Age-dependent changes in the association between sleep duration and impaired glucose metabolism

Table 4 Odds ratios (95%CIs) of sleep duration for prediabetes and diabetes in older subjects n (%)

Sleep duration	≤ 5 h	6 h	7 h	≥ 8 h	P value
Prediabetes
Cases	1259 (32.8)	6006 (32.8)	4788 (33.9)	2343 (37.4)
Model 1	0.95 (0.88-1.03)	0.95 (0.91-0.995)a	1 (reference)	1.16 (1.09-1.24)d	1.07 (1.05-1.10)d
Model 2	1.17 (1.09-1.27)d	1.08 (1.03-1.14)b	1 (reference)	0.97 (0.91-1.03)	0.94 (0.91-0.96)d
Model 3	1.12 (1.03-1.21)b	1.04 (0.99-1.11)	1 (reference)	0.97 (0.91-1.04)	0.96 (0.94-0.99)b
Diabetes
Cases	271 (7.1)	1379 (7.5)	1181 (8.4)	702 (11.2)
Model 1	0.83 (0.73-0.95)b	0.89 (0.82-0.97)b	1 (reference)	1.38 (1.25-1.52)d	1.19 (1.15-1.24)d
Model 2	1.10 (0.95-1.26)	1.06 (0.98-1.15)	1 (reference)	1.11 (1.00-1.22)	1.00 (0.96-1.04)
Model 3	1.02 (0.88-1.18)	1.01 (0.93-1.11)	1 (reference)	1.11 (1.02-1.25)a	1.03 (0.99-1.08)

Model 1: Unadjusted; Model 2: Adjusted for age and sex; Model 3: Model 2 plus adjustment for current smoking (vs non-smoking), daily alcohol consumption (vs infrequent/no alcohol consumption), regular exercise (vs no regular exercise), pharmacotherapy (hypertension, diabetes and dyslipidemia), BMI, WBC count, systolic blood pressure, triglycerides, HDL-cholesterol, duration of work (as continuous variables), and past history of CVD (vs no history). Triglyceride concentrations were log transformed before analysis. Regular exercise was defined as ≥ 30 min exercise per session at least twice a week.

^aP < 0.05,

^bP < 0.01,

^dP < 0.001. BMI: Body mass index; HDL: High-density lipoprotein; CVD: Cardiovascular disease; WBC: White blood cell.