DDT and Breast Cancer: Prospective Study of Induction Time and Susceptibility Windows

Abstract

Background: In a previous Child Health and Development Studies report, p, p'-DDT was associated with a fivefold increased risk of premenopausal (before age 50 years) breast cancer for women first exposed before puberty. Here we extend our observation to breast cancer diagnosed during early postmenopause (ages 50-54 years) to determine whether age at diagnosis modifies the interaction of DDT with age at exposure.

Methods: We conducted a second prospective, nested case-control study in the Child Health and Development Studies (153 incident breast cancer cases diagnosed at ages 50-54 years and 432 controls matched to cases on birth year). These were analyzed separately and pooled with our previous study (129 breast cancer cases diagnosed at ages 31-49 years and 129 controls matched on birth year). Blood samples were obtained during pregnancy (median age, 26 years), 1-3 days after delivery from 1959 to 1967 in Oakland, California. Serum was assayed for p, p'-DDT, o, p'-DDT, and p, p'-DDE. Odds ratios (ORs) below are given for doubling of serum p, p'-DDT. All statistical tests were two-sided.

Results: For early postmenopausal breast cancer, p, p'-DDT was associated with risk for all women (ORDDT 50-54 = 1.99, 95% CI = 1.48 to 2.67). This association was accounted for by women first exposed to DDT after infancy (ORDDT 50-54 for first exposure after infancy = 2.83, 95% CI = 1.96 to 4.10 vs ORDDT 50-54 for first exposure during infancy = 0.56, 95% CI = 0.26 to 1.19; Pinteraction DDT x age at first exposure = .01). In contrast, for premenopausal breast cancer, p, p'-DDT was associated with risk among women first exposed during infancy through puberty, but not after (ORDDT<50 for first exposure during infancy = 3.70, 95% CI = 1.22 to 11.26, Pinteraction DDT x age at first exposure x age at diagnosis = .03).

Conclusions: p, p'-DDT was associated with breast cancer through age 54 years. Risk depended on timing of first exposure and diagnosis age, suggesting susceptibility windows and an induction period beginning in early life. DDT appears to be an endocrine disruptor with responsive breast targets from in utero to menopause.

Figure 1.

Case and control distributions of serum p.p’-DDT by age at first exposure and age at diagnosis of breast cancer. A) Cases younger than age 50 years. B) Cases from ages 50 to 54 years. For both outcomes, least squares regression lines are depicted for cases (solid lines) and controls (dashed lines).

Figure 2.

p, p’-DDT associated breast cancer by age at first exposure and age at diagnosis. The y-axis corresponds to study birth cohorts. The x-axis corresponds to calendar year. The light grey area indicates years of DDT use showing that all birth cohorts were DDT-exposed but first exposure occurred at different ages. White boxes show age at first exposure for relevant birth cohorts. The black box corresponds to birth cohorts diagnosed with breast cancer before age 50 years who were first exposed to DDT from in utero to age 13 years. The dark grey area corresponds to birth cohorts diagnosed from ages 50-54 years who were first exposed after infancy. Findings support that intrauterine and early infant p, p’-DDT exposure increases risk of premenopausal breast cancer, whereas p, p’-DDT exposure after infancy increases breast cancer risk in the early menopausal years. ORs and 95% CIs were estimated from conditional logistic regression models as described in Table 2.