This Article Abstract Full Text (PDF) Alert me when this article is cited Alert me if a correction is posted Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Smedby, K. E. Articles by Melbye, M. Search for Related Content PubMed PubMed Citation Articles by Smedby, K. E. Articles by Melbye, M.

Childhood Social Environment and Risk of Non–Hodgkin Lymphoma in Adults

¹ Department of Medical Epidemiology and Biostatistics, Karolinska Institutet; ² Department of Pathology and Oncology, Karolinska Institutet and University Hospital, Stockholm, Sweden; ³ Department of Epidemiology Research, Statens Serum Institut, Copenhagen, Denmark; ⁴ Northern California Cancer Center, Fremont, California; ⁵ Division of Epidemiology, Department of Health Research and Policy, Stanford University School of Medicine, Stanford, California; ⁶ Department of Oncology, Radiology, and Clinical Immunology, University of Uppsala, Uppsala, Sweden; and ⁷ Department of Epidemiology, Harvard School of Public Health, Boston, Massachusetts

Requests for reprints: Karin Ekström Smedby, Department of Medical Epidemiology and Biostatistics, Karolinska Institutet, Box 281, SE-171 77 Stockholm, Sweden. Phone: 468-517-707-06; Fax: 468-517-793-04; E-mail: karin.ekstrom.smedby{at}ki.se.

	Abstract

Top Abstract Introduction Materials and Methods Results Discussion References

 
Better hygiene and sanitation and decreasing family size parallel the increasing incidence of non–Hodgkin lymphoma (NHL) in many populations around the world. However, whether sibship size, birth order, and crowding are related to adult NHL risk is not clear. We investigated how family structure and childhood social environment were related to the risk of NHL and NHL subtypes in a large Scandinavian population–based case control study with 6,242 participants aged 18 to 74 years. Detailed exposure information was obtained through telephone interviews. Odds ratios (OR) and 95% confidence intervals (95% CI) were calculated using logistic regression, and all statistical tests were two-sided. Having four or more siblings was associated with a moderately increased risk of NHL, compared with having no siblings (OR 1.34, 95% CI 1.11-1.62, P_trend < 0.001). Having four or more older siblings was associated with a similar risk increase (OR 1.33, 95% CI 1.12-1.59, P_trend = 0.003) compared with being the oldest, whereas number of younger siblings was unrelated overall. The associations were independent of other environmental exposures and did not vary by country, age, or sex. High household crowding was also positively associated with risk of NHL. Results were slightly stronger for diffuse large B-cell and T-cell lymphomas than for other major NHL subtypes. Our findings add to the evidence that large sibship size, late birth order, and childhood crowding are associated with an elevated risk of NHL. Effect mechanisms may be related to early age at onset and high frequency of specific infections or total microbial exposure in childhood. [Cancer Res 2007;67(22):11074–82]

	Introduction

Top Abstract Introduction Materials and Methods Results Discussion References

 
Over recent decades, changes in childhood social environment in much of the world, especially industrialized nations, have been characterized by increased hygiene and sanitation and smaller family size (1). According to the "hygiene hypothesis," these changes may affect the development of the immune system in early life through postponement of common infections and, as a result, may also influence the risk of immune-related disorders (2). Although recent findings suggest that the hygiene hypothesis is overly simplified (3), it provides a useful framework to examine how childhood exposures may be associated with risk of immunologic diseases in adulthood. Given that early microbial exposure is of importance for immune development (1) and that malignant lymphomas arise from immune system cells, childhood social environment may influence the risk of these malignancies.

Recent studies indicate that birth order and crowding (4–7), as well as frequency of and age at infections during childhood (6, 8), may be associated with risk of non–Hodgkin lymphoma (NHL) later in life. However, reported findings are inconsistent and contradictory. In a U.S. study (6), history of frequent infections up to age 8 or 15 years was associated with a reduced risk of NHL, whereas late birth order and crowding, presumably associated with early and frequent exposure to infections, were found to increase risk. Large sibship size and late birth order were similarly positively associated with risk of NHL in some studies (4, 5, 7), whereas others reported no (8–11) or an inverse association (12). Related factors, such as childhood socioeconomic status, housing, and parental age at birth have generally not been accounted for.

