University of Southern California Dissertations and Theses

Determining diabetes risk assessment in the elderly dental patient

(USC Thesis Other)

Determining diabetes risk assessment in the elderly dental patient

PDF

Download a page range

Download transcript

Copy asset link

Request this asset

Transcript (if available)

Content DETERMINING DIABETES RISK ASSESSMENT
IN THE ELDERLY DENTAL PATIENT
by
Rose Ann Mulligan
A Thesis Presented to the
LEONARD DAVIS SCHOOL OF GERONTOLOGY
UNIVERSITY OF SOUTHERN CALIFORNIA
In Partial Fulfillment of the
Requirements for the Degree
MASTER OF SCIENCE IN GERONTOLOGY
May 1987
Copyright 1987 Rose Ann Mulligan
UMI Number: EP58938
All rights reserved
INFORMATION TO ALL USERS
The quality of this reproduction is dependent upon the quality of the copy submitted.
In the unlikely event that the author did not send a complete manuscript
and there are missing pages, these will be noted. Also, if material had to be removed,
a note will indicate the deletion.
DissertaUofi F^ïbiismng
UMI EP58938
Published by ProQuest LLC (2014). Copyright in the Dissertation held by the Author.
Microform Edition © ProQuest LLC.
All rights reserved. This work is protected against
unauthorized copying under Title 17, United States Code
ProQuest LLC.
789 East Eisenhower Parkway
P.O. Box 1346
Ann Arbor, Ml 48106-1346
UHJVERSJTy Of SOUTHERN CALIFORNIA
LEONARD VAVIS SCHOOL Of GERONTOLOGV
UNJVERSny PARK
LOS MGELES, CALIFORNIA 90007
(^e.ron
V z
m S 9
po
TIvLi t h u t i , uvujtten by
_________Rose Ann Milligan
undeA the, dOizctoA. o i her The6t& ComnUXtee,
and app/iove,d by aJLt Jitfp mejnbe/u, ho6 be.cn pAc-
6cntcd to and accepted by the Vean oi The Leona/id
Vavts SchooZ GeAontoZogy, tn poA ttaZ iiiZ^ZlZment
o£ th e AequÂAejnent& ioA the degAce
Maste/l of Science^in Gerontology
Vean
Vote Q .
THESIS COmiTTEE
Ehatnman
11
TABLE OF CONTENTS
Page
1ST OF TABLES......................................... iv
iiNTRODUCTION........................................... 1
PART I. AN OVERVIEW OF DIABETES
Chapter
I. THE DISEASE OF DIABETES...................... 8
History of the Disease
Classification of Diabetes
II. CURRENT DETECTION AND MONITORING TECHNIQUES . . 12
Urine Testing
Capillary Blood Glucose Test
Fasting Plasma Glucose Test
Oral Glucose Tolerance Test
Glycosylated Hemoglobin Tests
III. DIABETES IN THE ELDERLY........................ 21
Diagnostic Levels
Symptoms
Complications
Treatment
IV. ORAL MANIFESTATIONS OF DIABETES................ 30
Salivary Gland Effects
Caries Rate
Periodontal Disease
Other Effects
Management of Oral Disease in the Diabetic
PART II AN EVALUATIVE STUDY OF THE
DIABETIC STATUS OF ELDERLY DENTAL PATIENTS
V. AIMS AND OBJECTIVES........................... 38
VI. METHODS........................................... 40
VII. RESULTS........................................... 47
Fisher's Exact Test
Correlation Coefficients
-----------Regrouping_of_Subjects________________________ ____
Ill
Linear Regression Analysis
Questionnaire Results
Regression Values
VIII. DISCUSSION...........................................65
Fisher's Exact Test
Correlation Coefficients
Linear Regression Analysis
Stepwise Logistic Regression Analysis
IX. CONCLUSIONS...................................... 73
APPENDIXES............................................... 75
BIBLIOGRAPHY............................................ 85
IV
LIST OF TABLES
'1. Age Distribution of Subjects...........................47
2. Descriptive Statistics ............................... 48
3. Table of Compliance (As Measured by VBG) Vs
Noncompliance .................................... 49
4. Table of Compliance (As Measured by HBF) Vs
Noncompliance................................... 50
5. Correlation Coefficients of Fasting Variables . . . 51
6. Correlation Coefficients of Non-Fasting Variables . 52
Comparisons of Correlational
Coefficients and Ranges ........................ 54
8. Linear Regression Analysis........................ 55^
9. Logistic Regression Analysis 1 ...................... 56
10. Logistic Regression Analysis 2 ...................... 51
11. Logistic Regression Analysis 3 . . . ................58
12. Logistic Regression Analysis 4...................... 6C
13. Logistic Regression Analysis 5 ...................... 61
14. Logistic Regression Analysis 6 ...................... 62
15. Logistic Regression Analysis 7 ...................... 63
INTRODUCTION
Using diagnostic criteria developed by the National
Diabetes Data Collection Group (1979), the prevalence of
diabetes in this country has been estimated to be between 2
to 5% (Owen et al.1983). Diabetes is however an age
dependent disease. Although people in the 20-39 year old
age bracket have a prevalence rate of 1.5%, for the 60 to
74 year old the prevalence rate increases to 17.1%
(Goldberg et al.1985). In individuals aged 85 and over,
the rate climbs to 25% (Williams 1983). An additional 10%
of the older age groups also have impaired glucose
tolerance. This is another form of hyperglycemia
identifying those who have glucose levels which are
elevated, but not to a great enough extent to be considered
diabetic.
Just as the risk of diabetes increases with age, such
is also the case with one of the major dental diseases,
periodontal disease. By the 5th decade, 80% of all
individuals have periodontal disease (Carranza 1984) which
becomes much more severe in the diabetic, progressing
rapidly in its destructive processes due to the decreased
neutrophil chemotaxis (Bartolucci et al.1981; and
Manouchehr-Pour et al, 1981) and altered blood vessel
permeability (Genco et al. 1982).
2
Recently cohorts of elderly have begun to retain their
teeth and seek sophisticated dental treatment requiring
invasive dental procedures. It is known that dental
treatment in the hyperglycemic of any age may result in
complications ranging from poor wound healing (Gotthelf et
al.1983) to life-threatening ketoacidosis (Cutler et al.
1982). Therefore it becomes obvious that invasive dental
care could further compromise the diabetic patient's status
if such an underlying systemic disease was undiagnosed or
uncontrolled.
Our interest specifically in the elderly diabetic
dental patient is additionally prompted by the fact that
epidemiologic studies indicate that not only are there 10
million diabetics in the United States (Murrah 1985) with a
like amount who are undiagnosed (Goldberg et al. 1985), but
also because the prevalence of diabetes is increasing
yearly (Zimmet 1982).
As a result of these trends and the implications that
they have for the proper delivery of dental care to this
at-risk population, it becomes very important that dental
practitioners become increasingly aware of the following:
1) The heightened prevalence of diabetes in the elderly; 2)
The roles that infections (such as periodontal disease) and
diabetes play in influencing the severity of each other
(Cutler et al.1982); and 3) The very real possibility that
undiagnosed diabetes may be present in large numbers of
3
elderly dental patients. The practitioner must be certain
to take precautions before treating a diabetic patient in
order to decrease the possibility of complications in an
individual with hyperglycemia.
Another equally important issue for the dental
clinician is the previously diagnosed diabetic patient's
adherence to his or her therapeutic regimen. It is known
that with many chronic diseases, adherence to the
therapeutic regimen decreases with the increasing
complexity of the regimen and the duration of treatment
(Mulligan 1985). Asking patients if they have followed a
prescribed therapeutic regimen will often result in a false
answer (Evans et al. 1983), the magnitude of which is
dependent upon the disease involved (O'Hanrahan et al.
1981). Since diabetics may require lifetime medications as
well as diet and exercise prescriptions to keep their
disease stable, patient adherence presents a significant
problem in the control of this disease (Cerkoney et al.
1980). Nor are the treating physicians the ultimate
resources in attempting to determine if patients are
adhering to their therapeutic regimens, as the physicians'
estimates of compliance have been shown to be no better
than chance at predicting whether or not a patient is
adhering to that regimen (Wing et al.1985). Therefore a
screening device which would uncover those individuals at
risk for diabetes, as well as the diabetics who do not have
4
their disease in control would be very important in helping
the dentist identify patients who may be demonstrating
clinical or subclinical dysfunction. This knowledge and
the subsequent action taken to resolve any problems
discovered would improve the ultimate treatment results, as
opposed to the inconsistent and at times dangerous outcomes
of treatment which can occur when diabetes remains
unidentified or uncontrolled in the dental patient. Thus
the quality of dental care provided to these patients would
be significantly improved.
In order to achieve these goals it appears that some
new directions are necessary regarding the collection and
analysis of the patient’s medical history in the dental
office. In particular there is a need to determine what
percentage of elderly patients seeking care in a dental
facility have symptoms of diabetes but have never been
diagnosed as being diabetic, as well as what percentage of
elderly patients with already diagnosed diabetes have their
disease in control when they seek dental care.
Therefore in the format of a scientific study, this
thesis explores the following questions; 1) How many
elderly diabetics have their disease in control when they
seek dental care? 2) Is an elderly diabetic's self-reported
level of compliance usually accurate? 3) How many
undiagnosed elderly diabetic dental patients are there? 4)
What is a reliable and cost effective technique to
5
recognize a patient with hyperglycemia in the dental
office? 5) Is a questionnaire a reliable device to help
predict an undiagnosed or uncontrolled elderly diabetic in
the dental office?
This thesis has been divided into two parts: Part I
will include an overview of the disease of diabetes; and.
Part II will present the scientific study that was
performed. A brief discussion of the contents of each
chapter of the thesis follows.
Part I, Chapter I exsimines the history and recognition
of the disease of diabetes as well as the current
classification system employed. Chapter II deals with
current detection and monitoring techniques presently in
use, including urine testing and the various types of blood
tests that can be performed.
In Chapter III the elderly diabetic patient is
discussed from the point-of-view of the changing diagnostic
criteria and symptoms which may manifest differently from a
younger patient population with this, same disease. In
addition, the symptoms, complications, and treatment of
diabetes specific to the elderly are reviewed.
Chapter IV discusses the specific oral problems that
occur due to the presence of diabetes according to the
categories of oral diseases seen. Also included is
information on the mechanism behind the pathology observed.
Reference is made to the possible sequelae that could occur
if an uncontrolled diabetic undergoes various dental
procedures.
Part II, the section describing the study, is composed
of five chapters. The first. Chapter V, outlines the study
hypotheses. Chapter VI presents the methods of the study.
Chapter VII exhibits the results observed while Chapter
VIII discusses those results. The last chapter of this
section and the thesis. Chapter IX presents the conclusions
of the study.
PART I
AN OVERVIEW OF DIABETES
CHAPTER I
THE DISEASE OF DIABETES
History of the Disease
Although diabetes was identified as long ago as 1500
B.C. (Saadoun 1980), it wasn't until 1921 that the
discovery of insulin by Canadians Banting and Best provided
diabetics with the ability to live near normal lives. Prior
to this time the onset of symptoms and the diagnosis of
diabetes was often quickly followed by weakness and death
(Fitzgerald et al. 1985).
The term diabetes no longer describes one disease but
a cluster of distinctly different types of chronic
disorders caused by anatomic and chemical problems relating
to an absolute or relative deficiency of insulin or its
function (Cahill 1985). It is characterized by glucose
intolerance and other metabolic irregularities in the
metabolism of proteins and fats (Owen et al.1983). There
is often a genetic component associated with the disease
(Fitzgerald et al.1985). Conditions such as accelerated
atherosclerosis and abnormalities such as retinopathy,
nephropathy, and neuropathy are also commonly found.
People with diabetes mellitus have an increased rate of
-cerebrovascular_accidents.,_myocardial_infarctions_,_and-----
9
peripheral vascular problems which are particularly
troublesome in the feet (American Diabetes Association
1984). Some infections occur more often in the diabetic,
particularly if the diabetes is poorly controlled
(Rabinowitz 1981).
Classification of Diabetes
In the past diabetics were placed into one of two
categories: the juvenile onset diabetic, and the maturity
onset diabetic. Over the years it has been observed that
each type of diabetes could occur outside of its usual
described age group (Marble et al. 1985). As a result an
international study group sponsored by the National
Diabetes Data Group (1979) of the National Institutes of
Health, established new classifications according to the
manifestations of the disease (Appendix A).
Type I or Insulin Dependent Diabetes Mellitus (IDDM)
is characterized by little or no insulin endogenously
produced, and the development of ketoacidosis. These
patients are usually lean and upon diagnosis, report a
recent weight loss. They must be given insulin
supplementation on a regular basis to survive (Marble et al,
1985). The average age of onset of IDDM is approximately
12 years (Cahill 1985), although it may occur at any age.
There appears to be an association between this type of
diabetes and certain genetic factors, however, in some
cases an autoimmune component (islet cell antibodies) and
10
in other cases viral infections have been implicated (Owen
et al.1983).
In Type II or the Non-Insulin Dependent Diabetes
Mellitus (NIDDM) onset usually occurs after 40 years of
age, however, this type of diabetes can also be found in
adolescents and young adults. NIDDM patients are not prone
to ketoacidosis except under stressful conditions caused by
surgery, trauma, or infections (American Diabetes
Association 1984). Initially they may have none of the
classic symptoms of diabetes (Kart et al.1978), although
the systemic complications of NIDDM are many (Eckel et al.
1982). Peripheral and autonomic neuropathy are often
present and retinopathy and nephropathy increase in both
severity and incidence. Atherosclerosis, such as coronary
heart disease or that which manifests as gangrene of the
lower extremities can be severe (Goldberg et al,1985).
One subset of Type II diabetes is characterized by
obesity in 80% of the group (Lipson et al.1984). As might
be expected, the treatment of this particular subset
emphasizes weight reduction which will decrease insulin
resistance. In addition to diet modifications, physical
activity and oral hypoglycemic agents are often used to
manage the Type II patient (Lipson et al*1984). Insulin
may be required to control the glucose level during periods
