LAUSR

Women may enter in breast cancer (BCa) treatment with poor sleep, or it may begin during treatment. We assessed how subjective and objective sleep changes during the first year of treatment for women with BCa. Further, we examined whether this differs between previously good and poor sleepers and whether there was agreement between subjective and objective measures of sleep. Sleep onset latency (SOL), wake after sleep onset (WASO), total sleep time (TST), and sleep efficiency (SE) were measured among 100 patients with newly diagnosed, non-metastatic BCa using 7 days of diary and actigraphy collected at 4 time points: pre-treatment, 4, 8, and 12 months. Women with a score ≥5 on the Pittsburgh Sleep Quality Index at treatment onset were classified as poor sleepers. A 4 (time: 0-, 4-, 8-, 12-months) by 2 (sleep measure: sleep diary, actigraphy) by 2 (group: good, poor sleepers) mixed model ANOVAs was performed for each sleep parameter. There was a time by sleep measure by group interaction for TST, [F(3,294)= 3.014, p = .03). Good sleepers reported greater TST on diaries- than actigraphy at pre-treatment and 12 months, whereas there were no differences in poor sleepers. There was a group by time effect for good vs. poor sleepers [F(3,294)= 2.909, p = .035]. Good sleepers experienced decreased TST and SE from pre-treatment through 4-mo, followed by increases. Poor sleepers showed the opposite pattern. Neither group returned to pre-treatment levels. Sleep diaries and actigraphy are concordant over time for TST, but not SOL, WASO, or SE. Sleep parameters worsen during the first year following onset of BCa and concordance between sleep diaries and actigraphy differ between good or poor sleepers. Dr. Garland is supported by a Scotiabank New Investigator Award and seed funding from the Beatrice Hunter Cancer Research Institute (BHCRI).