P.M. (1981). An investigation into the observed
sex difference in prevalence of unipolar depression. Journal of Abnormal
Psychology, 90, 1-13.
3. American Psychiatric Association. (1994). Diagnostic and Statistical Manual of
Mental Disorders. (4th ed.). Washington, DC.
4. Angst, J., & Dobler-Mikola, A. (1984). Do the diagnostic criteria determine the
sex ratio in depression? Journal of Affective Disorder. 7, 189-198.
5. Barsky, A.J. (1979). Patients who amplify bodily sensations. Annals of Internal
Medicine, 91, 63-70.
6. Beck, A.T., Steer, R.A., Garbin, M.G. (1988). Psychometric properties of the
Beck Depression Inventory: Twenty-five years of evaluation. Clinical Psychology
Review. 8, 77-100.
7. Beck, A.T., Ward, C.H., Mendelson, M., Mock, J., & Erbaugh, J. (1961). An
inventory for measuring depression. Archives of General Psychiatry, 4, 561-571.
8. Ben-Porath, Y.S. (1991). Contribution of MMPI-2 content scales to the
differential diagnosis of schizophrenia and manic depression. Psychological
Assessment: A Journal of Consulting & Clinical Psychology, 3, 634-640.
9. Bemdt, D.J. (1990). Inventories and scales. In E.D. B.B. Wolman, G. Strieker, et
al. (Eds.), Depressive disorders: Facts, theories, and treatment methods, (p. 436).
New York, NY: John Wiley & Sons.
10. Bolla-Wilson, K. & Blecker, M. L. (1989). Absence of depression in elderly
adults. Journal of Gerontology: Psychological Sciences, 44, 53-55.
11. Boyd, J.H., Weissman, M.M., Thompson, W.D., & Meyers, J.K. (1982).
Screening for depression in a community sample: Under-standing discrepancies
between depression symptom and diagnostic scales. Archives of General Psychiatry,
39, 1195-1200.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
92
12. Broadhead, W.E., Blazer, D.G., George, L.K., & Tse, C.K. (1990). Depression,
disability days, and days lost from work in a prospective epidemiological survey.
Journal of American Medical Association. 264, 2524-2528.
13. Butcher, J.N., Aldwin, C.M., Levenson, M.R., Ben-Porath, Y.S., Spiro, A., &
Bosse, R. (1991). Personality and aging: A study of the MMPI-2 among older
adults. Psychology and Aging, 6, 361-370.
14. Chino, A.F., & Funabiki, D. (1984). A cross-validation of sex differences in the
expression of depression. Sex Roles, 11, 175-187.
15. Chou, S.P. (1994). Sex differences in morbidity among respondents classified as
alcohol abusers and/or dependent: result of a national survey. Addiction, 89, 87-93.
16. Cleary, P.D. (1987). Gender differences in stress-related disorders. In E.D. R.C.
Barnett, L. Bierer, & G.K. Baruch (Eds.), Gender and Stress (pp. 39-72). New York:
Free Press.
17. Coulehan, J.L. Schulberg, H.C., Block, M.R., Janosky, J.E., & Arena, V.C.
(1990). Depressive symptomatology and medical co-morbidity in a primary care
clinic. International Journal of Psychiatry in Medicine, 20, 335-347.
18. Craig, T.J., & Van Natta, P.A. (1979). The influence of demographic
characteristics on two measures of depressive symptoms. Archives of General
Psychiatry, 36, 149-154.
19. Dahlstrom, W.G., Welsh, G.S., Dahlstrom, L.E. An MMPI handbook. Volume I:
Clinical interpretation. Minneapolis: University of Minnesota Press, 1972.
20. Dimatteo, M.R., Friedman, H.S. (1982). Social Psychology and Medicine.
Cambridge, MA: Oelgeschlager, Gunn and Hain.
21. Donald, C.A., Ware, J.E. (1984). The measurement of social support. In: J.R.
Greenley (ED). Research in Community and Mental Health, (p. 325). Greenwich,
CT: JAI Press.
22. Frank, E., Carpenter, L.L., & Kupfer, D.J., (1988). Sex differences in recurrent
depression: Are there any that are significant? American Journal of Psychiatry, 145,
41-45.
23. Friedman, H.S., Dimatteo, M.R. (1989). Health Psychology. Englewood Cliff,
New Jersey: Prentice Hall.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
93
24. Goldberg, D.P. & Bridges, K. (1988). Somatic presentations of psychiatric
illness in primary care setting. Journal of Psychosomatic Research, 32,137-144.
25. Hammen, C.L., DeMayo, R. (1982). Cognitive correlates of teacher stress
and depressive symptoms: Implications for attributional models of depression.
Journal of Abnormal Psychology. 91, 96-101.
26. Hammen, C.L. & padesky, C.A. (1977). Sex differences in the depression of
depressive responses on the Beck Depression Inventory. Journal of Abnormal
Psychology, 87, 322-332.
27. Hammen, C.L. & Peters, S.D. (1977). Differential responses to male and female
depressive relations. Journal of Consulting and Clinical Psychology, 45, 994-1001.
28. Hankin, J.R., Locke, B.Z. (1983). Extent of depressive symptomatology among
patients seeking care in a prepaid group practice. Psychological Medicine, 13(1),
121-129.
29. Harlow, S.D., Goldberg, E.L., & Comstock, G.W. (1991). A longitudinal study
of the prevalence of depressive symptomology in elderly widowed and married
women. Archives of General Psychiatry, 48, 1065-1068.
30. Harris, R., Lingoes, J. (1968). Subscales for the Minnesota Multiphasic
Personality Inventory (mimeograhed materials). Department of Psychology,
University of Michigan.
31. Hays, R.D., DiMatteo, M.R. (1987). A short-form measure of loneliness. Journal
of Personality Assessment, 51, 69-81.
32. Hays, R.D., Sherboume, C.D., & Mazel, R.M. (1993). The 36-Item Health
Survey. 1.0 Health Economics, 2: 217-227.
33. Hays, R.D. & Stewart, A.L. (1990) The structure of self-reported health in
chronic disease patients. Psychological Assessment: A Journal of Consulting and
Clinical Psychology, 2: 22-30.
34. Hunsley, J., Hanson, R. K, Parker, K. C. (1988). A summary of the reliability
and stability of MMPI scales. Journal of Clinical Psychology. 44,44-46.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
9 4
35. Johnson, J. Weissman, M.M., & Klerman, G.L. (1992). Service utilization and
social morbidity associated with depressive symptoms in the community. Journal of
American Medical Association; 267, 1478-1483.
