Annelies Smit¹, Trishika Binda², Diana van Heemst², Raymond Noordam², Elizabeth Winter³


(1) Department of Medicine, Division of Endocrinology & Einthoven Laboratory, Leiden University Medical Center, Leiden, The Netherlands. (2) Department of Internal Medicine, Section of Gerontology and Geriatrics, Leiden University Medical Center, Leiden, The Netherlands. (3) Department of Medicine, Division of Endocrinology, Einthoven Laboratory & Center for Bone Quality, Leiden University Medical Center, Leiden, Netherlands. 

Purpose:

Experimental studies suggest a link between sleep behavior and osteoporosis (OP), but epidemiological data are inconclusive. We explored associations and potential causality between sleep traits with OP incidence, as well as bone mineral density (BMD) in population-based cohorts.

Methods:

In the UK Biobank, we assessed associations between self-reported sleep behaviors and OP incidence using multi-variable adjusted Cox regression analysis among 402,570 individuals without prior OP (45% men, mean age 56 years (SD 8), mean follow-up 13.1 years (IQR 12.8-14.4)). Using linear regression, we analyzed associations between sleep traits with lumbar spine and femoral neck BMD T-scores of 41,557 individuals (48% men, mean age 64 years (SD 8)). Mendelian randomization evaluated potential causality between sleep traits and OP in the Finngen cohort (3,203 cases; 209,575 controls) and UK Biobank (1,976 cases; 461,034 controls), and between sleep traits and BMD in the GEFOS consortium (n=32,375 femoral neck; n=28,498 lumbar spine).

Results:

All sleep traits—sleeping 6 hours or less (HR 1.21, 95%CI[1.17,1.26]), sleeping over 9 hours (HR 1.19, 95%CI[1.13,1.26]), insomnia symptoms (HR 1.25, 95%CI[1.20,1.29]), daytime dozing (HR 1.47, 95%CI[1.35,1.61]), and evening chronotype (HR 1.05, 95%CI[1.02,1.09])—were associated with increased OP incidence. Conversely, only evening chronotype associated with lower femoral neck T-score (-0.03 95%CI[-0.05,-0.01]). Notably, individuals with multiple poor sleep traits (PS) had higher OP risk (1 PS: HR 1.13 95%CI[1.06,1.20]; 3 PS: HR 1.50 95%CI[1.39,1.61]), as well as lower T-scores of femoral neck (1 PS: -0.03 95%CI[-0.06,0.01]; 3 PS: -0.08 95%CI[-0.12,-0.04]) and lumbar spine (1 PS: -0.05[-0.10,0.07]; 3 PS: -0.08 [-0.15,-0.02]). However, Mendelian randomization did not support causality between sleep traits with OP risk, nor with BMD.

Conclusion:

All sleep traits were associated with OP risk, thus screening for sleep disturbances may identify individuals at risk. As we found no evidence for causality, mediating factors should be further explored.