University of Southern California Dissertations and Theses

Reproductive factors and risk of B-cell non-Hodgkin's lymphomas among women in Los Angeles

(USC Thesis Other)

Reproductive factors and risk of B-cell non-Hodgkin's lymphomas among women in Los Angeles

PDF

Download a page range

Download transcript

Copy asset link

Request this asset

Transcript (if available)

Content

REPRODUCTIVE FACTORS AND RISK OF B-CELL NON-HODGKIN’S
LYMPHOMAS AMONG WOMEN IN LOS ANGELES

by

Yani Lu
———————————————————————————

A Thesis Presented to the
FACULTY OF THE GRADUATE SCHOOL
UNIVERSITY OF SOUTHERN CALIFORNIA
In Partial Fulfillment of the
Requirements for the Degree
MASTER OF SCIENCE
(APPLIED BIOMETRY AND EPIDEMIOLOGY)

May 2008

Copyright 2008 Yani Lu

ii

ACKNOWLEDGEMENTS

I would like to express my greatest gratitude to my thesis committee:
Dr. Leslie Bernstein, committee chair, Dr. Wendy Cozen, and Dr. Wendy Mack
for their expert advice and guidance, and close supervision.
iii

TABLE OF CONTENTS

Acknowledgements ii
List of Tables iv
Abstract v
Introduction 1
Methods 4
Results 11
Discussion 20
References 30
iv
LIST OF TABLES

Table 1: Related risk factors and risk of B-cell NHL 12
in Los Angeles County females, 20-79 years of age

Table 2: Related risk factors and risk of NHL subtypes 14
in Los Angeles County females, 20-79 years of age

Table 3: Pregnancy history and risk of B-cell NHL 15
in Los Angeles County females, 20-79 years of age

Table 4: Breast feeding history and risk of B-cell NHL 16
in Los Angeles County females, 20-79 years of age

Table 5: Pregnancy conditions and risk of B-cell NHL 17
in Los Angeles County females, 20-79 years of age

Table 6: Reproductive factors and risk of NHL subtypes 19
in Los Angeles County females, 20-79 years of age

v

ABSTRACT

Study Aim: Reproductive patterns have been inconsistently linked to NHL risk. The aim
of this study is to identify if these factors affect B-cell NHL risk among women. Methods:
We conducted a population-based case-control study in Los Angeles County. A total of
555 histologically confirmed B-cell NHL patients and 555 individually matched control
subjects were included. Conditional logistic regression was used to evaluate these
associations. Results: Reproductive history factors such as gravidity, parity, and outcome
of first pregnancy were not associated with B-cell NHL risk overall or any NHL subtype.
A modest association was observed for use of lactation suppressants (OR=0.78, 95%
CI=0.56-1.08). The use of diethylstilbestrol (DES) during pregnancy increased the risk of
B-cell NHL (OR=2.10, 95% CI=0.71-6.23). Conclusion: Reproductive factors were not
associated with the risk of B-cell NHL overall or any subtype. Future studies with greater
numbers of NHL subtypes are needed to confirm our study results.

1
INTRODUCTION
During the second half of the 20th century, remarkable increases in the incidence
rates of non-Hodgkin’s lymphoma among both men and women were observed
worldwide (Bhurgri et al. 2005, 364-369; Groves et al. 2000, 1240-1251; Liu, Semenciw,
and Mao 2003, 57-66; Sandin et al. 2006, 1295-1300). Improvements in diagnostic
practice and the increase in AIDS-associated NHL contributed to the rise in NHL but are
not sufficient to explain entirely these dramatic increases (Eltom et al. 2002, 1204-1210;
Gail et al. 1991, 695-701; Hartge and Devesa 1992, 5566s-5569s). In the United States,
trends in incidence changed in the 1990s; since 1991, the increase in NHL incidence has
been confined to women (Cancer facts and figures 2007. 2007). However, one must be
cautious in interpreting NHL rates as a single disease entity, because previous descriptive
epidemiological studies have revealed striking differences in incidence patterns and
trends by histologic subtype (Groves et al. 2000, 1240-1251; Morton et al. 2006, 265-
276). One recent descriptive analysis of U.S. cancer statistics (Surveillance,
Epidemiology and the End Results) which used the World Health Organization (WHO)
subtype classification system indicated that males had higher rates than females for most
lymphoid neoplasm subtypes (Morton et al. 2006, 265-276). From 1992 to 2001, among
B-cell lymphomas, incidence rates rose significantly for marginal zone lymphoma,
mantle cell lymphoma, and Burkitt lymphoma/leukemia, and declined significantly for
diffuse large B-cell lymphomas (DLBCL) and small lymphocytic/chronic lymphocytic
leukemia (SLL/CLL) (Morton et al. 2006, 265-276).
2
Our inability to explain these demographic changes in incidence rates reflects our
limited knowledge of the risk factors for NHL and the etiologic heterogeneity among the
subtypes of NHL. During a pregnancy, a woman experiences changes in immune
function that permit her to establish and maintain a successful pregnancy; these
pregnancy associated changes in immune function could possibly affect the risk of NHL,
a disease in which immune function plays a role (Chaouat et al. 2004, 93-119; Luppi
2003, 3352-3357; Poole and Claman 2004, 161-170). The results from several
epidemiologic studies, however, have been inconsistent. One case-control study in
Connecticut found that as the number of births increases, NHL risk decreases (Zhang et
al. 2004, 766-773), whereas other studies reported no significant association with parity
or number of births (Adami et al. 1997, 155-158; Cerhan et al. 2002, 131-136; Frisch et al.
2006, 673-679; Nelson, Levine, and Bernstein 2001, 1381-1387) In a cohort study in
Denmark, early age at first full-term pregnancy was related to an increased risk of NHL
(Frisch et al. 2006, 673-679); in contrast, late age at first full-term pregnancy was
indicated as a risk factor of NHL in a hospital-based case-control study in Sweden
(Olsson, Olsson, and Ranstam 1990, 185-190). In another cohort study, the Iowa
Women’s Health Study, breast-feeding was associated with a reduced risk of NHL
(Cerhan et al. 2002, 131-136), but this association has not been confirmed in other studies
(Nelson, Levine, and Bernstein 2001, 1381-1387).
So far, few studies have investigated the association between reproductive factors
and the risk of NHL according to subtype. However, it is important to study NHL
subtypes as the etiology of the more than 30 classifications of NHL may differ.
3
We initiated a population-based, case-control study of B-cell NHL in Los Angeles
County to determine the relative impact of reproductive factors on the risk of NHL. We
excluded T-cell and NK-cell neoplasms since B-cell lymphomas account for more than
85% of NHL in the United States. Due to the possible differences in etiology of subtypes
of NHL, we further explored associations between the three main subtypes (DLBCL,
follicular, and SLL/CLL) as defined by the WHO modification to the REAL (Revised
European-American Lymphoma) classification of hematopoietic malignancies (Jaffe,E.S.,
Harris,N.L., Stein, H., and Vardiman, J.W. 2001).
4
METHODS
Case patients and Non-Hodgkin Lymphomas Subtype Classification
This population-based, case-control study identified women who were newly
diagnosed with a first primary B-cell NHL through rapid case-ascertainment (RCA)
procedures implemented by the Cancer Surveillance Program (CSP), the population-
based cancer registry for Los Angeles County that is part of the National Cancer
Institute’s Surveillance, Epidemiology, and End Results (SEER) Program and the
California Cancer Registry. All patients newly diagnosed between July 1, 2004, and
October 31, 2007, who were English-speaking, residents of Los Angeles County,
between the ages of 20 and 79 years at diagnosis and without previous diagnosis of NHL,
Hodgkin’s disease, multiple myeloma, or any type of leukemia were potentially eligible
for this study. We asked all potentially eligible participants a series of questions during
the initial telephone contact, including a question on HIV status, without specifying
which criterion would rule them out. We then screened out any HIV-positive women
since HIV is a known risk factor for NHL. During our screening call, we asked women
whether they had lived in Los Angeles County on a specific date (their diagnosis date)
and had no previous diagnosis of NHL, Hodgkin’s disease, multiple myeloma, or any
type of leukemia, AIDS, or HIV prior to that date. Of the 940 women who were
identified as eligible case patients, 555 (59%) patients were interviewed; 127 (14%)
women were deceased; 22 (2%) patients were too ill to be interviewed, 119 (13%)
patients refused to participate and 50 (5%) patients could not be interviewed within 18
months after diagnosis. We were unable to contact 67 (7%) patients for one of the
5
following reasons: their physicians refused to grant permission to contact the patient; the
patient could not be located; or the patient had moved out of Los Angeles County after
diagnosis and could not be interviewed in person.
In our study, we used the WHO modification of the REAL (Revised European-
American Lymphoma) classification system to classify major subtypes of B-cell NHL
(Jaffe,E.S., Harris,N.L., Stein, H., and Vardiman, J.W. 2001). The REAL classification
system was designed to classify tumors according to 5 properties: morphology
(histology), immunophenotype, genotype, normal cell counterpart and clinical features,
and the degree to which these properties contribute to a particular classification entity
vary (Harris et al. 1994, 1361-1392). In the late 1990s, WHO updated and refined the
REAL classification system. One important change is that the WHO classification
recognizes that CLL and SLL are the same disease as are precursor B lymphoblastic
leukemia and precursor B lymphoblastic lymphoma. Our main subtypes include: DLBCL
with the ICD-O morphology codes 9679, 9680, 9684, 9687, and 9826; follicular
lymphomas (follicular) with the morphology codes 9690, 9691, 9695, 9698; CLL or SLL
with morphology codes 9670 and 9823; marginal zone lymphomas (marginal zone) with
morphology codes 9689 and 9699; and a combined group of the remaining subtypes
classified as “other” (morphology codes 9590, 9591, 9671, 9673, 9833, 9835 and 9836).
All pathology reports were reviewed for eligibility by Dr. Wendy Cozen (expert
pathologist).

