TY - JOUR T1 - Life course social roles and women’s health in mid-life: causation or selection? JF - Journal of Epidemiology and Community Health JO - J Epidemiol Community Health SP - 484 LP - 489 DO - 10.1136/jech.2005.042473 VL - 60 IS - 6 AU - Anne McMunn AU - Mel Bartley AU - Rebecca Hardy AU - Diana Kuh Y1 - 2006/06/01 UR - http://jech.bmj.com/content/60/6/484.abstract N2 - Study objective: To investigate whether relations between social roles and health are explained by health selection into employment and parenthood by examining the influence of early health on relations between long term social role histories and health in mid-life. Design: Prospective, population based, birth cohort study. Participants and setting: Women from a national British cohort born in 1946, including 1171 women with a valid measure of self reported health at age 54 and valid work and family role measures at ages 26, 36, 43, and 53, as well as 1433 women with a valid body mass index (BMI) measure at age 53 and valid work and family role measures at ages 26, 36, 43, and 53. Outcome measures: Self reported health at age 54 and obesity at age 53, taken from objective height and weight measures conducted by a survey nurse during face to face interviews in respondents’ homes. Main results: Women who occupied multiple roles over the long term reported relatively good health at age 54 and this was not explained by early health. Women with weak long term ties to the labour market were more likely to be obese at age 53. Examination of body mass index (BMI) from age 15 showed that long term homemakers were larger than other women from age 26, but their mean BMI increased significantly more with age than that of other women. Conclusions: Relations between social roles and health were generally not explained by health selection into employment and parenthood, although some health selection may occur for obesity. ER -