American Journal of Epidemiology - current issue

Type 2 Diabetes Mellitus and Risk of Non-Hodgkin Lymphoma: A Systematic Review and Meta-Analysis

Chao, C., Page, J. H. — 2008-08-22

Type 2 diabetes mellitus is associated with altered immune function and chronic inflammation. Both of these immune conditions are implicated in the pathogenesis of non-Hodgkin lymphoma. The authors performed a systematic review to summarize findings from the current literature on the association between history of type 2 diabetes mellitus and risk of non-Hodgkin lymphoma. Ten case-control studies and three prospective cohort studies were included in this review. Meta-analysis found that a history of type 2 diabetes mellitus was positively associated with overall non-Hodgkin lymphoma risk. However, there was significant heterogeneity between studies. Study design was an important source of heterogeneity. The rate ratio between type 2 diabetes mellitus and non-Hodgkin lymphoma was found to be 1.18 (95% confidence interval: 0.99, 1.42) among case-control studies and 1.79 (95% confidence interval: 1.30, 2.47) among the prospective cohort studies. Weaknesses were identified in some of the included studies in the areas of case and control selection, measurement of covariates and non-Hodgkin lymphoma, and confounding control. Although a positive association between type 2 diabetes mellitus and risk of non-Hodgkin lymphoma was suggested, the evidence is inconclusive because of methodological limitations of the included case-control studies. More prospective studies with improved control of confounding are needed to confirm these findings.

Perceived Stress and Cause-specific Mortality among Men and Women: Results from a Prospective Cohort Study

Nielsen, N. R., Kristensen, T. S., Schnohr, P., Gronbaek, M. — 2008-08-22

The authors assessed the effect of psychological stress on total and cause-specific mortality among men and women. In 1981–1983, the 12,128 Danish participants in the Copenhagen City Heart Study were asked two questions on stress intensity and frequency and were followed in a nationwide registry until 2004, with <0.1% loss to follow-up. Sex differences were found in the relations between stress and mortality (p = 0.02). After adjustments, men with high stress versus low stress had higher all-cause mortality (hazard ratio (HR) = 1.32, 95% confidence interval (CI): 1.15, 1.52). This finding was most pronounced for deaths due to respiratory diseases (high vs. low stress: HR = 1.79, 95% CI: 1.10, 2.91), external causes (HR = 3.07, 95% CI: 1.65, 5.71), and suicide (HR = 5.91, 95% CI: 2.47, 14.16). High stress was related to a 2.59 (95% CI: 1.20, 5.61) higher risk of ischemic heart disease mortality for younger, but not older, men. In general, the effects of stress were most pronounced among younger and healthier men. No associations were found between stress and mortality among women, except among younger women with high stress, who experienced lower cancer mortality (HR = 0.51, 95% CI: 0.28, 0.92). Future preventive strategies may be targeted toward stress as a risk factor for premature death among middle-aged, presumably healthy men.

Invited Commentary: Stress and Mortality

Hotopf, M., Henderson, M., Kuh, D. — 2008-08-22

In this issue of the Journal, Nielsen et al. (Am J Epidemiol 2008;168:481–91) use data from a large Danish study to provide evidence that self-reported stress is associated with increased all-cause mortality over the next 20 years. The finding is remarkable. In this commentary, the authors explore what is really meant by stress; they argue that it would be naïve to view stress as reported in this way, with some external exposure. It has to be seen through the lens of the participant's personal experience, and this lens is likely to be clouded by personality, coping styles, and the common mental disorders—depression and anxiety. The authors discuss a wider literature concerning similar findings associating depression with mortality, suggesting three broad reasons for the association. First, the findings might be explained by the impact of stress or distress on well-established risk factors for cardiovascular disease and cancer. Second, there might be direct, underlying psychosomatic pathways by which stress or distress can affect immune or autonomic function. Third, there might be common causal pathways—shared genes or early adversities that predict both stress and mortality from other causes independently. The authors suggest that life course epidemiologic research is required to test these competing hypotheses.