of acute injury, infection, surgery, or if there is an
allergy or reaction to sulfonylurea drugs (American
11
Diabetes Association 1984). It has been estimated that 15%
to 20% of Type II diabetics will require insulin on a day
to day maintenance basis; however, insulin given to control
the elevated levels of blood glucose and the resulting
pathologies they cause, is not necessary for life as it is
in the IDDM (Cahill 1985).
A third designated category includes other types of
glucose intolerance which are secondary to a number of
conditions: pancreatic disease, endocrinopathies, drugs and
chemical agents, insulin-receptor abnormalities, genetic
syndromes, and certain miscellaneous conditions such as
malnutrition.
A fourth category called Impaired Glucose Tolerance
(IGT) describes individuals with plasma glucose levels
greater than normal but lower than what is considered
diagnostic for diabetes mellitus (Marble et al. 1985). This
group is further divided into obese and nonobese. It is
expected that 25% of people with impaired glucose tolerance
will develop overt diabetes over time (American Diabetes
Association 1984). Those individuals whose blood values
place them in this category are also at risk for many of
the complications experienced by the overtly diabetic
patient such as macrovascular complications including
coronary artery disease (Jarrett et al. 1982), cerebral and
peripheral vascular disease (Fitzgerald et al.1985), and
hyperosmolar coma (Greene 1986).
12
CHAPTER II
CURRENT DETECTION AND MONITORING TECHNIQUES
Glucose is the principal source of fuel for the
mammalian body. As a result, it is a part of a very complex
interplay of systems which involve the liver, kidney, gut,
endocrine, adipose tissue and muscle attempting to regulate
its balance (Burrin et al. 1985). Since this is the case,
the detection and quantification of glucose by laboratory
testing plays a very important role in the treatment of
carbohydrate imbalance diseases such as diabetes mellitus.
The most common types of tests to detect glucose include
tests of the urine, capillary blood, and venous blood. More
sophisticated tests measure the total glycosylated
hemoglobins or a specific hemoglobin component.
Urine Testing
Urine testing has been used in the past as a method to
manage patients with diabetes. Gray (1985) considers the
urine test an inexpensive, quick indicator of the level of
control of the disease which can be personally performed by
the patient. It is easily accomplished by dipping a glucose
oxidase reagent strip into freshly collected urine (Bonar
1980). The tape is removed immediately after wetting, and a
two minute waiting period is begun. If glucose is present
13
it reacts with the oxygen in the air (the glucose oxidase
serving as the catalyst). A resulting color change from the
original yellow to light green or blue develops on the
enzymatic testing strip, if glucose has been detected. If
no glucose is present the strip will remain yellow. The
st;rip is then compared to a standard visual chart supplied
with the test strips and the percent of glucose present is
determined.
The level of sensitivity of this test when used with
the elderly individual is quite often unacceptable. Even if
the Crede maneuver to empty the bladder completely, and
double voiding are employed as part of the technique (Gray
1985), there are still a number of reasons to avoid the use
of the urine test as a self-monitoring practice in the
elderly patient population. First, there is often a poor
correlation of glucosuria with blood levels of glucose
because of the lag time necessary for the urine to
accumulate. As a result, a urine sample does not provide
an accurate estimate of the concurrent blood glucose value.
Second, in the elderly, renal thresholds for excretion of
glucose are higher and inconsistent. Therefore significant
hyperglycemia (in some cases as high as 300 mg/dl) may
occur before glucose is spilled into the urine (Redmon
1984). For individuals with renal insufficiency effected
by Other concurrent diseases such as hypertension, the
problem is compounded. Hanuschak et al,(1985) enumerate a
14
number of other problems with this test when it is used
with the elderly. For example, many elderly are
incontinent or have significant joint stiffness,
incoordination or incapacitation thus complicating the
urine catch. If an elderly person has any mental
difficulties they may be unable to follow the procedure for
collecting and measuring the urine. If there are visual
problems, they may not be able to read or compare the
resulting color. Finally, if the blood sugars are less
than the threshold level, there is no way for this
information to register on the reference scale. As a
result, elderly diabetic patients will not know if they are
near hypoglycemia when they are using a urine test, which
is of course very important information.
Capillary Blood Glucose Test
The simplest blood glucose monitoring test is that
which is performed using capillary blood. This is obtained
by pricking the finger with a lancet and placing a few
drops of blood on a glucose oxidase reagent pad to cover it
completely. Newer products such as Chemstrip bG (Boehringer
Mannheim, Indianapolis, IN) do not require water to be used
to remove the blood from the pad after the developing time
has elapsed (in this case 60 seconds). Instead, the blood
is wiped off with a dry cotton ball after the first minute,
and then wiped twice more, while the color of the pad is
-allowed_to_develop_in_a_reaction_sequence_similar_tp_tha±—
-----Y5
which was described for the urine test reagent (Burrin
1985). The resultant color is then compared with a
reference color chart which may range from 20 to 240 mg/dl.
Reagent strips demonstrating results higher than 240 mg/dl
are allowed to develop an extra minute before the final
reading of 400 mg/dl or 800 mg/dl is read. Thus with this
type of test the full range of glycémie control can be
determined, from hypoglycemia to hyperglycemia.
Even if an elderly patient has visual or color
perception difficulties, a reflectance meter which gives a
digital readout of the strip can be used. Additionally,
some brands of strips can be saved for several days during
which time they may be read by others for corroboration
(Van Crombrugge et al.1985). Hanuschak (1985) again
describes dexterity problems as a result of strokes,
Parkinson's Disease, or arthritis all of which may limit an
elderly patient's ability to perform this test.
The previously described tests require minimal
equipment or skill for performance and for those reasons
can be administered at home as well as in an office or
clinic environment. The following tests described will only
be available where there is access to a we11-equipped
diagnostic laboratory.
Fasting Plasma Glucose Test
Determinations of fasting plasma glucose levels are
often recommended for screening tests for diabetes. This
16
is because the test is relatively unaffected by age
(approximately 1 mg/dl increase per decade of life) and is
reproducible from day to day in an individual whether they
have normal or abnormal results (Goldberg et al.1985).
Also standardizing the conditions of the phlebotomy (a set
amount of time since food intake) is easy. The test itself
is convenient and fast. The caloric intake or the
composition of the food prior to the fast does not
significantly affect the outcome. Although 115mg/dl has
been recommended as the upper limits of normal for this
test, in order to avoid a large number of false positive
results, only a fasting glucose value above 140 mg/dl, if
confirmed by a second test is diagnostic of diabetes
(Goldberg et al. 1985). If the fasting plasma glucose does
not reach that level (is between 115 and 140 mg/dl) but
there is a high clinical suspicion of diabetes, a
diagnostic oral glucose tolerance test should be performed.
The problem with recommending this test as a screening
measure for diabetes or impaired glucose tolerance in the
elderly is that hyperglycemia is often obvious only after
observing the results of postprandial plasma glucose levels
and not fasting levels (Lipson 1986). Therefore an elderly
subject who has fasting values within the normal range
could easily have postprandial plasma glucose values which
reach diagnostic levels for diabetes or IGT.
17
Oral Glucose Tolerance Test
The oral glucose tolerance test should be scheduled in
the morning, after a fast of 10 to 14 hours. After drawing
an initial fasting sample of blood, the patient is given 75
grams of a standard glucose solution. Blood samples are
then obtained at designated intervals which have been
outlined by the National Diabetes Data Group (1979). Three
values are then used to determine a diagnosis of diabetes.
The first value (fasting plasma glucose) must be less than
140 mg/dl. The two hour glucose value and an intermediate
value must be equal to or greater than 200 mg/dl. If the
glucose result at two hours is between 140 - 200 mg/dl with
a fasting plasma glucose of less than 140 mg/dl and an
intervening reading of 200 mg/dl or greater, the diagnosis
of impaired glucose tolerance is made.
Problems with the oral glucose tolerance test lie in
its lack of reproducibility due to the effects of drugs,
physical activity, preceding dietary intake including
composition and caloric level, infections, time of day,
other disease states, and age (Marble et al. 1985). Since
there are so many factors which can influence the oral
glucose tolerance test resulting in false readings, it
could have been predicted that a newer type of laboratory
test would be developed which would lessen the influence of
the previously mentioned conditions. Such is felt to be
the case with the glycosylated hemoglobin tests (see
18
below). Unfortunately, however, in spite of its outlined
difficulties, the standard test by which diabetes is
diagnosed remains the oral glucose tolerance test.
Glycosylated Hemoglobin Tests
Glycosylated hemoglobins are the result of simple
chemical reactions between molecules of hemoglobin and
sugar after the hemoglobin molecule has been synthesized
(Peacock 1984). They are formed throughout the usual 120
day life cycle of the erythrocyte in a nonenzymatic, post-
translational process. Kozak (1982) describes this
gycosylation as being irreversible for the life of the red
blood cell. If the plasma glucose is chronically elevated,
as would be the case in an uncontrolled diabetic, the
degree of glycosylation acts as a determinant of the
average glucose level over the half-life of the erythrocyte
(estimated to be 2 months) (Marble et al.1985). If, for
example, the glycosylated hemoglobin results of an
individual were near the normal level, it could be
concluded that that person had exhibited good metabolic
control of their glucose level over the previous several
weeks (Boden 1980).
Hemoglobin (made up of A^a2 ' ^Ib' ^Ic
components) comprises 90% of the hemoglobin of erythrocytes
in normal adults (Health & Public Policy Committee 1984).
Of the four minor components HbA]^^, is the most important as
_a_determinant of glucose control. It makes up 3-6% of the
13
total hemoglobin in normals. The total glycosylated
hemoglobins can be measured, the total fast hemoglobin A^s
(Ala, ^Ib' ^rid A^c) , or HbA^c alone.
The difficulty inherent in using the glycosylated
hemoglobin as a measurement of the level of glucose control
is the sophisticated analytical technique required to
process the blood sample (Peacock 1984). Technique
imprecision causing a 1% difference in the amount of
glycosylated hemoglobin level represents a change of
approximately 35% in the blood glucose level (Health &
Public Policy Committee 1984). In addition, hemoglobin A^
levels determined by chromatography can be effected by the
levels of other hemoglobins present (such as F, S, C, or
D), renal failure, iron deficiency and hemolytic anemias,
splenectomy, elevated triglycerides, alcohol, lead toxicity
and blood loss. Also a relatively brief period of poor
control can increase glycosylation in all generations of
the erythrocytes in the circulation very quickly
(Peacock,1984).
The effects of sex or obesity on glycosylated
hemoglobin assays are unknown (Health & Public Policy
Committee 1984). Although Arnetz et al.(1982) has shown
that age has a significant effect on the level of HbAlc.
Reliable reference standards for normals and diabetics
have not yet been developed for all the various
glycosylated hemoglobin tests. According to the Health and
20
Public Policy Committee of the American College of
Physicians (1984), the glycosylated hemoglobin test has
high specificity but low sensitivity with the sensitivity
directly related to the quality of the assay. A result
higher than normal usually indicates diabetes, although a
test lower than normal does not rule out impaired glucose
tolerance.
The glycosylated hemoglobin test is particularly
useful when the diabetic patient is not being compliant
with home testing or therapeutic instructions until a few
days before a check-up medical visit. It is now widely
becoming accepted as an index of long term diabetic control
if the specific caveats are borne in mind (Willey et al.
1984) .
21
CHAPTER III
DIABETES IN THE ELDERLY
It has been clearly demonstrated that the prevalence
of NIDDM and IGT increases in the elderly (Wilson et al.
1986). It is unclear however, whether this change is due
to an increasing incidence of glucose intolerance or is
characteristic of physiologic changes occurring with the
aging process. It is most probably the former for many
experts consider increasing glucose intolerance a normal
part of aging. As a result, there has been much discussion
in the literature as to whether the same diagnostic
criteria should be applied to the elderly as to younger
adults (Fitzgerald et al. 1985).
Diagnostic Levels
Since fasting plasma glucose values are only minimally
altered by age at the rate of Img/dl increase per decade
(Eckel et al. 1982), some recommend this test to be used to
diagnose diabetes in the elderly. Others recommend the use
of the oral glucose tolerance test (Kart et al. 1978),
however, when this test is used, 10 mg/dl should be added
to the results obtained for every decade the patient is
aged above 50 years (Marble 1985). Justification for the
age level adjustments of the latter test is to compensate
22
for the recognized decrease in the sensitivity of the
tissues to insulin seen with aging (Goldberg et al. 1985).
Using such age adjusted criteria for the two hour plasma
glucose level, the diagnosis of diabetes in the middle aged
and particularly the elderly has decreased from a
unrealistically high prevalence of 50% of the entire
elderly population (Williams 1983). Those who argued in
support of this elevated cutoff for diagnosis in the oldest
age groups did so based on the fact that although mortality
rates in the overtly diabetic were much higher than the
rest of the population (Kart et al.1978), evidence
regarding the severity of complications in individuals with
impaired glucose tolerance were not as well quantified
(Bennett 1984). However, recently published results of the
Bedford Survey reveal that even those with impaired glucose
tolerance have macrovascular disease and mortality rates
greater than individuals with normoglycemia (Jarrett et al.
1982). Since it appears that borderline glucose intolerance
