Estimating model-adjusted risks, risk differences, and risk ratios from complex survey data.

There is increasing interest in estimating and drawing inferences about risk or prevalence ratios and differences instead of odds ratios in the regression setting. Recent publications have shown how the GENMOD procedure in SAS (SAS Institute Inc., Cary, North Carolina) can be used to estimate these parameters in non-population-based studies.

Occupation and breast cancer in women 20-44 years of age (United States).

Objective: To examine the relation between breast cancer risk and job history among women 20-44 years of age who participated in a multi-center, population-based, case-control study.

Methods: Participants consisted of women newly diagnosed with breast cancer (1642) and controls identified by random-digit dialing (1494). Details about the three longest jobs were collected and coded by an industrial hygienist.

Health and professional characteristics of lesbian and heterosexual women physicians.

Objective: To determine whether lesbian and heterosexual female physicians differ on health, professional, and demographic characteristics.

Methods: The Women Physicians' Health Study (WPHS), a mailed questionnaire sample survey, was completed by 4501 women physicians (59% response rate), of whom 115 were identified as lesbians and 4177 were identified as heterosexuals. Lesbian and heterosexual were defined by response to items about self-identification and sexual behavior.

Risk of breast cancer classified by joint estrogen receptor and progesterone receptor status among women 20-44 years of age.

To gain insight into whether breast cancer tumors jointly classified by estrogen receptor (ER) and progesterone receptor (PR) status represent diseases with differing etiologies, data from a population-based case-control study of US women 20-44 years of age were analyzed. Cases included 1,556 women diagnosed between 1990 and 1992. Age- and geographic-frequency-matched controls included 1,397 women identified by random digit dialing.

Exploration of threshold analysis in the relation between stressful life events and preterm delivery.

Biologic evidence suggests that the hormones activated by stress affect gestational length, but the results of epidemiologic investigations are inconsistent. The authors of this paper know of no threshold models that have been studied; these models assume that stress does not affect preterm delivery until a certain amount of stress has been experienced but that each unit of stress above the threshold adds to the risk of preterm delivery. By using standard logistic regression, the authors compared threshold and nonthreshold models of the relation between number of stressful life events and preterm delivery in 11 US states.