Nielsen et al. Respond to "Stress and Mortality"

Nielsen, N. R., Kristensen, T. S., Schnohr, P., Gronbaek, M. — 2008-08-22

Adult Height and the Risk of Mortality in South Korean Women

Song, Y.-M., Sung, J. — 2008-08-22

To evaluate the association between adult height as a surrogate marker of childhood circumstances and the risk of mortality, 344,519 South Korean women aged 40–64 years categorized into six height groups were prospectively followed for mortality between 1994 and 2004. In Cox proportional hazards regression with adjustment for behavioral and biologic risk factors, there was an inverse association between height and total mortality; mortality risk decreased 7% for each 5-cm increment in height. The association did not materially change after adjustment for behavioral factors and adulthood socioeconomic factors or after full adjustment for all available covariates. When height-associated risks of death from specific causes were evaluated in a fully adjusted analysis, a 5-cm increment in height was associated with lower risks of death from respiratory diseases, stroke, diabetes mellitus, and external causes (hazard ratios were 0.84 (95% confidence interval (CI): 0.74, 0.96), 0.84 (95% CI: 0.80, 0.88), 0.87 (95% CI: 0.80, 0.96), and 0.88 (95% CI: 0.83, 0.94), respectively) and with a higher risk of death from cancer (hazard ratio = 1.05, 95% CI: 1.02, 1.09). Given that adult height reflects early-life conditions, the independent associations between height and mortality from all causes and specific causes support the view that early-life circumstances significantly influence health outcomes in adulthood.

Mobility Disability and the Urban Built Environment

Clarke, P., Ailshire, J. A., Bader, M., Morenoff, J. D., House, J. S. — 2008-08-22

Research on the effects of the built environment in the pathway from impairment to disability has been largely absent. Using data from the Chicago Community Adult Health Study (2001–2003), the authors examined the effect of built environment characteristics on mobility disability among adults aged 45 or more years (n = 1,195) according to their level of lower extremity physical impairment. Built environment characteristics were assessed by using systematic social observation to independently rate street and sidewalk quality in the block surrounding each respondent's residence in the city of Chicago (Illinois). Using multinomial logistic regression, the authors found that street conditions had no effect on outdoor mobility among adults with only mild or no physical impairment. However, among adults with more severe impairment in neuromuscular and movement-related functions, the difference in the odd ratios for reporting severe mobility disability was over four times greater when at least one street was in fair or poor condition (characterized by cracks, potholes, or broken curbs). When all streets were in good condition, the odds of reporting mobility disability were attenuated in those with lower extremity impairment. If street quality could be improved, even somewhat, for those adults at greatest risk for disability in outdoor mobility, the disablement process could be slowed or even reversed.

Water Disinfection By-Products and Prelabor Rupture of Membranes

Joyce, S. J., Cook, A., Newnham, J., Brenters, M., Ferguson, C., Weinstein, P. — 2008-08-22

The causes of term prelabor rupture of membranes (term PROM) remain poorly defined. The authors conducted a record-based prevalence study to explore a possible relation between disinfection by-products in drinking water and term PROM in an Australian community with spatially variable trihalomethane and nitrate levels. A multilevel statistical model was used to examine the relation between factors operating at the levels of the individual, district, and water distribution zone and the prevalence of PROM at term among 16,229 women in Perth, Western Australia (2002–2004). Adjusted odds ratios for term PROM increased with increasing tertiles of nitrate exposure (moderate exposure: odds ratio = 1.23, 95% confidence interval: 1.03, 1.52; high exposure: odds ratio = 1.47, 95% confidence interval: 1.20, 1.79), but there was no significant relation with exposure to trihalomethanes. This study raises the possibility that water contaminants may promote the development of PROM at term.

Maternal Smoking during Pregnancy and Children's Cognitive and Physical Development: A Causal Risk Factor?

Gilman, S. E., Gardener, H., Buka, S. L. — 2008-08-22

There remains considerable debate regarding the effects of maternal smoking during pregnancy on children's growth and development. Evidence that exposure to maternal smoking during pregnancy is associated with numerous adverse outcomes is contradicted by research suggesting that these associations are spurious. The authors investigated the relation between maternal smoking during pregnancy and 14 developmental outcomes of children from birth through age 7 years, using data from the Collaborative Perinatal Project (1959–1974; n = 52,919). In addition to adjusting for potential confounders measured contemporaneously with maternal smoking, the authors fitted conditional fixed-effects models among siblings that controlled for unmeasured confounders. Results from the conditional analyses indicated a birth weight difference of –85.63 g associated with smoking of ≥20 cigarettes daily during pregnancy (95% confidence interval: –131.91, –39.34) and 2.73 times' higher odds of being overweight at age 7 years (95% confidence interval: 1.30, 5.71). However, the associations between maternal smoking and 12 other outcomes studied (including Apgar score, intelligence, academic achievement, conduct problems, and asthma) were entirely eliminated after adjustment for measured and unmeasured confounders. The authors conclude that the hypothesized effects of maternal smoking during pregnancy on these outcomes either are not present or are not distinguishable from a broader range of familial factors associated with maternal smoking.