6
Control subjects
One female control subject was individually matched to each interviewed case
patient based on age (within 5 year age groups), race/ethnicity (Hispanic white, non-
Hispanic white, black, or Asian or Pacific Islander) and the CSP socioeconomic status
(SES) classification of the neighborhood of residence of the case at diagnosis. The
neighborhoods of all case patients interviewed at the time a control was sought were
eligible if no control had previously been identified in the case’s neighborhood and the
SES of the case for whom the control was sought had the same SES classification as the
neighborhood selected. To do this, we identified a residence with a fixed relationship to
the residence of the case living in the neighborhood and conducted walks based on a
detailed algorithm within the neighborhood that blocked out the block and surrounding
homes where the case lived at diagnosis. At each individual address, if no-one was at
home, an explanatory letter and postage-paid return envelope were left and the address
and details were recorded. The control identification procedure continued until an eligible
control subject was identified and interviewed. To be eligible, control subjects had to be
English-speaking, HIV-negative with no prior diagnosis of NHL, Hodgkin’s disease,
multiple myeloma, or any type of leukemia. We identified 653 eligible control subjects,
and we successfully interviewed 555 (85%) of these. For 83.2% of the case subjects, the
first eligible control subject agreed to be interviewed; for 14.8%, the interviewed control
was the second eligible person in the sequence; and for 2.0%, three or more eligible
control subjects were approached before obtaining an interview. We could not interview
98 eligible women (15%) because the women refused to participate (n=37), we were
7
unable to schedule an interview after several contacts (n=41), the women were sick (n=7),
or we were unable to locate the women (n=15).
Data Collection
All participants were interviewed in person with a standardized, structured
questionnaire, which combined elements of the questionnaire used for our previous NHL
study (1989-1992) (Nelson, Levine, and Bernstein 2001, 1381-1387) and the Women’s
Contraceptive and Reproductive Experiences (CARE) Study (Marchbanks et al. 2002,
213-221). This study protocol was approved by the University of Southern California
Institutional Review Board. Prior to the in-person interview, an informed consent and
HIPAA authorization were obtained from each subject. During interviews, we collected
information on each subject’s reproductive history including pregnancy and pregnancy
outcomes, medications used in pregnancy, and history of gynecologic surgery up to their
reference date. The date of diagnosis was defined as the reference date for each case
patient, and this same date was used for the control subject who was individually
matched the case. A calendar of life events, created during the interview, was used to
facilitate the recall of reproductive history and hormone use. We also collected other
lifestyle information such as smoking history and alcohol intake history. In this report,
we used all the information mentioned above up to 12 months before participants’
reference dates.
Exposure definitions
Information on reproductive factors including gravidity (ever pregnant vs. never
pregnant), parity (no full-term pregnancy vs. full-term pregnancy), number of
8
pregnancies, number of full-term pregnancies, age at first full-term pregnancy, outcome
of first pregnancy (full-term vs. short-term), history of stillbirth, miscarriage, and induced
abortion, treatment for nausea or vomiting during pregnancy, diethylstilbestrol (DES) use
during pregnancy, and use of lactation suppressants to dry up the milk or breast swelling
were collected on all participants. Full-term pregnancy was defined as any pregnancy that
lasted more than 26 weeks (6 months). Age at first full-term pregnancy was defined as
the age at which that pregnancy ended. We also collected information on breastfeeding
duration for those women who breastfed at least one live birth baby for at least one week.
Smoking history was recorded after completion of the calendar of life events.
Participants were defined as “ever smokers” if they had ever smoked a total of 100
cigarettes and had ever smoked cigarettes at least once a month for at least six months.
Participants who smoked at least 100 cigarettes but less than six months, who smoked
less than 100 cigarettes or who never smoked before the reference date were defined as
“non-smokers.”
Among ever smokers, we further documented details of their usual smoking
patterns and how their patterns had changed throughout their lives. The start and the end
of each pattern were defined as one episode. Starting with the first episode in which a
participant smoked at least once a month for six months or more, we recorded all
smoking pattern episodes up to her reference date. In each episode, we collected
information on the age at which the woman started and stopped that smoking pattern and
the number of cigarettes she usually smoked each day, week, or month. The participants
who were smokers on the reference date or who quit smoking less than twelve months
9
were classified as “current smokers.” The participants who quit smoking at least twelve
months prior to the reference date were classified as “former smokers” since more recent
cessation of smoking habits could be related to onset of the disease. The cumulative
lifetime exposure to cigarette smoking was obtained by adding up the pack-years (pack-
years=number of packs smoked per day x number of years smoked) across all smoking
episodes. Continuous exposure variables, including pack-years and total years of smoking,
were categorized according to the distribution among controls.
Alcohol variables were created in a fashion similar to the process used for
smoking. “Non-drinkers” were defined as those who did not drink more than 1 drink of
alcoholic beverage such as beer, wine, or hard liquor per month for at least 6 months. For
“ever drinkers,” we recorded lifetime alcohol consumption patterns and how these
patterns had changed up to their reference dates. Drinking status one year before the
reference date was used to define “current drinkers” or “former drinkers”. From the age
of first exposure to alcohol, we asked how many drinks (one drink was defined as 12 oz.
of beer, 4 oz. of wine, or 1.5 oz. of liquor) each woman usually drank each day, week or
month for all episodes defining changes in drinking pattern up to the reference age. We
calculated the number of drinks consumed per week for each year of age, for each
beverage, and for all beverages combined. Total numbers of drinking years (duration)
until the reference date, and average number of all types of drinks per week from age 15
years to 1 year before the reference date were calculated to measure cumulative lifetime
exposure.

10
Statistical Analysis
Conditional logistic regression was used to estimate odds ratios (ORs; relative
risk estimates) and corresponding 95% confidence intervals (95% CIs). Based on the
variable distributions among control subjects, prior research or common use, we
categorized continuous variables into categorical variables for variables of interest and
for potential confounders. Potential confounding variables included in the final
multivariate model were smoking status (current, former or never smokers), drinking
status (current, former or never drinkers) and body mass index (BMI) (<25, 25-29.9, and
≥30 kg/m
2
) at five years before the reference date. These variables were selected a priori
because of their fairly strong association with NHL risk in our data or in prior studies.
We also assessed other potential confounding variables such as lifetime smoking duration
(smoking years), smoking pack-years, and lifetime average drinks per week from age 15
to 1 year before the reference date. We did not include these variables in the final
multivariate models as they did not result in material changes for the observed
association. Likelihood ratio tests were used for tests for trend (p values) by scoring the
categories and entering these ordinal variables