36. Katon, W., Kleinman, A. & Rosen, G. (1982). Depression and somatization: a
review. Part 1. American Journal of Medicine, 72, 127-134.
37. Katz, M.M., Wetzler, S., Cloitre, M., Swann, A., Secunda, S., Mendels, J.,
Robins, E. (1993). Expressive characteristics of anxiety in depressed men and
women. Journal of Affective Disorder, 28: 267-277.
38. Klerman, G.L., & Weissman, M.M. (1992). The course, morbidity and costs of
depression. Archives of General Psychiatry, 49, 831-834.
39. Kessler, R.C., Brown, R.L., & Broman, C.L. (1981). Sex differences in
psychiatric help-seeking: evidence from four large-scale surveys. Journal of Health
Social Behavior, 22,49-64.
40. Kessler, R.C., McGonagle, K.A., & Zhao, S. (1994). Lifetime and 12-month
prevalence of DSM-IU-R psychiatric disorders in the United States: Results from
the National Comorbidity Survey. Archives of General Psychiatry, 51, 8-19.
41. Lehman, A.F., Ward, N.C., & Linn, L.S. (1982). Chronic mental patients: the
quality of life issues. American Journal of Psychiatry, 139, 1271-1276.
42. Marcus, A.C., & Seeman, T.E. (1981). Sex differences in reports of illness and
disability: a preliminary test of the “fixed role obligations” hypothesis. Journal of
Health Social Behavior, 22, 174-182.
43. Mazor, K.M. (1998). Using statistical procedures to identify differentially
functioning test items. Educational Measures: Issues and Practice, 18, 31-44.
44. McHomey C.A., Ware, I.E., Rogers, W.H., Raczek, A. & lu, J.F.R. (1992). The
validity and relative precision of MOS Short- and Long-Form Health Status Scales
and Dartmouth COOP Charts: Results from the Medical Outcomes Study. Medical
Care: 30 (suppl.): MS253-MS265.
45. Morgan, B.S. (1976). Intimacy of Disclosure topics and sex differences in self
disclosure. Sex Roles, 2, 161-66.
46. Myers, J.K., et al. (1984). Six-month prevalence of psychiatric disorder in three
communities: 1980 to 1982. Archives of General Psychiatry, 41(10), 959-967.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
9 5
47. Myers, J.K., Weissman, M.M., Tischler, G.L., Holzer m, C.E., Leaf, P.J.,
Orvaschel, H., Anthony, J.C., Boyd, J.H., Burke, J.D., Kramer, M. & Stoltzman, R.
(1984). Six month prevalence of psychiatric disorders in three communities.
Archives of General Psychiatry, 41, 959-967.
48. Newman, J.P. (1984). Sex differences in symptoms of depression: clinical
disorder or normal distress? Journal of Health and Social Behavior. 25, 136-159.
49. Oliver, S.J., & Toner, B.B. (1990). The influence of gender role typing on the
expression of depressive symptoms. Sex Roles, 22, 775-790.
50. Orlando, M., Sherboume, C., & Thissen, D. (2000). Summed-score linking
using Item Response Theory: Application to depression measurement.
Psychological Assessment, 12, 354-359.
51. Padesky, C.A. & Hammen, C.L. (1981). Sex differences in depressive symptom
expression and help-seeking among college students. Sex Roles, 7, 309-320.
52. Parker, K.C., Hanson, R.K., Hunsley, 1.(1988). MMPI, Rorschach, and WAIS:
A meta-analytic comparison of reliability, stability, and validity. Psychological
Bulletin. 103, 367-373
53. Potts, M.K., Bumam, M.A., & Wells, K.B. (1991). Gender differences in
depression detection: a comparison of clinician diagnosis and standardized
assessment. Psychological Assessment: A Journal of Consulting and Clinical
Psychology, 3, 609-615.
54. Radloff, L.S. (1977). The CES-D scale: A self-report depression Scale for
research in general population. Applied Psychological Measurement, 1, 385-401.
55. Rapp, S.R., Parisi, S.A., Walsh, D.A. & Wallace, C.E. (1988). Detecting
depression in elderly medical inpatients. Journal of Consulting and Clinical
Psychology, 56, 509-513.
56. Repetti, R. L., Crosby, F. (1984). Gender and depression: Exploring
the adult-role explanation. Journal of Social & Clinical Psychology,
2, 57-70.
57. Robb, J., Young, L.T., Cooke, R.G., Joffes, R.T. (1998). Gender differences in
patients with bipolar disorder influence outcome in the medical outcome survey
(SF-20) subscale scores. Journal of Affective Disorder, 49, 189-193.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
9 6
58. Roberts, R.E., Andrews, J.A., Lewinsohn, P.M., Hops, H. (1990). Assessment of
depression in adolescents using the Center for Epidemiological Studies Depression
Scale. Psychological Assessment. 2. 122-128.
59. Ross, C.E., & Mirowsky, J. (1984). Components of depressed mood in married
men and women. American Journal of Epidemiology, 119, 997-1004.
60. Samejima, F. (1969). Estimation of latent ability using a response pattern of
graded scores. Psychometric Monograph No. 17. Iowa City, IA: Psychometric
Society.
61. Santor, D.A., Zuroff, D.C., Cervantes, P., Palacios, J., Ramsay, J.O. (1995).
Examining scale discriminability in the BDI and CES-D as a function of depressive
severity. Psychological Assessment. 7, 131-139.
62. Schoenbeck, V.J., Kaplan, B.H., Grimson, R.C., et al. (1982) Use of symptom
scale to study the prevalence of a depressive syndrome in young adolescents.
American Journal of Epidemiology. 116, 791-800.
63. Schulberg, H.C., et al. (1985). Assessing depression in primary medical and
psychiatric practices. Archives of General Psychiatry, 42(12), 1164-1170.
64. Sheppard, D., Smith, G.T., Rosenbaum, G. (1988). Use of MMPI
subtypes in predicting completion of a residential alcoholism treatment program.
Journal of Consulting & Clinical Psychology. 56, 590-596.
65. Steer, R.A., Beck, A.T., Brown, G. & Berchick, R.J. (1987). Self-reported
depressive symptoms differentiating major depression from Dysthymic disorders.
Journal of Clinical Psychology, 43, 246-250.
66. Steer, R.A., Beck, A.T., Riskind, J. & Brown, G. (1986). Differentiation of
depressive disorders from generalized anxiety by the Beck Depression Inventory.
