Meta-analysis often requires pooling of correlated estimates to compute regression slopes (trends) across different exposure or treatment levels. The authors propose two methods that account for the correlations but require only the summary estimates and marginal data from the studies. These methods provide more efficient estimates of regression slope, more accurate variance estimates, and more valid heterogeneity tests than those previously available. One method also allows estimation of nonlinear trend components, such as quadratic effects. The authors illustrate these methods in a meta-analysis of alcohol use and breast cancer. Am J Epidemiol 1992; 135: 1301–9

We discuss methods for summarizing epidemiologic studies of dose-response. The data from such a study typically appear as a series of dose-specific relative risks, with one category serving as the common reference group. We present methods for estimating the dose-response parameters from single and multiple study reports, for assigning levels to exposure categories when modeling relative risks, and for investigating the effects of study design and subject characteristics on dose-response curves. Finally, we discuss the choice of fixed vs random effects models. We illustrate our points with data from case-control studies of the relation between duration of oral contraceptive use and risk of breast cancer.

BACKGROUND: Obesity may reduce fecundity. We examined the obesity–fecundity association in relation to menstrual cycle regularity, parity, smoking habits and age to gain insight into mechanisms and susceptible subgroups. METHODS: Data were provided by 7327 pregnant women enrolled in the Collaborative Perinatal Project at 12 study centres in the United States from 1959 to 1965. Prepregnancy body mass index (BMI) was analysed continuously and categorically [underweight (<18.5 kg/m²), optimal weight (18.5–24.9 kg/m²), overweight (25.0–29.9 kg/m²) and obese (≥30.0 kg/m²)]. Adjusted fecundability odds ratios (FORs) were estimated using Cox proportional hazards modelling for discrete time data. RESULTS: Fecundity was reduced for overweight [OR = 0.92, 95% confidence interval (95% CI): 0.84, 1.01] and obese (OR = 0.82, 95% CI: 0.72, 0.95) women compared with optimal weight women and was more evident for obese primiparous women (OR = 0.66, 95% CI: 0.49, 0.89). Fecundity remained reduced for overweight and obese women with normal menstrual cycles. Neither smoking habits nor age modified the association. CONCLUSIONS: Obesity was associated with reduced fecundity for all subgroups of women and persisted for women with regular cycles. Our results suggest that weight loss could increase fecundity for overweight and obese women, regardless of menstrual cycle regularity, parity, smoking habits and age.

▪ Abstract Organochlorines are a diverse group of persistent synthetic compounds, some of which are detectable in nearly everyone. Many organochlorines are endocrine disruptors or carcinogens in experimental assays. p,p′-DDE (dichlorodiphenyl- dichloroethene) and PCBs (polychlorinated biphenyls) comprise the bulk of orga- nochlorine residues in human tissues. We reviewed relevant human data cited in the 1991–1995 Medline database and elsewhere. High-level exposure to selected organochlorines appears to cause abnormalities of liver function, skin (chloracne), and the nervous system. Of more general interest, however, is evidence suggesting insidious effects of background exposure. Of particular concern is the finding of neonatal hypotonia or hyporeflexia in relation to PCB exposure. The epidemiologic data reviewed, considered in isolation, provide no convincing evidence that organochlorines cause a large excess number of cancers. A recent risk assessment that considered animal data, however, gives a cancer risk estimate for background exposure to dioxin and dioxin-like compounds (e.g. some PCBs) with an upper bound in the range of 10⁻⁴ per year.

The evidence of an association of lactation with a reduction in the risk of breast cancer among women has been limited and inconsistent. The effect of lactation appears to be confined to premenopausal women with a history of long lactation, but most studies of this relation have been limited in statistical power. We conducted a multicenter, population-based, case-control study with a sample large enough for us to describe more precisely the association between lactation and the risk of breast cancer. Patients less than 75 years old who had breast cancer were identified from statewide tumor registries in Wisconsin, Massachusetts, Maine, and New Hampshire. Controls were randomly selected from lists of licensed drivers if the case subjects were less than 65 years old, and from lists of Medicare beneficiaries if they were 65 through 74 years old. Information on lactation, reproductive history, and family and medical history was obtained by means of telephone interviews. After the exclusion of nulliparous women, 5878 case subjects and 8216 controls remained for analysis. After adjustment for parity, age at first delivery, and other risk factors for breast cancer, lactation was associated with a slight reduction in the risk of breast cancer among premenopausal women, as compared with the risk among women who were parous but had never lactated (relative risk, 0.78; 95 percent confidence interval, 0.66 to 0.91); the relative risk of breast cancer among postmenopausal women who had lactated, as compared with those who had not, was 1.04 (95 percent confidence interval, 0.95 to 1.14). With an increasing cumulative duration of lactation, there was a decreasing risk of breast cancer among premenopausal women (P for trend <0.001) but not among postmenopausal, parous women (P for trend = 0.51). A younger age at first lactation was significantly associated with a reduction in the risk of premenopausal breast cancer (P for trend = 0.003). As compared with parous women who did not lactate, the relative risk of breast cancer among women who first lactated at less than 20 years of age and breast-fed their infants for a total of six months was 0.54 (95 percent confidence interval, 0.36 to 0.82). There is a reduction in the risk of breast cancer among premenopausal women who have lactated. No reduction in the risk of breast cancer occurred among postmenopausal women with a history of lactation.