in the regression analyses.
To assess whether the association between B-cell NHL and reproductive factors
varies by disease subtypes, we further analyzed our data for the three subtypes: diffuse,
follicular, and CLL/SLL. We did not attempt analysis of the marginal subtype due to the
limited number of patients in this subtype.
P values less than 0.05 were considered statistically significant, and all p values
reported were 2-sided.
11
RESULTS
The average age at diagnosis for B-cell Non-Hodgkin’s Lymphomas case patients
in this study was 59.0 years (Standard deviation [SD] =13.8 years); the average age at
reference date for control subjects was 58.6 years (SD=13.8 years). 934 (75%)
participants were non-Hispanic white, 116 (10%) were Hispanic white, 86 (8%)
participants were black, and 74 (7%) participants were Asian or Pacific Islander. Among
all subtypes of case patients, 164 patients (30%) were diagnosed with the diffuse large B-
cell subtype; 132 (53%) patients were diagnosed of the follicular subtype; 114 (21%)
patients were diagnosed with the CLL/SLL subtype; 66 (12%) patients were diagnosed of
the marginal subtype; and 79 (14%) patients had other subtypes of NHL. The average
age at diagnosis was 56.6 years (SD=15.1 years) for diffuse large B-cell lymphomas, 59.1
years (SD=12.3 years) for follicular, 63.8 years (SD=11.4 years) for CLL/SLL, and 59.1
years (SD=14.1 years) for marginal case patients.
B-cell NHL risk associated with Alcohol, tobacco use and BMI
Our potential confounders (smoking, alcohol, and body mass index) have a strong
association with the risk of B-cell NHL in this study (Table 1). After adjustment for
alcohol and body mass index, NHL risk was almost 40% higher among women who were
former smokers, and 80% higher among current smokers compared with those who had
never smoked (for former smokers: OR=1.39 with 95% CI=1.05-1.84; for current
smokers: OR=1.79 with 95% CI=1.18-2.73). Furthermore, NHL risk estimates
moderately increased with increasing duration and pack-years of smoking (p for trend <
0.01 and p = 0.01, respectively). Compared to women who had never smoked, risk
12
Table 1. Related risk factors and risk of B-cell NHL in Los Angeles County females, 20-79 years of
age
Cases
(555)
Controls
(555)
Crude OR (95%CI) Adjusted OR (95%CI)*
Smoking status
Never 306 350 1.00 1.00
Former 182 161 1.29 (0.99-1.68) 1.39 (1.05-1.84)
Current 67 44 1.72 (1.14-2.58) 1.79 (1.18-2.73)
Lifetime smoking (years)
Never smoking 306 350 1.00 1.00
<10 54 50 1.22 (0.79-1.86) 1.42 (0.91-2.22)
11-20 62 50 1.42 (0.94-2.13) 1.51 (0.99-2.32)
21-35 65 59 1.26 (0.86-1.85) 1.33 (0.90-1.98)
36+ 68 46 1.69 (1.12-2.55) 1.68 (1.10-2.56)
p for trend 0.006 0.005
Pack years‡
Never smoking 306 350 1.00 1.00
<=5 87 75 1.30 (0.92-1.82) 1.47 (1.03-2.11)
5.01-20 82 64 1.51 (1.05-2.19) 1.59 (1.08-2.33)
>20 79 65 1.39 (0.96-2.00) 1.39 (0.95-2.04)
p for trend 0.013 0.011
Drinking status
Never 260 244 1.00 1.00
Former 105 92 1.05 (0.75-1.49) 0.92 (0.64-1.32)
Current 190 219 0.79 (0.59-1.05) 0.74 (0.54-0.996)
Lifetime average drink/week from age 15 to reference date‡
Never drink 260 244 1.00 1.00
<2 154 158 0.90 (0.67-1.22) 0.83 (0.61-1.14)
2-6.9 98 114 0.81 (0.58-1.13) 0.72 (0.50-1.03)
>=7 40 38 0.97 (0.60-1.59) 0.88 (0.52-1.48)
p for trend 0.40 0.16
Body mass index (BMI) 5 years before reference date‡
<=24.9 276 299 1.00 1.00
25-29.9 154 177 0.90 (0.67-1.21) 0.90 (0.67-1.21)
>=30 123 88 1.49 (1.06-2.08) 1.44 (1.02-2.03)
p for trend 0.13 0.19
*Mutually adjusted for smoking status (never, former, and current), alcohol drinking status (never,
former, and current), and body mass index (BMI) at five years before reference date (date at
diagnosis for cases, date at diagnosis of cases for matched controls) (BMI<=24.9, 25-29.9, and
>=30).
‡Two patients did not remember BMI at five years before reference date; one patient and one
control subject did not remember the smoking amount; three patients and one control subject did
not remember the drinking amount; these subjects and their matched pairs were excluded from all
the multivariate models.
13
estimates were elevated for the longest duration of smoking (≥36 years; OR=1.68, 95%
CI=1.10-2.56) and women who had smoked 20 pack-years in their lifetime (OR=1.39,
95% CI=0.95-2.04).
Contrary to the effect of smoking which increased the risk for B-cell NHL in this
study, currently drinking alcohol was associated with reduced risk of B-cell NHL after
adjustment for smoking and BMI. The risk of B-cell NHL was 26% lower among
women who were current drinkers, compared with women who had never used alcohol
(OR=0.74, 95% CI=0.54-0.996). However, we did not find a statistically significant dose
effect when we looked at the amount of alcohol consumption (p for trend = 0.16), but
found the most protective effect among moderate amount consumers (2-6.9 drink/week,
OR=0.72, 95% CI=0.50-1.03, Table 1). Body mass index was also a risk factor for NHL.
Compared to women whose body mass index (BMI) was under 25 kg/m
2
, risk was not
increased among overweight (BMI 25-29.9 kg/m
2
) women (OR=0.90, 95% CI=0.67-
1.21), but was associated with a statistically significant increased risk among obese
women after adjustment for alcohol and smoking (OR=1.44, 95% CI=1.02-2.03).
When evaluating by NHL subtype, we found that smoking had a non-significant
positive association with all B-cell NHL subtypes, with the highest risk among current
smokers (Table 2). The risk of follicular NHL risk was 1.70 times high among former
smokers (95% CI=0.92-3.12) and 2.38 times high among current smokers (95% CI=0.97-
2.85) when compared with women who had never smoked. Alcohol consumption was
consistently protective against all B-cell NHL subtypes and a significantly lower risk was
observed among former drinkers for the follicular lymphomas (OR=0.39, 95% CI=0.17-
14
0.89) and among current drinkers for the SLL/CLL lymphomas (OR=0.48, 95% CI=0.24-
0.94). No association between body mass index and NHL risk was observed among
DLBCL and the follicular risk lymphomas, whereas a strong adverse association of
obesity was observed on the risk of SLL/CLL lymphomas. The risk of developing
SLL/CLL lymphomas was 2.55 (95% CI=1.06-6.15) times as high in obese women as
women whose BMI was under 25 kg/m
2
.
Table 2. Related risk factors and risk of NHL subtypes in Los Angeles County females, 20-79 years of
age
Diffuse large B cell (164) Follicular (132) SLL/CLL(114)
Cases
/controls
OR
(95%CI)*
Cases
/controls
OR
(95%CI)*
Cases
/controls
OR
(95%CI)*
Smoking status
Never 95/110 1.00 72/84 1.00 61/64 1.00
Former 48/40 1.60
(0.91-2.81)
43/38 1.70
(0.92-3.12)
37/41 1.11
(0.61-2.02)
Current 21/14 1.73
(0.84-3.59)
17/10 2.38
(0.97-5.85)
16/9 2.41
(0.91-6.36)

Drinking status
Never 77/73 1.00 58/50 1.00 59/48 1.00
Former 30/26 0.97
(0.49-1.92)
24/32 0.39
(0.17-0.89)
22/15 0.96
(0.39-2.37)
Current 57/65 0.73
(0.42-1.26)
50/50 0.61
(0.30-1.23)
33/51 0.48
(0.24-0.94)

BMI 5 years before reference date
<=24.9 87/86 1.00 62/69 1.00 57/60 1.00
25-29.9 43/49 0.76
(0.41-1.41)
40/41 1.21
(0.64-2.27)
28/38 0.75
(0.38-1.50)
>=30 34/29 1.03
(0.56-1.88)
30/22 1.75
(0.82-3.74)
29/16 2.55
(1.06-6.15)
* Mutually adjusted for smoking status (never, former, and current), alcohol drinking status (never, former,
and current), and body mass index (BMI) at five years before the reference date (date at diagnosis for cases,
date at diagnosis of cases for matched controls) (BMI<=24.9, 25-29.9, and >=30).