Journal of Clinical Psychology, 40, 475-478.
67. Stewart, A.L. and Ware, J.E., editors. (1992) Measuring Functioning and well
being: The Medical Outcomes Study Approach. Durham, NC: Duke University
Press.
68. Stewart, A.L., Ware, J.E., Brook, R.H. (1977). A study of the reliability,
validity, and precision of scales to measure chronic functional limitations due to
poor health. Santa Monica, CA: RAND Corp.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
97
69. Tessler, R.; Mechanic, D. (1987). Psychological distress and perceived health
status. Journal of Health & Social Behavior, 19(3), 254-262.
70. Thissen, D. Pommerich, M., Billeaud, K., Williams, V.S.L. (1995). Item
response theory for scores on tests including polytomous items with ordered
responses. Applied Psychological Measurement, 19(1), 39-49.
71. Van Grootheest, D.S., Beekman, A.T.F., Van Groenou, M.I.B., Deeg, DJ.H.
(1999). Sex differences in depression after widowhood: Do men suffer more? Social
Psychiatry & Psychiatric Epidemiology. 34, 391-398.
72. Veit, C.T., Ware, J.E. (1983). The structure of psychological distress and well
being in general populations. Journal of Consulting and Clinical Psychology; 5:
730-742.
73. Verbrugge, L. Sex Differences in morbidity and mortality in the United States.
(1976). Sociological Biology, 23, 275-296.
74. Verbrugge, L.M. (1979). Comments on Walter R. Gove and Michael Hughes,
Possible causes of the Apparent Sex Differences in Physical Health. American
sociological Review, 23, 507-513.
75. Verbrugge, L.M. (1986). Role burdens and physical health of women and
men. Women & Health, 11, 47-77.
76. Vredenburg, K., Krames, L. & Fleet, G. (1986). Sex differences in the clinical
expression of depression. Sex Roles, 14, 37-49.
77. Wan, T.T.H., Livieratos, B. (1978). Interpreting a general index of subjective
well-being. Health and Society, 56, 531-56.
78. Ware, J.E., Davies-Avery, A., Brook, R.H. et al. (1978a). Associations Among
Psychological Well Being and Other Health Status Constructs. Santa Monica, CA:
RAND Corp.
79. Ware, J.E., Davies-Avery, A., Donald, C.A. (1978b). The Conceptualization and
Measurement of Health for Adults in the Health Insurance Study. Volume V,
General Health Perceptions. Santa Monica, CA: RAND Corp.
80. Ware, J.E., Johnston, S.A., Davies-Avery, A. et al. (1979). Conceptualization
and measurement of health for adults in Health Insurance Study: Vol. M, Mental
Health. Santa Monica, CA: RAND Corp.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
98
81. Ware, J.E., Kosinski, M. & Keller, S.D. (1994) SF-36 Physical and Mental
Summary scales: A user’s Manual. Boston, MA: The Health Institute.
82. Ware, J.E., Kosinski, M. & Keller, S.D. (1995) SF-12 Physical and Mental
Summary scales. Boston, MA: The Health Institute, New England Medical Center.
83. Ware, I.E., Kosinski, M., Byliss, M.S., McHomey, C.A., Rogers, W.H., &
Raczek, A. (1995). Comparison of methods for scoring and statistical analysis of
SF-36 health profiles and summary measures: summary of results from the Medical
Outcomes Study. Medical Care, 33, AS264-AS279.
84. Ware, I.E., Sherboume, C.D. (1992). The MOS 36-Item Short-Form Health
Survey (SF-36): I. Conceptual framework and item selection. Medical Care, 30,
473-483.
85. Weissman, M.M., & Klerman, G.L. (1977). Sex differences in the epidemiology
of depression. Archives of General Psychiatry, 34, 98-111.
86. Weissman, M.M., Sholomskas, D., Pottenger, M., Prusoff, B.A., & Locke, B.Z.
(1977). Assessing depressive symptoms in five psychiatric populations: A validation
study. American Journal of Epidemiology; 106, 203-213.
87. Wells, K.B., Bumam, M.A., Rogers, W., Hays, R., et al. (1992). The course of
depression in adult outpatients: Results from the Medical Outcomes Study. Archives
of General Psychiatry, 49, 788-794.
88. Wells, K.B., Golding, J., & Bumam, M.A. (1988) Psychiatric disorder and
limitations in physical functioning in a general population. American Journal of
Psychiatry. 145,712-717.
89. Wells, K.B., Hays, R.D., Bumam, M.A., Rogers, W., Greenfield, S., & Ware, J.
(1989). Detection of depressive disorder for patients receiving prepaid or free-for-
service care: Results from the Medical Outcome Study. Journal of the American
Medical Association. 262, 3298-3302.
90. Wells, K.B., Manning, W.G., Duan, N., Newhouse, J.P. & Ware, J.E. (1986).
Use of outpatient mental health services by a general population with insurance
coverage. Hospital and Community Psychiatry, 37, 1119-1125.
91. Wells, K.B., Stewart, A.L., Hays, R.D., & Bumam, A., Rogers, W., Daniels, M.,
Berry, S.D., Greenfield, S. & Ware, I.E., Jr. (1989). The functioning and well-being
of depressed patients: Results from the Medical Outcomes Study. Journal of the
American Medical Association, 262, 914-919.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
9 9
92. Wells, K.B., Strum, R., Sherboume, C.D., Meredith, L.S. (1996). Caring for
Depression. Harvard University Press, Cambridge, MA.
93. World Health Organization. (1997). Composite International Diagnostic
Interview (CEDI) (Ver. 2.1). Sydney, Australia.
94. Wilson, R.W., Drury, T.F. (1984). Interpreting trends in illness and disability:
health statistics and health status, Annual Rev Public Health. 15, 83-106.
95. Young, M.A., Scheftner, W.A., Fawcett, J., & Klerman, G.L. (1990). Gender
differences in the clinical features of unipolar major depressive disorder. Journal of
Nervous and Mental Disease, 178, 200-203.
96. Zeng, L. & Kolen, M.J. (1955). An alternative approach for IRT observed-score
equating of number correct score. Applied Psychological Measurement, Vol. 19,
No. 3, 231-240.
97. Zimmerman, M., Coryell, W. (1994). Screening for major depressive disorder in
the community: A comparison of measures. Psychological Assessment, 6(1), 71-74.
