Skip Navigation

This Article
Right arrow Abstract Freely available
Right arrow FREE Full Text (PDF) Freely available
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in ISI Web of Science
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Add to My Personal Archive
Right arrow Download to citation manager
Right arrow Search for citing articles in:
ISI Web of Science (14)
Right arrowRequest Permissions
Right arrow Disclaimer
Google Scholar
Right arrow Articles by Glaser, S. L.
Right arrow Articles by Dorfman, R. F.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Glaser, S. L.
Right arrow Articles by Dorfman, R. F.
Social Bookmarking
Add to CiteULike   Add to Connotea   Add to Del.icio.us  
What's this?

Am J Epidemiol 2003; 158:553-563.
Copyright © 2003 by Johns Hopkins Bloomberg School of Public Health

ORIGINAL CONTRIBUTIONS

Reproductive Factors in Hodgkin’s Disease in Women

Sally L. Glaser1 , Christina A. Clarke1, Rebecca A. Nugent1, Cynthia B. Stearns1 and Ronald F. Dorfman2

1 Northern California Cancer Center, Union City, CA.
2 Division of Surgical Pathology, Stanford University Medical Center, Stanford, CA.

Received for publication January 29, 2003; accepted for publication April 9, 2003.


   ABSTRACT
TOP
 ABSTRACT
INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 
Reproductive factors have been suggested to have an impact on the development of Hodgkin’s disease (HD) in women. In the San Francisco Bay Area, the authors conducted a population-based case-control study addressing the effects of reproductive experience and hormone use on HD risk. Cases were 370 women with HD diagnosed at ages 19–79 years between July 1988 and December 1994. Controls were 450 community women found through random digit dialing. Among the 312 cases and 325 controls interviewed, HD risk was related to parity versus nulliparity but only among never nursers (odds ratio (OR) = 2.2, 95% confidence interval (CI): 1.0, 5.0). Risk was marginally related to having uterine fibroids (OR = 0.6, 95% CI: 0.5, 1.0) and long-term versus short-term hormone use (OR = 0.7, 95% CI: 0.4, 1.0) and was significantly related to recurrent miscarriage (OR = 2.8, 95% CI: 1.1, 7.4). Among women aged 35–54 years, for whom the sex difference in incidence is largest, nursing decreased risk; among never nursers, a parity of 1 lowered risk and higher parity increased risk; long-term hormone use lowered risk; and recurrent miscarriage increased risk. Among women under age 35 years, endometriosis lowered HD risk; the lack of significant findings for most other variables may reflect selection bias in controls. Among older women, no significant associations were observed, although hormone use appeared to be protective. These data suggest that steroid hormones may affect HD development.

case-control studies; endometriosis; Hodgkin disease; lactation; parity

Abbreviations: Abbreviations: CI, confidence interval; OR, odds ratio; Th1, T helper cell type 1; Th2, T helper cell type 2.


   INTRODUCTION
TOP
 ABSTRACT
 INTRODUCTION
MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 
Hodgkin’s disease is an uncommon lymphoma whose etiology remains incompletely understood (1). Age-specific variation in incidence rates initially suggested an infectious origin in young adults (2), and subsequent associations of risk with higher childhood social class implicated the timing of exposure to common infections (3, 4). Risk has also been linked to certain occupations and to genetic factors, including human leukocyte antigen type (5, 6). However, few other leads regarding causation have been identified. A new direction for exploring etiology was suggested by the observation that the male excess in Hodgkin’s disease incidence rates was particularly marked during the childbearing years (7). This incidence pattern—in conjunction with lower Hodgkin’s disease rates observed among women who were married or of higher parity, an association of parity with Hodgkin’s disease risk in two early analytical studies, and a protective effect of pregnancy on lymphoma incidence in rats—raised the possibility that childbearing or related reproductive factors are involved in Hodgkin’s disease development in women (8). To investigate this hypothesis, we undertook a population-based case-control study of Hodgkin’s disease risk in the San Francisco Bay Area, considering a range of factors encompassing reproductive potential and experience and exposure to exogenous hormones and examining the possibility of confounding by the social class characteristics shown to be risk factors for Hodgkin’s disease (3, 4, 9–11).


   MATERIALS AND METHODS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
RESULTS
 DISCUSSION
 REFERENCES
 
Subject identification and contact
The study protocol was reviewed and approved by the Northern California Cancer Center Human Subjects Review Committee. Cases were identified through the population-based Greater Bay Area Cancer Registry, which ascertains all incident cancers in Alameda, Contra Costa, Marin, Monterey, San Benito, San Francisco, San Mateo, Santa Clara, and Santa Cruz counties as part of the Surveillance, Epidemiology, and End Results Program and the California Cancer Registry. Eligible cases were female area residents newly diagnosed at ages 19–79 years with histologically confirmed Hodgkin’s disease (International Classification of Diseases for Oncology, Second Edition, morphology codes 9650–9667 (12)) between mid-July 1988 and the end of December 1994 (n = 395) who were alive at the time of report to the Registry (n = 376) and spoke English (n = 370). Uniform diagnosis and histologic subclassification of Hodgkin’s disease were established through expert re-review (performed by R. F. D.) of diagnostic pathology slides (13). There was agreement between the review diagnosis and the original diagnosis for 245 (68 percent) of the 362 reviewed cases, but agreement varied by histologic subtype (nodular sclerosis, 95 percent; lymphocyte predominance, 69 percent; mixed cellularity, 58 percent; lymphocyte depletion, 0 percent; not otherwise specified, 40 percent).