does cause complications leading to a higher degree of
morbidity and mortality, therapeutic intervention is
indicated (Lipson 1986).
The current National Diabetes Data Group (NDDG)
standards, however, are based on a conservative reading of
the blood glucose values needed for diagnosis, and may be
too high for young adults and too low for very old adults
(Goldberg et al*1985). This was intentional however, as
23
the NDDG preferred to recommend standards with high
specificity and low sensitivity. Even though these
standards will ensure that no normals are labeled as being
diabetic, it is unfortunate that they will also result in
some individuals with the disease not being diagnosed and
receiving treatment.
To further complicate matters, a number of factors
which occur with much greater frequency in the elderly can
influence the conventional venous blood test used to assess
the level of glucose. When individuals of any age are
confined to bed, have any chronic situation which limits
physical activity or nutritional status, or becomes
deprived of potassium, they will exhibit changes in the way
their body metabolizes glucose (American Diabetes
Association 1984). Medication interactions, various
endocrine diseases, other co-existing diseases (acute or
chronic), and genetic syndromes may also interfere with
glucose metabolism. The altered body composition, decreased
physical fitness, changes in dietary consumption, and
sympathetic nervous system hyperactivity seen in the
elderly may change the insulin-mediated glucose response
compounding the above factors (Marble et al 1985).
Whether this decrease is called normal aging or
pathologic, the mechanism responsible for the decline in
the elderly relates to the decreased insulin biosynthesis
and/or secretion (Lipson 1986) and the increased resistance
24
of peripheral tissues to insulin, which is attributed to a
post-receptor defect in insulin action at the target
tissues (Goldberg 1985).
Symptoms
It is often the case that the symptoms of diabetes in
the elderly are not like those usually found in younger
individuals. Plasma glucose levels may exceed 250mg/dL
before glucosuria is present due to the increased renal
threshold for glucose (Goldberg et al.1985). Similarly
polydipsia, and polyuria may not be present, again because
of the increased renal threshold for glucose which may
easily exceed 300 mg/dl and result in a significant
hyperglycemia before polydipsia and polyuria will occur
(Redmon 1984).
Other symptoms are often vague. An elderly patient
does not present with clearcut complaints but may be
concerned about visual deterioration, increasing
paresthesias, generalized weakness, and weight loss (Redmon
1984). Recurrent infections that involve the skin or
urinary tract and which tend to resolve slowly, the
presence of obesity, and family histories of diabetes or
poor obstetrical histories in women (spontaneous abortions,
miscarriages, etc.) may provide other clues as to the
possible presence of diabetes (Goldberg et al. 1985;
Fitzgerald et al.1985).
25
Complications
Elderly diabetics are subject to a number of
complications as a result of their disease which some feel
mimic conditions seen in premature aging (Eckel et al.
1982). Vascular disease as a result of atherosclerosis is
the most serious of the conditions. Lipid and cholesterol
deposits in far greater numbers then are usually found in
matched nondiabetics, cause changes in the walls of the
blood vessels, resulting in accelerated atherosclerotic
lesions occurring at earlier ages. Silent heart attacks or
strokes are often the outcome (Levin 1983). Such
macrovascular disease accounts for 75% of deaths in
diabetics (Eckel et al. 1982).
Peripheral vascular disease (PVD) involves the small
arteries and arterioles causing outright occlusion as is
seen in gangrene of the feet and toes, or atrophic skin
changes (Levin 1983). Diabetes is a primary cause of PVD
although smoking, hypertension, and lipid abnormalities
compound the problem (Eckel et al.1982). Presently 50% -
70% of all amputations performed in the U.S. are on
diabetics (Levin 1983).
The causes of visual loss in the diabetic include
cataracts, glaucoma, and optic neuropathy (Liang et al.
1980). Diabetic retinopathy is the result of both an
accumulation of lipids in the macula, superimposed on a
compromised vascular system (Levin 1983). Such a condition
26
causes a decrease in central vision often resulting in
legal blindness. The course of diabetic retinopathy and
visual impairment occurs much more quickly and severely in
the elderly diabetic. Diabetics have a 13 times greater
chance of being blind than nondiabetics (Kart et al. 1978).
Nephropathy in the diabetic is usually manifested as a
change in the glomeruli with proteinuria being one of the
earliest signs (Eckel et al.1982). Hypertension will often
occur as a late sign in patients with nephrosclerosis and
pyelonephritis (Levin 1983). Urinary tract infections are
more commonly severe in the elderly diabetic and may be the
result of loss of glycémie control (Fitzgerald et al.1985).
Neuropathy can cause significant complications
particularly as a result of the hypoesthesia of the lower
extremities. Such a condition causes an insensitivity to
any type of trauma to the feet (Levin 1983) and compounds
the damage incurred through peripheral vascular disease. As
a result, injuries of even the most inconsequential nature
can lead to the ulcerations, infection, gangrene and
amputation of the toes and feet as previously described.
Further, autonomic neuropathy in the elderly can manifest
as impotence, neurogenic bladder, alterations in
gastrointestinal motility, and orthostatic hypotension
(Eckel et al. 1982). It is not uncommon for the development
of signs and symptoms of neuropathy to occur even in a
well-controlled diabetic.
27
Ketoacidosis occurs at a much lower rate in the
elderly then in younger age groups, however, the number of
fatal outcomes is greater (Goldberg et al.1985).
Infections which may be of a trivial nature are the most
likely precipitating events of the ketoacidosis although
stroke, trauma, and myocardial infarction have also been
implicated (Fitzgerald et al. 1985). Goldberg et al. (1985)
states that infections which more commonly effect the
elderly diabetic and which run a more florid course are
gingivitis, vulvovaginitis, urinary tract infections (in
females), and infected foot ulcers. The usual cause of
death from ketoacidosis will be from infection, arterial
thrombosis, and shock.
Nonketotic hyperglycemic hyperosmolar coma is a more
frequent complication of diabetes in the elderly than is
ketoacidosis. The fatality rate is quite high with this
condition as well, reaching 50% or more (Redmon 1984).
Circulating insulin, sufficient to suppress fatty acid
mobilization and subsequent ketogenesis but insufficient to
control glucose metabolism is present, causing severe
hyperglycemia in the absence of ketoacidosis (Goldberg et
al. 1985). Profound dehydration and neurologic signs
ranging from confusion to coma occur (Fitzgerald et al.
1985). The development of this type of coma has been linked
to physiologic stress (Redmon 1984), high carbohydrate
28
intake (Fitzgerald et al.1985), or certain medications and
medical procedures (Goldberg et al. 1985).
Treatment
For many NIDDM individuals, diet control is the
mainstay of the therapeutic regimen (Eckel et al. 1982).
Achieving an idealized weight may normalize the glucose
tolerance (Lipson et al. 1984). A balanced diet tailored to
the individual's economic and cultural background and
containing 45% - 55% complex carbohydrates as its core
should be emphasized (Redmon 1984). Total caloric reduction
should be encouraged in the obese as should exercise
(Marble 1985). Since exercise improves glucose tolerance
and may often decrease the need for medications (Lipson
1986), it should be encouraged for all diabetics. An
activity such as walking is an excellent form of exercise
which is well tolerated by many of the elderly (Kart et al,
1978) .
When the symptoms of diabetes are uncontrolled by diet
and exercise, a sulfonylurea medication is added to the
daily routine. The modes of action of the sulfonylurea
medications are four: 1) Enhancement of glucose stimulated
insulin release ; 2) Reduction of hepatic glucose
production; 3) Growth in the number of insulin receptors ;
and 4) Improvement of the intracellular action of insulin
(Lipson 1986). Most of the above actions, however, are
contingent upon the prolonged use of the medication. If
29
after a suitable trial period, the upper limits of the
medication dosages are reached and the fasting plasma level
still exceeds 180 mg/dl, insulin therapy should be
instituted (Lipson et al. 1984). In the elderly a single
daily injection of a premixed, intermediate acting insulin
preparation may be preferred (Fitzgerald et al. 1985) for
ease of administration. However, when such a regimen is
unsuccessful at alleviating the hyperglycemia, a split dose
routine whereby the patient injects two-thirds of the dose
before breakfast and one-third before dinner may be begun
(Shuman 1984). Initially treatment is initiated with a
small dose of insulin and is adjusted during weekly or
biweekly visits to the physician until a fasting glucose
between 120 to 140 mg/dl and a postprandial glucose of
between 180 and 220 mg/dl are achieved (Shuman 1984).
30
CHAPTER IV
ORAL MANIFESTATIONS OF DIABETES
Diabetes mellitus is associated with several changes
in the oral tissues. As a result, the uncontrolled
diabetic in particular often has complaints of a sore
mouth, burning tongue, loose teeth, gingiva of a violaceous
hue, and even a foul taste. One of the recognized cardinal
symptoms of diabetes, which is in part due to the osmotic
diuresis caused by the hyperglycemia, is polydipsia.
Diabetics, particularly those whose disease is out of
control often experience an uncomfortable xerostomia which
they attempt to relieve by increasing their fluid intake.
This symptom of dry mouth, however, is a result of the
salivary glands attempting to adjust their outflow to
conserve fluids, and is not rectified unless the cause of
the diuress is eliminated.
Salivary Gland Effects
A decrease in the number of acinar cells of the
salivary glands (particularly the parotid) and their
replacement with fat cells alters the capacity for
production and excretion of saliva (Gottsegen 1983). A non
inflammatory, non-neoplastic enlargement of the salivary
glands which has been labeled sialadenosis, has been
31
observed in the diabetic (Cutler et al.1982). Such changes
may be due to degeneration of the intraglandular nerves
that regulate secretion (Murrah 1985). The combination of
all of the above factors results in a diminished flow of a
thick, mucoid type of saliva, which is unable to perform
the normal salivary functions of lubrication of the
tissues, remineralization of the dentition, mechanical
cleansing of debris from the hard and soft tissues, and
antimicrobial activity directed against oral organisms. As
a result, the uncontrolled diabetic can suffer from
symptoms of burning tissues, dysphonia, dysphagia, altered
taste sensation, increased caries activity, and problems in
denture wearing (Cutler et al. 1982). Further, upon
examination, an increased incidence of fungal infections
(Saadoun 1980), and fragility of the oral mucosa may be
observed (Rothwell et al.1984).
Caries Rate
Increased levels of glucose have been detected in the
saliva of diabetics, particularly that saliva which comes
from the parotid gland (Sharon et al, 1985). In spite of
this finding, and although glucose is a simple sugar and
therefore might be expected to be cariogenic, there is no
evidence that the caries rate routinely increases in the
diabetic. Years ago in the pre-insulin era this was the
usual finding (Gottsegen 1983). It is hypothesized that
-the_dietarv_restrictions observed by the diabetic__________
32
contributes to the low caries rate, as diabetic diets
commonly have low cariogenic potential (Murrah 1985). For
children who are insulin dependent diabetics this has
proven to be the case, as they have much lower caries
activity than do age and sex matched controls (Genco et al.
1982) .
In the uncontrolled elderly diabetic when xerostomia
is a problem, the level of caries often increases due to
the diminished salivary flow and the absence of the
cleansing, remineralizing, and antimicrobial effects of
saliva. The destruction of the hard tissues is compounded
if the patient participates in caries promoting behaviors
(e.g. sucking on sticky or cariogenic foods, or the
frequent intake of foods) to reduce the dry mouth symptoms.
Periodontal Disease
Uncontrolled IDDM and severe NIDDM have been shown to
increase the severity of periodontal disease causing
inflamed gingiva with increased alveolar bone loss and
tooth mobility (Gotthelf et al. 1983). Local abscess
formation in such cases is not uncommon and has been known
to lead to an infiltrating cellulitis (Cutler et al. 1982).
Local factors such as bacterial plaque and calculus
encourage the progress of periodontal disease in all
individuals, however, in a diabetic, the progress is more
rapid and destructive, particularly in the uncontrolled
diabetic, and appears to be out of proportion to the
33
quantity of bacterial plaque or calculus present. (Genco et
al. 1982). Evasti et al. (1985) has shown that in diabetics
with similar levels of etiologic factors, the difference in
the amount of gingival bleeding between poorly controlled
diabetics and diabetics with good or moderate control was
statistically significant. Similarly in a study by
Tervonen et al. (1986) which examined periodontal conditions
in a population of diabetics, there was a statistically
significant increase in the prevalence of periodontal
pockets as the control of the diabetes worsened.
The role of the polymorphonuclear neutrophilic
leukocyte has been described as extremely important in
providing protection against the bacteria which are the
causative factors of periodontal disease (Manouchehr-Pour
et al. 1981). The hyperglycemia and its resulting
hyperosmolarity, however, has been shown to decrease
activities of the neutrophil which affect its ability to
control bacterial invasion including phagocytosis,
diapedesis, and intracellular bacteriocidal action thus
altering this protective role (Saadoun 1980). McMullen et
al. (1981) demonstrated that patients with abnormal glucose
tolerance (prediabetics) exhibit a decrease in chemotaxic
response. Manouchehr-Pour et al. (1981) proved the same
thing in diabetics. Furthermore, in the latter study the
level of inhibition of chemotaxic activity was not
34
correlated with the fasting blood glucose levels, insulin
doses, or duration of diabetes.
Ketoacidosis also affects the ability of the host to
fight infection by further decreasing phagocytosis,
interfering with the mobilization of the neutrophil, and
decreasing antibody production in those who are frequently
in a ketoacidotic state (Genco et al. 1982).
Finally, vascular insufficiency caused by capillary
basement membrane thickening, narrowing of the lumen, and
periendothelial thickening results in a decreasing flow of
blood to the tissues thus interfering with neutrophil
mobilization (Gottsegen 1983). McMullen et al.(1967) found
that prediabetics with lesions involving the eye, kidney,