98. Zola, I. (1966). Culture and Symptoms: An analysis of patients’ processing
complaints. American Sociological Review, 31, 615-630.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
1 0 0
Appendix A. Study Design
: - No Consent
- Assess e l i g i b i l i t y
- Age: 18-96
- Care a t the c l i n i c
- Visiting the study cli ni c ia n
- No medical emergency
- Right insurance plan
_____________________ Screening Questionnaire
- Demographic I n f o .
- Health Status
- Assess Depression
- 5 CIDI Items
- 2 additional items
assessing mood i n
the l a s t month
- WHO (CIDI 2.0 -12 month)
- DSM MD 12 month
- DSM Dysthymia 12 month
- Subthreshold Depression
(Screened + on screener
no current D/o)
- If e l i g i b l e ; get consent
- Patient Assessment Quest.
- Symptoms (CES-D)
- DSM items
- Functioning & well-being
______________________ - CIDI (DSM & ICD
(Diagnosis and s e v er i ty )
Participants: 1,187.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
1 0 1
Appendix B. CES-D & DSM Questions
How often have you f e l t or behaved t h i s way during the l a s t week?
(Circle One Number On Each Line)
Rarely or Some or a Occasionally Most or a l l
none of the l i t t l e of (3-4 days) of the time
time ( l e s s the time (5-7 days)
than 1 day) (1-2 days)
1 I did not f e e l l i k e e ating; my
appetite was poor
1 2 3 4
2 I f e l t depressed. 1 2 3 4
3 I f e l t everything was an e f f o r t . 1 2 3 4
4 My sleep was r e s t l e s s . 1 2 3 4
5 I was happy. 1 2 3 4
6 I enjoyed l i f e . 1 2 3 4
7 I f e l t sad. 1 2 3 4
8 I could not "get going”. 1 2 3 4
9 I f e l t t h a t I could not shake off
the blues even with help from
my family or f r i e n d s .
1 2 3 4
10 I considered s u i c id e . 1 2 3 4
11 Nothing was fun or enjoyable. 1 2 3 4
12 I had l e s s i n t e r e s t in my usual
a c t i v i t i e s .
1 2 3 4
13 I was eating way too much. 1 2 ■ 3 4
14 I s lept too much. 1 2 3 4
15 I moved much slower than
normal.
1 2 3 4
16 I was r e s t l e s s or j i t t e r y . 1 2 3 4
17 I f e l t that I was j ust as good as
other people.
1 2 3 4
18 I blamed myself for the way
things a r e .
1 2 3 4
19 I had trouble keeping my mind
on what I was doing.
1 2 3 4
20 I could not concentrate. 1 2 3 4
21 I thought a l o t about death. 1 2 3 4
22 I f e l t hopeful about the f u t u r e . 1 2 3 4
23 I had crying s p e l l s . 1 2 3 4
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Appendix C. SF-36 Questionnaire
THE MOS 36-ITEM SHORT-FORM HEALTH SURVEY (SF-36)
INSTRUCTIONS: This survey asks for your views about your health. This information
will help keep track of how you feel and how weH you are able to do your usual activities.
Answer every question by marking the answer as indicated. If you are unsure about how
to answer a question, please give the best answer you can.
1. In general, would you say your health is:
(circle one)
Excellent................................................................................................ 1
Very g o o d ............................................................................... 2
Good .....................................................................................................3
Fair.......................................................................................................... 4
P o o r .......................................................................................................5
2. Compared . to one .year ago, how would you rate your health in general ugw?
(circle one)
Much better now than one year ago .......................................... 1
Somewhat better now than one year a g o ............... 2
About the same as one year- a g o ........................... 3
Somewhat worse now than one year a g o .............................. 4
Much worse now than one year ago ............... 5
Copyright® 1992 Nsw England MsdlcsS Gsntsr HaspHate. Inc.
A N rights r ts m c d .
U.K. Vsrsion of St&nd&d $F<38 Heafth Survey s/93
Source: SF-36 Manual
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
3. The following questions are about activities you might do during a typical day.
Does vour health now lim it you in these activities? If so, how much?
(circle one number on each line)
ACTIVITIES
Yes,
Limited
A Lot
Yes,
Limited
A Little
NO, Not
Limited
At All
a. Vigorous activities, such as running, lifting
heavy objects, participating in strenuous
sports
1 2 3
b. Moderate activities, such as moving a
table, pushing a vacuum cleaner, bowling,
or playing golf
1 2 3
c. Lifting or carrying groceries ' 1 2 3
d. Climbing several flights of stairs 1 2 3
e. Climbing one flight of stairs 1 2 3
f. Bending, kneeling, or stooping 1 2 3
g. Walking more than a mile 1 2 3
h. Walking half a mile 1 2 3
i. Walking one hundred yards 1 2 3
J . Bathing or dressing yourself 1 2 3
4. During the past 4 weeks, have you had any of the following problems with your work
or other regular daily activities as a result of vour physical health?
_______ (circle one number on each line)
YES NO
a. Cut down on the amount of time you spent on
work or other activities
1 2
b. Accomplished less than you would like
1 2
c. Were limited in the kind of work or other
activities
1 2
d. Had difficulty performing the work or other
activities (for example, it took extra effort)
1 2
Copyright® 199S Ne«* England M e a l Cantai MoapiMs, Inc.
A ft fights fassfvsd.
UK. Version of Standard SF-36 Hsaim Survvy 5/93
Source: SF-36 Manual
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
5. During the past 4 weeks, have you had any of the following problems with your work
or other regular daily activities as a result of anv emotional problems (such as feeling
depressed or anxious)?
_________________________ (circle one number on each line)
YES NO
a. Cut down on the amount of time you spent on work or other
activities
1 2
b. Accomplished less than you would like 1 2
c. Didn’ t do work or other activities as carefully as usual 1 2
6. During the past 4 weeks, to what extent has your physical health or emotional
problems interfered with your normal social activities with family, friends, neighbours,
or groups?
(circle one)
Not at a l l .......................................................................... . ......................1
Slightly .......................................................................... 2
Moderately ..............................................................................................3
Quite a b it ............................................................................ 4
Extremely ................... 5
7. How much bodily pain have you had during the oast 4 weeks?
(circle one)
None ............................................ 1
Very m ild ..................................................................................................2
Mild............................................................................................................. 3
Moderate ................. 4
Severe........................................................................................................ 5
Very se v e r e ..................................... 6
Copyright® 1982 Now England Maenad Cam®; Hospitals, Inc.
M rights rw rved.
U.K. Version of Standard SF-3S Hoallh Survey s/93
Source: SF-36 M anual
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
8. During the past 4 weeks, how much did pain interfere with your normal work
(including both work outside the home and housework)?