Controls were English-speaking women without a history of Hodgkin’s disease residing in the same nine-county area who were identified through random digit dialing, as described elsewhere (14). Briefly, calls were made to 12,970 random telephone numbers, yielding 830 nonanswers, 2,787 businesses or pay phones, 3,762 disconnected numbers, 945 electronic responses, 18 second telephone lines, and 4,628 residences. Among the residences, 10 percent of respondents refused to participate in the household enumeration, which targeted adult females by age and race/ethnicity; 58 percent were ineligible on the basis of sex, age, race/ethnicity, or language use; and 32 percent were eligible for future contact. From this last group, we randomly selected and attempted to recruit one woman per household as needed for frequency-matching to cases on 5-year age group and race/ethnicity.

Physicians of eligible case women were first queried about contraindications to contacting the patient. Cases without physician refusal and eligible controls were invited, initially by letter and then by telephone, to participate in the study’s in-person interview. Participating subjects were required to sign a consent form before being interviewed.

The standardized interview inquired about reproductive potential (age at menarche, years from menarche to commencement of regular menses, average duration of menstrual cycles, number of menstrual periods missed (aside from pregnancies and lactation), history of and age at hysterectomy and oophorectomy, age at last menstrual period); reproductive experience (age at, duration of, and outcome of each pregnancy; duration of each associated period of nursing); gynecologic conditions (history of and age at diagnosis of endometriosis, uterine fibroids, and ovarian cysts); and use of exogenous hormones (dates and duration of each episode of use for oral contraceptives and for other hormones, including use of estrogens with or without progestins in menopause-associated hormone replacement therapy). The interview also addressed childhood and adult social class measures. All questions pertained to the year prior to diagnosis for cases and a corresponding time before interview for controls; controls were frequency-matched to cases according to this interval. Subjects completed a life events calendar to facilitate recall of study exposures (15). Thirty subjects who had moved from the area were interviewed by telephone using the same questionnaire and mailed interview materials.

The reliability of both random digit dialing enumeration and selected questionnaire responses was addressed through reinterview of a random sample of participants and was found to be good (14, 16). Potential bias resulting from the exclusion of deceased cases was evaluated using questionnaire responses obtained from the next of kin of 19 of the 38 deceased case women through a self-administered, abbreviated version of the study instrument.

Of the 370 study-eligible cases, two (<1 percent) were not contacted at their physician’s request, 10 (3 percent) could not be contacted within 2 years of diagnosis (a cutoff established because of concerns about recall accuracy), two (<1 percent) had left the United States, 19 (5 percent) died before interview, five (1 percent) could not be located, 12 (3 percent) refused to participate, and 320 (87 percent) were interviewed. Among the 310 interviewed patients for whom pathology specimens were obtained, 302 (97 percent) were considered to have Hodgkin’s disease on histologic and/or immunohistochemical grounds, and 259 (86 percent) of these were found to have the most common histologic subtype, nodular sclerosis. Given the very low rate of disease misclassification overall, the 10 interviewed patients without pathology re-review were assumed to have Hodgkin’s disease. Thus, 312 cases were included in the analyses. Among the 450 women selected for recruitment as controls, 49 could not be relocated, 76 refused, and 325 (72 percent) were interviewed.

Statistical analysis
To examine case-control differences, we first compared relative frequency distributions with chi-squared or Fisher’s exact tests. We evaluated normally distributed continuous variables with t tests to compare mean values; we categorized data on all other variables into tertiles based on the distributions of control responses. We calculated unadjusted odds ratios and corresponding 95 percent confidence intervals to estimate relative risks. For reproductive potential, we calculated the cumulative numbers of menstrual cycles and reproductive years using age at menarche and age at the last menstrual period or in the reference year, considering cessation of menstrual periods only if it occurred at least 1 year prior to the reference year to minimize any disease-related effect on menses. For reproductive experience, we calculated the total numbers of pregnancies (gravidity), livebirths (parity), therapeutic abortions, miscarriages (<=20 weeks’ gestation), and stillbirths (>20 weeks’ gestation); the total numbers of months pregnant, months since last delivery, and months of nursing; and age at first full-term pregnancy. Recurrent miscarriage, evaluated among women with at least one fetal loss, was defined as two or more unexplained miscarriages because of the low prevalence of this outcome. Women were considered to have nursed if their lifetime duration of lactation was at least 1 month. For exogenous hormones, we summed periods of use to obtain total duration of exposure; long-term users were compared with both never users (data presented in the tables) and short-term users (data presented in the text). The prevalence of oophorectomy was deemed too low to warrant joint examination with age at surgery.

Analyses were first conducted among women of all ages for evaluation of the overall effects of risk factors and for consistency with prior studies (17–21). Because of age-specific variation in male-female incidence rate differences (figure 1), the strong association of age with some reproductive characteristics (e.g., parity, hormone replacement therapy), and our prior findings of age-specific variation in social class risk factors (10, 11), we also examined case-control differences separately in women aged 19–34, 35–54, and 55–79 years after recategorizing continuous variables and social class indicators into tertiles based on age-specific distributions of control responses.