As published results are inconsistent and their implications are unclear, we investigated interrelated factors, such as sibship and family size, birth order, parental age at birth, housing type, and other variables of early childhood social environment in association with risk of NHL overall and four major NHL subtypes, in a large population-based case control study in Denmark and Sweden.

	Materials and Methods

Top Abstract Introduction Materials and Methods Results Discussion References

 
Participants. The Scandinavian lymphoma etiology study has been described previously (13). In brief, residents of Denmark and Sweden were enrolled in this population-based case control study from June 2000 through August 2002 in Denmark and October 1999 through April 2002 in Sweden. The study population was restricted to individuals with ages 18 to 74 years, with sufficient knowledge of the Danish or Swedish language and without a history of organ transplantation, HIV infection, or previous hematopoietic malignancy. Patients with a first, newly diagnosed NHL, including chronic lymphocytic leukemia, were identified through a rapid case ascertainment network involving 157 hospital clinics in both countries, with back-up from nationwide tumor registries.

Control subjects were randomly sampled from updated population registries every 6 months during the study period. The number of controls was frequency-matched on sex and age (in 10-year intervals) to the expected distribution of NHL patients in each country. Altogether, 3,055 patients with NHL and 3,187 controls chose to participate (81% among eligible patients and 71% among eligible controls). The study was approved by regional ethics committees, and informed consent (oral and/or written) was obtained from each participant before interview.

Tumor material of the study patients was reviewed, and NHL subtypes were classified according to WHO classification (14). In Denmark, the review was done within the National Lymphoma Registry Organization (15), wherein all but 20% of the cases were evaluated by a senior hematopathologist. In Sweden, six expert hematopathologists or cytologists were specially appointed for tumor review, and slides were evaluated for all but 1.5% of the cases, for whom the written results of the primary investigation was evaluated instead. The Swedish study evaluation resulted in a change to a different or more detailed NHL subtype diagnosis from that of the primary clinical report in 3% of the cases and to a less detailed subtype diagnosis in 4%.

Exposure assessment. Participants were interviewed by telephone regarding their early childhood social environment and other possible risk factors for lymphoma based on a standardized questionnaire. Specifically, we interrogated for sibship size and birth order, birth year of siblings, parental age at birth of the index subject, parental education (9, 10-12, or >12 years), and living conditions during the first 6 years of childhood. The latter included type of housing (single-family house, farm, townhouse, or apartment), number of rooms (excluding kitchen and bathroom), total number of persons sharing a household, family socioeconomic status (average, higher, or lower than average), urban or rural residential area, contact with domestic animals, and day care attendance (ever/never, age at start, and number of children in the group). Townhouses are single-family residences, often situated in cities or suburban areas, that are unconnected but share adjoining walls. If any of these variables changed during the first 6 years of childhood, subjects were asked to report the situation that lasted longest.

We also assessed personal history of mononucleosis and tuberculosis (ever/never and age at diagnosis) and tuberculosis in first-degree relatives (ever/never), as these infections have been implicated in the etiology of NHL (14, 16), and because their occurrence may be related to socioeconomic level. We constructed a variable to describe household density or crowding by dividing the number of household members by the number of household rooms. Parental age at the subject's birth was categorized a priori into quartiles based on the distribution in the control population.

Statistical analysis. We used unconditional logistic regression to estimate odds ratios (OR) and 95% confidence intervals (95% CI) as measures of relative risk. The regression model included adjustment for the matching variables, age (in 10-year intervals), sex, and country, as independent variables and interaction terms. Potential confounders were considered based on prior knowledge of risk factors for NHL (17) and on changes of effect estimates comparing models with and without additional covariates. Heterogeneity of associations by country, sex, or age (with cutoffs at 40 or 60 years of age) was evaluated using a likelihood ratio test for the significance of an interaction term between each exposure variable and either country, sex, or age. Assessment of heterogeneity by ethnic group was not meaningful due to the large predominance of Caucasian participants (>95%). We did internal checks for consistency, for example, comparing reported maternal and paternal age at subject's birth with reported birth order.