and cutaneous tissues, also demonstrated a thickening of
the small blood vessels in the alveolar mucosa. It is
theorized that this type of microangiopathy may begin many
years before diabetes is actually detected (Genco et al.
1982).
These changes in the blood vessel permeability may
reduce the metabolic exchange (particularly of oxygen) in
the periodontal tissues, thus allowing an increase in
anaerobic organisms (Gottsegen 1983). This is a
significant finding since most organisms associated with
destructive periodontal disease are anaerobic. A reduced
oxygen supply may also contribute to pathologic bone
remodeling (Genco et al.1982). The accessibility of the
35
tissues to plasma antibody and complement are also affected
by the vascular insufficiency resulting in a periodontium
more susceptible to the pathogenesis of periodontal disease
(Genco et al. 1982).
Other Effects
It has been noted that the filiform papillae in the
central part of the tongue of the diabetic may be atrophied
while at the same time the fungiform papillae of the tongue
may be enlarged although the reason for such changes is
unknown (Gotthelf et al. 1983). Many investigators have
described a relationship between oral lichen planus
(particularly the erosive type) and diabetes mellitus
(Lundstrom 1983), although a recent report by Lozada-Nur et
al. (1985) challenges this finding.
Management of Oral Disease in the Diabetic
Periodontal disease is much more difficult to manage
in a diabetic and is often a contributing factor in the
loss of control of the diabetes (Rothwell et al. 1984). It
is not, however, our sole concern. Because of the abundant
flora present in the oral cavity, and the decrease in the
inherent protective mechanisms of the neutrophils,
particularly during periods of hyperglycemia, there is an
increased risk of sequelae for any kind of oral surgical
procedure, from extractions to gingival surgery (Gotthelf
et al.1983). Delayed healing is a further complication
36
which can result in a number of post operative problems
including localized osteitis (Cutler et al.1982). For these
reasons it is important for the dental practitioner to know
whether or not an elderly patient may in fact be an
uncontrolled diabetic prior to initiating invasive dental
treatment. In addition, since the presence of oral
infections can drastically affect the day to day status of
the hyperglycemic control by altering glucose metabolism
and in the insulin controlled, the dosage needed; and
because oral infections (including periodontal disease) can
be much more severe in the diabetic, the physician treating
an elderly diabetic patient needs to stress the importance
of good oral health and regular preventive dental visits to
patients who have hyperglycemia (Cutler et al,1982).
37
PART II
EVALUATIVE STUDY OF THE DIABETIC STATUS
OF ELDERLY DENTAL PATIENTS
38
CHAPTER V
AIMS AND OBJECTIVES
The four aims and objectives of this research project
are set forth in the following statements:
1. Test the validity of the assertion by elderly diabetic
patients (those aged 50 or above) attending the USC
Geriatric Dental Clinics that their diabetes is in control
(disease control will be verified by fasting plasma glucose
and glycosylated hemoglobin levels).
2. Determine the accuracy of capillary blood glucose tests
compared with plasma glucose tests (both fasting and
postprandial) to monitor the hyperglycemic status of
diabetic and nondiabetic elderly patients (aged 50 years or
more) attending the USC Geriatric Dental Clinics.
3. Determine the accuracy of a questionnaire (Appendix C)
which asks classic questions about the signs and symptoms
of diabetes compared with the plasma glucose and
glycosylated hemoglobin tests to detect the presence of
undiagnosed or uncontrolled hyperglycemia (using the
criteria of the National Diabetes Data Group, Appendix A)
in the elderly patient (aged 50 years or more) attending
the USC Geriatric Dental Clinics.
4. Determine the prevalence of undiagnosed diabetes or IGT
(as defined by the National Diabetes Data Group criteria.
39
Appendix A) in elderly patients (aged 50 years or more)
presenting for dental treatment at the USC Geriatric Dental
Clinics as verified by laboratory test (oral glucose
tolerance test).
40
CHAPTER VI
METHODS
From March 4, 1985 until April 30, 1986, all
individuals who visited either of the two extramural
geriatric dentistry clinics operated by the University of
Southern California School of Dentistry and who were not
acutely ill, in the process of actively receiving medical
treatment to stabilize a chronic condition, or having a
decreased mental status were asked if they wished to be
included in this study. The sample was drawn from a
multiethnic potential population of 1400 individuals living
in two congregate housing facilities for the elderly in the
greater metropolitan area of Los Angeles. The diagnosed
diabetics were placed in Group A, and a second group (Group
B) was established for those who had never been previously
diagnosed as diabetic. The protocol of the study was
explained and medical history forms were filled out. The
medical history form used for the collection of data was
the standard form in use throughout the University of
Southern California Dental School Clinics. A consent form
explaining the purposes and possible problems of the study
was signed by the each of the participants before they were
enrolled as study subjects (Appendix B). The study
subjects were then interviewed using a signs and symptoms
41
questionnaire (Appendix C). The signs and symptoms
questionnaire was developed through the use of a signs and
symptoms table published by Goldberg et al (1985). This
questionnaire consisted of a total of 17 questions. Eleven
of the questions were asked of each patient. Three
additional questions relating to a patient's obstetric and
gynecologic history were asked of the female subjects.
Each question was stated in such a way that it would be
easily understood by the patient, and no complicated
medical terminology was used. The examiner of the patient
summarized the patient's medical history on this form and
also completed two additional questions : Question 15 asked
the examiner if the subject was obese (that is fulfilled
the standard of being at least 20% over ideal body weight).
A positive response by the examiner to question 16
indicated that the patient had some type of history of
arterial disease (macroangiopathy) whether it be
cardiovascular, cerebrovascular, or of the peripheral
vasculature. A final 17th question, asking if the subject
felt that his or her diabetes was currently in control was
asked only of the previously diagnosed diabetics.
Once all of the above data was derived from the
initial interview, subjects were reappointed for the
laboratory testing. Written and verbal instructions were
given requesting the subjects to fast for at least ten
hours prior to their blood being drawn. Those who were
42
taking medications for their medical conditions were told
to continue to take their medications as prescribed except
for those diagnosed diabetics who were taking oral
hypoglycemics or insulin. For these patients, because their
schedules invariably called for early morning doses of
their medications, very early blood drawing appointments
were scheduled. In this way the patients would be able to
resume their usual morning medication routines shortly
after their blood was drawn, thereby causing minimal
disruption to their medication schedule. Detailed dietary
instructions were spelled out with regard to the exclusion
of all liquids except water.
When the subjects returned to the clinic for their
phlebotomy they were questioned about their food,
medication, and liquid intake. (Those who were noncompliant
were reinstructed and given new appointments.)
While the patients were either sitting or reclining,
blood samples were obtained using an aseptic technique for
the following blood tests: fasting venous blood glucose,
glycosylated hemoglobin, and a capillary blood glucose
test. For both of the venous blood collections, 5 cc
evacuated blood collection tubes were used, one with
ethylenediaminotetraacetate (EDTA) and the other with
potassium oxalate as the preservatives. The capillary blood
was collected via a prick of the pad of the ring finger of
either hand (patient’s choice). The prick was delivered
43
with a sterile lancet mounted in a spring activated device
(Autoclix, Biodynamics, Inc., Indianapolis, IN). Such
devices are recommended because they are fast and they
control the depth of the puncture.
After these fasting blood samples were drawn, subjects
who had never been diagnosed as being diabetic were asked
to drink a 75 gram noncarbonated glucose (dextrose USP)
beverage prepared for postprandial and glucose tolerance
testing (Dextrol, American Scientific Products, Irvine,
CA). Subjects were cautioned about not eating or drinking
anything except water during the subsequent intervals. At
one hour and two hours later additional blood samples were
drawn. Another capillary test was also run at the two hour
mark. The already diagnosed diabetics were not given the
carbohydrate load and subsequent blood work since the
purpose of the extended glucose testing was to ascertain
the numbers of undiagnosed diabetics which might exist in
the control population.
The results of the capillary blood glucose test were
determined in the treatment clinic immediately after the
finger prick, by placement of a drop of capillary blood on
the pad of a glucose detecting reagent strip (Chemstrip bG,
Boehringer Mannheim Diagnostics, Inc., Indianapolis, IN).
Each strip was read twice, first with the unaided eye, and
second with a reflectance meter (Accu-Chek bG, Boehringer
Mannheim Diagnostics, Inc., Indianapolis, IN.). The
44
reflectance meter was used to verify the accuracy of the
unaided evaluation of the reagent strips. All evaluations
by both machine and unaided were made by the same
individual throughout the entire study.
All blood samples were processed in the following
manner immediately after being drawn: The evacuated blood
collection tube which had the potassium oxalate
preservative was labeled and centrifuged for five minutes
at 2000g. The supernatant plasma was withdrawn with a
sterile pipette and decanted into 3cc tubes, labeled and
frozen at -16° C. The sediment was discarded. The specimen
was transported on ice in this state to the laboratory
within 96 hours where analysis was performed using an
automated glucose oxidase hydrogen peroxide method
(Boehringer Mannheim Diagnostics, Indianapolis, IN.).
The second sample of venous blood, collected in the
evacuated blood collection tube with EDTA as the
preservative was also immediately labeled, and centrifuged
at 2000g. The supernatant plasma was withdrawn from the
tubes as previously described and discarded. The sediment
was stored at 1° C in the original collection tube. The
specimens were transported to the laboratory on ice within
96 hours. There the erythrocyte sediment was washed twice
in normal saline (0.14 mol/1 NaCl) and resuspended in an
equal volume of normal saline. It was incubated overnight
at 0° C. After a second centrifugation, an aliquot of
45
erythrocytes was hemolyzed in five volumes of distilled
water. Hemoglobin electrophoresis was performed on
cellulose acetate at alkaline pH and glycosylated fractions
were adjusted in the presence of abnormal hemoglobins. The
glycosylated hemoglobin fraction was separated by ion
exchange chromatography on Bio-Rex 70 according to a
modification of the method of Trivelli et al (1971).
Prefilled minicolumns were used to measure total
glycosylated hemoglobin (Isolab, Inc., Akron, Ohio). The
hemoglobin fractions were measured in a Gilman
Spectrophotometer at 415 nm, and the quantity of the
glycosylated fraction was expressed as a percentage of the
total amount of hemoglobin. The mini column procedures have
been shown to correlate quite closely with the Trivelli et
al (1971) macrocolumn method (Abraham et al 1978;
Schellekens et al 1981). Inter and intra-assay variation
was 3.9% and 4.6% respectively.
Descriptive statistics was used to define, compare and
contrast the two groups of subjects. Fisher's Exact Test
(2-tail) was applied for the analysis of the self-report of
disease control of the diabetic group. Correlation analyses
and simple linear regression were performed to examine the
relationships between the different types of blood test
results. Stepwise logistic regression was performed on the
use of a number of variables including the signs and
46
symptoms questionnaire as predictors of uncontrolled or
undiagnosed diabetes.
47
CHAPTER VII
RESULTS
A total of 133 subjects were recruited from a
potential patient population of 1400 individuals living at
the sites of the USC Geriatric Dentistry Clinics. Of that
number 27 of the subjects were dropped from the study.
Eighteen of those dropped had never been previously
diagnosed as diabetic while nine of them had had such a
diagnosis. Three of these subjects were discontinued when
it became impossible to successfully complete their one and
two hour venipunctures ; the remainder were dropped when
they repeatedly failed their appointments for the blood to
be drawn. Of the 106 subjects who completed the study, 42
were male and 64 were female. The subjects who had already
been diagnosed as diabetics numbered 33 (Table 1). In this
Table 1
Age Distribution of Subjects
Age Groups 51-60 61-70 71-80 81-90 Total
Diabetics 1 8 17 7 33
Non Diabetics 4 20 30 19 73
group (Group A), four subjects were using oral hypoglycemic
medic at ions_to_control_the.i r diabetes, eleven were on_____
48
insulin, and the remainder were diet controlled. The second
group (Group B) consisting of those subjects who had never
been previously diagnosed as diabetic, numbered 73.
The age range for the diabetics was 57 to 90 years while
that for the nondiabetics was 54 to 90 with the mean ages
for each group being 74.3 and 74.8 respectively.
Table 2
Descriptive Statistics
Group A Group B
Variable N Mean S.D. N Mean S.D.
Fast. Visual
Capillary (VC) 33 170.6 69.5 72 115.7 20.3
Fast. Machine
Capillary (MC) 33 170.7 62.0 72 120.9 16.6
Fast. Venous
Blood (VBG) 33 180.9 62.7 71 105.4 14.5
Fast. Glycosylated
Hemoglobin (HBF) 33 9.5 2.7 73 7.7 1.4
1 Hr. P.P. Venous
(VBGl) 73 168.4 59.6
2 Hr. P.P. Visual
Capillary (VC2) 72 162.1 37.1
2 Hr. P.P. Machine
Capillary (MC2) 73 153.2 30.3
2 Hr. P.P. Venous
(VBG2) 72 136.0 46.2
The sample size. means and standard deviations for
each variable by group are presented in Table 2.
49
Fisher's Exact Test
Since it is important to know if already diagnosed
diabetics (Group A) are compliant with their prescribed
therapeutic regimen to control the disease of diabetes
before initiating dental care, an analysis of the data was
performed to see how accurately subjects’ self-report the
control of their diabetes (Table 3). It can be seen from
Table 3
Table of Compliance (As Measured by VBG) Vs Noncompliance
COMP-VBG
Frequency
Percent
Col Pet
Under
Control
Yes
Under
Control
No
COMP-SR (Self-Report)
Under Control
Yes I No
17
60.71
11
39.29
2
40.00
3
60.00
Total
19
14
Total 28 33
Fisher's Exact Test (2 - tail) p value = 0.63
these results that among the 28 subjects who reported their
diabetes to be under control, 11 (39%) were not under
control according to their fasting glucose values (VBG).