Reproductive factors and overall risk of B-cell NHL
We found no evidence that specific aspects of pregnancy history were associated
with overall B-cell NHL risk (Table 3) in either univariate models or multivariate models.
15
Table 3. Pregnancy history and risk of B-cell NHL in Los Angeles County females, 20-79 years of age
Cases (555) Controls (555) Crude OR (95%CI) Adjusted OR
(95%CI)*
Ever Pregnant
No 86 87 1.00 1.00
Yes 469 468 1.02 (0.72-1.43) 1.01 (0.71-1.43)
Number of pregnancies
0 86 87 1.00 1.00
1 68 63 1.10 (0.68-1.75) 1.07 (0.66-1.73)
2 119 122 0.99 (0.66-1.48) 1.01 (0.67-1.53)
3 111 110 1.02 (0.68-1.53) 1.05 (0.70-1.60)
4+ 171 173 1.00 (0.68-1.47) 0.95 (0.64-1.41)
Age at first pregnancy†
Never pregnant 86 87 0.93 (0.63-1.36) 0.95 (0.64-1.40)
<20 117 109 1.02 (0.73-1.42) 0.98 (0.69-1.39)
20-24 191 180 1.00 1.00
25-29 106 121 0.83 (0.60-1.15) 0.86 (0.61-1.20)
30+ 54 58 0.91 (0.58-1.41) 1.02 (0.65-1.60)
Outcome of first pregnancy
Never pregnant 86 87 1.00 1.00
Short-term 125 125 1.01 (0.68-1.51) 1.01 (0.67-1.52)
Full-term 344 343 1.02 (0.71-1.45) 1.01 (0.70-1.46)
Ever full-term pregnancy(FTP)
Never Pregnant 86 87 1.00 1.00
Yes 423 426 1.00 (0.71-1.42) 1.00 (0.71-1.43)
Only short-term
pregnancies
46 42 1.12 (0.65-1.91) 1.07 (0.62-1.85)
Number of full-term pregnancies(FTP)
Never Pregnant 86 87 1.00 1.00
1 84 69 1.23 (0.79-1.92) 1.19 (0.75-1.87)
2 150 154 0.98 (0.67-1.45) 1.03 (0.69-1.526)
3 99 115 0.86 (0.56-1.30) 0.87 (0.57-1.34)
4+ 90 88 1.01 (0.65-1.60) 0.93 (0.58-1.48)
Only short term
pregnancies
46 42 1.13 (0.66-1.93) 1.09 (0.63-1.88)
Age at first full-term pregnancy(FTP)†
Never Pregnant 86 87 0.99 (0.67-1.47) 0.99 (0.66-1.48)
<20 90 80 1.14 (0.78-1.66) 1.06 (0.72-1.56)
20-24 162 163 1.00 1.00
25-29 112 114 0.99 (0.71-1.38) 1.01 (0.72-1.42)
30+ 58 69 0.85 (0.55-1.31) 0.89 (0.57-1.39)
Only short-term
pregnancies
46 42 1.10 (0.67-1.82) 1.05 (0.63-1.76)
* Adjusted for smoking status (never, former, and current), alcohol drinking status (never, former, and
current), and body mass index (BMI) at five years before reference date (date at diagnosis for cases, date at
diagnosis of cases for matched controls) (BMI<=24.9, 25-29.9, and >=30).
* Two patients did not remember the BMI at five years before the reference date, these subjects and their
matched controls were excluded from all the multivariate models.
† One patient did not like to provide the information on age at first pregnancy, this subject and her matched
control were excluded from the analyses regarding age at first pregnancy and age at first full-term
pregnancy.

16
Using women who were never pregnant as the reference group, no effects were observed
for gravity, parity, outcome of first pregnancy, age at first pregnancy, age at first full-
term pregnancy, number of pregnancies, or number of full-term pregnancies. Women
who only had short-term pregnancies had a higher risk of B-cell NHL than women who
had never been pregnant although this effect was not statistically significant (Table 3).
A history of breast-feeding was not associated with B-cell NHL risk. As observed
in Table 4, the odds ratio was 0.83 (95% CI= 0.61-1.12) for women who had ever breast-
fed compared to women who had at least one full-term pregnancy but never breast-fed.
Further exploration of the duration of breast-feeding showed no duration-response trend
(Table 4).

Table 4. Breast feeding history and risk of B-cell NHL in Los Angeles County females, 20-79 years of age
Cases (555) Controls (555) Crude OR (95%CI) Adjusted OR (95%CI)*
Ever breast fed
No 148 127 1.00 1.00
Yes 276 299 0.78 (0.58-1.05) 0.83 (0.61-1.12)
Never and short term
pregnancies
131 129 0.86 (0.60-1.24) 0.89 (0.61-1.29)

Breast feeding duration (weeks)
0 148 127 1.00 1.00
1-16 79 93 0.71 (0.48-1.05) 0.74 (0.50-1.10)
17-60 102 101 0.86 (0.60-1.24) 0.90 (0.62-1.31)
61+ 94 105 0.75 (0.51-1.10) 0.82 (0.55-1.21)
Never and short term
pregnancies
131 129 0.87 (0.61-1.25) 0.90 (0.62-1.30)
* Adjusted for smoking status (never, former, and current), alcohol drinking status (never, former, and
current), and body mass index (BMI) at five years before the reference date (date at diagnosis for cases,
date at diagnosis of cases for matched controls) (BMI<=24.9, 25-29.9, and >=30).
* Two patients did not remember BMI at five years before reference date, these subjects and their matched
controls were excluded from all the multivariate models

Table 5 presents the odds ratios for the association between B-cell NHL and
treatments or conditions during pregnancy using only information up to the date that was
12 months prior to a woman’s reference date. Treatment for nausea by drugs or
17
hospitalization was not associated with an increased risk of B-cell NHL. Use of DES
appeared to increase NHL risk, but the estimate was imprecise as a result of the small
number of women who had ever used this drug (OR=2.10, 95% CI=0.71-6.23). Lactation
suppressants were found to have a non-statistically significant protective effect against B-
cell NHL in this study. Compared to women who once had at least one full-term
pregnancy but never used any lactation suppressants, women who had used the drug to
dry up milk production or for breast swelling had 22% lower risk (OR=0.78, 95%
CI=0.56-1.08).
Table 5. Pregnancy conditions and risk of B-cell NHL in Los Angeles County females, 20-79 years of age
Cases (555) Controls (555) Crude OR (95%CI) Adjusted OR (95%CI)*
Ever treated for nausea by drugs or hospitalization
No 392 409 1.00 1.00
Yes 64 53 1.24 (0.85-1.80) 1.16 (0.79-1.71)
Do not know 13 6 2.22 (0.84-5.84) 2.10 (0.79-5.60)
Never Pregnant 86 87 1.02 (0.72-1.44) 1.01 (0.71-1.44)

Number of pregnancies ever been treated for nausea by drugs or hospitalization
0 392 409 1.00 1.00
1 45 40 1.16 (0.75-1.79) 1.06 (0.68-1.65)
2+ 19 13 1.48 (0.72-3.01) 1.50 (0.72-3.12)
Do not know 13 6 2.20 (0.84-5.80) 2.08 (0.78-5.53)
Never Pregnant 86 87 1.02 (0.72-1.44) 1.02 (0.71-1.45)

Ever take DES (diethylstilbestrol) during pregnancy
No 433 446 1.00 1.00
Yes 10 5 2.07 (0.70-6.07) 2.10 (0.71-6.23)
Do not know 26 17 1.59 (0.85-3.00) 1.77 (0.91-3.44)
Never Pregnant 86 87 1.01 (0.72-1.43) 1.02 (0.72-1.45)

Ever used lactation suppressants
No 274 266 1.00 1.00
Yes 117 134 0.83 (0.61-1.14) 0.78 (0.56-1.08)
Do not know 32 26 1.19 (0.69-2.04) 1.09 (0.63-1.89)
Only short term
pregnancies
132 129 1.00 (0.73-1.36) 0.96 (0.70-1.32)
* Adjusted for smoking status (never, former, and current), alcohol drinking status (never, former, and
current), and body mass index (BMI) at five years before the reference date (date at diagnosis for cases,
date at diagnosis of cases for matched controls) (BMI<=24.9, 25-29.9, and >=30).
* Two patients did not remember BMI at five years before reference dates, these subjects and their matched
controls were excluded from all the multivariate models.

18
Reproductive factors and risk of subtypes B-cell NHL
Table 6 shows the relative risk of the diffuse large B-cell, the follicular and the
SLL/CLL lymphomas and reproductive factors. Despite the considerable size of the study,
interpretation of this analysis among subtypes was severely hampered by lack of
statistical power. However Table 6 shows that pregnancy history, including gravity,
parity, number of full-term pregnancies, and outcome of first pregnancy had an inverse
association with the diffuse large B-cell lymphomas risk but a positive association with
the follicular lymphomas risk and the SLL/CLL lymphomas risk although none of the
effects were statistically significant and the confidence intervals are quite wide. Similarly
to those pregnancy variables, breast-feeding was associated with a non-statistically
significant protective effect on the risk of diffuse large B-cell lymphomas (OR=0.67,
95% CI=0.36-1.24), but a non-statistically significant positive association with the
follicular lymphomas risk (OR=1.53, 95% CI=0.80-2.94), and no effect on the risk of
SLL/CLL lymphomas subtype (OR=1.0, 95% CI=0.50-1.99). Finally, we observed
inverse associations between use of lactation suppressants and risk of these three NHL
subtypes but with less precise estimation.

