Skip to main content
. 2021 May 11;19:104. doi: 10.1186/s12916-021-01973-y

Table 2.

Characteristics of the Canadian Longitudinal Study on Aging cohort by baseline sleep apnoea status

Variable Sleep apnoea
N = 3609
Non-sleep apnoea
N = 20,464
OR (95% CI)3
Sex Women 1102 (30.5%) 10,912 (53.3%) 1 [reference]
Men 2507 (69.5%) 9552 (46.7%) 2.60 (2.41–2.80)
Age at recruitment Mean (SD), years 62.08 ± 9.31 62.69 ± 10.16 0.94 (0.91–0.98) per 10-year increase
Ethnicity White 2492 (69.0%) 14,970 (73.2%) 1 [reference]
Asian 34 (0.9%) 201 (1.0%) 1.02 (0.69–1.44)
African 18 (0.5%) 110 (0.5%) 0.98 (0.58–1.58)
Others 1065 (29.5%) 5183 (25.3%) 1.23 (1.14–1.33)
Smoking Current 344 (9.5%) 1608 (7.9%) 1 [reference]
Never 1546 (42.8%) 10,071 (49.2%) 0.72 (0.63–0.82)
Previous 1719 (47.6%) 8785 (42.9%) 0.91 (0.81–1.04)
Social standing1 Mean (SD) 6.18 ± 1.93 6.27 ± 1.80 0.97 (0.95–0.99)
WHR Mean (SD) 0.95 ± 0.09 0.90 ± 0.10 1.84 (1.77–1.92)
Obesity No 652 (18.1%) 8065 (39.5%) 1 [reference]
Yes 2946 (81.9%) 12,355 (60.5%) 2.95 (2.70–3.23)
SBP Mean (SD), mm Hg 129.99 ± 18.60 127.43 ± 19.54 1.07 (1.05–1.09) per 10-mmHg increase
HDL cholesterol Mean (SD), mmol/L 1.34 ± 0.43 1.53 ± 0.48 0.37 (0.34–0.40)
Total cholesterol Mean (SD), mmol/L 4.92 ± 1.13 5.19 ± 1.09 0.80 (0.77–0.82)
Presence of diabetes No 3074 (85.2%) 18,825 (92.0%) 1 [reference]
Yes 535 (14.8%) 1639 (8.0%) 2.00 (1.80–2.22)
Presence of heart disease No 3048 (84.6%) 18,377 (89.9%) 1 [reference]
Yes 556 (15.4%) 2060 (10.1%) 1.63 (1.47–1.80)
Incident AMD2 No 3268 (97.5%) 18,783 (98.0%) 1 [reference]
Yes 83 (2.5%) 374 (2.0%) 1.28 (1.00–1.61)
Incident glaucoma2 No 3312 (97.3%) 18,876 (98.0%) 1 [reference]
Yes 93 (2.7%) 393 (2.0%) 1.35 (1.07–1.69)

AMD, age-related macular degeneration; CI, confidence interval; OR, odds ratio; SBP, systolic blood pressure; SD, standard deviation; WHR, waist-to-hip ratio

1The self-rated social standing ranges between 1 and 10, where higher rating means better socioeconomic status

2AMD and glaucoma cases represent incident cases after excluding prevalent cases

3The ORs were calculated using univariate logistic regression models