Predictors of longitudinal change in sleep-disordered breathing in a nonclinic population

To quantify and identify the determinants of the 5-year change in the respiratory disturbance index (RDI).

Longitudinal cohort study (Cleveland Family Study). Multivariate analyses were used to quantify baseline RDI and RDI change.

Community-based study.

486 cohort members (62% from families of probands with sleep-disordered breathing [SDB])--mean age 31.6 +/- 17.9 (SD) years, 60% female, 21% Black ethnicity--who underwent 2 assessments over 5.3 +/- 0.9 (SD) years.

NA.

The RDI was measured twice over approximately 5 years with in-home monitoring. Symptoms and medical histories were obtained from standardized questionnaires, and weight, height, and blood pressure were measured.

The prevalence of SDB (defined by a RDI > or = 15), increased from 13.7% to 23.4% (P < 0.01) in men and from 8.3% to 11.4% (P = 0.13) in females. Baseline and follow-up mean RDIs were 6.0 +/- 10.0 and 8.6 +/- 14.3; both were higher in older individuals, in men, and in those with a higher body mass index. Median 5-year change in RDI varied nonlinearly with age (-0.1, 1.1, 2.3, and 0.9, for those < 18, 19-40, 41-54, and > or = 55 years, respectively) and obesity (2.8 vs -0.1, for the top versus lowest body mass index quartile). The effects of changing weight and aging varied in population subgroups. At any given age and weight, the RDI increased less in women.

Longitudinal change in the RDI varies nonuniformly with age, sex, and weight. Older heavier men may experience the highest rate of increase in RDI over time and, thus, may benefit most from prospective monitoring.