View larger version (26K):
[in this window]
[in a new window]
  		FIGURE 1. Average annual age-specific incidence rates of Hodgkin’s disease per 100,000 population, by sex, Surveillance, Epidemiology, and End Results Program, 1988–1994. Vertical lines demarcate study age groups. CI, confidence interval.

 
We used unconditional logistic regression to explore the simultaneous effects of reproductive characteristics suggestively associated with case-control status while controlling for the potentially confounding effects of age, race/ethnicity, and social class indicators. For the latter, prior analyses had identified increased risk of Hodgkin’s disease in young adults with family-owned versus rented childhood homes; single bedrooms versus shared bedrooms at age 11 years among the US-born; and, for women with two or more livebirths, smaller versus larger childhood households (10). We explored the interaction between nursing and parity, because nursing was significantly associated with Hodgkin’s disease and the association varied with parity; no other interactions were considered. We constructed parsimonious models to obtain adjusted odds ratios and corresponding 95 percent confidence intervals for variables that were significant at p < 0.10.


   RESULTS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
DISCUSSION
 REFERENCES
 
Reflecting Hodgkin’s disease incidence patterns (figure 1), the 312 case women were predominantly young (23 percent aged 19–24 years, 20 percent aged 25–29 years, 19 percent aged 30–34 years, 9 percent aged 35–39 years, 10 percent aged 40–44 years, 6 percent aged 45–49 years, 4 percent aged 50–59 years, 6 percent aged 60–69 years, and 3 percent aged 70–79 years) and of non-Hispanic White race/ethnicity (80 percent White, 8 percent African-American, 6 percent Hispanic, and 6 percent other). There were no significant differences in age distribution between participating cases and controls overall (p = 0.84) or within racial/ethnic groups (p = 0.88 for Whites, p = 0.97 for African Americans, p = 0.89 for Hispanics, and p = 0.97 for others).

Distributions of selected study variables among cases and controls are shown in table 1. For reproductive potential, there were no case-control differences in the indicators shown or in average age at menarche (12.8 years vs. 13.0 years, p = 0.22), average duration of the menstrual cycle (28.4 days vs. 28.7 days, p = 0.25), proportion of subjects who had missed more than six menstrual periods (8.3 percent vs. 8.6 percent, p = 0.89), or, among postmenopausal women, average age at the last menstrual period (42.7 years vs. 45.0 years, p = 0.16). Similar proportions of cases and controls reported reproductive surgery (9 percent vs. 11 percent (p = 0.63) for hysterectomy; 4 percent vs. 5 percent (p = 0.40) for oophorectomy), but cases were somewhat younger than controls at surgery (mean age at oophorectomy, 39.6 years vs. 46.7 years (p = 0.05); mean age at hysterectomy, 38.8 years vs. 42.6 years (p = 0.11)).


View this table:
[in this window]
[in a new window]
  		TABLE 1. Distribution of cases and controls and odds ratios for Hodgkin’s disease according to reproductive variables among women aged 19–79 years, San Francisco Bay Area, 1988–1994
 
Regarding reproductive experience, cases and controls did not differ significantly in terms of proportion ever gravid (62 percent vs. 64 percent, p = 0.58) or proportion ever parous (44 percent vs. 49 percent, p = 0.15). Women with one livebirth had a marginally lower unadjusted risk of Hodgkin’s disease than nulliparous women, but there were no associations for higher-order births. Nursing was associated with a lower unadjusted risk of Hodgkin’s disease, apparently irrespective of parity (for one birth, odds ratio (OR) = 0.6, 95 percent confidence interval (CI): 0.3, 1.5; for two births, OR = 0.8, 95 percent CI: 0.3, 1.8; for three or more births, OR = 0.6, 95 percent CI: 0.3, 1.3) and duration (among nursers, the unadjusted odds ratio for each additional month of lactation was 1.0). There were no case-control differences in age at first birth, years since last pregnancy, or abortion history. Among women experiencing fetal loss, cases were more likely than controls to have had two or more miscarriages.

Uterine fibroids were reported by slightly fewer cases than controls, but endometriosis and ovarian cysts were not associated with Hodgkin’s disease risk. The highest tertile of cumulative oral contraceptive use (>=5.3 years) was suggestively linked with lower disease risk in comparison with the lowest tertile of use (<2.2 years) (unadjusted OR = 0.6, 95 percent CI: 0.4, 1.0). Cases were less likely than controls to have ever used hormone replacement therapy (9.9 percent vs. 15.1 percent, p = 0.05; OR = 0.6, 95 percent CI: 0.4, 1.0), and longer-term users were at approximately half the Hodgkin’s disease risk of short-term users (OR = 0.7, 95 percent CI: 0.4, 1.0).

In multivariate analyses (table 1), parity was not associated with Hodgkin’s disease among women who had nursed. Among never-nursing women, those with three or more births were suggested to have an elevated risk of Hodgkin’s disease in comparison with nulliparous women. Women who had had two or more miscarriages had a nearly threefold higher risk of Hodgkin’s disease than women with one miscarriage. Marginally protective effects remained for a history of uterine fibroids and for long-term use of oral contraceptives as compared with short-term use (for >=5.3 years of use vs. <2.2 years, adjusted OR = 0.7, 95 percent CI: 0.4, 1.0) but not for long-term use of hormone replacement therapy versus short-term use (adjusted OR = 0.5, 95 percent CI: 0.1, 1.6).