19
Table 6. Reproductive factors and risk of NHL subtypes in Los Angeles County females, 20-79 years of age
Diffuse large B cell (164) Follicular (132) SLL/CLL(114)
Cases
/controls
OR
(95%CI)*
Cases
/controls
OR
(95%CI)*
Cases
/controls
OR
(95%CI)*
Ever Pregnant
No 35/32 1.00 20/23 1.00 13/17 1.00
Yes 129/132 0.87
(0.49-1.56)
112/109 1.27
(0.63-2.55)
101/97 1.21
(0.49-3.01)
Ever full-term pregnancy (FTP)
Never 35/32 1.00 20/23 1.00 13/17 1.00
Yes 115/115 0.91
(0.51-1.64)
102/98 1.29
(0.64-2.61)
91/90 1.20
(0.49-2.98)
On ly sh ort term
pregnancies
14/17 0.64
(0.25-1.67)
10/11 1.08
(0.34-3.46)
10/7 1.61
(0.40-6.51)
Number of full-term pregnancies (FTP)
Never 35/32 1.00 20/23 1.00 13/17 1.00
1 25/20 1.03
(0.48-2.19)
18/17 1.40
(0.53-3.68)
16/9 1.88
(0.50-7.01)
2+ 90/95 0.87
(0.47-1.62)
84/81 1.27
(0.62-2.60)
75/81 1.18
(0.48-2.92)
On ly sh ort term
pregnancies
14/17 0.65
(0.25-1.68)
10/11 1.08
(0.34-3.46)
10/7 1.75
(0.43-7.21)
Outcome of first pregnancy
Never 35/32 1.00 20/23 1.00 13/17 1.00
Short-term 36/38 0.78
(0.39-1.59)
32/29 1.40
(0.59-3.33)
27/28 1.27
(0.45-3.54)
Full-term 93/94 0.92
(0.50-1.69)
80/80 1.22
(0.59-2.52)
74/69 1.19
(0.46-3.04)
Ever breast fed
No 37/28 1.00 31/37 1.00 34/30 1.00
Yes 78/87 0.67
(0.36-1.24)
72/61 1.53
(0.80-2.94)
57/60 1.00
(0.50-1.99)
Never and
short term
pregnancies
49/49 0.71
(0.35-1.44)
29/34 1.04
(0.48-2.23)
23/24 1.01
(0.43-2.34)
Ever used lactation suppressants†
No 76/73 1.00 73/63 1.00 53/53 1.00
Yes 26/33 0.76
(0.38-1.51)
23/30 0.63
(0.32-1.23)
29/34 0.76
(0.37-1.60)
Never and
short term
pregnancies
49/49 0.90
(0.51-1.59)
30/34 0.65
(0.33-1.26)
23/24 1.15
(0.50-2.63)
* Adjusted for smoking status (never, former, and current), alcohol drinking status (never, former, and
current), and body mass index (BMI) at five years before the reference date (date at diagnosis for cases,
date at diagnosis of cases for matched controls) (BMI<=24.9, 25-29.9, and >=30).
* Two patients did not remember BMI at five years before reference date, these subjects and their matched
controls were excluded from all the multivariate models.
† Thirteen patients and nine control subjects in diffuse large B cell subtype; six patients and five control
subjects in follicular subtype; and nice patients and three control subjects did not know if they ever used
lactation suppressants, these subjects and their matched pairs were excluded from the analyses.