(circle one)
Not at a l l ..................................................................................................1
A little b it ..................................................................................................2
Moderately .............................................................................................3
Quite a b it ........................................................................... 4
Extremely ............................................................... 5
9. These questions are about how you feel and how things have been with you during
the past 4 weeks. For each question, please give the one answer that comes closest
to the way you have been feeling. How much of the time during the past 4 weeks -
(circle one number on each line)
All
of the
Time
Most
of the
Tim®
A Good
Bitot
the
Time
Som e
ot the
Tim®
A
Little
of the
Time
None
of the
Time
a. Did you feel full of life? 1 2 3 4 5 6
b. Have you been a very
nervous person?
2 3 4 5 6
c. Have you felt so down
in the clumps that
nothing could cheer you
up?
1 2 3 4 5 6
d. Have you felt calm and
peaceful?
1 2 3 4 5 6
e. Did you have a lot of
energy?
1 2 3 4 5 6
f. Have you felt
downhearted and tow?
1 2 3 4 5 6
g. Did you feel worn out? 1 2 3 4 5 6
h. Have you been a happy
person?
1 2 3 4 5 6
i. Did you feel tired? 1 2 3 4 5 6
Copyright ® 198S New England I M M Csntw HsspM s, Ine.
M rights raeoivod.
U.K. V m isn o l SWMtesri SF-33 HssaWi Survey 6 /8 3
Source: SF-36 Manual
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
10. During the oast 4 weeks, how much of the time has your physical health or emotional
problems interfered with your social activities (like visiting with friends, relatives, etc.)?
(circle one)
A ll of the time .........................................................................................1
Most of the time ....................................................................................2
Some of the t im e ................................................................................... 3
A little of the time .................................................................................4
None of the time ....................................................................................5
11. How TRUE or FALSE is each of the following statements for you?
____________________________ (circle one number on each line)
Definitely
True
Mostly
True
Don’t
Know
Mostly
False
Definitely
False
a. 1 seem to get ill more
easily than other
people
1 2 3 4 5
b. 1 am as healthy as
anybody I know
1 2 3 4 5
c. 1 expect my health to
get worse
1 2 3 4 5
d. My health is excellent 1 2 3 4 5
Copyright ® 1892 Now England Medical Cantor Hospitals. Inc.
tot rights reserved.
U.K. Vterskm of Standard SF-36 Health Survey S/93
Source: SF-36 Manual
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
107
Appendix D: SF-36 Measurement Model
Hems
Scales
Summary
Measures
3a. Vigorous ActMttes
3b. Moderate AdMtles
3c. Lift. Catty Groreulos
3d. Climb Severs! Flights
3e. Climb One Fight
31. Bend. Kneel
3g. Walk Mite
S R :
Walk Several Blocks.
31. Walk One Block
3|. Bathe. Dress
4a. Cut Down Time
4b. Accomplslwd Less
4c. Limited H Kind
4d. Had DHIIcuPy
Paln-M»gn|tude -
P a ln - ln te r f e r a —
1. EVGFP Paling
I to. Sick Easter
1 lb. As Healthy
11c I lesllh To Gel Worse -
11d. Health Excellent -
Physical Functioning (PF)
Role-Physical (RP)
; Bodily Pain (BP)
General Health (GH)
Physical
Health
9a. Pep/LHe -
Be. Energy----
9g. Worn Out -
6. SocbrC xlanl-
10. Social-Tim® —
; Vitality (VT)V
5a. Cut Down lim e ------
5b. AccompPsbed Less -
5c. Not C areful-----------
Social Functioning (SF)*-
: Role-Emotional (RE) -
Mental
Health
9b. Nervous
9c. Dawn In Dumps
9d. Peaceful
91. DkrelSad —
9h. Happy
Mental Health (M H )-
Significant correlation with other sum m ary m easure.
: SF-36 Manual
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
108
Appendix E. SF-36 Scoring
FILENAME IN ’ C:\MANUAL\RAWDATA’ ;
4 =
* PROGRAM: SF36SUMM *
* PURPOSE: SAS SCORING PROGRAM FOR THE SF-36 *
* *
* SF-36 SCALE SCORING EXERCISE (SECOND EDITION). *
* COPYRIGHT 1992, 1994 MEDICAL OUTCOMES TRUST. *
* ALL RIGHTS RESERVED. *
* *
* SF-36 IS A REGISTERED TRADEMARK OF MEDICAL OUTCOMES
TRUST. *
* *
* SAS IS A REGISTERED TRADEMARK OF SAS INSTITUTE, INC., CARY
NC. *
9
?
STEP 1: INPUT DATA ^ ***;
DATA SF36DATA;
INFILE IN;
INPUT ID $ 1-3
@ 5 (GH1 HT PF01-PF10 RP1-RP4 RE1-RE3 SF1
BP1-BP2 VT1 MH1 MH2 MH3 VT2 MH4 VT3 MH5
VT4 SF2 GH2 GH3 GH4 GH5) (1.);
RUN;
jf c j{c i f c # ^ :j s r jc 5fc ^ > j c >j< ^ % . % >jc >Jc sfc ;f s Jfc >fs ^ ■ % ^ >fc ^ s}c s}c : j c 5{c s j ; # # ?fc 3fc > £ >jc >jc ^ >jc >j« >Jc ^ # s |c .
9
*** STEP 2: SF-36 SCALE CONSTRUCTION , ***;
?|C ?JC ?jc ^ ?ji ?fC ?jt ^ jjc 5|C £jc jJ. /Jc )[% jjc *jc 5j£ 9|C ^ ? |\ ^ *{£ 5jc ^ ^ 5jc 5j* ?|C ?j> jjc ?|C 5|* ^jl 5ji ^ jf ; ^ 3|? 5|C jjt 5ji 5}» *{• 5j» ^ •
9
* « t ^ * « , ^ ,4 . J> 4* 4/ 4« 4/ k l* 4^ 4 k 4* 4k 4k 4k 4* 4 k k lk 4k * 4 * 4k 4 k k U * 4 # 4k 4 k 4k 4s 4k 4k ■ 4 '* v tk 4. 4 ^ v L * .J. * J > 4k 4 k 4k 4k 4 k 4k 4k 4 k k L . 4 * 4. 4. 4 k 4k k^ k jk 4 k k J > k ik 4 k * 4 k 4k 4f 4k 4 k
k ] k k |k k jk 2 J. k j^ k jk kj. ^k k jk k jk k ] k k^ ^k k ] k kj. ^k k jk k jk ^k ^k ^k ^ J > k j, k f * k jk k f f c ^ ^ k^ k jk > J > k^ k { k k^ - k j k ^ k f k > f k k [ k k j. k ^ i ^k ^k k jv k jk k j k , . k j k . k ^ k k ^ k 4 * k f k k |k k * J .