Adjustment for personal educational level, body mass index, smoking, history of autoimmune or allergic disorders, occupational exposure to pesticides or organic solvents, sun bathing habits, or family history of cancer or lymphoma (self-reported or registered; ref. 18) changed estimates of <10% and were not included in the final model. Given the apparent close relationship of several of the variables analyzed, we computed Spearman's rank correlation coefficients, r_s (19). Mutual adjustment for weakly correlated analyzed variables affected some estimates at 10% and are shown separately. In the multivariate model, a separate category for missing data was included for maternal age, paternal education, and urban/rural residence, respectively, after checking that the missing categories were unrelated to the outcome.

We tested for trend across exposure categories by assigning equally spaced values (e.g., 1, 2, 3, and 4) to the categories and treating them as continuous variables in the logistic regression analysis. We assessed potential heterogeneity of the effect of exposures across NHL subtypes in a polytomous regression model, with the NHL subtypes and control status as outcomes. Each variable in the model was the log odds ratio of an NHL subtype outcome compared with control status for the exposure levels compared with a reference level, and heterogeneity was tested with a Wald test. All statistical tests were two-sided. Analyses were done using the SAS System software, release 8.2 and 9.1 (SAS Institute, Inc.; 1999-2001).

	Results

Top Abstract Introduction Materials and Methods Results Discussion References

 
General characteristics of the participants are shown in Table 1 . Results for family structure and household crowding, parental age at birth, and history of mononucleosis or tuberculosis are shown in Table 2 . Subjects with four or more siblings were at a statistically significant 30% increase in risk of NHL compared with subjects with no siblings (OR 1.34, 95% CI 1.11-1.62, P_trend < 0.001; Table 2). A risk increase of similar magnitude was observed for subjects having four or more older siblings compared with first-born subjects (P_trend = 0.003), whereas number of younger siblings was unrelated to risk of NHL overall (Table 2). In an analysis restricted to individuals with at least one sibling, having four or more older siblings was still associated with an increased risk (P_trend = 0.02), whereas no such trend was seen for younger siblings (P_trend = 0.30). Among subjects with older siblings, the age difference in years to the nearest older sibling (<3, 3-5, or 6 years) did not appreciably affect NHL risk (data not shown).

With regard to NHL subtypes, a risk increase with increasing sibship size was indicated for all types analyzed (Table 2). Effect estimates were slightly more pronounced for the diffuse large B-cell and T-cell NHL subtypes, but results did not differ statistically among subtypes (P_{heterogeneity by subtype} = 0.63). Similarly, the pattern of positive associations with number of household members and with number of older siblings, but not with younger siblings, was most evident for diffuse large B-cell and T-cell lymphoma. High household crowding was associated with a statistically significant 40% increase in risk of diffuse large B-cell lymphoma.

Older maternal or paternal age at the subject's birth, self-reported personal history of mononucleosis or tuberculosis, and history of tuberculosis in close relatives were not significantly associated with risk of overall NHL or NHL subtypes (Table 2). There was no statistically significant heterogeneity by country, sex, or age in risk of NHL overall for any exposure listed in Table 2.

Results for childhood socioeconomic status and environment are shown in Table 3 . For several of these variables, we found significant heterogeneity of risk by country, and results are therefore shown separately for Denmark and Sweden. Swedish controls were more likely than Danish controls to have parents with a lower level of education and to have resided in a rural area and/or a townhouse. Danish controls were more likely than Swedes to have had higher family socioeconomic status and to have resided in an apartment in a large city. Overall, parental education and family socioeconomic status were not associated with risk of all NHL (Table 3) or NHL subtypes (data not shown). Although results went in different directions in the two countries, there were no clear trends in either nation with increasing levels of education.

In Table 4 , we show the effect of some exposures on NHL risk using slightly varying sets of adjustment variables to aid the interpretation of effect of each determinant. Additional inclusion in the model of interaction terms between country and paternal education and between country and urban/rural residence reduced the heterogeneity by country for all other variables in Table 3. Large sibship size, late birth order, and farm or townhouse residence remained positively associated with NHL risk after adjustments, and the indication of a positive association with crowding was strengthened (Table 4). Associations with domestic animal contact and daycare attendance and risk of NHL disappeared (Table 4). In an analysis restricted to individuals without siblings, daycare attendance was still not associated with NHL risk (data not shown).