20
DISCUSSION
In this population-based, individually matched case-control study, we found no
evidence that reproductive factors play a strong role in the development of B-cell NHL
when NHL is treated as a homogeneous disease. Furthermore, our data suggested that
pregnancy or birth have no association with diffuse large B-cell lymphomas, follicular
lymphomas or SLL/CLL lymphomas. Breast feeding is not associated with overall NHL
risk or with different NHL subtypes risks. Women who had ever taken DES while
pregnant were at a higher risk of NHL, but the 95% confidence interval included 1.0,
possibly due to the small number of women who had used DES. Treatment for nausea
was not associated with increased risk, but the use of lactation suppressants was
somewhat associated with a decreased risk, particularly after multivariate adjustment.
The results from previous epidemiologic studies on the relationship between
reproductive factors and NHL have been inconsistent. A cohort study of 1573 women
patients in Denmark and a case-control study conducted in Los Angeles reported no
significant associations of NHL risk with parity or number of live births (Frisch et al.
2006, 673-679; Nelson, Levine, and Bernstein 2001, 1381-1387). In contrast,
investigators in a recent case-control study from Connecticut reported a significant
negative association with parity (p for trend=0.008) (Zhang et al. 2004, 766-773).
Women who had four or more live births during their lifetime were at a 40% lower risk
compared to nulliparous women. One case-control study nested within a Swedish
nationwide fertility registry also found a weak, negative relationship between the number
of births and the risk of NHL (p for trend=0.11) (Adami et al. 1997, 155-158). Whereas,
21
in this nested case-control study, the odds ratios for the number of births did not decrease
in a monotonic fashion, with the odds ratio lowest for women who had had three births
(OR=0.79, 95% CI=0.64-0.96) and somewhat higher for women who had four (OR=0.92,
95% CI=0.70-1.20) or five births (OR=0.93, 95% CI=0.64-1.34) when compared with
women who had had only one birth. Furthermore, no effect of parity was found. Results
from the Iowa Women’s Health Study, showed a potential increase in risk of NHL in
women who had at least one pregnancy (Cerhan et al. 2002, 131-136). Although similar
results were found in an Italian hospital-based case-control study among women under 50
years old (Tavani et al. 1997, 885-888), this study only included 50 patients and the
confidence limits were quite wide (≥3 numbers of birth versus nulliparae: OR=1.4, 95%
CI=0.3-6.4).
The cohort study in Danish women observed that women who had their first child
at age 15-19 years had the highest risk of NHL compared to women who had their first
child at age 20-24 years (Frisch et al. 2006, 673-679). No other studies found age at first
birth or full-term pregnancy was associated with NHL risk (Adami et al. 1997, 155-158;
Cerhan et al. 2002, 131-136; Nelson, Levine, and Bernstein 2001, 1381-1387; Tavani et
al. 1997, 885-888; Zhang et al. 2004, 766-773).
Two studies have investigated the effect of breast-feeding on the risk of NHL
(Cerhan et al. 2002, 131-136; Nelson, Levine, and Bernstein 2001, 1381-1387). The Iowa
Women’s Health Study found a protective effect of breast-feeding; however, this finding
was probably due to chance as stated by the authors and was not confirmed in other
studies. The earlier case-control study in Los Angeles also did not find a protective
22
effect of breast-feeding. This study restricted case eligibility to intermediate- and high-
grade B-cell NHL subtypes which were classified by the Working Formulation.
Nelson et al. reported that lactation suppressants were protective factors for
intermediate- or high-grade B-cell NHL (Nelson, Levine, and Bernstein 2001, 1381-
1387). Among parous women, those who had used lactation suppressants were 50% less
likely to develop NHL than unexposed women (OR=0.50, 95% CI=0.29-0.85). In our
study, we found a similar but weaker association between lactation suppressant use and
B-cell NHL risk (a decrease of 22% in risk). The drugs used for suppressing lactation
have changed during the past several decades (Kochenour 1980, 1045-1059; Rayburn
1996, 69-71). Before 1980, long-acting estrogens (quinestrol, diethylstilbestrol, and
chlorotrianisene) or those in combination with an androgen were mainly used to inhibit
the production of prolactin. As the result of their side effects, which increased the
incidence of postpartum thromboembolic disease, they were widely replaced by
Bromocriptine (Parlodel). Bromocriptine takes effect both immediately post-partum and
after lactation by inhibiting prolactin secretion with synthetic ergot alkaloids. The Food
and Drug Administration (FDA) removed the indication that bromocriptine was safe and
effective in 1995 as a consequence of concerns of the increased risks of heart attack,
seizure and stroke (Iffy et al. 1996, 309-312; Rayburn 1996, 69-71). In the Nelson et al.
study, the lactation suppressants used were all likely to be estrogenic compounds since all
but one of their exposed participants had initiated use prior to 1980. We did not collect
information on the type of lactation suppressant used in our study and can not separate
the effect of estrogenic suppressants from bromocriptine.
23
Nelson et al. also found a significant protective effect of oral contraceptives (OC)
and a slightly protective effect of hormone therapy (Nelson, Levine, and Bernstein 2001,
1381-1387). Several other studies also investigated the role of exogenous estrogens, and
the conclusions were inconsistent (Adami et al. 1997, 155-158; Cerhan et al. 2002, 1466-
1471; Zhang et al. 2004, 766-773). The nested case-control study by Adami et al. found a
weak, protective effect of OC use (OR: 0.69, 95% CI=0.43-1.11). The researchers also
found a significant increased risk for women who had used hormone therapy (HT) for
more than 12 months when compared with women who had never used HT (OR=1.58,
95% CI=1.09-2.29). The results in the Iowa Women’s Health Study showed a weak,
increased risk among current HT users, and this increased risk was only significant
among the follicular subtype, a subtype not included in the study by Nelson and
colleagues. No effects of oral contraceptive or HT use were found in the case-control
study by Zhang et al. Whether or not exogenous estrogens have an association with NHL
risk needs to be investigated further, particularly by subtype of NHL.
No prior studies have reported results on the use of DES and the risk of NHL.
DES is a potent synthetic estrogen which was first synthesized in 1938 and was
prescribed by physicians to prevent miscarriage from the 1940s through the 1960s
(Dodds EC, Goldberg L, Lawson W, Robinson R. 1938, 247-8). The strong association
between DES use and risk of clear cell adenocarcinoma (CCA) of the vagina and cervix
in young women is well known (Herbst, Ulfelder, and Poskanzer 1971, 878-881). New
evidence also shows that breast cancer risk is increased in mothers given DES during
pregnancy and in women exposed to DES in utero (Greenberg et al. 1984, 1393-1398;
24
Hatch et al. 1998, 630-634; Palmer et al. 2002, 753-758; Palmer et al. 2006, 1509-1514;
Troisi et al. 2007, 356-360). Our study indicated that DES has a moderate effect on the
risk of NHL (OR=2.10, 95% CI=0.71-5.53). Unfortunately, this analysis was based on
few exposed subjects and is therefore less robust.
Because of their diversity, the Non-Hodgkin’s Lymphomas are difficult to study
epidemiologically. Few studies have provided information on the distribution of NHL
subtypes (using any classification) of participating patients. The lack of consistency of
the associations between reproductive factors and NHL risk from study to study may be
due, in part, to the differences in the distributions of subtypes across these various studies.
We evaluated the associations according to three main subtypes and found odds ratios for
reproductive factors were mostly below 1.00 for diffuse large B-cell, but were all above
1.00 for follicular lymphomas and SLL/CLL lymphomas. Although none of the
confidence intervals for the odds ratios excluded 1.00, our data showed the possibility of
different etiologies for NHL subtypes. No extant studies have evaluated the reproductive
factors with specific subtypes of NHL by using the WHO/REAL classification system,
thus we are not able to directly compare our results with other studies. However, the Iowa
Women’s Health Study, which used the second edition of the ICD-O classification
system, showed that the effect of HT varied according to NHL subtype (Cerhan et al.
2002, 1466-1471). While HT use had no association with diffuse or small lymphocytic
NHL, they found a strong positive association with follicular NHL. The risk of follicular
NHL was 3.3 (95% CI=1.6-6.9) times as high for current users and 2.6 (95% CI=1.8-8.6)
25
times as high for former users when compared with women who had never used HT. This
result suggests the Non-Hodgkin’s Lymphomas have different etiologies.
The concept that pregnancy is an immunosuppressive state permitting the fetal
allograft to implant and grow has held center stage for years. Thus it raises the question
of whether pregnancy has a long-term effect on the risk of NHL. The immunological
changes during pregnancy are often explained by the Th1/ Th2 paradigm which was first
proposed by Wegmann and colleagues (Wegmann et al. 1993, 353-356). This paradigm
states that the high level of estrogen during pregnancy enhances T-helper 2-type cytokine
production but lessens the T-helper 1-type cytokine production. This Th1 response,
which leads to cell-mediated immunity, is characterized by the production IL-2,
interferon-γ, and lymphotoxin; on the other hand, the Th2 response, which leads to
humoral immunity, is characterized by the release of IL-4, IL-6, IL-10 and IL-13 (Dealtry,
O'Farrell, and Fernandez 2000, 107-119). Clinically, the severity of diseases associated
with the Th1 immune response (i.e. sclerosis or rheumatoid arthritis) ameliorate during
pregnancy but return to pre-pregnancy levels after birth. In contrast, pregnancy may
worsen a disease like systemic lupus erythematosus by further enhancing its ongoing Th2
immune response (Whitacre, Reingold, and O'Looney 1999, 1277-1278). A series of
studies indicate that the Th1/ Th2 paradigm is over-simplistic and may not be appropriate
as a general explanation for the immunologic changes occurring in pregnancy (Chaouat et
al. 2004, 93-119). In fact, modifications occur in virtually every facet of the immune
response and achieve a general balance between enhancement and suppression (Chaouat
et al. 2004, 93-119; Luppi 2003, 3352-3357). More importantly, little evidence has
26
showed particular elements of immune response change postpartum. Thus, pregnancy by
itself is unlikely to be a strong risk factor for NHL, as indicated in our study.
In this study, we assessed the associations of tobacco, alcohol consumption and
BMI with the risk of B-cell NHL in our determination of potential confounding factors..
Smoking had a positive association with B-cell NHL risk; risk increased with increasing
lifetime years of smoking and with total pack-years of smoking. Risks varied across NHL
subtypes, although all odds ratios were above 1.0 for current smokers compared to non-
smokers. Smoking appears to be most positively associated with follicular NHL risk.
This finding is consistent with a recent pooled analysis of 6,594 cases and 8,892 controls
by Morton et al., in which the only NHL subtype that was significantly associated with
current smoking was follicular lymphoma (OR=1.31, 95% CI=1.12-1.52)(Morton et al.
2005, 925-933).
The inverse association between alcohol consumption and B-cell NHL risk found
in our study is consistent with the majority of studies that have evaluated alcohol
consumption in relation to risk of NHL. In a pooled analysis of 9 case-control studies,
Morton et al. reported that risk estimates were lower for current drinkers (OR=0.73, 95%
CI=0.64-0.84) than for former drinkers (OR=0.95, 95% CI=0.80-1.14) when compared
with non-drinkers, but risk did not decrease with increasing amount of alcohol consumed
(Morton et al. 2005, 469-476); this was also seen in our analysis. The protective effect of
alcohol varied across NHL subtypes. Although some odds ratios were statistically
different from 1.0, our results could be due to chance because of the small numbers in
each subtype. A cohort study of women also confirmed the results from case-control
27
studies (Chiu et al. 1999, 1476-1482). In this cohort study, authors found a statistically
significant inverse association between alcohol consumption and NHL risk (p for trend =
0.03), and this association was observed with different NHL grades and types. The
potential protective effect of light to moderate alcohol consumption may play its role by
improving cellular and humoral immune response (Diaz et al. 2002, S50-3).
Evidence in most recent studies suggests that NHL risk increases with increasing
BMI, and the risk differs between different NHL subtypes (Cerhan et al. 2005, 1203-1214;
Fernberg et al. 2006, 2298-2302; Pan et al. 2005, 1162-1173; Willett et al. 2005, 811-816;
Wolk et al. 2001, 13-21). In our study, NHL risk only increased among obese women (≥
30 kg/m
2
) but not among overweight women (25-29.9 kg/m
2
), and this association was
most pronounced for SLL/CLL NHL. However, results from Cerhan et al. showed that
BMI was only positively associated with diffuse NHL (OR=1.73 for 35+ versus <25
kg/m
2
, 95% CI=1.15-2.59), but was not associated with other subtypes or all NHL
combined (Cerhan et al. 2005, 1203-1214). In contrast, the study by Willett et al.
indicated that obesity (over 30 kg/m
2
) at five years before diagnosis was associated with
an increased risk of NHL (OR=1.5, 95% CI=1.1-2.1), especially for diffuse large B-cell
lymphoma (OR=1.9, 95% CI=1.3-2.8)(Willett et al. 2005, 811-816). One reason for these
inconsistencies could be the different study populations, since we only included women
in our study. More studies are needed to further evaluate the effect of BMI on the risk of
different NHL subtypes.
One of the strengths of our study is the utilization of the new WHO classification
system to define the main subtypes of B-cell Non-Hodgkin’s Lymphomas. We found
28
only two studies that have investigated the role of reproductive factors on the risks of
different histologic subtypes (Nelson, Levine, and Bernstein 2001, 1381-1387; Zhang et
al. 2004, 766-773). Both studies however, used the Working Formulation classification
system, which categorizes NHL as high grade, intermediate grade, and low grade and was
designed to classify NHL according to prognosis and not etiologically relevant subgroups
(Anonymous1982, 2112-2135).
Our study has additional strengths. We used a rapid reporting case ascertainment
process which provides case identification within, at most, one month of diagnosis. The
response rate of case patients (59%) is fairly high compared with other studies in NHL.
Additionally, the strict individual matching of the case patients and control subjects
increased the statistical power. Furthermore, the population-based subject ascertainment
decreased the possibility of selection bias. Last but not least, all the pathology reports
with unclear diagnosis were reviewed by our pathology expert, which reduced the
misclassification of NHL histology due to differences in different pathologic practices.
However, some potential limitations should be considered in our study. First,
although our study sample size is fairly large when we treated NHL as a homogeneous
disease, we nevertheless had limited statistical power to detect modest associations
between reproductive factors and the risk of each subtype of lymphoma. Secondly, as
with all case-control studies, we could not rule out of the possibility of recall bias.
However, we used a calendar of life events to assist subjects in recalling their
reproductive histories. Finally, we probably did not totally exclude participants who were
HIV-positive since we relied only on a screening question instead of using a blood test to
29
confirm HIV status. On the other hand, if we were to include HIV testing, it would more
likely to decrease participation rates.
In summary, our study suggests that characteristics of reproductive history are not
associated with B-cell NHL risk or risk of diffuse, follicular, or CLL/SLL NHL. To better
understand the etiology of each specific subtype of Non-Hodgkin’s Lymphoma, future
studies with greater numbers of each NHL subtype are clearly needed.

