S_ary A case-control study of 689 breast cancer patients seen at Tata Memorial Hospital during the period 1980-84 was carried out. During the same period 711 females who attended the hospital without a history of benign breast lesions or gynaecological complaints were selected as controls. Patients were interviewed by trained investigators to collect data on reproductive factors, menstrual history, tobacco smoking and chewing habit, dietary practices (vegetarian and non-vegetarian diet) and alcohol consumption. Cases and controls were stratified into four age groups (<35 years, [35][36][37][38][39][40][41][42][43][44][45][46][47][48][49][50][51][52][53][54] (Bombay, Maharashtra, others). Odds ratios were calculated by univariate methods as well as by stratified analysis. The Mantel and Haenzel (1959) summary chi-squared test was used for testing statistical significance and a test-based estimation procedure was used for calculation of confidence intervals for odds ratios (Kleinbaum et al., 1982).
ResultsGeneral features of breast cancer cases and controls are shown in Table I. The average age of cancer patients was 46.2 years, whereas it was 42.8 years for controls. The religious distribution between cases and controls did not differ and hence is not adjusted for in the analysis. Reproductive factors in cases and controls are presented in Table II. Factors such as age at menarche, age at marriage, age at first pregnancy and number of pregnancies appeared to be similar between the cancer cases and controls.The relative risks (RRs) for factors studied are presented in Table III. Cases and controls were stratified by four age groups and three places of residence. In our study, unmarried women had a 2.3 times higher risk of developing breast cancer than married women. The nulliparous women had 2.2 times the risk of parous women (P<0.OO1). Breast feeding,