Recreational Physical Activity and Outcomes After Breast Cancer in Women at High Familial Risk

• Published in: JNCI cancer spectrum Vol. 5; no. 6
• Main Authors: Kehm, Rebecca D, MacInnis, Robert J, John, Esther M, Liao, Yuyan, Kurian, Allison W, Genkinger, Jeanine M, Knight, Julia A, Colonna, Sarah V, Chung, Wendy K, Milne, Roger, Zeinomar, Nur, Dite, Gillian S, Southey, Melissa C, Giles, Graham G, McLachlan, Sue-Anne, Whitaker, Kristen D, Friedlander, Michael L, Weideman, Prue C, Glendon, Gord, Nesci, Stephanie, Phillips, Kelly-Anne, Andrulis, Irene L, Buys, Saundra S, Daly, Mary B, Hopper, John L, Terry, Mary Beth
• Format: Journal Article
• Language: English
• Published: England Oxford University Press 01.12.2021
• Subjects: Adult; Age Factors; Aged; Breast Neoplasms - genetics; Breast Neoplasms - mortality; Cause of Death; Editor's Choice; Exercise - statistics & numerical data; Female; Follow-Up Studies; Genes, BRCA1; Genes, BRCA2; Genetic Predisposition to Disease; Humans; Middle Aged; Neoplasm Recurrence, Local - epidemiology; Neoplasm Recurrence, Local - genetics; Neoplasms, Second Primary - epidemiology; Neoplasms, Second Primary - genetics; Proportional Hazards Models; Recreation Therapy - statistics & numerical data; Time Factors; Young Adult
• ISSN: 2515-5091; 2515-5091
• DOI: 10.1093/jncics/pkab090

Recreational physical activity (RPA) is associated with improved survival after breast cancer (BC) in average-risk women, but evidence is limited for women who are at increased familial risk because of a BC family history or and pathogenic variants ( PVs). We estimated associations of RPA (self-reported average hours per week within 3 years of BC diagnosis) with all-cause mortality and second BC events (recurrence or new primary) after first invasive BC in women in the Prospective Family Study Cohort (n = 4610, diagnosed 1993-2011, aged 22-79 years at diagnosis). We fitted Cox proportional hazards regression models adjusted for age at diagnosis, demographics, and lifestyle factors. We tested for multiplicative interactions (Wald test statistic for cross-product terms) and additive interactions (relative excess risk due to interaction) by age at diagnosis, body mass index, estrogen receptor status, stage at diagnosis, / PVs, and familial risk score estimated from multigenerational pedigree data. Statistical tests were 2-sided. We observed 1212 deaths and 473 second BC events over a median follow-up from study enrollment of 11.0 and 10.5 years, respectively. After adjusting for covariates, RPA (any vs none) was associated with lower all-cause mortality of 16.1% (95% confidence interval [CI] = 2.4% to 27.9%) overall, 11.8% (95% CI = -3.6% to 24.9%) in women without PVs, and 47.5% (95% CI = 17.4% to 66.6%) in women with / PVs (RPA* multiplicative interaction = .005; relative excess risk due to interaction = 0.87, 95% CI = 0.01 to 1.74). RPA was not associated with risk of second BC events. Findings support that RPA is associated with lower all-cause mortality in women with BC, particularly in women with / PVs.