The_Fisher_Ls_Exac.t__Test (2 tail) was not significant with a
50
p = 0.63. That is to say there is no statistical
significant association between the subject's self-report
of control of their diabetes (Comp-SR) and the actual
control as measured by a fasting blood glucose test (Comp-
VBG) .
The same analysis was run using the glycosylated
hemoglobin values as the laboratory indicator of compliance
Table 4
Table of Compliance (As Measured by HBF) Vs Noncompliance
COMP-HBF
Frequency
Percent
Row Pet
Under
Control
Yes
Under
Control
No
COMP-SR (Self-Report)
Under Control
Yes ! No
10
35.71
18
64.29
2
40.00
3
60.00
Total
12
21
Total 28 33
Fisher's Exact Test (2 - tail) p value = 1.00
(Table 4). In this case 18 out of 28 subjects of those who
by self-report were in control were actually out of control
by their laboratory results, or 64%. The Fisher's Exact
Te St _(_2 _t a il )__wa s_non significant with a p = 1.00.__________
51
Table 5
Correlation Coefficients of Fasting Variables
MC VBG HBF
VC r = 0.96 0.86 0.41
Similar to the previous test, the results of this
analysis using the glycosylated hemoglobin (HBF) values in
comparison to the subjects self-reported estimate of
compliance again reveals that there is no correlation
(p >•05).
Correlation Coefficients
Correlation coefficients were determined for the
fasting variables (Tables 5 and 6). The capillary tests
MC
VBG
P
n
r
P
n
r
P
n
0.0001
104
0.0001
103
0.88
0.0001
103
0.0001
105
0.42
0.0001
105
0.51
0.0001
104
visually (VC) and machine read (MC) were very strongly
correlated, (r = 0.96, p = .0001). The fasting blood
glucose (VBG) test results also correlated strongly with
both the VC score (r = 0.86, p = .0001) and the MC value (r
= 0.88, p= .0001). There were moderate but statistically
significant correlations between the fasting blood glucose
52
(VBG) test and the glycosylated hemoglobin (HBF) test (r =
0.51,
P =
.0001), and between the glycosylated hemoglobin
(HBF) test and the visually read (VC) and machine read (MC)
capillary tests ( r = 0.41, p = .0001 and r = .42, p =
Table 6
Correlation Coefficients of Non-Fasting Variables
VBGl VC 2 MC2 VBG2
VC r =
P =
n =
0.36
0.0016
72
0.30
0.0111
72
0.30 0.29
0.0095 0.0128
72 71
MC r =
P =
n =
0.25
0.0325
72
0.30
0.0097
71
0.31 0.24
0.0086 0.0412
72 71
VBG r =
P =
n =
0.65
0.0001
71
0.40
0.0007
70
0.46 0.65
0.0001 0.0001
71 70
HBF r =
P =
n =
0.11
0.3336
73
0.06
0.6016
72
0.09 0.10
0.4723 0.4273
73 72
VBGl r =
P =
n=
0.46
0.0001
72
0.50 0.65
0.0001 0.0001
73 72
VC2 r =
P =
n =
0.90 0.71
0.0001 0.0001
72 71
MC2 r =
P =
n =
0.80
0.0001
72
.0001 respectively). The tests performed with blood
samples drawn at one and two hours (Table 6), did not
53
correlate as strongly with each other as the fasting values
had: VC2 correlated to MC2, r = 0.90, p = .0001; VC2
correlated to VBG2, r = 0.71, p = .0001; MC2 correlated to
VBG2, r = 0.80, p = .0001.
The correlation of the fasting tests to the one and
two hour tests were even weaker, the highest correlation
being between VBG and VBGl (r = 0.65, p = .0001). All
correlations were in a positive direction, however, with
p < .05 for all correlations except those between the
glycosylated hemoglobin (HBF) and the nonfasting blood
tests (VBGl, VC2, MC2, VBG2) as displayed in Table 6.
Regrouping of Subjects
Individuals who, according to the National Diabetes
Data Group (1979), fit the criteria (Appendix A) for
impaired glucose tolerance or diabetes as interpreted by
the results of their glucose tolerance tests were grouped
with those who had already been diagnosed as diabetics
(Group A) to form a new set of subjects labeled Group C.
All other subjects were then placed in Group D. Re
examination of the data and comparisons with the previous
findings from Groups A and B are displayed in Table 7.
Comparing Group A with Group B and Group C with Group
D, it can be seen that the correlation coefficients between
the groups are quite different. When Group A is compared to
Group C and Group B is compared to Group D, there is very
54
little difference between the results. Further inspection
Table 7
Comparisons of Correlational Coefficients and Ranges
Diagnosed
Yes
as Diabetics
No
Real Diabetics or IGT
Yes No
Tests Group A Group B Group C Group D
n = 33 n = 73 n = 47 n = 59
VBG-MC r = 0.90 0.36 0.90 0.18
VBG-VC r = 0.87 0.45 0.86 0.21
VBG-HBF r = 0.38 0.12 0.44 0.01
VC-MC r = 0.96 0.82 0.96 0.76
Ranges
VBG 96 - 358 83 - 148 96 - 358 83 - 136
HBF 4.7-15.0 4.4-12.5 4.7-15.0 4.4-12.5
shows that the difference between Groups C and D is greater
than the difference between Groups A and B.
Linear Regression Analysis
Linear regression analysis was performed using VC as
the independent variable and MC and VBG as the dependent
variables respectively (Table 8). With MC serving as the
dependent variable, r 2 = 0.916. This regression explained
91.6% of the total variability. When the dependent
variable was changed to VBG, R^ =
variability was explained by VC.
0.740, 74% of the
55
total
Table 8
Linear Regression Analysis
Intercept (SE) Slope(SE) r2
Indep Var = VC
Dep Var = MC 22.78 (3.64)*** 0.86 (0.03)** 0.916
Dep Var = VBG 11.07 (7.46) 0.89 (0.05)** 0.740
Indep Var = VC2
Dep Var = MC2 34.99 (7.24)*** 0.74 (0.04)** 0.803
Dep Var = VBG2 -6.05 (17.66) 0.88 (0.11)*** 0.498
** Statistically significant at .01 level.
*** Statistically significant at .001 level.
Linear regression analyses were also performed with
VC2 serving as the independent variable (Table 8). With MC2
as the dependent variable, = 0.803, thus VC2 explaining
80.3% of the total variability. With VBG2 as the dependent
variable, R^ equals 0.498, VC2 explaining only 49.8% of the
total variability.
Questionnaire Results
Although initially it was expected that the data from
the Signs and Symptoms Questionnaire would be interpreted
according to the original two study groups, the already
diagnosed diabetics (Group A), and those who had never been
diagnosed-as~such—(-Group-B-)-,_-it_was_clear_f.rom_the_results_
56
of the data that there were quite a number of subjects in
Group B who actually met the criteria for overt diabetes or
impaired glucose tolerance. Therefore the resulting
analysis of the data was accomplished using the two new
designations previously described: Group C and Group D.
Regression Values
The data was first analyzed using the newly defined
Group C or Non Group C (Group D) as the dependent variable,
with age, sex, and questions SI through S16 as the
independent variables (Table 9). Since some of the
questions on the questionnaire were for females only, and
since the statistical analysis program automatically
excludes those subjects where the data is not complete, the
resulting analysis reflects the responses of 62 female
subjects.
Table 9
Logistic Regression Analysis 1
Dep Var = Group C or non Group C (Grp D) N = 62
Indep Var = age, sex, SI - SI6
% of Correct Prediction: Group C = 80%
non Group C = 57.5%
Term Coefficient Standard Coefficient
Error S.E.
SI (Hist) 2.52 0.72 3.521
S16 (Macro) 1.99 0.76 2.63I
Constant -2.75 0.79 -3.46^
Âsi-gnif icant-at—.01—level_f or_2—S-ided_Z_tes-t..
57
From the results of this analysis, it appears that
females who are diabetic or glucose intolerant, whether
previously diagnosed or not are more likely to have a
family history of diabetes (question SI), as well as a
history of major artery (macrovascular) disease (question
S16) .
The same data were rerun eliminating questions S12,
S13, S14 which are those which pertain to women only (Table
10). In this case the dependent variable was again Group C
or non Group C (Group D), and the independent variables
were age, sex, SI - Sll, S15 and S16. The total number of
respondents who had no missing data for this analysis was
103.
Table 10
Logistic Regression Analysis 2
Dep Var = Group C or non Group C (Group D) N = 103
Indep Var = age, sex, SI - Sll, S15, S16
% of Correct Prediction: Group C = 80%
non Group C = 57.5%
Term Coefficient Standard Coefficient
Error S.E.
S16 (Macro) 1.23 0.50 2.481
S6 (Vision) 1.16 0.48 2.421
SI (Hist) 0.99 0.48 2.O5I
S2 (Polydip) -0.87 0.48 -1.80
Constant -1.70 0.47
-3.6o2
^Significant at .05 level for 2 sided Z test.
^Significant at .01 level for 2 sided Z test.
53
In this analysis, S16 (macrovascular disease) and SI
(a history of diabetes in the family) are again present as
they were when the data from the females only were
analyzed. In addition, positive responses to question S6
regarding blurred vision also appears to be significant.
Although the term S2 (polydipsia) was in the model, it was
not significant at the .05 level.
Further analysis was performed using a new variable
called "yes". This variable is the percentage of positive
answers to the total non-missing questions asked of each
subject (excluding question 17 which is a self-report of
the control of the diagnosed disease and was only asked of
the diabetics). The dependent variable remain Group C or
non Group C (Group D) and the independent variables were
age, sex, and the "yes" variable (Table 11).
Table 11
Logistic Regression Analysis 3
Dep Var = Group C and non Group C (Group D) N = 106
Indep Var = age, sex, "yes"
% of Correct Prediction: Group C = 80%
non Group C = 42.5%
Term Coefficient Standard Coefficient
Error S.E.
yes
Constant
3.41
-1.49
1.22 2.795I
0.50 -2.9881
^Significant at .01 level for 2 sided Z test.
59
From this data it appears that there is a correlation
between the number of positive responses to the signs and
symptoms questionnaire and the likelihood that an
individual will have diabetes.
In an effort to reduce the number of variables, and
also to attempt to evaluate the impact of the patient's
current medical history in addition to their responses on
the signs and symptoms questionnaire, an additional data
sheet was prepared for each subject (Appendix D). This data
sheet consisted of a total of seven major categories of
problems often found in the diabetic individual. If a
subject had a positive response on the questionnaire (e.g.
S5, numbness or tingling of the hands or the feet) or had a
corresponding medical problem which would be of the same
category (to be paired with S5, the problem would have to
be a neurological one such as heart rhythm irregularities),
a positive response would be recorded for question 3 (Q3).
In this way data was collapsed into a total of seven
categories which is a modification of the eight groupings
suggested by Goldberg et al (1985). Stepwise logistic
regression was used to analyze the data in its collapsed
state.
For this analysis, the dependent variable remains
Group C or non Group C (Group D ), however, the independent
variables are now age, sex, and Q1-Q7 (Table 12). Variable
Q4 concerns the presence of disease in the smaller blood
60
vessels (microangiopathy) as determined by eye or renal
Table 12
Logistic Regression Analysis 4
Dep Var = Group C or non Group C (Group D) N = 106
Indep Var = age, sex, Q1 - Q7
% of Correct Prediction: Group C = 80%
non Group C = 60%
Term Coefficient Standard Coefficient
Error S.E.
Q4 (Micro) 1.3 5 0.44 3.071
Q7 (Macro) 0.98 0.46
2.152
Constant -1.63 0.44 -3.72^
^Significant at .01 level for 2 sided Z test.
^Significant at .05 level for 2 sided Z test.
disorders. Variable Q7 addresses problems of the larger
blood vessels (macroangiopathy). Both of these questions
were found to have statistical significance.
The difficulties inherent in providing dental care to
diabetics relate to the average control of the blood
glucose. Therefore a further issue to be analyzed in this
study was an investigation of the level of control of the
blood glucose in those individuals who have the disease,
attempting to relate that level of control to their present
signs and symptoms.
A new variable, COMPV, was designated for the subjects
in Group C. This variable was defined as negative if the
fasting venous blood glucose (VBG) test was 180 mg/dl or
61
greater (for the previously diagnosed diabetics), or all
three criteria were met for the diagnosis of diabetes using
the glucose tolerance test as outlined by the National
Diabetes Data Group (if the subjects had never been
previously diagnosed as diabetic). COMPV, representing
those individuals who were hyperglycemic or non COMPV,
representing those who were not hyperglycemic was the
dependent variable in this series of inquiries. In the
first analysis, the independent variables are age, sex, and
Q1-Q7 (Table 13).
Table 13
Logistic Regression Analysis 5
Dep Var = COMPV or non COMPV N = 47
Indep Var = age, sex, Q1 - Q7
% of Correct Prediction; Group C = 80%
non Group C = 57.5%
Term Coefficient Standard Coefficient
Error S.E.
Q2 (card) -2.62 1.12 -2.34^
Q7 (Macro) 2.07 1.00 2.061
Q1 (Hist) 1.96 1.07 1.84
Constant -1.09 1.02 -1.07
^Significant at .05 level for 2 tailed t-test.
The results of this analysis reveals that the better
the subjects' fasting venous blood glucose values represent
a control of the diabetes, the less likely cardinal
symptoms of the disease, polyuria, polydipsia, nocturia
621
(Q2) will exist. The presence of macroangiopathy (Q7)
however is more likely to be found in subjects who are
controlling their diabetes. In this analysis, although term
Q1 (histories of family incidence of diabetes and obstetric
and gynecologic problems) was in the model, it was not
significant at the p <.05 level.
A second analysis using COMPV or non COMPV as the
dependent variables and age, sex, and questions Si - SI6
was run (Table 14). Since all the questions were used,
only the female subjects who had COMPV values were analyzed
(n = 27).
Table 14
Logistic Regression Analysis 6
Dep Var = COMPV or non COMPV N = 27
Indep Var = age, sex, SI - S16
% of Correct Prediction: Group C = 80%
non Group C = 65%
Term Coefficient Standard Coefficient
Error S.E.
S2(Dry Mouth) -2.54 1.17 -2.17^
S9(Oral Bleed) 2.16 1.31 1.65
Constant 0.00 0.56 0.17
^Significant at .05 level for 2 tailed t-test.
The results of this analysis indicate that if subjects
have their blood glucose under control, it is less likely
that they will demonstrate signs and symptoms of dry mouth
63
(S2). Although term S9 (which is concerned with bleeding
gums or denture sores) was in the model, it was not
statistically significant at the p < .05 level.
An additional analysis was performed again using COMPV
or non COMPV as the dependent variable and age, sex, and
questions SI - Sll, and S15 and S16 as the dependent
variables (Table 15).
Table 15
Logistic Regression Analysis 7
Dep Var = COMPV or non COMPV N = 45
Indep Var = age, sex, SI - Sll, S15, S16
% of Correct Prediction: Group C =
non Group C =
80%
52.5%
Term Coefficient Standard
Error
Coefficient
S.E.
S16(Macro) 4.78 1.70 2.811
S3 (Polyuria) -2.88 1.16
-2.472
S15(Obese) -1.97 0.97
-2.032
Sll (Teeth) 1.88 1.00 1.88
Constant -2.17 1.19 -1.82
^Significant at .01 level for 2 tailed t-test.
^Significant at .05 level for 2 tailed t-test.
In this analysis those individuals who had their blood
glucose in control are more likely to have macroangiopathy
(S16). The presence of polyuria (S3) or obesity (S15) is
less likely with better control of the blood glucose.
Although Sll (which pertains to missing teeth) was in the
64
model, it was not statistically significant at the p < .05
level.
An attempt was made to continue rerunning the data
using COMPV or non COMPV as the dependent variable,
duplicating the same analyses as were done with Group C and
Group D as the dependent variables. However, when the
