“…Additionally, since the women’s chronotype (originally assessed in 5 categories: definite or more likely morning type, definitive or more likely evening type, and neither, which we grouped into 3 categories of morning type, evening type, and neither) was only reported once (2009), it was considered as a time-invariant factor. Time-varying factors were updated throughout follow-up and included physical activity (originally calculated as a continuous variable; we grouped it into previously used categories 39 of <3, 3 to <9, 9 to <18, 18 to <27, 27+ metabolic equivalent hours/week; assessed in 2001, 2005, 2009, and 2013), alcohol use (originally calculated as a continuous variable; we grouped it into previously used categories 39 of 0, 1 to <5, 5 to <10, 10 to <20, and 20+ g/day; assessed in 1999, 2003, 2007, 2011, and 2015), smoking status (originally categorized into 15 groups; we regrouped it into previously used categories 39 of never smokers, former smokers, or current smokers of 1–14, 15–24, or 25+ cigarettes/day; assessed biennially between 2001 and 2017), and menopausal status (originally categorized into 4 groups). We regrouped it into categories of premenopausal or postmenopausal (assessed biennially between 2001 and 2017), shiftwork history (we grouped it into categories of never, <5, 5 to <10, and 10+ years; assessed in 2001, 2005, 2009, 2011, and 2013), diagnosis of sleep apnea (2013 and 2017), diet quality based using Dietary Approaches to Stop Hypertension 40,41 (assessed in 1999, 2003, 2007, and 2011), and body mass index (kg/m 2 ; using updated weight, which was assessed biennially between 2001 and 2017).…”