Send to

Choose Destination
J Am Med Dir Assoc. 2014 Jan;15(1):30-5. doi: 10.1016/j.jamda.2013.11.003.

Low testosterone levels, depressive symptoms, and falls in older men: a cross-sectional study.

Author information

Department of Healthcare Epidemiology, Graduate School of Medicine and Public Health, Kyoto University, Kyoto, Japan.
Department of Urology, Juntendo University School of Medicine, Tokyo, Japan; Department of Urology, Teikyo University School of Medicine, Tokyo, Japan.
Department of Orthopedic Surgery, Fukushima Medical University School of Medicine, Fukushima, Japan.
Institute for Health Outcomes and Process Evaluation Research (i-Hope International), Kyoto, Japan.
Department of Preventive Medicine and Epidemiologic Informatics, National Cerebral and Cardiovascular Center, Osaka, Japan.
Department of Community Health Sciences, Kobe University Graduate School of Health Sciences, Kobe, Japan.
Department of Healthcare Epidemiology, Graduate School of Medicine and Public Health, Kyoto University, Kyoto, Japan; Center for Innovative Research in Clinical Evaluative Science (CiRCLE), Fukushima Medical University, Fukushima, Japan. Electronic address:



While several studies have cited a potential association between testosterone deficiency and risk of falls among community-dwelling older men, evidence for such an association is conflicting. Depressive symptoms, which occasionally accompany testosterone deficiency but which are often neglected as associated symptoms, may actually provoke falls independent of or jointly with testosterone deficiency. We examined the association between testosterone levels, depressive symptoms, and falls, and assessed the joint effect of testosterone levels and depressive symptoms on falls among older men.


Data for this cross-sectional study were obtained from 869 men aged over 60 years who participated in health check-ups conducted in 2010 from 2 Japanese municipalities. Salivary testosterone (sT) levels were measured using an enzyme-linked immunosorbent assay, and depressive symptoms were assessed via the short form of the Center for Epidemiologic Studies Depression Scale.


Self-reported "any fall" over the 1-month period.


Among the total of 482 participants analyzed (median age, 70 years), 10.8% reported any fall. On comparison between 90th percentile sT levels and lower levels, our logistic regression model with restricted cubic splines showed that lower sT levels were associated with an increased likelihood of suffering any fall after adjustment for sociodemographic characteristics, comorbidities, and mobility function. For example, 5th percentile sT was associated with any fall [adjusted odds ratio (OR), 4.23; 95% confidence interval (CI), 1.66-10.8]. Depressive symptoms were also strongly associated with any fall [adjusted OR, 3.49 (95% CI, 1.52-8.04)]. We noted no apparent interaction of sT and depressive symptoms with falls (P = .079), suggesting that the joint effect of testosterone deficiency and depressive symptoms on falls was multiplicative. Indeed, compared with a combination of 90th percentile sT values and no depressive symptoms, adjusted OR for any fall in a combination involving 5th percentile sT and depressive symptoms was 14.8-fold (95% CI, 3.76-58.0).


Our findings indicated that both relatively low testosterone levels and presence of depressive symptoms were independently associated with falls among older men. Causality of these associations should be confirmed in future prospective studies.


Testosterone; depression; fall

[Indexed for MEDLINE]
Free full text

Supplemental Content

Full text links

Icon for Elsevier Science Icon for Kyoto University Research Information Repository - Articles
Loading ...
Support Center