independent variables of age, sex, and "yes" (which is the
value related to all the yes answers on the questionnaire)
were input, none of the variables was selected in the
stepwise logistic regression.
Similar attempts were made to rerun the data using the
categories of COMPH or non COMPH which are parallel
variables to COMPV, but use the glycosylated hemoglobin
value as the determining criteria. This logistic
regression model did not seem to fit the data well.
65
CHAPTER VIII
DISCUSSION
Fisher's Exact Test
The analyses of the self-report of compliance with the
therapeutic regimen which would result in a controlled
glycémie state and thus blood glucose and glycosylated
hemoglobin values within normal limits, reveals that there
is no relationship between the self-report of the diabetic
and the actual level of control of glycemia as determined
by a laboratory test (Tables 3 and 4). This is consistent
with the compliance literature as reviewed by Becker et al.
(1985). For both Fisher's Exact Tests, p> 0.05. Therefore
the null hypothesis, that there is no relationship between
the answer to the question "Is your diabetes in control?"
and the actual laboratory values testing that control must
be accepted.
Correlation Coefficients
The correlation coefficient obtained when comparing
the visually (VC) and machine read (MC) fasting capillary
blood values in this study of r = 0.96 was better than the
r = 0.9 correlation recommended by Peterson (1985) as
necessary in order to have clinical relevance. Gross et al.
(1985) in their attempt to train patients to evaluate their
blood values, achieved a correlation of r = 0.91 increasing
66
to r = 0.94 by the end of the study. The stronger
correlations in the current study may reflect the fact that
in this study all interpretations were made by one
observer, unlike the method in the study by Gross et al.
According to Burrin and Price (1985) in fasting
subjects the values for venous and capillary glucose levels
are virtually the same. Therefore it is not unexpected that
the correlations in this study of fasting plasma glucose
(VBG) with the two fasting capillary tests (VC and MC)
would be strongly positive (r = 0.86, p = 0.0001; and r =
0.88, p = 0.0001 respectively). The correlations of these
values were not as great as the VBG to VC and the VBG to MC
results of r = 0.97, p < 0.0001, and r = 0.99, p < 0.0001
respectively obtained by Van Crombrugge et al.(1985). The
difference in the two sets of results may be due to the
fact that the sample obtained for the plasma glucose test
was not analyzed immediately as were the capillary tests.
Although Van Crombrugge et al. (1985) did not state the
length of time between the phlebotomy and the laboratory
analysis of the plasma samples, it was noted in other
studies that the samples were processed immediately after
collection (Wing et al. 1985; Boden 1980). Since in the
present study, samples were collected at a site remote from
the laboratory, it is possible that the storage and
transportation of those samples, which varied per batch
within the scheduled timetable, may have allowed some
67
degradation of the sample to occur and thus been the source
of the slight decrease in correlation seen. Bonar (1980)
states that blood glucose in samples can disappear at a
rate of up to 5% per hour if not properly cooled.
The correlations of glycosylated hemoglobin (HBF) and
plasma glucose (VBG) (Table 7) obtained in this study (r =
0.44, p = 0.002) for Group C were not as strong as those
obtained by Graf et al. (1978) for diabetics (r = 0.85, p,
0.001). It is important to note, however, that the
population used in this study for this analysis was Group
C, or all those with previously diagnosed diabetes or
present hyperglycemia which would place them in the
impaired glucose tolerance or diabetes category according
to their oral glucose tolerance test results. Of the 47
subjects comprising Group C, 14 had never been previously
diagnosed as having hyperglycemia. As a result they were
under no treatment regimens for the control of their
disease. In comparison, the subjects in the study by Graf
et al.(1978) were considered clinically stable diabetics.
In addition, 10% of the diabetics in the latter study were
on oral hypoglycemic agents and none on insulin. This
compares to 15% on oral hypoglycemic agents and 33% on
insulin in the present study. The work by Czech et al.
(1983) indicates that poorer correlations are found between
plasma glucose and glycosylated hemoglobin when the
patients are Type II diabetics being maintained on insulin
68
(r = 0.10). Other studies which looked at the diabetic on
insulin have correlations of the plasma glucose and
glycosylated hemoglobin ranging from r = 0.74, p <0.001
(Gonen et al.1977) to r = 0.49, p < 0.01 (Elkeles et al.
1978). In a study of 130 diabetic patients by Lim et al.
(1985) the correlations between the plasma glucose and
glycosylated hemoglobin were higher in the non-insulin
treated diabetics (r = 0.61, p < 0.001) than in the insulin
treated individuals (r = 0.41, p = 0.01-0.02); also those
who were symptomatic had higher correlations (r = 0.64, p =
0.001-0.01) than those who were asymptomatic (r = 0.54, p <
0.001).
The minimal correlation (r = 0.01) between the fasting
plasma glucose (VBG) and the glycosylated hemoglobin (HBF)
for what could be considered the group with normal glycemia
(Group D) had a nonsignificant p value (Table 7). This
result is similar to that found by Dods et al (1979) for a
group of normals, r = 0.07, with no statistical
significance.
Although they correlated moderately well, the one and
two hour blood test values were not as highly correlated
with themselves or with the fasting values. This is to be
expected for according to Burrin and Price (1985) after
meals, rapid glucose uptake can occur in the periphery thus
reflecting different values for the tests obtained from
capillary versus venous blood. The correlation value of
69
fasting plasma glucose (VBG) with the two hour postprandial
plasma glucose (VBG2) in this study was r = 0.65, p = .0001
which compares favorably to the correlation of those two
variables found in the work of Simon et al.(1985) of r =
0.788, p < 0.001.
The lack of statistical significance of the fasting
HBF correlations to the various post carbohydrate loading
tests, VBGl, VC2, MC2, VBG2 (r = 0.11, p = 0.33; r = 0.06,
p = 0.60; r = 0.09, p = 0.47; and r= 0.10, p = 0.43
respectively) is to be expected and relates to the nature
of the components being measured. The latter tests are
sensitive to the carbohydrate load regardless of the
glycosylated hemoglobin value which is dependent upon the
glucose level for the preceding 90 days for its value.
Linear Regression Analysis
Of the four linear regression analyses performed
(Table 8), the R^ value achieved with the capillary reading
read by machine (MC) as the dependent variable and the
visually determined capillary value (VC) as the independent
variable explained the greatest amount of variability
(92%). Even though the two values were not identical, this
high explanation of variability reveals that the two
variables correlate very well. Although the slope of the
VBG regress on VC and VBG2 regress on VC2 were
statistically significant (p = .01 and p = .001
respectively), the intercepts for each were not. The
70
resulting values of 0.740 and 0.498 do not explain
enough of the variability to be used as diagnostic tests in
this context. According to Peterson (1985), tests to
determine blood glucose levels must have sufficient
accuracy and precision to achieve a correlation of greater
than r = 0.9 to be used routinely as measuring instruments.
Since for the MC2 regress on VC2, value only equaled
0.80, even though the slope and the intercept were
statistically significant (p = 0.01 and p = 0.001
respectively), these values do not have the accuracy of the
capillary (MC and VC) variables.
Stepwise Logistic Regression Analysis
Four different stepwise logistical regression analysis
were run with Group C or non Group C (Group D) serving as
the dependent variable and four sets of independent
variables as described previously: age, sex. Si - S16
grouped in four different ways (Si - S16; SI - S12, S15,
S16; Q1 - Q7; and yes). In three of these analyses,
macroangiopathy (S16 or Q 7) entered into the regression
with a diagnosis of diabetes or an abnormal glucose
tolerance test (p = .01, .01, .05). The presence of
microangiopathy (Q 4) or vision disorders which may be due
to microangiopathy (S6) was significantly correlated in two
of the analyses (p = .01, .05). A history of diabetes in
the family entered into the regression in two analyses (p =
711
0.01 and p = 0.05). The "yes" variable also entered into
the regresssion (p = .01).
Among these four analyses the variables of a history
of diabetes in the family (Si) and the presence of
macroangiopathy (S16) are the predictors of diabetes or
glucose intolerance which have the most importance for
female subjects. For all subjects, however, the presence
of macroangiopathy (S16), vision problems (S6), and a
family history of diabetes (SI) are of most importance in
determining those who have impaired glucose tolerance or
are diabetics. The percentage of correctness indicates how
well the model can help us predict the status of the
dependent variable. If the correct percentage of predicting
diabetic patients is set at 80%, we can then compare the
percentage of correct predictions for nondiabetic patients.
It can be seen that using the variables SI - S16 or Q1 - Q7
can give the percentage of correct predictions for
nondiabetic patients about 60% of the time. While if we
collapse these variables to a single "yes" variable, the
correct prediction percentage for nondiabetic patients
decreases to 42.5% Therefore the sheer numbers of
positive responses to signs and symptoms, as investigated
via the "yes" variable had the weakest percentage of
correctness, indicating that just taking the average of
positive responses is not as effective a method of
72!
predicting the presence of diabetes or impaired glucose
tolerance.
That the micro and macroangiopathy variables were
important was confirmed in an additional analysis using
COMPV or non COMPV as the dependent variable. In these
analyses (Tables 13 through 15) it was observed that the
more likely the subjects were to have their disease in
control the less likely they were to experience the
transient signs of hyperglycemia such as polyuria,
polydipsia, and nocturia (Q2, S2, and S3). Obesity (S15)
was also less likely to be found in those maintaining good
control of their disease, however, macroangiopathy (Q7 and
S16) correlated positively with the disease when other
explanatory variables stated above are in the model, just
as it did in a number of the previous analyses.
73
CHAPTER IX
CONCLUSIONS
As a result of this study a number of conclusions can
be drawn;
1) Asking an older patient who has been diagnosed as a
diabetic if his or her diabetes is in control will most
likely not provide a dentist with reliable information upon
which to base a treatment plan and initiate care.
2) Since the MC and VC tests can be totally controlled and
varified for precision in the office environment using
reflectance meters with precalibrated reagent strips, and
since for fasting subjects a high degree of precision can
be achieved when there is good quality control, it is
recommended that dental offices use a capillary blood type
of test for blood glucose determinations in their elderly
diabetic dental patients. For best results, this test
should be administered after a fast.
3) Since the level of undiagnosed hyperglycemia is so high
in this patient population, the capillary test should be
routinely used with all elderly patients. Even if no
invasive procedure is contemplated, the detection and
recognition of hyperglycemia and referral of a patient with
with positive signs to a physician for confirmation and
management of the hyperglycemia will help a great deal in
74
allaying the morbidity and mortality associated with the
disease in this patient population.
4) Although a questionnaire assessing the signs and
symptoms of hyperglycemia would be helpful in predicting
which patients were at risk for hyperglycemia, certain
aspects of the patients medical history, including the
presence of macroangiopathy, microangiopathy, and a history
of diabetes in the family should be given a great amount
of attention. The dentist should not be mislead by the
absence of what has been considered to be the cardinal
symptoms of diabetes: polyurea, polydipsia, or nocturia in
the elderly patient population at risk for Type II
diabetes.
In summary, given the increasing risk of diabetes in
the elderly and the fact that many elderly are undiagnosed,
and that these same individuals are requiring more invasive
dental procedures, the outcome of which could effect and be
effected by the presence of hyperglycemia, and given the
ease of administration, the minimal cost and the precision
of the capillary glucose test, it would be prudent for the
dentist to utilize this test in the dental office with all
elderly dental patients about to undergo dental care on a
routine basis, and particularly prior to invasive
treatment.
APPENDIXES
76
APPENDIX
DIAGNOSTIC CRITERIA FOR DIABETES AND IGT
77
DIAGNOSTIC CRITERIA FOR DIABETES MELLITUS AND IMPAIRED
GLUCOSE TOLERANCE*
Diabetes Mellitus
The presence of one of the following is necessary for a
jdiagnosis of diabetes mellitus:
A random plasma glucose level of 200 mg/dl or greater
îplus classic symptoms of diabetes mellitus.
Or
A fasting plasma glucose of 140 mg/dl or greater on at
least two occasions.
Or
A fasting plasma glucose level of less than 140 mg/dl
plus sustained elevated plasma glucose levels during at
least two oral glucose tolerance tests. The 2 hour sample
and at least one other between 0 and 2 hours after the 75
gram glucose dose should be 200 mg/dl or greater.
Impaired Glucose Tolerance
The diagnosis of Impaired Glucose Tolerance is restricted
to those individuals having all of the following:
A fasting plasma glucose of less than 140 mg/dl.
And
A 2 hour oral glucose tolerance test plasma glucose
level between 140 and 200 mg/dl.
And
An intervening oral glucose tolerance test plasma
glucose of 200 mg/dl or greater.
78
APPENDIX B
INFORMED CONSENT FOR STUDY PARTICIPATION
79
DIABETES RISK ASSESSMENT IN ELDERLY DENTAL PATIENTS
Informed Consent
Purpose :
You are invited to participate in a study designed to
evaluate a laboratory test and an interview process for
Sdentists to use with their elderly dental patients. The
reason why we are developing this procedure is so that we
'can learn how many elderly patients who want dental care
have diabetes, and how many elderly diabetic patients have
jtheir diabetes in control when they come to our clinic
requesting dental care. You were selected as a possible
participant in this study because you are 50 years or older
in age, have come to our clinic requesting a dental
examination, and have not had any recent changes in your
medical history.
Procedure:
If you decide to participate we will ask you to answer some
questions about your medical history. This interview will
take about 30 minutes. Then we will ask you to fast from
midnight on a certain day. The morning of your fast, you
will be met in the dental clinic and a small amount of
blood will be drawn for blood tests. These are routine