Table 2 presents relative frequency distributions and odds ratios for cases and controls aged 19–34, 35–54, and 55–79 years for variables associated with Hodgkin’s disease risk in at least one of the age groups. For women aged 19–34 years, recurrent miscarriage was linked to elevated risk of Hodgkin’s disease, although confidence intervals around the adjusted odds ratios were wide, and endometriosis was suggestively associated with a reduced risk of Hodgkin’s disease after adjustment for confounders. Among women aged 35–54 years, cases were half as likely as controls to have nursed, but this association appeared to vary with parity (for one birth, unadjusted OR = 2.6, 95 percent CI: 0.3, 26.1; for two births, OR = 0.5, 95 percent CI: 0.2, 1.4; for three or more births, OR = 0.2, 95 percent CI: 0.0, 1.1). In multivariate analysis, ever nursing seemed to be protective regardless of parity. Among never-nursing women, those with one child also had a substantially lower risk of Hodgkin’s disease than nulliparous women, but those with higher parity appeared to have an elevated risk. Recurrent miscarriage remained a significant predictor of Hodgkin’s disease risk. Longer-term hormone use was associated with lower Hodgkin’s disease risk, suggestively for oral contraceptives in comparison with shorter-term use (adjusted OR = 0.5, 95 percent CI: 0.2, 1.2) and significantly for hormone replacement therapy in comparison with no use. For both of these age groups, findings were quite similar among the 169 and 70 nodular sclerosis cases (comprising 87 percent and 86 percent of all cases aged 19–34 years and 35–54 years, respectively) compared with their respective controls (data not shown).


View this table:
[in this window]
[in a new window]
  		TABLE 2. Relative frequency distributions of case and control women and odds ratios for Hodgkin’s disease according to reproductive variables, by age group, San Francisco Bay Area, 1988–1994
 
Cases aged 55–79 years were more likely than older controls to have had irregular menses, but no case-control differences occurred for parity, nursing, or recurrent miscarriages. Among the nine cases and 12 controls reporting ovary removal, cases were younger at surgery (mean age 39.8 years vs. 48.7 years, p = 0.05). Endometriosis was suggested to increase Hodgkin’s disease risk and hormone use to lower risk. However, none of these factors predicted Hodgkin’s disease at a statistically significant level in the multivariate analysis.


   DISCUSSION
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
REFERENCES
 
Several studies on Hodgkin’s disease risk and reproductive factors have now been published (table 3) (17–25). Most have described a slight-to-moderate decrease in Hodgkin’s disease risk with higher parity, with some but not all finding a more protective effect among women of reproductive age (19–22, 25) and one finding an effect only among better-educated women (24). Greater age at first birth has been weakly linked to higher disease risk, particularly among women under age 45 years (19, 21, 22); time since last birth has been associated inconsistently with risk (21, 22, 25); and one study reported a protective effect for abortion (21). The inconsistency of findings is not surprising, given the differences in the studies’ designs, age groups, variables included, and sample sizes, and the absence of pathology re-review, which can cause disease misclassification in older subjects (13, 26).


View this table:
[in this window]
[in a new window]
  		TABLE 3. Published findings from studies of the relation between reproductive factors and Hodgkin’s disease risk in women
 
Our population-based case-control study of pathologically re-reviewed Hodgkin’s disease considered a variety of reproductive exposures in a sample of sufficient size for multivariate analysis controlling for social class confounders. We found that Hodgkin’s disease risk overall was related to recurrent miscarriage and marginally to higher parity among never nursers and that it was suggested to be inversely associated with a history of uterine fibroids and with long-term hormone use. However, associations appeared to vary across the age groups corresponding to sex differences in incidence rates. For women under age 35 years, recurrent miscarriage and endometriosis were the only significant risk factors for Hodgkin’s disease. The absence of an association with childbearing in this age group might be expected from the similarity of incidence rates for males and females under age 30 and the relatively low proportions of ever-parous and ever-nursing women (30 percent and 20 percent in this study population). Similarly, low exposure prevalence may help explain the lack of a strong reproductive risk for Hodgkin’s disease reported by others, since a substantial proportion of all adult female patients are likely to be under age 25 years, given the bimodal age pattern of Hodgkin’s disease incidence rates in many populations (1). However, our null findings regarding parity in women under age 35 years could also reflect the selection bias we detected for parity among controls based on age- and race-specific 1990 US Census data for this age group in the nine-county study region (unpublished data). In analyses weighted to correct for this bias, unadjusted odds ratios were significantly protective for parity versus nulliparity (OR = 0.6, 95 percent CI: 0.4, 0.8), for two births versus no births (1 vs. 0: OR = 0.7, 95 percent CI: 0.4, 1.3; 2 vs. 0: OR = 0.4, 95 percent CI: 0.2, 0.7; >=3 vs. 0: OR = 0.6, 95 percent CI: 0.2, 1.3), and for ever nursing versus never nursing (OR = 0.5, 95 percent CI: 0.3, 0.7); this suggests effects of reproductive experience in younger women pointing in the same direction as those in women of concurrent or recent reproductive age (35–54 years).