	Discussion

Top Abstract Introduction Materials and Methods Results Discussion References

Our findings are in agreement with observations from five smaller population-based case control studies (4–7, 10). In a U.S. study (6), large sibship size (7 siblings) and childhood residence in crowded homes were both associated with 50% increase in risk of NHL, but birth order did not modify the effect. A similar association between large sibship size and risk of NHL was observed in a subset of heterosexual participants (4) and among homosexual men (7) in the U.S. study. From Australia, Grulich et al. reported a reduced risk of NHL in only-children and first-born subjects compared with subjects born fourth or later (5). Large sibship size was also weakly positively associated with NHL risk in a German study (10).

In contrast, in two Italian studies, a population-based (8) and a hospital-based, (11), and one American-nested case control study (9), there was no association between sibship size (8, 9, 11) or birth order (11) and NHL risk. Similarly, family size and birth order were not associated with risk of malignant lymphomas in a large Swedish cohort study (20). However, in the cohort study that was restricted to subjects below 55 years of age, NHL and Hodgkin lymphoma patients were considered together. As the association of sibship size with NHL may be in the opposite direction of that with Hodgkin lymphoma, results for the two lymphoma groups may have cancelled each other out (21–23). In a hospital-based U.K. study (12), only-children were at a 40% increased risk of adult NHL, compared with individuals with siblings. Thus, although previous findings are mixed, most population-based studies indeed support a positive association between large sibship size and risk of NHL (4–7, 10).

Biological mechanisms for the suggested association are not clear, but a role for infections is feasible. Large sibship size, late birth order, and more crowded homes are associated with early childhood exposure to several specific infectious agents, including Helicobacter pylori (24), Meningococci (25), Parvovirus B 19 (26), and hepatitis A virus (27), and likely also favor exposure to common infections in general (28). Older siblings probably influence age at first common infection and number of infections more than younger ones in early childhood, as older siblings bring home contagious diseases from school or day care (21, 28). Opposite to NHL, Hodgkin lymphoma has been associated with small sibship size and early birth order, especially among young adults (21–23). This has been attributed at least in part to postponement of childhood infections, in particular with EBV (29, 30). EBV is also related to NHL development in strongly immunosuppressed individuals (17). Given our findings with regard to sibship size, an inverse association between postponement of EBV infection, i.e., infectious mononucleosis, and sporadic NHL (not coupled with strong immunosuppression) could have been expected but was not shown here or in a previous report (31).

Two case control studies directly investigated the potential importance of age at onset and frequency of common childhood infections and risk of NHL (6, 8). Bracci et al. (6) observed, in contrast to their findings on sibship size, that a history of several specified infections up to age of 8 or 15 years was associated with a reduced risk of NHL. Vineis et al. (8) observed that old age (4 years) at first infection (from a list of specified infections) was associated with an increased risk of NHL, but only in small families. These results apparently contradict the hypothesis that a large sibship size augments NHL risk through increased frequency of infections and/or earlier age at infectious disease onset. However, it is not certain that the specific infections assessed in these two studies (6, 8) or the age periods chosen were the most relevant for NHL development nor that they accurately reflect the total burden of infectious disorders in early childhood. It is further possible that adulthood recall of infections in early childhood is subject to large misclassification. Sibship size and birth order, important and constant aspects of family structure throughout life, are less likely subject to differential recall and misclassification.

Still, other mechanisms for the presented findings must also be considered. Various biological, socioeconomic, and cultural factors are potentially relevant for large families. For example, late birth order is positively associated with high birth weight (32) and with duration of maternal breast feeding (33), of which the effects on adult NHL risk are unknown. According to recent census data, large family size (3 children, currently comprising 4% of all households in Denmark and Sweden) was associated with lower socioeconomic status and rural residence.⁸ However, it is uncertain whether this was true in the middle of the 20th century, during the early childhood of most of our study participants.