blood tests which you have probably had before. If you
have never in the past been told that you have diabetes,
you will be given a sugary liquid to drink. At one and two
hours later you will have the same blood samples taken as
just described. This visit to draw blood will take about 10
minutes.
Risks :
The risks and potential discomfort which usually
accompanies having blood drawn is all that is anticipated
in this study.
Benefits :
The benefits of the study include a better understanding of
how many elderly dental patients there are with diabetes
and how well those who have already been diagnosed as
diabetic keep their disease in control. By knowing this
information we can develop better methods of evaluating the
elderly diabetic patient in order to keep complications of
dental treatment from occurring.
Alternatives :
No standard treatment is being withheld from you if you
-take—part—in—this_study.________________________________ _
80
Confidentiality:
Any information that is obtained in connection with the
study and that can be identified with you will remain
confidential.
Offer to Answer Questions:
If you have any questions now or later, please ask us. You
will be given a copy of this form to keep.
Coercion and Withdrawal Statement:
Your decision whether or nor to participate will not
interfere with your future care in this clinic. If you
decide to participate you are free to withdraw your consent
and to discontinue your participation at any time.
Physical Injury Statement:
If you require medical treatment as a result of injury
arising from your participation in this study, the
financial responsibility for such care will be your.
YOUR SIGNATURE INDICATES THAT YOU HAVE DECIDED TO
PARTICIPATE HAVING READ THE INFORMATION PROVIDED ABOVE.
Patient Signature Date
Witness
81
APPENDIX C
SIGNS AND SYMPTOMS QUESTIONNAIRE
82
DIABETES RESEARCH STUDY
SIGNS AND SYMPTOMS INTERVIEW QUESTIONNAIRE
Name Date
SocSecurity#
Age_ Sex_______ Race_______ Height Weight
Current Medications & Dose
Ask the subjects the following questions:
1. Is there a history of diabetes in your family? Yes No
2. Do you often have a dry mouth? Yes No
3. Do you urinate more than 6 times per day? Yes No
4. Do you get up at night to urinate? Yes No
5. Do you have any numbness or tingling
in your hands or your feet? Yes No
6. Does your vision ever blur or give you trouble? Yes No
7. Do you get skin infections, particularly in
your lower limbs? Yes No
8. Do you have wounds that heal slowly? Yes No
9. Do your gums bleed or do you have denture sores?Yes No
10. Have you ever been told that you have gum
disease, or do you think you have it? Yes No
11. Have you ever lost any teeth due to gum disease?Yes No
FOR WOMEN ONLY
12. Have you every had any miscarriages, stillbirths,
or spontaneous abortions? Yes No
13. Have you given birth to any large babies
(10 lbs or more)? Yes No
14. Do you often have vaginal infections? Yes No
TO INTERVIEWER
15. Is the subject obese (overweight by 20%
ideal weight)? Yes No
16. Does the subject have any medical history
of (macroangiopathy)? Yes No
FOR THE PREVIOUSLY DIAGNOSED DIABETIC ONLY
17. Do you feel your diabetes is currently
in control? Yes No
OTHER SIGNIFICANT MEDICAL FINDINGS
1. 2._____________________
3. 4.
83
APPENDIX D
DATA SUMMARY SHEET
84
DIABETES RESEARCH STUDY
DATA SUMMARY SHEET
Subject Name
Age__________ Response
Ql. Combines questionnaire responses from
Hist questions SI, S12, S13, (a family
history of diabetes or obstetric or
gynecologic problems.) +
Q2. Combines questionnaire responses from
Card questions S2, S3, S4, (the cardinal
signs of diabetes). +
Q3. Combines questionnaire response S5 with
Neuro any history of hypotension, heart rhythm
irregularities, or pacemaker use
(neurologic dysfunction). +
Q4. Combines questionnaire response S6 with
Micro any history of cataracts, glaucoma, or
kidney insufficiency (microangiopathy). +
Q5. Combines questionnaire responses S7, S8,
Infect and S14 and any history of recurrent
infections. +
Q6. Combines questionnaire response S15 and
Endo any history of endocrine - metabolic
complications including thyroid disease. +
Q7. Is there a current history of heart disease.
Macro strokes, or peripheral vascular disease
(macroangiopathy)? +
85
BIBLIOGRAPHY
86
BIBLIOGRAPHY
Abraham, E.C.; Huff, T.A.; Cope, M.D.; Wilson, J.B.;
Bransome, E.D.; and Huisman, T.H.J. Determination of
the glycosylated hemoglobins (HbAl) with a new micro
column procedure. Diabetes, 1978, 2%, 931-937.
American Diabetes Association. The physician's guide to
type II diabetes (NIDDM): Diagnosis and treatment.
New York: American Diabetes Association, 1984.
Arnetz, Bengt B.; Kallner, Anders; and Theorell, Tores The
influence of aging on hemoglobin Ale (HbAlc), J
Gerontol, 1982, 37, 648-650.
Bartolucci, Enrico G. and Parks, Roger B. Accelerated
periodontal breakdown in uncontrolled diabetes. Oral
Surg. , 1981, 387-390.
Beck, Marshall H. and Janz, Nancy. The health belief model
applied to understanding diabetes regimen compliance.
Diabetes Educator, 1985, Spring, 41-47,
Bennett, Peter H. Diabetes in the elderly: diagnosis and
epidemiology. Geriatrics, 1984, 3_9, 37-41.
Boden, Guenther; Master, Ravidatt W.; Gordon, Sandra, S.;
Shuman, Charles R.; and Owen, Oliver E. Monitoring
metabolic control in diabetic outpatients with
glycosylated hemoglobin, Ann Int Med. , 1980, 9_2, 357-
360.
Bonar, J.R. Diabetes, a clinical guide, 2nd ed. Garden
City: Medical Examination Pub. Co. Inc., 1980.
Burrin, J. M. and Price, C. P. Measurement of blood
glucose, Ann Clin Biochem., 1985, 22, 327-342.
Cahill, George F. "Current concepts of diabetes." In
Joslin's diabetes mellitus, 12th ed., pp. 1-11.
Edited by Alexander Marble, Leo P. Krall, Robert F.
Bradley, A. Richard Christlieb, and J. Stuart
Soeldner. Philadelphia: Lea and Febiger, 1985.
Carranza, Fermin A. Glickman's clinical periodontology, 6th
ed. Philadelphia: W.B. Saunders Co., 1984.
87
Cerkoney, Kathy A. and Hart, Laura K. The relationship
between the health belief model and compliance of
persons with diabetes mellitus. Diabetes Care, 1980,
3, 594-598.
Cutler, Leslie S. and Valachovic, Richard W. "The oral
cavity: Manifestations of the metabolic abnormality."
In Diabetes mellitus: Problems in management, pp.
392-340. Edited by J.D. Schnatz. Menlo Park: Addison-
Wesley Pub. Co., 1982.
Czech, Anna and Taton, Jan. Glycosylated hemoglobin (HbAl)
as an indicator of therapy effects in different
clinical types of diabetes, J Chron Dis., 1983, 36,
803-810.
Dods, Richard F. and Bolmey, Carlos. Glycosylated
hemoglobin assay and oral glucose tolerance test
compared for detection of diabetes mellitus, Clin
Chem., 1979, 25^ 764-768.
Eckel, Robert H. and Hofeldt, Fred D. "Endocrinology and
metabolism in the elderly." In Clinical internal
medicine in the aged, pp. 222-255. Edited by Robert
W. Schrier. Philadelphia: W.B. Saunders Co., 1982.
Elkeles, R.S.; Wu, J.; and Hambly, J. Haemoglobin A1, blood
glucose, and high-density lipoprotein cholesterol in
insulin-requiring diabetics. Lancet, 1978, 2, 547-
548.
Ervasti, Tellervo; Knuuttila, Matti; Pohjamo, Leena; and
Haukipuro, Kyosti. Relation between control of
diabetes and gingival bleeding. J Periodontol.,
1985, 154-157.
Evans, Larry and Spelman, Michael. The problem of non-
compliance with drug therapy. Drugs, 1983, 25, 63-76.
Fitzgerald, M.G. and Kilvert, A. "The endocrine system -
diabetes." In Textbook of geriatric medicine and
gerontology, pp. 715-730. Edited by J.C.
Brocklehurst. Edinburgh; Churchill Livingstone, 1985.
Genco, Robert J. and McMullen, James A., "The oral
complications of diabetes." In Diabetes mellitus in
obesity, pp.640-650. Edited by B.N. Brodoff and S.J.
Bleichor. Baltimore : Williams and Wilkins, 1982.
88
Goldberg, Andrew; Andres, Reubin; and Bierman, Edwin L.,
"Diabetes mellitus in the elderly." In Principles of
geriatric medicine, pp. 750-763. Edited by Reubin
Andres, Edwin L. Bierman, and William R. Hazzard. New
York; McGraw Hill, 1985.
Gonen, Boas ; Rubenstein, Arthur H.; Rockman, Hyman; Tanega,
Sonia P.; and Horwitz, David L. Haemoglobin Al: An
indicator of the metabolic control of diabetic
patients. Lancet, 1977, 2, 734-736.
,botthelf, Terry S. and Rose, Louis F., "Dental
correlations; Diabetes mellitus." In Internal
medicine for dentistry, pp. 1294-1295. Edited by
Louis F. Rose and Donald Kaye. St. Louis: C.V. Mosby
Co., 1983.
Gottsegen, Robert. "Dental and oral aspects of diabetes
mellitus." In Diabetes mellitus theory and practice,
pp.895-906. Edited by Max Ellenberg and Harold
Rifkin. New Hyde Park: Medical Examination Pub. Co.
Inc., 1983.
Graf, Ronald J.; Halter, Jeffrey B.; and Porte, Daniel,
Glycosylated hemoglobin in normal subjects and
subjects with maturity-onset diabetes. Diabetes,
1978, 27, 834-839.
Gray, Deanna L. Elderly diabetics and urine testing.
Geriatric Nursing, Nov/Dec 1985, pp. 332-334.
Greene, Douglas A. Acute and chronic complications of
diabetes mellitus in older patients. Amer J Med,
1986, 39-53.
Gross, Alan M.; Magalnick, Lauren; Richardson, Pat; and
Davidson, Paul C. Increasing the accuracy of blood
glucose values using visually read chemstrips.
Diabetes Educator, Spring 1985, pp. 50-52.
Jarrett, R.J.; McCartney, P.; and Keen, H., The Bedford
survey: Ten year mortality rates in newly diagnosed
diabetics, borderline diabetics and normoglycaemic
controls and risk indices for coronary heart disease
in borderline diabetics, Diabetologia, 1982, 22^, 79-
84.
Hanuschak, Lee N. and Duncan, Theodore G. Why having older
diabetics self-monitor blood sugars pays off.
Geriatrics, 1985, 40, 91-98.
- - 39
îealth and Public Policy Committee, American College of
Physicians 1984. Glycosylated hemoglobin assays in
the management and diagnosis of diabetes mellitus.
Ann Inter Med., 1984, 101, 710-713.
Kart, Cary S.; Metress, Eileen S.; and Metress, James F.
Aging and health; Biologic and social perspectives.
Menlo Park: Addison - Wesley Pub. Co., 1978.
Kozak, George P. "Why treat diabetes?" In Clinical diabetes
mellitus, pp. 43-51. Edited by G.P. Kozak.
Philadelphia: W.B. Saunders Co., 1982.
Levin, Marvin E. Diabetes: geriatric complications.
Geriatrics, 1983, 39-47.
Liang, James C.; and Goldberg, Morton F. Review: Treatment
of diabetic retinopathy. Diabetes, 1980, 2^, 841-851.
Lim, H. S.; Chew, W.L.S.; Tan, Y.T.; and Tan, F.B.Y. Fast
fractions haemoglobin (HbAl) in the assessment of
diabetes control, Ann Acad Med., 1985, lA, 203-208.
Lipson, Loren G. Diabetes in the elderly: Diagnosis,
pathogenesis, and therapy. Amer J Med, 1986, 80^, 10-
21.
Lipson, Loren G. and Lipson, Morris. The therapeutic
approach to the obese maturity-onset diabetic
patient. Arch Intern Med., 1984, 144, 135-138.
Lozada-Nur, F.; Luangjarmekorn, L.; Silverman, S.; and
Karam, J. Assessment of plasma glucose in 99 patients
with oral lichen planus, J Oral Med., 1985, 60-
61.
Lundstrom, Inger M. Incidence of diabetes mellitus in
patients with oral lichen planus, Int J Oral Surg.,
1983, 12, 147-152.
Manouchehr-Pour, M.; Spagnuolo, P.J.; Rodman, H.M.; and
Bissada, N. F. Comparison of neutrophil chemotactic
response in diabetic patients with mild and severe
periodontal disease, J Periodontol, 1981, 52_, 410-
415.
Marble, Alexander. "Diabetes in the elderly." In Joslin* s
Diabetes mellitus. 12th ed., pp. 520-525. Edited by
Alexander Marble, Leo P. Krall, Robert F. Bradley, A.
Richard Christlieb, J. Stuart Soeldner. Philadelphia:
Lea and Febiger, 1985.
90
Marble, Alexander and Ferguson, B. Dan, "Diagnosis and
classification of diabetes mellitus and the
nondiabetic meliturias." In Joslin's Diabetes
mellitus. 12th ed., pp. 520-525. Edited by Alexander
Marble, Leo P. Krall, Robert F. Bradley, A. Richard
Christlieb, J. Stuart Soeldner. Philadelphia: Lea and
Febiger, 1985.
yicMullen, J.A. ; Legg, M. ; Gottsegen, R. ; and Camerini-
Davalos, R. Microangiopathy within the gingival
tissues of diabetic subjects with special reference
to the prediabetic state, J Am Soc Periodont, 1967,
61, 1967.
McMullen, J.A.; Van Dyke, T.E.; Horoszewicz, H.U.; and
Genco, R.J. Neutrophil chemotaxis in individuals with
advanced periodontal disease and a genetic
predisposition to diabetes mellitus, J Periodontol,
1981, 167-173.
Mulligan, R. A. Medical regimen compliance in the elderly
and its implications for the dental practitioner,
Gerodontics, 1985, If 231-235.
Murrah, V.A. Diabetes mellitus and associated oral
manifestations : A review, J Oral Path, 1985, lA, 271-
281.
National Diabetes Data Group. Classification and diagnosis
of diabetes mellitus and other categories of glucose
intolerance. Diabetes, 1979, %8, 1039-1057.
O'Hanrahan, M. and O'Malley, K. Compliance with drug
treatment, British Medical J, 1981, 283, 298-300.
Owen, Oliver E.; Shuman, Charles, R.; Boden, Guenther,; and
Hoeldthe, Robert D. "Pathogenesis and diagnosis of
diabetes mellitus." In Internal medicine for
dentistrv, pp. 1259-1264. Edited by Louis F. Rose and
Donald Kaye. St. Louis: C.V. Mosby Co., 1983.
Peacock, Ian. Glycosylated haemoglobin: Measurement and
clinical use, J Clin Pathol, 1984, 3^, 841-851.
Peterson, Charles M. Quality control for glucose monitoring
programs. Diabetes Educator, 1985, Spring, 19-21.
Rabinowitz, Stanley G. "Infection in the diabetic patient."
In Diabetes mellitus vol. V, pp. 213-217. Edited by
Harold Rifkin and Philip Raskin. Bowie, MD: Robert J.
Brady Co., 1981.
91
Redmon, James E. Treatment of diabetes in the elderly.
Comprehensive Therapy, 1984, IC^, 24-28.
Rothwell, Bruce R. and Richard, Eldon L. Diabetes mellitus:
Medical and dental considerations. Special Care
Dentistrv, 1984, A, 58-65.
Saadoun, Andre. Diabetes and periodontal disease : A review
and update, J Western Soc Periodont/Periodont Abs,
1980, 28, 116-139.
Schellekens, A.P.M.; Sanders, G.T.B.; Thronton, W.; and
Groenestein, T. Sources of variation in the column-
chromatographic determination of glycohemoglobin
(HbAl), Clin Chem, 1981, 27, 94-99.
Sharon, Avigdor; Ben-Aryeh, Hannah; Itzhak, Biran; Yoram,
Kanter; Szargel, Raymonde; and Gutman, David,
Salivary composition in diabetic patients. J Oral
Med, 1985, 40, 23-26.
Shuman, Charles R. Optimum insulin use in older diabetics.
Geriatrics, 1984, 39, 71-82, 84, 89.
Tervonen, Tellervo and Knuuttila, Matti. Relation of
diabetes control to periodontal pocketing and
alveolar bone level. Oral Surg, 1986, 61^, 346-349.
Trivelli, Liliana A.; Ranney, Helen M.; and Lai, Hong-Tien.
Hemoglobin components in patients with diabetes
mellitus, N Engl J Med, 1971, 284^ 353-357.
Van Crombrugge, P.; Rottiers, R.; and Vermeulen, A. Self
blood glucose evaluation of Haemo-glukotest 20-800R,
Visidex I, and Reflolux, Diabete Métabolisme (Paris),
1985, 11, 210-215.
Willey, D.G.: Rosenthal, M.A.; and Caldwell, S.
Glycosylated haemoglobin and plasmaglycoprotein
assays by affinity chromatography, ,Diabetologia,
1984, 27, 56-58.
Wilson, Peter W.F.; Anderson, Keaven M. ; and Kannel,
William B. Epidemiology of diabetes mellitus in the
elderly. Amer J Med, 1986, 3-9.
Wing, Rena R.; Epstein, Leonard, H.; Norwalk, Mary P.;
Scott, Nancy; and Koeske, Randi, Diabetes Care, 1985,
8, 456-460.
Williams, T. Franklin. "Diabetes mellitus in older people."
-------In_C1inical_aspects_of_aging^,_2nd_ed., pp. 411-415.
92
Edited by William Reichel. Baltimore: Williams and
Wilkins, 1983.
Zimmet, P. Type 2 (non-insulin dependent) diabetes - An
epidemiological overview, Diabetologia, 1982, 22,
399-411.