In women aged 35–54 years, the effect of parity was modified by nursing, such that higher parity appeared to be associated with increased Hodgkin’s disease risk among never-nursing women, while nursing appeared to decrease risk irrespective of parity. Since, to our knowledge, no previous studies of Hodgkin’s disease risk have examined nursing, an associated protective effect could explain some of the lower disease risk reported for parity (19, 25, 26), and the variation in parity effect estimates across studies could reflect differences in the population prevalences of nursing. However, such confounding could only explain the inverse association of parity and Hodgkin’s disease risk (18, 24, 25) if a nursing effect were cumulative, which it was not in our data but recently was found to be in a study of non-Hodgkin’s lymphoma (27). Among women aged 35–54 years, Hodgkin’s disease risk also was significantly elevated with recurrent miscarriage and diminished with hormone use.

For women over age 54 years, Hodgkin’s disease risk was suggestively associated only with irregular menses, endometriosis, and hormone use. Analyses for this age group were hampered by the small sample size, which was consistent with Hodgkin’s disease incidence patterns and with the low interview rate (53 percent of all Hodgkin’s disease patients over age 54 years) due to mortality, although inclusion of next-of-kin data for 12 of the 21 deceased older cases produced no marked changes in associations beyond a stronger effect of oral contraceptive use (OR = 0.3, 95 percent CI: 0.1, 0.8). In addition, data were missing for 9 percent of older subjects as compared with 0.1 percent of women aged 35–54 years, and older cases are more heterogeneous in terms of tumor histologic types, which exhibit distinctive epidemiologic patterns (28).

In general, our data are consistent with some effect of steroid hormones or other hormones on the development of Hodgkin’s disease. A protective role of higher levels of hormone exposure is suggested by most of our findings, including reduced risks with parity, nursing, and exogenous hormone use, a marginally reduced risk with uterine fibroids (whose growth is driven by steroid hormones (29)), and increased risk with younger age at reproductive surgery (which involves loss of hormone production). Because lymphocytes are thought not to exhibit hormone receptors (30), this protective influence probably operates indirectly through the well-established effects of estrogens, progesterone, and other hormones on immune function. Although their precise immunologic effects remain poorly understood, estrogen, progesterone, and prolactin have been noted to influence regulation of T helper cell type 1 (Th1) and T helper cell type 2 (Th2) cytokine production and lymphopoiesis (31). Physiologic changes associated with several study exposures (increases in circulating estrogens and progesterone seen during pregnancy, in laboratory and animal studies (31–34), and with use of oral contraceptives and hormone replacement therapy (31, 35, 36); prolactin increases seen during nursing (37); and immunologic changes associated with endometriosis (29, 38)) have also been associated with changes in the relative predominance of Th1/Th2 cytokines, reductions in the number of natural killer cells, and inhibition of B-cell production (32, 34, 39). In addition, nursing elevates levels of parathyroid hormone-related protein, which has been shown to suppress the growth of lymphocyte cell lines infected with human T-cell lymphotropic virus type 1 (40, 41). Recurrent miscarriage is characterized by elevated levels of natural killer cell activity (29), consistent with the elevation of Hodgkin’s disease risk that we found for this factor. Thus, steroid hormones and other hormones may affect development of Hodgkin’s disease, now thought to be a B-cell lymphoma, through their impact on B-cell production. Postmenopausal use of exogenous hormones may reduce risk of intermediate or high-grade non-Hodgkin’s lymphoma by suppressing production of interleukin-6, a Th2 cytokine known to act as a paracrine growth factor for lymphoma (42). However, use of hormone replacement therapy was recently associated with increased risk of follicular lymphoma (43).

Some of our findings for parity are inconsistent with a suppressive effect of hormones on lymphomagenesis. Among women aged 35–54 years, the limitation of lower Hodgkin’s disease risk associated with childbearing to a parity of 1 is puzzling, as is the increased risk for higher parity. Only one prior study noted an increase in Hodgkin’s disease risk with parity, and it did not report any protective influence of low parity (17). Given the accumulated epidemiologic evidence for an inverse association of Hodgkin’s disease with number of births (table 3) and the absence of selection bias for our controls aged 35–54 years, this effect of parity on Hodgkin’s disease risk in our population may be a chance finding. An alternative explanation involves social correlates of parity (6, 24). Hodgkin’s disease risk in young adults has been linked to childhood infection delayed until adolescence (6). Among women under age 45 years in our study, those from smaller childhood households (a proxy for less early contact with children and thus delayed infection exposure) were at four times the Hodgkin’s disease risk (adjusted OR = 4.0, 95 percent CI: 1.3, 12.9) of those from larger childhood households if they also had had two or more births (10). Thus, among women not otherwise protected from Hodgkin’s disease development by nursing, an elevated Hodgkin’s disease risk with higher levels of parity could reflect an overriding consequence of late exposure to infection via their offspring. Small sample size prevented a test of this hypothesis in our data.

