Recreational Physical Activity Is Associated with Reduced Breast Cancer Risk in Adult Women at High Risk for Breast Cancer: A Cohort Study of Women Selected for Familial and Genetic Risk

Abstract

Although physical activity is associated with lower breast cancer risk for average-risk women, it is not known if this association applies to women at high familial/genetic risk. We examined the association of recreational physical activity (self-reported by questionnaire) with breast cancer risk using the Prospective Family Study Cohort, which is enriched with women who have a breast cancer family history (N = 15,550). We examined associations of adult and adolescent recreational physical activity (quintiles of age-adjusted total metabolic equivalents per week) with breast cancer risk using multivariable Cox proportional hazards regression, adjusted for demographics, lifestyle factors, and body mass index. We tested for multiplicative interactions of physical activity with predicted absolute breast cancer familial risk based on pedigree data and with BRCA1 and BRCA2 mutation status. Baseline recreational physical activity level in the highest four quintiles compared with the lowest quintile was associated with a 20% lower breast cancer risk (HR, 0.80; 95% confidence interval, 0.68-0.93). The association was not modified by familial risk or BRCA mutation status (P interactions >0.05). No overall association was found for adolescent recreational physical activity. Recreational physical activity in adulthood may lower breast cancer risk for women across the spectrum of familial risk. SIGNIFICANCE: These findings suggest that physical activity might reduce breast cancer risk by about 20% for women across the risk continuum, including women at higher-than-average risk due to their family history or genetic susceptibility.See related commentary by Niehoff et al., p. 23.

Figure 1.. Joint association of baseline and adolescent recreational physical activity with breast cancer risk in the Prospective Family Study Cohort (N=14,619)

N=14,619 (cases=852) because 931 participants were missing data on recreational physical activity during adolescence. Hazard ratios (HRs) and 95% CIs (confidence intervals) are adjusted for race/ethnicity, study center, lifetime familial BC risk, education, parity and breastfeeding, and use of alcohol, cigarettes, hormonal birth control and menopausal hormone therapy, and body mass index; stratified by birth cohort. Reference group = inactive (defined as lowest quintile (Q1) of age-adjusted recreational physical activity) in adolescence and baseline. Active is defined as highest four quintiles (Q2–Q5) of age-adjusted recreational physical activity for given exposure period. Refer to eTable 1 in the supplemental materials for the minimum number of metabolic equivalents (METs) required to be classified as active for a given age at baseline.

Figure 2.. Association of baseline recreational physical activity with breast cancer risk by baseline characteristics of the Prospective Family Study Cohort (N=15,550)

GED=general education degree. Reference group = inactive at baseline, defined as lowest quintile (Q1) of age-adjusted recreational physical activity. Active is defined as highest four quintiles (Q2–Q5) of age-adjusted recreational physical activity. Refer to eTable 1 in the supplemental materials for the minimum number of metabolic equivalents (METs) required to be classified as active for a given age at baseline. Hazard ratios (HRs) and 95% confidence intervals (CI) are adjusted for race/ethnicity, study center, lifetime familial breast cancer risk, education, parity and breastfeeding, and use of alcohol, cigarettes, hormonal birth control and menopausal hormone therapy, and body mass index; stratified by birth cohort. The interaction p-value is the Wald test statistic used for the interaction term between physical activity and the baseline characteristic predicting breast cancer risk.

Figure 3.. Adolescent recreational physical activity and breast cancer risk by percentiles of full lifetime familial breast cancer risk estimated by the Breast and Ovarian Analysis of Disease Incidence and Carrier Estimation Algorithm (BOADICEA) in the Prospective Family Study Cohort (N=14,619)

Point estimates reflect hazard ratios comparing active (defined as highest four quintiles (Q2–Q5) of age-adjusted recreational physical activity) to inactive (defined as lowest quintile (Q1) of age-adjusted recreational physical activity) in adolescence by percentiles of full lifetime risk (interaction p-value = 0.03). The percent full lifetime familial breast cancer risk corresponding to each percentile is provided below the X axis. Refer to eTable 1 in the supplemental materials for the minimum number of metabolic equivalents (METs) required to be classified as active for a given age at baseline. Hazard ratios and 95% confidence intervals are adjusted for race/ethnicity, study center, education, parity and breastfeeding, use of alcohol, cigarettes, hormonal birth control and menopausal hormone therapy, and body mass index; stratified by birth cohort.

Figure 4.. Associations of baseline and adolescent recreational physical activity with breast cancer risk by BRCA1 and BRCA2 mutation carrier status in the Prospective Family Study Cohort

Point estimates reflect hazard ratios comparing active (defined as highest four quintiles (Q2–Q5) of age-adjusted recreational physical activity) to inactive (defined as lowest quintile (Q1) of age-adjusted recreational physical activity) for the given exposure period. Refer to eTable 1 in the supplemental materials for the minimum number of metabolic equivalents (METs) required to be classified as active for a given age at baseline. Hazard ratios and 95% confidence intervals are adjusted for race/ethnicity, study center, education, parity and breastfeeding, use of alcohol, cigarettes, hormonal birth control and menopausal hormone therapy, and body mass index; stratified by birth cohort. The interaction p-value is the Wald test statistic used for the interaction term between physical activity and mutation carrier status. Non-carriers include women who were tested and not known to carry pathogenic mutations, as well as women who did not undergo genetic testing. The combined cohort includes prevalent breast cancer cases diagnosed within 5 years prior to study enrollment.