Statistical analysis
Descriptive statistics were calculated for all exposures, outcomes, and
covariates. Differences between study variables by short sleep duration
were reported by using Chi-square statistics for categorical variables
and t-tests for continuous variables. Modified Poisson regression models
with robust error variance28 were used to examine
associations between phenotypes of short sleep duration and health
outcomes. Separate models were conducted for each phenotype of short
sleep and each health outcome, while adjusting for age at evaluation,
sex, race/ethnicity, age at diagnosis, BMI, physical inactivity and
condition-specific treatment exposures. When examining the associations
between phenotypes of short sleep duration and the composite burden
score of physical health outcomes, multinomial logistic regression was
employed. Burden models were adjusted for age at evaluation, sex,
race/ethnicity, BMI, and physical inactivity. All analyses were
completed using SAS v9.4. All tests were two-sided and a p-value ≤ .05
was considered statistically significant.