Some of our study findings are also consistent with nonhormonal mechanisms of Hodgkin’s disease risk. Recurrent miscarriage, associated with a substantially elevated risk of Hodgkin’s disease in women under age 55 years, has been associated with certain human leukocyte antigen class II alleles and haplotypes (44, 45), as has risk for the nodular sclerosis histologic subtype of Hodgkin’s disease, which comprised the majority of our cases under age 55 (5, 46). Thus, recurrent miscarriage and Hodgkin’s disease could be related partly through a common genetic susceptibility, and the absence of an association between miscarriage and Hodgkin’s disease in older women could reflect the much lower prevalence of the nodular sclerosis subtype in these women. Endometriosis, which affected Hodgkin’s disease risk in an age-modified pattern, manifests differently in younger and older women in terms of clinical presentation and other features (47–49), suggesting differences in etiology by age. The protective effect of endometriosis on Hodgkin’s disease risk among younger women in our study is consistent with its associated changes in immune function (29, 38). Endometriosis also is associated with low parity, hormone treatment, family aggregation, and, in one small study, infectious mononucleosis—all risk factors for Hodgkin’s disease (50)—but its persistent association with Hodgkin’s disease risk in our multivariate analysis makes confounding an unlikely explanation.

Finally, our findings may have been affected by error. Because all patient information was self-reported, its validity is uncertain, although the reliability of responses to questions on reproductive experience was very good ({kappa} = 0.65–0.94) (16). For medical conditions with diagnostic imprecision (e.g., endometriosis), some misclassification probably occurred, but the similar mean years of education for cases (14.1 years) and controls (14.3 years) would tend to rule out differential misreporting based on medical access or knowledge. While recall bias is also a possibility, the opposing associations for miscarriage and endometriosis make this a less likely explanation for our findings. In some subgroups, the small numbers of subjects produced unstable odds ratios, and the low parity of our population limited our statistical power to detect effects that might be apparent only with higher-order births (25). Finally, we cannot rule out the possibility that our findings are due to uncontrolled social correlates of reproductive experience rather than to hormonal effects.

On balance, our population-based examination of a range of reproductive and related factors supports the possibility of a transient effect of steroid hormones and other hormones on the development of Hodgkin’s disease in women. Given the established if incompletely understood interaction between hormones and immune function, additional exploration of this understudied aspect of Hodgkin’s disease etiology in a larger population could be informative.


   ACKNOWLEDGMENTS
 
This work was supported in part by National Cancer Institute grant R29 CA50381 (S. L. G., R. F. D.). Cancer incidence data were collected by the Northern California Cancer Center under contract N01-PC-65107 with the National Cancer Institute and with the support of the California Cancer Registry (a project of the Cancer Surveillance Section, California Department of Health Services) under subcontract 1000891 with the Public Health Institute (Oakland, California).

The authors thank Barbara Dabney, Janet Kettlehut, Virginia Zemlin, Marta Zahn, Rita Leung, Dr. Susan Stewart, Alice Dantsuka, and Connie Cady for their contributions to this project.

The content of this publication does not necessarily reflect the views or policies of the US Department of Health and Human Services or the California Department of Health Services, nor does the mention of trade names, commercial products, or organizations imply endorsement by the US government or the state of California.


   NOTES
 
Correspondence to Dr. Sally L. Glaser, Northern California Cancer Center, 32960 Alvarado-Niles Road, Suite 600, Union City, CA 94587 (e-mail: sglaser{at}nccc.org). Back