An immunologic influence of specific infectious agents or of the total infectious disease burden in early childhood would go along with established disease mechanisms in NHL. Although the immunologic basis for the "hygiene hypothesis" is uncertain, evidence support that postponement of common infections prevents the normal shift of a T-helper cell type 2 to a T-helper cell type 1 phenotype (1). A recent alternative view suggests that a lower microbial burden decreases the activity of T-regulatory cells (1). The biological hypothesis that follows would be that a T-helper cell type 1 phenotype and/or a high T-regulatory cell activity increase risk of adult NHL. However, unfortunately, the present results, along with recent trends in society toward smaller families, do not help us explain the past increase in incidence of NHL.

In our study, family socioeconomic status and parental education level, parental age at birth, and tuberculosis were unrelated to risk of NHL in univariate analyses, and day care and domestic animals were unassociated in the multivariate model. However, we were unable to examine associations with very early age at start of day care attendance (<2 years). Early-life social class was also unrelated to lymphoma mortality in adults in a large Swedish registry-based study (34). However, previous reports on tuberculosis and NHL do not exclude a small excess risk conferred by this agent (16, 35). With regard to geographic residence and type of housing, the effect modification by country rendered the interpretation of these findings difficult. Childhood farm residence was associated with an increased risk in both Denmark and Sweden, but was not associated with NHL risk in two previous studies (10, 12). However, our finding of a similarly increased risk with townhouse residence puts the specificity of the association with farms in question. Environmental childhood exposures that are assessed through self-reports in adulthood may be subject to differential and/or nondifferential misclassification, which may call in question the exactness of some results. However, given the only marginal effects upon adjustment of these exposures, it seems unlikely that residual confounding by, for example, parental age or childhood living conditions could offer an explanation for our main findings.

The strengths of our study include the large size that permitted analysis of risk of major NHL subtypes, the population-based design, and the simultaneous assessment of various related exposures in childhood social environment and other potential risk factors for NHL. A limitation is the significant heterogeneity of effect by country for a few exposures that could be due to chance or to varying participation rates among control subjects between countries, resulting in different degrees of selection bias. Control subjects that choose to participate in health studies are often more highly educated (36), which could create false inverse associations or obscure true positive associations between socioeconomic level and disease risk. Alternatively, the different results by country in our study may reflect true effects of childhood living conditions that, although they are labeled the same way, may represent different life-styles, the details of which we were unable to discern. For example, Denmark is much more densely populated than Sweden, and life-style differences with regard to tobacco smoking and diet observed in recent years^9,10 may imply that some life-style patterns have differed also historically.

To conclude, the results from this large population-based study add strongly to the evidence that having more siblings, especially older ones, is associated with an increased risk of NHL. This finding, in combination with a positive association with childhood residential crowding, could indicate that age at onset and number of infections during the first few years of childhood are relevant mechanisms in the pathogenesis of NHL.

	Acknowledgments

 
Grant support: NIH 5R01 grant CA69269-02 (M. Melbye) and Swedish Cancer Society grant 04 6420 (H-O. Adami).

The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

	Footnotes

 
⁸ Scb.se [homepage on the internet]. Stockholm: Statistics Sweden. [cited 2007 Aug 9]. Available from http://www.scb.se/.

⁹ Scb.se [homepage on the internet]. Stockholm: Statistics Sweden. [cited 2007 Aug 9]. Available from http://www.scb.se/.

¹⁰ Danmarks statistik-statistikbanken.dk [database on the internet]. Copenhagen (Denmark): Statbank Denmark [cited 2007 Aug 9]. Available from: http://www.dst.dk/.

Received 5/11/07. Revised 8/22/07. Accepted 9/21/07.

	References

Top Abstract Introduction Materials and Methods Results Discussion References

 