* USING THE SAS DATASET CREATED IN PART 1, CHANGE OUT-OF-
RANGE
* VALUES TO MISSING FOR EACH ITEM. RECODE AND RECALIBRATE
ITEMS
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
109
* AS NEEDED. AN R ’ PREFIX MEANS THE VARIABLE IS RECODED.
9
DATA SF36SCAL;
SET SF36DATA;
* THE SF-36 PHYSICAL FUNCTIONING INDEX.
* ALL ITEMS ARE POSITIVELY SCORED - THE HIGHER THE ITEM
* VALUE, THE BETTER THE PHYSICAL HEALTH.
*
* THIS SCALE IS POSITIVELY SCORED.
* THE HIGHER THE SCORE THE BETTER THE PHYSICAL FUNCTIONING.
9
ARRAY PFI(IO) PF01-PF10;
DOI=l TO 10;
IF PFI(I) < 1 OR PFI(I) > 3 THEN PFI(I) = .;
END;
PFNUM = N(OF PF01-PF10);
PFMEAN = MEAN(OF PF01-PF10);
DO I = 1 TO 10;
IF PF1(I)= • THEN PFT(I) = PFMEAN;
END;
IF PFNUM GE 5 THEN RAWPF = SUM(OF PF01-PF10);
PF = ((RAWPF - 10)/(30-10)) * 100;
LABEL PF = ’ SF-36 PHYSICAL FUNCTIONING (0-100)’
RAWPF = RAW SF-36 PHYSICAL FUNCTIONING’ ;
* THE SF-36 ROLE-PHYSICAL INDEX.
* ALL ITEMS ARE POSITIVELY SCORED - THE HIGHER THE ITEM
VALUE,
* THE BETTER THE ROLE-PHYSICAL FUNCTIONING.
*
* THIS SCALE IS POSITIVELY SCORED.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
1 1 0
* THE HIGHER THE SCORE THE BETTER THE ROLE-PHYSICAL.
ARRAY RPA(4) RP1-RP4;
DO I = 1 TO 4;
IF RPA(I) < 1 OR RPA(I) > 2 THEN RPA(I) = .;
END;
ROLPNUM = N(OF RP1-RP4);
ROLPMEAN = MEAN(OF RP1-RP4);
DO 1=1 TO 4;
IF RPA(I) = . THEN RPA(I) = ROLPMEAN;
END;
IF ROLPNUM GE 2 THEN RAWRP = SUM(OF RP1-RP4);
RP = ((RAWRP - 4)/(8-4)) * 100;
LABEL RP = ’ SF-36 ROLE-PHYSICAL (0-100)’
RAWRP = RAW SF-36 ROLE-PHYSICAL’ ;
* THE SF-36 PAIN ITEMS.
* ITEM RECODING DEPENDS ON WHETHER BOTH PAIN1 AND PAIN2
* ARE ANSWERED OR WHETHER ONE OF THE ITEMS HAS MISSING
DATA.
* AFTER RECODING, ALL ITEMS ARE POSITIVELY SCORED - THE
HIGHER
* THE SCORE, THE LESS PAIN (OR THE MORE FREEDOM FROM PAIN).
*
* THIS SCALE IS POSITIVELY SCORED. THE HIGHER THE
* SCORE THE LESS PAIN OR THE MORE FREEDOM FROM PAIN.
,
IFBP1 < 1 OR BP1 > 6 THEN BP1 = .;
IF BP2 < 1 OR BP2 > 5 THEN BP2 = .;
* RECODES IF NEITHER BP1 OR BP2 HAS A MISSING VALUE;
IFBP1NE. AND BP2 NE . THEN DO;
IF BP1 = 1 THEN RBP1 = 6;
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
IFBP1 =2 THEN RBP1 = 5.4
IF BP1 = 3 THEN RBP1 = 4.2
IFBP1 = 4 THEN RBP1 = 3.1
IF BP1 = 5 THEN RBP1 = 2.2
IF BP1 = 6 THEN RBP1 = 1;
IFBP2
IFBP2
IF BP2
IFBP2
IFBP2
IFBP2
END;
1 AND BP1 = 1 THEN RBP2 = 6;
1 AND 2 LE BP1 LE 6 THEN RBP2 = 5;
2 AND 1 LE BP1 LE 6 THEN RBP2 = 4
3 AND 1 LE BP1 LE 6 THEN RBP2 = 3
: 4 AND 1 LE BP1 LE 6 THEN RBP2 = 2
: 5 AND 1 LE BP1 LE 6 THEN RBP2 = 1
RECODES IF BP1 IS NOT MISSING AND BP2 IS MISSING;
IFBP1N E. AND BP2
IF BP
IF BP
IF BP
IF BP
IF BP
IF BP
RBP2
END;
= 1 THEN RBP
= 2 THEN RBP
= 3 THEN RBP
= 4 THEN RBP
= 5 THEN RBP
= 6 THEN RBP
= RBP1;
. THEN DO;
= 6;
= 5.4;
= 4.2;
= 3.1;
= 2.2;
= 1;
* RECODES IF BP1 IS MISSING AND BP2 IS NOT MISSING;
IF BP1 = . AND BP2 NE . THEN DO;
IF BP2 = 1 THEN RBP2 = 6;
IF BP2 = 2 THEN RBP2 = 4.75;
IF BP2 = 3 THEN RBP2 = 3.5;
IF BP2 = 4 THEN RBP2 = 2.25;
IF BP2 = 5 THEN RBP2 = 1; '
RBP1 = RBP2;
END;
BPNUM = N(BP1,BP2);
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
1 1 2
IF BPNUM GE 1 THEN RAWBP = SUM(RBP1,RBP2);
BP = ((RAWBP - 2)/(12-2)) * 100;
LABEL BP = ’ SF-36 PAIN INDEX (0-100)’
RAWBP = ’ RAW SF-36 PAIN INDEX’ ;
* THE SF-36 GENERAL HEALTH PERCEPTIONS INDEX.
* REVERSE TWO ITEMS AND RECALIBRATE ONE ITEM. AFTER
RECODING
* AND RECALIBRATION, ALL ITEMS ARE POSITIVELY SCORED - THE
* HIGHER THE SCORE, THE BETTER THE PERCEIVED GENERAL
HEALTH.