Linked assets

University of Southern California Dissertations and Theses

Conceptually similar

PDF

Dental care of the aged patient: Implications of attitudes toward aging among dental students and dental hygiene students

PDF

The effectiveness of nutritional counseling in nutritional status and behavior in the elderly

PDF

Health care personnel needs for long-term care in California: Projections through the year 2020

PDF

Nursing and gerontology: A study of professionalism

PDF

Public administration and gerontology: Toward professionalization?

PDF

Professionalism of adult education and gerontology in California: Interrelationships, training, and implications

PDF

A study of professionalism as it relates to the field of counselor education and gerontology

PDF

Educational status and trends in gerontology and medicine

PDF

Survey of registered nurses' attitudes toward elderly patients

PDF

Institutionalized elderly: design, community, and management effects on life quality

PDF

Social work and gerontology: A study in professionalism

PDF

A comparative study of older and younger adults who have received job-specific classroom training to determine if there are significant differences as they relate to job placement

PDF

A comparison of training methods in cavity preparations in primary teeth for preclinical dental students

PDF

Comparison of gene expression of SCG10 and Stathmin/p19 in aging rat brain: an in situ hybridization study

PDF

The use of cubic splines for estimating the prognostic effect of age at diagnosis in childhood acute lymphoblastic leukemia

PDF

Crisis and values at Angel Crest Manor: a case study in long-term care management

PDF

Role accommodation: A study of optometrists in a prepaid medical group as perspective for viewing relations between two professiosl occupations

PDF

The reorganization of a junior high guidance program

PDF

The use of complaints against nursing homes as an index of quality care

PDF

Alcoholism among the elderly: Fact or fallacy

Asset Metadata

Creator Mulligan, Rose Ann (author)

Core Title Determining diabetes risk assessment in the elderly dental patient

School Leonard Davis School of Gerontology

Degree Master of Science

Degree Program Gerontology

Degree Conferral Date 1987-05

Publisher University of Southern California (original), University of Southern California. Libraries (digital)

Tag health and environmental sciences,OAI-PMH Harvest,social sciences

Format application/pdf (imt)

Language English

Contributor Digitized by ProQuest (provenance)

Permanent Link (DOI) https://doi.org/10.25549/usctheses-c37-410224

Unique identifier UC11657601

Identifier EP58938.pdf (filename),usctheses-c37-410224 (legacy record id)

Legacy Identifier EP58938.pdf

Dmrecord 410224

Document Type Thesis

Format application/pdf (imt)

Rights Mulligan, Rose Ann

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