1. Romagnani S. The increased prevalence of allergy and the hygiene hypothesis: missing immune deviation, reduced immune suppression, or both? Immunology 2004;112:352–63.[CrossRef][Medline]
2. Strachan DP. Hay fever, hygiene, and household size. BMJ 1989;299:1259–60.[Free Full Text]
3. Schaub B, Lauener R, von Mutius E. The many faces of the hygiene hypothesis. J Allergy Clin Immunol 2006;117:969–77; quiz 978.[CrossRef][Medline]
4. Holly EA, Lele C, Bracci PM, McGrath MS. Case-control study of non-Hodgkin's lymphoma among women and heterosexual men in the San Francisco Bay Area, California. Am J Epidemiol 1999;150:375–89.[Abstract/Free Full Text]
5. Grulich AE, Vajdic CM, Kaldor JM, et al. Birth order, atopy, and risk of non-Hodgkin lymphoma. J Natl Cancer Inst 2005;97:587–94.[Abstract/Free Full Text]
6. Bracci PM, Dalvi TB, Holly EA. Residential history, family characteristics and non-Hodgkin lymphoma, a population-based case-control study in the San Francisco Bay Area. Cancer Epidemiol Biomarkers Prev 2006;15:1287–94.[Abstract/Free Full Text]
7. Holly EA, Lele C. Non-Hodgkin's lymphoma in HIV-positive and HIV-negative homosexual men in the San Francisco Bay Area: allergies, prior medication use, and sexual practices. J Acquir Immune Defic Syndr Hum Retrovirol 1997;15:211–22.[Medline]
8. Vineis P, Miligi L, Crosignani P, et al. Delayed infection, family size and malignant lymphomas. J Epidemiol Community Health 2000;54:907–11.[Abstract/Free Full Text]
9. Paffenbarger RS, Jr., Wing AL, Hyde RT. Characteristics in youth predictive of adult-onset malignant lymphomas, melanomas, and leukemias: brief communication. J Natl Cancer Inst 1978;60:89–92.[Medline]
10. Becker N, Deeg E, Nieters A. Population-based study of lymphoma in Germany: rationale, study design and first results. Leuk Res 2004;28:713–24.[CrossRef][Medline]
11. Chatenoud L, Gallus S, Altieri A, et al. Number of siblings and risk of hodgkin's and other lymphoid neoplasms. Cancer Epidemiol Biomarkers Prev 2005;14:552.[Free Full Text]
12. Cartwright RA, McKinney PA, O'Brien C, et al. Non-Hodgkin's lymphoma: case control epidemiological study in Yorkshire. Leuk Res 1988;12:81–8.[CrossRef][Medline]
13. Smedby KE, Hjalgrim H, Melbye M, et al. Ultraviolet radiation exposure and risk of malignant lymphomas. J Natl Cancer Inst 2005;97:199–209.[Abstract/Free Full Text]
14. Jaffe ES, Harris NL, Stein H, Vardiman JW, editors. Pathology and genetics of Tumours of hematopoietic and Lymphoid Tissues. 1st ed. Lyon: IARC Press; 2001.
15. d'Amore F, Christensen BE, Brincker H, et al. Clinicopathological features and prognostic factors in extranodal non-Hodgkin lymphomas. Danish LYFO Study Group. Eur J Cancer 1991;27:1201–8.[Medline]
16. Vineis P, Crosignani P, Sacerdote C, et al. Haematopoietic cancer and medical history: a multicentre case control study. J Epidemiol Community Health 2000;54:431–6.[Abstract/Free Full Text]
17. Ekstrom-Smedby K. Epidemiology and etiology of non-Hodgkin lymphoma-a review. Acta Oncol 2006;45:258–71.[CrossRef][Medline]
18. Chang ET, Smedby KE, Hjalgrim H, Glimelius B, Adami HO. Reliability of self-reported family history of cancer in a large case-control study of lymphoma. J Natl Cancer Inst 2006;98:61–8.[Abstract/Free Full Text]
19. Altman D. Practical statistics for medical research. 1st ed. London: Chapman & Hill; 1997.
20. Hemminki K, Mutanen P. Birth order, family size, and the risk of cancer in young and middle-aged adults. Br J Cancer 2001;84:1466–71.[CrossRef][Medline]
21. Chang ET, Montgomery SM, Richiardi L, Ehlin A, Ekbom A, Lambe M. Number of siblings and risk of Hodgkin's lymphoma. Cancer Epidemiol Biomarkers Prev 2004;13:1236–43.[Abstract/Free Full Text]
