The demographic, polysomnographic, anthropometric, metabolic, inflammatory, and BMD characteristics of the patients are shown in . The patients were stratified into three groups according to their AHI. There were no statistical differences in age, height, weight, BP, and BMI between groups. Except for the HDL-C values, significantly lower in the group with severe OSA, no significant differences in metabolic syndrome parameters were observed. Fibrinogen was significantly higher in the severe OSA group than in the moderate and mild OSA groups; CRP values were significantly lower in the moderate OSA group compared with both the mild and severe OSA groups; serum ESR levels were not statistically different between groups.
Descriptive characteristics of subjects with varying severities of obstructive sleep apnea (OSA)
Among the 115 patients, 55.65% had normal BMD (n = 64), 40% had osteopenia (n = 46), and 4.35% had osteoporosis (n = 5). In the group with mild OSA, 15/50 patients were osteopenic (30%) and 3/50 patients were osteoporotic (6%); in the group with moderate OSA, 17/33 patients were osteopenic (51.52%) and 0/33 patients were osteoporotic (0%); in the group with severe OSA, 14/32 patients were osteopenic (43.75%) and 2/32 patients were osteoporotic (6.25%). The BMD values measured at the lumbar spine, total hip, and femoral neck levels were not significantly different among groups. There were no statistically different BMD values between men and women (data not shown).
The sub-cohort of menopausal women (32 patients) showed a lower BMD than the pre-menopausal women (27 patients) at all sites analyzed [L2–L4 BMD, 0.971 (0.89–1.07) g/cm2 vs 1.09 (1.05–1.11) g/cm2, P < 0.05; hip BMD, 0.956 (0.85–1.08) g/cm2 vs 1.03 (0.95–1.1) g/cm2, P < 0.05; femoral neck BMD, 0.77 (0.639–0.907) g/cm2 vs 0.88 (0.8–0.95) g/cm2, P < 0.05].
However, independent from the menopause status, no significant differences in BMD were found when the patients were distributed in the mild, moderate, and severe OSA groups, given that the increasing severity of OSA was not accompanied by a reduction in BMD ().
Lean body mass was significantly higher in men than in women, both in kg [men 65.23 (59.38–70.29) vs women 51.52 (48.19–56.53); P < 0.001] and in percentage [men 63.48 (57.56–68.38) vs women 54.06 (49.66–58.56); P < 0.001]. Moreover, we observed a higher lean mass in severe OSA groups compared to the group with mild OSA ().
shows a linear correlation between the variables. Spearman’s coefficients showed a lack of association between AHI and BMI, body weight, WC, FPG, CRP, triglycerides, and BMD at all bone sites. Indeed, lumbar spine, total hip, and femoral neck BMD were positively correlated to weight (P ≤ 0.001) and inversely correlated to age (P ≤ 0.001). Moreover, there was a correlation between total hip and lumbar BMD and lean mass, but not between femoral neck BMD and lean mass.
Spearman’s correlation coefficient analysis for all the covariates
A significant association of AHI with ESR and fibrinogen was found; ESR correlated with weight, BMI, WC, diastolic blood pressure, and CRP; while fibrinogen associated with weight, WC, CRP, and ESR. CRP did not associate with any of the covariates except BMI. Finally, the lean mass showed a significant positive correlation with AHI, weight, WC, and fibrinogen, and a negative correlation with HDL-C and age.