   REFERENCES
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 

  1. Mueller N, Grufferman S. The epidemiology of Hodgkin’s disease. In: Mauch PM, Armitage JO, Diehl V, et al, eds. Hodgkin’s disease. Philadelphia, PA: Lippincott Williams and Wilkins, 1999:61–77.
  2. MacMahon B. Epidemiology of Hodgkin’s disease. Cancer Res 1966;26:1189–201.[ISI][Medline]
  3. Gutensohn N, Cole P. Childhood social environment and Hodgkin’s disease. N Engl J Med 1981;304:135–40.[Abstract]
  4. Gutensohn NM. Social class and age at diagnosis of Hodgkin’s disease: new epidemiologic evidence for the "two-disease hypothesis." Cancer Treat Rep 1982;66:689–95.[ISI][Medline]
  5. Taylor GM, Gokhale DA, Crowther D, et al. Increased frequency of HLA-DPB1*0301 in Hodgkin’s disease suggests that susceptibility is HVR-sequence and subtype-associated. Leukemia 1996;10:854–9.[ISI][Medline]
  6. Mueller N. Hodgkin’s disease. In: Schottenfeld D, Fraumeni JF Jr, eds. Cancer epidemiology and prevention. New York, NY: Oxford University Press, 1996:893–919.
  7. Grufferman S, Delzell E. Epidemiology of Hodgkin’s disease. Epidemiol Rev 1984;6:76–106.[Free Full Text]
  8. Glaser SL. Reproductive factors in Hodgkin’s disease in women: a review. Am J Epidemiol 1994;139:237–46.[Free Full Text]
  9. Westergaard T, Melbye M, Pedersen JB, et al. Birth order, sibship size and risk of Hodgkin’s disease in children and young adults: a population-based study of 31 million person-years. Int J Cancer 1997;72:977–81.[CrossRef][ISI][Medline]
  10. Glaser SL, Clarke CA, Nugent RA, et al. Social class and risk of Hodgkin’s disease in young-adult women in 1988–94. Int J Cancer 2002;98:110–17.[CrossRef][ISI][Medline]
  11. Glaser SL, Clarke CA, Stearns CB, et al. Age variation in Hodgkin’s disease risk factors in older women: evidence from a population-based case-control study. Leuk Lymphoma 2001;42:997–1004.[ISI][Medline]
  12. Percy C, Van Holten V, Muir C. ICD-O: international classification of diseases for oncology. 2nd ed. Geneva, Switzerland: World Health Organization, 1990.
  13. Glaser SL, Dorfman RF, Clarke CA. Expert review of the diagnosis and histologic classification of Hodgkin’s disease in a population-based cancer registry: interobserver reliability and impact on incidence and survival rates. Cancer 2001;92:218–24.[CrossRef][ISI][Medline]
  14. Glaser SL, Stearns CB. Reliability of random-digit dialing screening calls to enumerate an adult female population. Am J Epidemiol 2002;155:972–5.[Abstract/Free Full Text]
  15. Wingo PA, Ory HW, Layde PM, et al. The evaluation of the data collection process for a multicenter, population-based, case-control design. Am J Epidemiol 1988;128:206–17.[Abstract/Free Full Text]
  16. Lin SS, Glaser SL, Stewart SL. Reliability of self-reported reproductive factors and childhood social class indicators in a case-control study of women. Ann Epidemiol 2002;12:242–7.[CrossRef][ISI][Medline]
  17. Miller AB, Barclay TH, Choi NW, et al. A study of cancer, parity and age at first pregnancy. J Chronic Dis 1980;33:595–605.[CrossRef][ISI][Medline]
  18. Kravdal O, Hansen S. Hodgkin’s disease: the protective effect of childbearing. Int J Cancer 1993;55:909–14.[ISI][Medline]
  19. Lambe M, Hsieh C-C, Tsaih S-W, et al. Childbearing and the risk of Hodgkin’s disease. Cancer Epidemiol Biomarkers Prev 1998;7:831–4.[Abstract]
  20. Abramson JH, Pridan H, Sacks MI, et al. A case-control study of Hodgkin’s disease in Israel. J Natl Cancer Inst 1978;61:307–14.[ISI][Medline]
  21. Tavani A, Pregnolato A, La Vecchia C, et al. A case-control study of reproductive factors and risk of lymphomas and myelomas. Leuk Res 1997;21:885–8.[CrossRef][ISI][Medline]
  22. Kravdal O, Hansen S. The importance of childbearing for Hodgkin’s disease: new evidence from incidence and mortality models. Int J Epidemiol 1996;25:737–43.[Abstract/Free Full Text]
  23. Franceschi S, Bidoli E, La Vecchia C. Pregnancy and Hodgkin’s disease. (Letter). Int J Cancer 1994;58:465–6.[ISI][Medline]
  24. Zahm SH, Hoover RN, Fraumeni JF Jr. Hodgkin’s disease and parity. (Letter). Int J Cancer 1995;62:362–3.[ISI][Medline]
  25. Zwitter M, Zakelj MP, Kosmelj K. A case-control study of Hodgkin’s disease and pregnancy. Br J Cancer 1996;73:246–51.[ISI][Medline]
  26. Glaser SL, Swartz WG. Time trends in Hodgkin’s disease incidence: the role of diagnostic accuracy. Cancer 1990;66:2196–204.[CrossRef][ISI][Medline]
  27. Cerhan JR, Habermann TM, Vachon CM, et al. Menstrual and reproductive factors and risk of non-Hodgkin lymphoma: The Iowa Women’s Health Study (United States). Cancer Causes Control 2002;13:131–6.[CrossRef][ISI][Medline]
  28. Glaser SL, Jarrett RF. The epidemiology of Hodgkin’s disease. Bailliere’s Clin Haematol 1996;9:401–16.
  29. Somigliana E, Vigano P, Vignali M. Endometriosis and unexplained recurrent spontaneous abortion: pathological states resulting from aberrant modulation of natural killer cell function? Hum Reprod Update 1999;5:40–51.[Abstract/Free Full Text]
  30. Maia DM, Sciarrotta J, Abendroth K, et al. Sex steroid receptors in Hodgkin’s disease. Leuk Lymphoma 2000;39:365–71.[ISI][Medline]
  31. Giltay EJ, Fonk JC, von Blomberg BM, et al. In vivo effects of sex steroids on lymphocyte responsiveness and immunoglobulin levels in humans. J Clin Endocrinol Metab 2000;85:1648–57.[Abstract/Free Full Text]
  32. Kincade PW, Medina KL, Smithson G. Sex hormones as negative regulators of lymphopoiesis. Immunol Rev 1994;137:119–34.[CrossRef][ISI][Medline]
  33. Kincade PW, Medina KL, Payne KJ, et al. Early B-lymphocyte precursors and their regulation by sex steroids. Immunol Rev 2000;175:128–37.[CrossRef][ISI][Medline]
  34. Medina KL, Strasser A, Kincade PW. Estrogen influences the differentiation, proliferation, and survival of early B-lineage precursors. Blood 2000;95:2059–67.[Abstract/Free Full Text]
  35. Baker DA, Salvatore W, Milch PO. Effect of low-dose oral contraceptives on natural killer cell activity. Contraception 1989;39:119–24.[CrossRef][ISI][Medline]
  36. Kamada M, Irahara M, Maegawa M, et al. Transient increase in the levels of T-helper 1 cytokines in postmenopausal women and the effects of hormone replacement therapy. Gynecol Obstet Invest 2001;52:82–8.[CrossRef][ISI][Medline]
  37. Matera L, Mori M, Geuna M, et al. Prolactin in autoimmunity and antitumor defence. J Neuroimmunol 2000;109:47–55.[CrossRef][ISI][Medline]
  38. Rier SE, Yeaman GR. Immune aspects of endometriosis: relevance of the uterine mucosal immune system. Semin Reprod Endocrinol 1997;15:209–20.[Medline]
  39. Medina KL, Smithson G, Kincade PW. Suppression of B lymphopoiesis during normal pregnancy. J Exp Med 1993;178:1507–15.[Abstract/Free Full Text]
  40. Moseley JM, Gillespie MT. Parathyroid hormone-related protein. Crit Rev Clin Lab Sci 1995;32:299–343.[ISI][Medline]
  41. McCauley LK, Rosol TJ, Merryman JI, et al. Parathyroid hormone-related protein binding to human T-cell lymphotropic virus type I-infected lymphocytes. Endocrinology 1992;130:300–6.[Abstract]
  42. Nelson RA, Levine AM, Bernstein L. Reproductive factors and risk of intermediate- or high-grade B-cell non-Hodgkin’s lymphoma in women. J Clin Oncol 2001;19:1381–7.[Abstract/Free Full Text]
  43. Cerhan JR, Vachon CM, Habermann TM, et al. Hormone replacement therapy and risk of non-Hodgkin lymphoma and chronic lymphocytic leukemia. Cancer Epidemiol Biomarkers Prev 2002;11:1466–71.[Abstract/Free Full Text]
  44. Christiansen OB. Epidemiological, immunogenetic and immunotherapeutic aspects of unexplained recurrent miscarriage. Dan Med Bull 1997;44:396–424.[ISI][Medline]
  45. Christiansen OB. The possible role of classical human leukocyte antigens in recurrent miscarriage. Am J Reprod Immunol 1999;42:110–15.
  46. Klitz W, Aldrich CL, Fildes N, et al. Localization of predisposition to Hodgkin disease in the HLA class II region. Am J Hum Genet 1994;54:497–505.[ISI][Medline]
  47. Leyendecker G, Kunz G, Noe M, et al. Endometriosis: a dysfunction and disease of the archimetra. Hum Reprod Update 1998;4:752–62.[Abstract/Free Full Text]
  48. Parazzini F, Vercellini P, Panazza S, et al. Risk factors for adenomyosis. Hum Reprod 1997;12:1275–9.
  49. Sangi-Haghpeykar H, Poindexter AN 3rd. Epidemiology of endometriosis among parous women. Obstet Gynecol 1995;85:983–92.[Abstract]
  50. Lamb K, Nichols TR. Endometriosis: a comparison of associated disease histories. Am J Prev Med 1986;2:324–9.[ISI][Medline]