Gender Differences in HIV Progression to AIDS and Death in Industrialized Countries: Slower Disease Progression Following HIV Seroconversion in Women

2008-08-22

To evaluate sex differences in human immunodeficiency virus (HIV) disease progression before (pre-1997) and after (1997–2006) introduction of highly active antiretroviral therapy, the authors used data from a collaboration of 23 HIV seroconverter cohort studies from Europe, Australia, and Canada restricted to the 6,923 seroconverters infected through injecting drug use and sex between men and women. Within a competing risk framework, they used Cox proportional hazards models allowing for late entry to evaluate sex differences in time from HIV seroconversion to death, to acquired immunodeficiency syndrome (AIDS), and to each first AIDS-defining disease and death without AIDS. While no significant sex differences were found before 1997, from 1997 onward, women had a lower risk of AIDS (adjusted cumulative relative risk (aCRR) = 0.76, 95% confidence interval (CI): 0.63, 0.90) and death (adjusted hazard ratio = 0.68, 95% CI: 0.56, 0.82) than men did. Compared with men, women also had lower risks of AIDS dementia complex (aCRR = 0.23, 95% CI: 0.07, 0.74), tuberculosis (aCRR = 0.60, 95% CI: 0.39, 0.92), Kaposi's sarcoma (aCRR = 0.27, 95% CI: 0.07, 0.99), lymphomas (aCRR = 0.47, 95% CI: 0.23, 0.96), and death without AIDS (aCRR = 0.74, 95% CI: 0.56, 0.98). Sex differences in HIV disease progression have become larger and statistically significant in the era of highly active antiretroviral therapy, supporting a stronger impact of health interventions among women.

Making the Most of Case-Mother/Control-Mother Studies

Shi, M., Umbach, D. M., Vermeulen, S. H., Weinberg, C. R. — 2008-08-22

The prenatal environment plays an important role in many conditions, particularly those with onset early in life, such as childhood cancers and birth defects. Because both maternal and fetal genotypes can influence risk, investigators sometimes use a case-mother/control-mother design, with mother-offspring pairs as the unit of analysis, to study genetic factors. Risk models should account for both the maternal genotype and the correlated fetal genotype to avoid confounding. The usual logistic regression analysis, however, fails to fully exploit the fact that these are mothers and offspring. Consider an autosomal, diallelic locus, which could be related to disease susceptibility either directly or through linkage with a polymorphic causal locus. Three nested levels of assumptions are often natural and plausible. The first level simply assumes Mendelian inheritance. The second further assumes parental mating symmetry for the studied locus in the source population. The third additionally assumes parental allelic exchangeability. Those assumptions imply certain nonlinear constraints; the authors enforce those constraints by using Poisson regression together with the expectation-maximization algorithm. Calculations reveal that improvements in efficiency over the usual logistic analysis can be substantial, even if only the Mendelian assumption is honored. Benefits are even more marked if, as is typical, information on genotype is missing for some individuals.

An Augmented Data Method for the Analysis of Nosocomial Infection Data

Cooper, B. S., Medley, G. F., Bradley, S. J., Scott, G. M. — 2008-08-22

The analysis of nosocomial infection data for communicable pathogens is complicated by two facts. First, typical pathogens more commonly cause asymptomatic colonization than overt disease, so transmission can be only imperfectly observed through a sequence of surveillance swabs, which themselves have imperfect sensitivity. Any given set of swab results can therefore be consistent with many different patterns of transmission. Second, data are often highly dependent: the colonization status of one patient affects the risk for others, and, in some wards, repeated admissions are common. Here, the authors present a method for analyzing typical nosocomial infection data consisting of results from arbitrarily timed screening swabs that overcomes these problems and enables simultaneous estimation of transmission and importation parameters, duration of colonization, swab sensitivity, and ward- and patient-level covariates. The method accounts for dependencies by using a mechanistic stochastic transmission model, and it allows for uncertainty in the data by imputing the imperfectly observed colonization status of patients over repeated admissions. The approach uses a Markov chain Monte Carlo algorithm, allowing inference within a Bayesian framework. The method is applied to illustrative data from an interrupted time-series study of vancomycin-resistant enterococci transmission in a hematology ward.