22. Gutensohn N, Cole P. Childhood social environment and Hodgkin's disease. N Engl J Med 1981;304:135–40.[Abstract]
23. Westergaard T, Melbye M, Pedersen JB, Frisch M, Olsen JH, Andersen PK. 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][Medline]
24. Santos IS, Boccio J, Santos AS, et al. Prevalence of Helicobacter pylori infection and associated factors among adults in Southern Brazil: a population-based cross-sectional study. BMC Public Health 2005;5:118.[CrossRef][Medline]
25. Goldacre MJ. Space-time and family characteristics of meningococcal disease and haemophilus meningitis. Int J Epidemiol 1977;6:101–5.[Abstract/Free Full Text]
26. Valeur-Jensen AK, Pedersen CB, Westergaard T, et al. Risk factors for parvovirus B19 infection in pregnancy. JAMA 1999;281:1099–105.[Abstract/Free Full Text]
27. Green MS, Zaaide Y. Sibship size as a risk factor for hepatitis A infection. Am J Epidemiol 1989;129:800–5.[Abstract/Free Full Text]
28. Kuper H, Hsieh C, Stuver SO, et al. Birth order, as a proxy for age at infection, in the etiology of hepatocellular carcinoma. Epidemiology 2000;11:680–3.[CrossRef][Medline]
29. Hjalgrim H, Askling J, Rostgaard K, et al. Characteristics of Hodgkin's lymphoma after infectious mononucleosis. N Engl J Med 2003;349:1324–32.[Abstract/Free Full Text]
30. Hjalgrim H, Ekstrom Smedby K, Rostgaard K, et al. Infectious mononucleosis, childhood social environment, and risk of hodgkin lymphoma. Cancer Res 2007;67:2382–8.[Abstract/Free Full Text]
31. Vajdic CM, Grulich AE, Kaldor JM, et al. Specific infections, infection-related behavior, and risk of non-Hodgkin lymphoma in adults. Cancer Epidemiol Biomarkers Prev 2006;15:1102–8.[Abstract/Free Full Text]
32. Harvey NC, Poole JR, Javaid MK, et al. Parental determinants of neonatal body composition. J Clin Endocrinol Metab 2007;92:523–6.[Abstract/Free Full Text]
33. Oddy WH, Kendall GE, Blair E, et al. Breast feeding and cognitive development in childhood: a prospective birth cohort study. Paediatr Perinat Epidemiol 2003;17:81–90.[CrossRef][Medline]
34. Lawlor DA, Sterne JA, Tynelius P, Davey Smith G, Rasmussen F. Association of childhood socioeconomic position with cause-specific mortality in a prospective record linkage study of 1,839,384 individuals. Am J Epidemiol 2006;164:907–15.[Abstract/Free Full Text]
35. Askling J, Ekbom A. Risk of non-Hodgkin's lymphoma following tuberculosis. Br J Cancer 2001;84:113–5.[CrossRef][Medline]
36. Jackson R, Chambless LE, Yang K, et al. Differences between respondents and nonrespondents in a multicenter community-based study vary by gender ethnicity. The Atherosclerosis Risk in Communities (ARIC) Study Investigators. J Clin Epidemiol 1996;49:1441–6.[CrossRef][Medline]

This article has been cited by other articles:

				Y. Lu, J. Prescott, J. Sullivan-Halley, K. D. Henderson, H. Ma, E. T. Chang, C. A. Clarke, P. L. Horn-Ross, G. Ursin, and L. Bernstein Body Size, Recreational Physical Activity, and B-Cell Non-Hodgkin Lymphoma Risk Among Women in the California Teachers Study Am. J. Epidemiol., November 15, 2009; 170(10): 1231 - 1240. [Abstract] [Full Text] [PDF]

				G. J. Tranah, P. M. Bracci, and E. A. Holly Domestic and Farm-Animal Exposures and Risk of Non-Hodgkin's Lymphoma in a Population-Based Study in the San Francisco Bay Area Cancer Epidemiol. Biomarkers Prev., September 1, 2008; 17(9): 2382 - 2387. [Abstract] [Full Text] [PDF]

This Article Abstract Full Text (PDF) Alert me when this article is cited Alert me if a correction is posted Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Smedby, K. E. Articles by Melbye, M. Search for Related Content PubMed PubMed Citation Articles by Smedby, K. E. Articles by Melbye, M.