*
* THIS SCALE IS POSITIVELY SCORED.
* THE HIGHER THE SCORE THE BETTER THE HEALTH PERCEPTIONS.
?
ARRAY GHP(5) GH1-GH5;
DO 1= 1 TO 5;
IF GHP(I) < 1 OR GHP(I) > 5 THEN GHP(I) = .;
END;
IF GH1 = 1 THEN RGH1 = 5;
IF GH1 = 2 THEN RGH1 = 4.4;
IF GH1 = 3 THEN RGH1 = 3.4;
IF GH1 = 4 THEN RGH1 = 2;
IF GH1 = 5 THEN RGH1 = 1;
RGH3 = 6 - GH3;
RGH5 = 6 - GH5;
GHNUM = N(GH1 ,GH2,GH3,GH4,GH5);
GHMEAN = MEAN (RGH1 ,GH2,RGH3 ,GH4,RGH5);
ARRAY RGH(5) RGH1 GH2 RGH3 GH4 RGH5;
DO 1= 1 TO 5;
IF RGH(I) = . THEN RGH(I) = GHMEAN;
END;
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
113
IF GHNUM GE 3 THEN RAWGH = SUM(RGH1 ,GH2,RGH3,GH4,RGH5);
GH = ((RAWGH - 5)/(25-5)) * 100;
LABEL GH = ’ SF-36 GENERAL HEALTH PERCEPTIONS (0-100)’
RAWGH = RAW SF-36 GENERAL HEALTH PERCEPTIONS’ ;
* THE SF-36 VITALITY ITEMS.
* REVERSE TWO ITEMS. AFTER ITEM REVERSAL, ALL ITEMS ARE
* POSITIVELY SCORED - THE HIGHER THE SCORE, THE LESS THE
FATIGUE
* AND THE GREATER THE ENERGY.
*
* THIS SCALE IS POSITIVELY SCORED.
* THE HIGHER THE SCORE THE GREATER THE VITALITY.
5
ARRAY VI(4) VT1-VT4;
DO I = 1 TO 4;
IF VI(I) < 1 OR VI(I) > 6 THEN VI(I) =.;
END;
RVT1 = 7-VT1;
RVT2 = 7-VT2;
VITNUM = N (VT1, VT2, VT3, VT 4);
VITMEAN = MEAN(RVT1,RVT2,VT3,VT4);
ARRAY RVI(4) RVT1 RVT2 VT3 VT4;
DO I = 1 TO 4;
IF RVI(I) = . THEN RVI(I) = VITMEAN;
END;
IF VITNUM GE 2 THEN RAWVT= SUM(RVT1,RVT2,VT3,VT4);
VT = ((RAWVT-4)/(24-4)) * 100;
LABEL VT = ’ SF-36 VITALITY (0-100)’
RAWVT = RAW SF-36 VITALITY’ ;
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
114
* THE SF-36 SOCIAL FUNCTIONING INDEX.
* REVERSE ONE ITEM SO THAT BOTH ITEMS ARE POSITIVELY SCORED
* THE HIGHER THE SCORE, THE BETTER THE SOCIAL FUNCTIONING.
* ■
* THIS SCALE IS POSITIVELY SCORED.
* THE HIGHER THE SCORE THE BETTER THE SOCIAL FUNCTIONING.
>
ARRAY SOC(2) SF1-SF2;
DO I = 1 TO 2;
IF SOC(I) < 1 OR SOC(I) > 5 THEN SOC(I) = .;
END;
RSF1 = 6 - SF1;
SFNUM = N(SF1,SF2);
SFMEAN = MEAN(RSF1,SF2);
ARRAY RSF(2) RSF1 SF2;
DO I = 1 TO 2;
IF RSF(I) = . THEN RSF(I) = SFMEAN;
END;
IF SFNUM GE 1 THEN RAWSF = SUM(RSF1,SF2);
SF = ((RAWSF - 2)/(10-2)) * 100;
LABEL SF = ’ SF-36 SOCIAL FUNCTIONING (0-100)’
RAWSF = RAW SF-36 SOCIAL FUNCTIONING’ ;
* THE SF-36 ROLE-EMOTIONAL INDEX.
* ALL ITEMS ARE POSITIVELY SCORED - THE HIGHER THE ITEM
VALUE,
* THE BETTER THE ROLE-EMOTIONAL FUNCTIONING.
*
* THIS SCALE IS POSITIVELY SCORED.
* THE HIGHER THE SCORE, THE BETTER THE ROLE-EMOTIONAL.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
115
ARRAY RM(3) RE1-RE3;
DO 1= 1 TO 3;
IF RM(I) < 1 OR RM(I) > 2 THEN RM(I) =
END;
ROLMNUM = N(OF RE1-RE3);
ROLMMEAN = MEAN(OF RE1-RE3);
DO 1= 1 TO 3;
IF RM(I) = . THEN RM(I) = ROLMMEAN;
END;
IF ROLMNUM GE 2 THEN RAWRE = SUM(OF RE1-RE3);
RE = ((RAWRE - 3)/(6-3)) * 100;
LABEL RE = ’ SF-36 ROLE-EMOTIONAL (0-100)’
RAWRE = RAW SF-36 ROLE-EMOTIONAL’ ;
* THE SF-36 MENTAL HEALTH INDEX.
* REVERSE TWO ITEMS. AFTER ITEM REVERSAL, ALL ITEMS ARE
* POSITIVELY SCORED - THE HIGHER THE SCORE, THE BETTER THE
* MENTAL HEALTH.
*
* THIS SCALE IS POSITIVELY SCORED.
* THE HIGHER THE SCORE THE BETTER THE MENTAL HEALTH.
?
ARRAY MHI(5) MH1-MH5;
DO 1= 1 TO 5;
IF MHI(I) < 1 OR MHI(I) > 6 THEN MHI(I)=.;
END;
RMH3 = 7-MH3;
RMH5 = 7-MH5;
MHNUM=N(MH1 ,MH2,MH3,MH4,MH5);
MHMEAN=MEAN(MH1 ,MH2,RMH3,MH4,RMH5);
ARRAY RMH(5) MH1 MH2 RMH3 MH4 RMH5;
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
116
DO 1=1 TO 5;
IF RMH(I) = . THEN RMH(I) = MHMEAN;
END;
IF MHNUM GE 3 THEN RAWMH = SUM(MH1 ,MH2,RMH3,MH4,RMH5);
MH = ((RAWMH-5)/(30-5)) * 100;
LABEL MH = ’ SF-36 MENTAL HEALTH INDEX (0-100)’
RAWMH = RAW SF-36 MENTAL HEALTH INDEX’ ;
* THE SF-36 HEALTH TRANSITION ITEM.