30
REFERENCES
Adami, H. O., S. Tsaih, M. Lambe, C. Hsieh, J. Adami, D. Trichopoulos, M. Melbye, and
B. Glimelius. 1997. Pregnancy and risk of non-Hodgkin's lymphoma: A prospective
study. International Journal of Cancer 70, (2) (Jan 17): 155-8.
Bhurgri, Y., S. Pervez, A. Bhurgri, N. Faridi, A. Usman, L. A. Kazi, R. Ahmed, N.
Kayani, and S. H. Hasan. 2005. Increasing incidence of non-Hodgkin's lymphoma in
Karachi, 1995-2002. Asian Pacific Journal of Cancer Prevention 6, (3) (Jul-Sep):
364-9.
Cancer facts and figures 2007. 2007. Cancer facts and figures 2007. Atlanta: American
Cancer Society.
Cerhan, J. R., L. Bernstein, R. K. Severson, S. Davis, J. S. Colt, A. Blair, and P. Hartge.
2005. Anthropometrics, physical activity, related medical conditions, and the risk of
non-Hodgkin lymphoma. Cancer Causes & Control 16, (10) (Dec): 1203-14.
Cerhan, J. R., T. M. Habermann, C. M. Vachon, S. D. Putnam, W. Zheng, J. D. Potter,
and A. R. Folsom. 2002. Menstrual and reproductive factors and risk of non-
Hodgkin lymphoma: The Iowa Women's Health Study (United States). Cancer
Causes & Control 13, (2) (Mar): 131-6.
Cerhan, J. R., C. M. Vachon, T. M. Habermann, S. M. Ansell, T. E. Witzig, P. J. Kurtin,
C. A. Janney, et al. 2002. Hormone replacement therapy and risk of non-Hodgkin
lymphoma and chronic lymphocytic leukemia. Cancer Epidemiology, Biomarkers &
Prevention 11, (11) (Nov): 1466-71.
Chaouat, G., N. Ledee-Bataille, S. Dubanchet, S. Zourbas, O. Sandra, and J. Martal. 2004.
TH1/TH2 paradigm in pregnancy: Paradigm lost? Cytokines in pregnancy/early
abortion: Reexamining the TH1/TH2 paradigm. International Archives of Allergy
and Immunology 134, (2) (Jun): 93-119.
Chiu, B. C., J. R. Cerhan, S. M. Gapstur, T. A. Sellers, W. Zheng, C. T. Lutz, R. B.
Wallace, and J. D. Potter. 1999. Alcohol consumption and non-Hodgkin lymphoma
in a cohort of older women. British Journal of Cancer 80, (9) (Jul): 1476-82.
Dealtry, G. B., M. K. O'Farrell, and N. Fernandez. 2000. The Th2 cytokine environment
of the placenta. International Archives of Allergy and Immunology 123, (2) (Oct):
107-19.
Diaz, L. E., A. Montero, M. Gonzalez-Gross, A. I. Vallejo, J. Romeo, and A. Marcos.
2002. Influence of alcohol consumption on immunological status: A review.
European Journal of Clinical Nutrition 56 Suppl 3, (Aug): S50-3.
31
Dodds EC, Goldberg L, Lawson W, Robinson R. 1938. Estrogenic activity of certain
synthetic compounds. Nature 141: 247-8.
Eltom, M. A., A. Jemal, S. M. Mbulaiteye, S. S. Devesa, and R. J. Biggar. 2002. Trends
in Kaposi’s sarcoma and non-Hodgkin’s lymphoma incidence in the United States
from 1973 through 1998. Journal of the National Cancer Institute 94, (16) (Aug 21):
1204-10.
Fernberg, P., A. Odenbro, R. Bellocco, P. Boffetta, Y. Pawitan, and J. Adami. 2006.
Tobacco use, body mass index and the risk of malignant lymphomas--a nationwide
cohort study in Sweden. International Journal of Cancer. Journal International Du
Cancer 118, (9) (May 1): 2298-302.
Frisch, M., B. V. Pedersen, J. Wohlfahrt, H. Hjalgrim, and R. J. Biggar. 2006.
Reproductive patterns and non-Hodgkin lymphoma risk in Danish women and men.
European Journal of Epidemiology 21, (9): 673-9.
Gail, M. H., J. M. Pluda, C. S. Rabkin, R. J. Biggar, J. J. Goedert, J. W. Horm, E. J.
Sondik, R. Yarchoan, and S. Broder. 1991. Projections of the incidence of non-
Hodgkin's lymphoma related to acquired immunodeficiency syndrome. Journal of
the National Cancer Institute 83, (10) (May 15): 695-701.
Greenberg, E. R., A. B. Barnes, L. Resseguie, J. A. Barrett, S. Burnside, L. L. Lanza, R.
K. Neff, M. Stevens, R. H. Young, and T. Colton. 1984. Breast cancer in mothers
given diethylstilbestrol in pregnancy. The New England Journal of Medicine 311,
(22) (Nov 29): 1393-8.
Groves, F. D., M. S. Linet, L. B. Travis, and S. S. Devesa. 2000. Cancer surveillance
series: Non-Hodgkin's lymphoma incidence by histologic subtype in the United
States from 1978 through 1995. Journal of the National Cancer Institute 92, (15)
(Aug 2): 1240-51.
Harris, N. L., E. S. Jaffe, H. Stein, P. M. Banks, J. K. Chan, M. L. Cleary, G. Delsol, C.
De Wolf-Peeters, B. Falini, and K. C. Gatter. 1994. A revised European-American
classification of lymphoid neoplasms: A proposal from the international lymphoma
study group. Blood 84, (5) (Sep 1): 1361-92.
Hartge, P., and S. S. Devesa. 1992. Quantification of the impact of known risk factors on
time trends in non-Hodgkin's lymphoma incidence. Cancer Research 52, (19 Suppl)
(Oct 1): 5566s-9s.
Hatch, E. E., J. R. Palmer, L. Titus-Ernstoff, K. L. Noller, R. H. Kaufman, R. Mittendorf,
S. J. Robboy, et al. 1998. Cancer risk in women exposed to diethylstilbestrol in utero.
JAMA: The Journal of the American Medical Association 280, (7) (Aug 19): 630-4.
32
Herbst, A. L., H. Ulfelder, and D. C. Poskanzer. 1971. Adenocarcinoma of the vagina.
Association of maternal stilbestrol therapy with tumor appearance in young women.
The New England Journal of Medicine 284, (15) (Apr 15): 878-81.
Iffy, L., J. Lindenthal, J. J. Mcardle, and V. Ganesh. 1996. Severe cerebral accidents
postpartum in patients taking bromocriptine for milk suppression. Israel Journal of
Medical Sciences 32, (5) (May): 309-12.
Jaffe,E.S., Harris,N.L., Stein, H., and Vardiman, J.W. 2001. WHO Classification of
Tumours. Pathology and Genetics of Tumours of Haematopoietic and Lymphoid
Tissues. Lyon, France: IARC Press.
Kochenour, N. K. 1980. Lactation suppression. Clinical Obstetrics and Gynecology 23,
(4) (Dec): 1045-59.
Liu, S., R. Semenciw, and Y. Mao. 2003. Increasing incidence of non-Hodgkin's
lymphoma in Canada, 1970-1996: Age-period-cohort analysis. Hematological
Oncology 21, (2) (Jun): 57-66.
Luppi, P. 2003. How immune mechanisms are affected by pregnancy. Vaccine 21, (24)
(Jul 28): 3352-7.
Marchbanks, P. A., J. A. Mcdonald, H. G. Wilson, N. M. Burnett, J. R. Daling, L.
Bernstein, K. E. Malone, et al. 2002. The NICHD women's contraceptive and
reproductive experiences study: Methods and operational results. Annals of
Epidemiology 12, (4) (May): 213-21.
Morton, L. M., P. Hartge, T. R. Holford, E. A. Holly, B. C. Chiu, P. Vineis, E. Stagnaro,
et al. 2005. Cigarette smoking and risk of non-hodgkin lymphoma: A pooled analysis
from the international lymphoma epidemiology consortium (interlymph). Cancer
Epidemiology, Biomarkers & Prevention: A Publication of the American Association
for Cancer Research, Cosponsored by the American Society of Preventive Oncology
14, (4) (Apr): 925-33.
Morton, L. M., S. S. Wang, S. S. Devesa, P. Hartge, D. D. Weisenburger, and M. S. Linet.
2006. Lymphoma incidence patterns by WHO subtype in the United States, 1992-
2001. Blood 107, (1) (Jan 1): 265-76.
Morton, L. M., T. Zheng, T. R. Holford, E. A. Holly, B. C. Chiu, A. S. Costantini, E.
Stagnaro, et al. 2005. Alcohol consumption and risk of non-Hodgkin lymphoma: A
pooled analysis. The Lancet Oncology 6, (7) (Jul): 469-76.