Add to CiteULike CiteULike   Add to Connotea Connotea   Add to Del.icio.us Del.icio.us    What's this?


This article has been cited by other articles:


Home page
 Cancer Epidemiol. Biomarkers Prev.Home page
T. H.M. Keegan, S. L. Glaser, C. A. Clarke, R. F. Dorfman, R. B. Mann, J. A. DiGiuseppe, E. T. Chang, and R. F. Ambinder
Body Size, Physical Activity, and Risk of Hodgkin's Lymphoma in Women.
Cancer Epidemiol. Biomarkers Prev., June 1, 2006; 15(6): 1095 - 1101.
[Abstract] [Full Text] [PDF]

Home page
 Cancer Epidemiol. Biomarkers Prev.Home page
C. A. Clarke, S. L. Glaser, T. H.M. Keegan, and A. Stroup
Neighborhood Socioeconomic Status and Hodgkin's Lymphoma Incidence in California
Cancer Epidemiol. Biomarkers Prev., June 1, 2005; 14(6): 1441 - 1447.
[Abstract] [Full Text] [PDF]

Home page
 Am J EpidemiolHome page
J. H. E. Promislow, B. C. Gladen, and D. P. Sandler
Maternal Recall of Breastfeeding Duration by Elderly Women
Am. J. Epidemiol., February 1, 2005; 161(3): 289 - 296.
[Abstract] [Full Text] [PDF]

This Article
Right arrow Abstract Freely available
Right arrow FREE Full Text (PDF) Freely available
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in ISI Web of Science
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Add to My Personal Archive
Right arrow Download to citation manager
Right arrow Search for citing articles in:
ISI Web of Science (14)
Right arrowRequest Permissions
Right arrow Disclaimer
Google Scholar
Right arrow Articles by Glaser, S. L.
Right arrow Articles by Dorfman, R. F.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Glaser, S. L.
Right arrow Articles by Dorfman, R. F.
Social Bookmarking
Add to CiteULike   Add to Connotea   Add to Del.icio.us  
What's this?