* THIS ITEM SHOULD BE ANALYZED AS CATEGORICAL DATA,
* PENDING FURTHER RESEARCH.
3
IF HT < 1 OR HT > 5 THEN HT = .;
LABEL HT=RAW SF-36 HEALTH TRANSITION ITEM’ ;
RUN;
*** STEP 3: SF-36 SCALE CONSTRUCTION ***;
DATA SF36INDX;
SET SF36SCAL;
?
* purpose: create physical and mental health index scores
* standardized but not normalized
* and standard deviations calculated with vardef=wdf
,
COMPUTE Z SCORES - OBSERVED VALUES ARE SAMPLE DATA
MEAN AND SD IS U.S GENERAL POPULATION
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
1 1 7
FACTOR ANALYTIC SAMPLE
N=2393: HAVE ALL EIGHT SCALES
9
PF_Z=(PF-84.52404)/22.89490;
RP_Z=(RP-81.19907)/33.79729;
BP_Z=(BP-75.49196)/23.55879;
GH_Z=(GH-72.21316)/20.16964;
VT_Z=( VT-61.05453)/2Q.86942;
SF_Z=(SF-83.59753)/22.37642;
RE_Z=(RE-81.29467)/33.02717;
MH_Z=(MH-74.84212)/18.01189;
rjx ^|c ^ ^ ^ ^ jj# vjj* ^ ^ ^ ^ ^ »[' ^ ^ ^ |/ fcj* % J ;J jj/ ^ k|« ^ ^ ^ jjj
COMPUTE SAMPLE RAW FACTOR SCORES
Z SCORES ARE FROM ABOVE
SCORING COEFFICIENTS ARE FROM U.S. GENERAL POPULATION
FACTOR ANALYTIC SAMPLE N=2393: HAVE ALL EIGHT SCALES
jJj-J. - * - v - ^
praw=(PF_Z * ,42402)+(RP_Z * ,35119)+(BP_Z * ,31754)+(SF_Z * -.00753)+
(MH_Z * -,22069)+(RE_Z * -,19206)+(VT_Z * .02877)+(GH_Z * .24954);
mraw=(PF_Z * -,22999)+(RP_Z * -,12329)+(BP_Z * -,09731)+(SF_Z * .26876)+
(MH_Z * ,48581)+(RE_Z * ,43407)+(VT_Z * ,23534)+(GH_Z * -.01571);
COMPUTE STANDARDIZED SCORES
^ ^ jj* jj* s jj ^ ^ jj/ ^ ^ J tJjJ jj/ ^
J
PCS = (praw*10) + 50;
MCS = (mraw*10) + 50;
label PCS=’ STANDARDIZED PHYSICAL COMPONENT SCALE-00’
MCS=’ STANDARDTZED MENTAL COMPONENT SCALE-00’ ;
Run;
Source: SF-36 Manual
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Linked assets
University of Southern California Dissertations and Theses
Conceptually similar
PDF
Gender role conflict, personality, and help -seeking in adult men
PDF
Development and validation of the Cooper Quality of Imagery Scale: A measure of vividness of sporting mental imagery
PDF
A daily diary approach to compare the accuracy of depressed and nondepressed participants' estimation of positive and negative mood: A test of the depressive realism hypothesis
PDF
Gender differences in motivation for sexual intercourse: Implications for risky sexual behavior and substance use in a university and community sample
PDF
Integration of science and practice: A collective case study of scientist -practitioner programs in counseling psychology
PDF
Attachment style, interpersonal guilt, parental alcoholism, parental divorce and eating disordered symptomatology in college women
PDF
An investigation of a new diagnostic sub-type: Post traumatic stress disorder with psychotic features
PDF
Heterosexuals' attitudes toward lesbians and gay men and the attitude functions they serve: Correlates, stability and gender differences
PDF
Intergenerational conflict, family functioning, and acculturation experienced by Asian American community college students
PDF
Assessment of racial identity and self -esteem in an Armenian American population
PDF
Considering the impact of culture on the therapeutic relationship: A look at Latino school-age children in a special education setting
PDF
Anger expression in men with elevated blood pressure: Relationships to assertiveness
PDF
A cluster analysis of the differences between expert and novice counselors based on real time training
PDF
Effects of test interpretation style and need for cognition on elaboration, favorability, and recall
PDF
Dementia caregiving and ethnicity: African American caregivers and the sociocultural stress and coping model
PDF
An integrative analysis of racism and quality of life: A comparison of multidimensional moderators in an ethnically diverse sample
PDF
A two-dimensional model of cognitive empathy: An empirical study
PDF
Effects of cultural values and behaviors in sociocultural stress and coping among Korean-American and Korean caregivers
PDF
Examining the differences in family functioning and meaning in life between Reform, Conservative, and Orthodox Jewish families
PDF
Family functioning, intergenerational conflict, and psychological symptomology of Asian American college students
Asset Metadata
Creator
Nasserbakht, Afsaneh
(author)
Core Title
Gender differences in symptom presentation of depression in primary care settings
School
Graduate School
Degree
Doctor of Philosophy
Degree Program
Counseling Psychology
Publisher
University of Southern California
(original),
University of Southern California. Libraries
(digital)
Tag
OAI-PMH Harvest,psychology, clinical
Language
English
Contributor
Digitized by ProQuest
(provenance)
Advisor
Goodyear, Rodney (
committee chair
), Hoffman, Kaaren (
committee member
), Silverstein, Merril (
committee member
)
Permanent Link (DOI)
https://doi.org/10.25549/usctheses-c16-105236
Unique identifier
UC11326912
Identifier
3027757.pdf (filename),usctheses-c16-105236 (legacy record id)
Legacy Identifier
3027757.pdf
Dmrecord
105236
Document Type
Dissertation
Rights
Nasserbakht, Afsaneh
Type
texts
Source
University of Southern California
(contributing entity),
University of Southern California Dissertations and Theses
(collection)
Access Conditions
The author retains rights to his/her dissertation, thesis or other graduate work according to U.S. copyright law. Electronic access is being provided by the USC Libraries in agreement with the au...
Repository Name
University of Southern California Digital Library
Repository Location
USC Digital Library, University of Southern California, University Park Campus, Los Angeles, California 90089, USA
Tags
psychology, clinical