33
National cancer institute sponsored study of classifications of non-Hdgkin's lymphomas:
Summary and description of a working formulation for clinical usage. The non-
Hdgkin's lymphoma pathologic classification project. 1982. Cancer 49, (10) (May
15): 2112-35.
Nelson, R. A., A. M. Levine, and L. Bernstein. 2001. Reproductive factors and risk of
intermediate- or high-grade B-cell non-Hodgkin’s lymphoma in women. Journal of
Clinical Oncology: Official Journal of the American Society of Clinical Oncology 19,
(5) (Mar 1): 1381-7.
Palmer, J. R., E. E. Hatch, C. L. Rosenberg, P. Hartge, R. H. Kaufman, L. Titus-Ernstoff,
K. L. Noller, et al. 2002. Risk of breast cancer in women exposed to
diethylstilbestrol in utero: Prelimiinary results (United States). Cancer Causes &
Control: CCC 13, (8) (Oct): 753-8.
Palmer, J. R., L. A. Wise, E. E. Hatch, R. Troisi, L. Titus-Ernstoff, W. Strohsnitter, R.
Kaufman, et al. 2006. Prenatal diethylstilbestrol exposure and risk of breast cancer.
Cancer Epidemiology, Biomarkers & Prevention: A Publication of the American
Association for Cancer Research, Cosponsored by the American Society of
Preventive Oncology 15, (8) (Aug): 1509-14.
Pan, S. Y., Y. Mao, A. M. Ugnat, and Canadian Cancer Registries Epidemiology
Research Group. 2005. Physical activity, obesity, energy intake, and the risk of non-
Hodgkin’s lymphoma: A population-based case-control study. American Journal of
Epidemiology 162, (12) (Dec 15): 1162-73.
Poole, J. A., and H. N. Claman. 2004. Immunology of pregnancy. Implications for the
mother. Clinical Reviews in Allergy & Immunology 26, (3) (Jun): 161-70.
Rayburn, W. F. 1996. Clinical commentary: The bromocriptine (parlodel) controversy
and recommendations for lactation suppression. American Journal of Perinatology
13, (2) (Feb): 69-71.
Sandin, S., H. Hjalgrim, B. Glimelius, K. Rostgaard, E. Pukkala, and J. Askling. 2006.
Incidence of non-Hodgkin’s lymphoma in Sweden, Denmark, and Finland from 1960
through 2003: An epidemic that was. Cancer Epidemiology, Biomarkers &
Prevention: A Publication of the American Association for Cancer Research,
Cosponsored by the American Society of Preventive Oncology 15, (7) (Jul): 1295-
300.
Tavani, A., A. Pregnolato, C. La Vecchia, and S. Franceschi. 1997. A case-control study
of reproductive factors and risk of lymphomas and myelomas. Leukemia Research
21, (9) (Sep): 885-8.

34
Troisi, R., E. E. Hatch, L. Titus-Ernstoff, M. Hyer, J. R. Palmer, S. J. Robboy, W. C.
Strohsnitter, R. Kaufman, A. L. Herbst, and R. N. Hoover. 2007. Cancer risk in
women prenatally exposed to diethylstilbestrol. International Journal of Cancer.
Journal International Du Cancer 121, (2) (Jul 15): 356-60.
Wegmann, T. G., H. Lin, L. Guilbert, and T. R. Mosmann. 1993. Bidirectional cytokine
interactions in the maternal-fetal relationship: Is successful pregnancy a TH2
phenomenon? Immunology Today 14, (7) (Jul): 353-6.
Whitacre, C. C., S. C. Reingold, and P. A. O'Looney. 1999. A gender gap in
autoimmunity. Science (New York, N.Y.) 283, (5406) (Feb 26): 1277-8.
Willett, E. V., C. F. Skibola, P. Adamson, D. R. Skibola, G. J. Morgan, M. T. Smith, and
E. Roman. 2005. Non-Hodgkin’s lymphoma, obesity and energy homeostasis
polymorphisms. British Journal of Cancer 93, (7) (Oct 3): 811-6.
Wolk, A., G. Gridley, M. Svensson, O. Nyren, J. K. McLaughlin, J. F. Fraumeni, and H.
O. Adam. 2001. A prospective study of obesity and cancer risk (Sweden). Cancer
Causes & Control: CCC 12, (1) (Jan): 13-21.
Zhang, Y., T. R. Holford, B. Leaderer, P. Boyle, S. H. Zahm, B. Zhang, K. Zou, et al.
2004. Menstrual and reproductive factors and risk of non-Hodgkin’s lymphoma
among Connecticut women. American Journal of Epidemiology 160, (8) (Oct 15):
766-73.

Abstract (if available)

Abstract Study Aim: Reproductive patterns have been inconsistently linked to NHL risk. The aim of this study is to identify if these factors affect B-cell NHL risk among women. Methods: We conducted a population-based case-control study in Los Angeles County. A total of 555 histologically confirmed B-cell NHL patients and 555 individually matched control subjects were included. Conditional logistic regression was used to evaluate these associations. Results: Reproductive history factors such as gravidity, parity, and outcome of first pregnancy were not associated with B-cell NHL risk overall or any NHL subtype. A modest association was observed for use of lactation suppressants (OR=0.78, 95% CI=0.56-1.08). The use of diethylstilbestrol (DES) during pregnancy increased the risk of B-cell NHL (OR=2.10, 95% CI=0.71-6.23). Conclusion: Reproductive factors were not associated with the risk of B-cell NHL overall or any subtype. Future studies with greater numbers of NHL subtypes are needed to confirm our study results.

Linked assets

University of Southern California Dissertations and Theses

Conceptually similar

PDF

Personal hair dye use and risk of B-cell non-Hodgkin’s lymphomas among adult women in Los Angeles County

PDF

Study of risk factors of B-cell non-Hodgkin lymphoma (NHL) in the California Teachers Study

PDF

Common immune-related factors and risk of non-Hodgkin lymphomy

PDF

Red and processed meat consumption and colorectal cancer risk: meta-analysis of case-control studies

PDF

Associations between exercise and quality of life in adult survivors of non-Hodgkin's lymphoma and colorectal cancer

PDF

Risk factors associated with smoking initiation among Chinese adolescents: a matched case-control study

PDF

Survival of children and adolescents with low-risk non-rhabdomyosarcoma soft tissue sarcomas (NRSTS) treated with surgery only: an analysis of 234 patients from the Children’s Oncology Group stud...

PDF

Non-steroidal anti-inflammatory drugs and risk of esophageal and gastric adenocarcinomas in Los Angeles county

PDF

Arm lymphedema in a multi-ethnic cohort of female breast cancer survivors

PDF

Carcinogen metabolism genes, meat intake, and colorectal cancer risk

PDF

Postmenopausal hormone therapy, risk of heart disease and total mortality among women in the California Teachers Study

PDF

Trends and charateristics of causes of death among persons with AIDS in Los Angeles County, 1990-2002

PDF

sFLT-1 gene polymorphisms and risk of severe-spectrum hypertensive disorders of pregnancy

PDF

Obesity paradox in acute heart failure decompensation

PDF

Native American ancestry among Hispanic Whites is associated with higher risk of childhood obesity: a longitudinal analysis of Children’s Health Study data

PDF

Diet quality and pancreatic cancer incidence in the multiethnic cohort

PDF

The evaluation of the long-term effectiveness of zero/low fluoroscopy workflow in ablation procedures for the treatment of paroxysmal and persistent atrial fibrillation

PDF

Sex hormones and atherosclerosis in postmenopausal women

PDF

Distant metastatic nonrhabdomyosarcoma soft tissue sarcomas in children and adolescents: clinical features and survival outcome among patients in Children's Oncology Group Phase 3 Study ARST0332

PDF

Comparative study of the POG and FNCLCC grading systems in non-rhabdomyosarcoma soft tissue sarcomas

Asset Metadata

Creator Lu, Yani (author)

Core Title Reproductive factors and risk of B-cell non-Hodgkin's lymphomas among women in Los Angeles

School Keck School of Medicine

Degree Master of Science

Degree Program Applied Biostatistics

Publication Date 04/16/2008

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

Tag non-Hodgkin's lymphomas,OAI-PMH Harvest,reproductive factors

Place Name Los Angeles County (city or populated place)

Language English

Advisor Bernstein, Leslie (committee chair), Cozen, Wendy (committee member), Mack, Wendy J. (committee member)

Creator Email yanilu@usc.edu

Permanent Link (DOI) https://doi.org/10.25549/usctheses-m1104

Unique identifier UC1280694

Identifier etd-Lu-20080416 (filename),usctheses-m40 (legacy collection record id),usctheses-c127-51836 (legacy record id),usctheses-m1104 (legacy record id)

Legacy Identifier etd-Lu-20080416.pdf

Dmrecord 51836

Document Type Thesis

Rights Lu, Yani

Type texts

Source University of Southern California (contributing entity), University of Southern California Dissertations and Theses (collection)

Repository Name Libraries, University of Southern California

Repository Location Los Angeles, California

Repository Email cisadmin@lib.usc.edu