Abstract
Introduction
Low total Mini-Mental State Examination (MMSE) score might significantly increase risk of hip fractures. This study was to investigate the effects of MMSE subdomains on the risk of hip fractures with a sex and age matched case control study.
Materials & Methods
A total of 217 patients with first low-trauma hip fractures were matched with 215 hospitalised controls. Seven MMSE subdomains were analysed using conditional logistic regression with adjustment of five important clinical confounders: education level, ADL, physical activities, body mass index and bone mineral density. ROC curve analyses were further used to investigate the predictability of the independent subdomains.
Results
In univariate analyses, low score of all MMSE subdomains significantly increased the risk of hip fracture. However, only time orientation and visual construction domain remained significant in multivariate analyses. In the ROC curve analyses, the AUC of these two independent subdomains along with five clinical confounders was significantly larger (p=0.008) than that of clinical confounders alone. The AUC of summed scores of these two subdomains were significantly higher than that of total MMSE score (p=0.009).
Conclusions
Time orientation and visual construction subdomains were significant independent risk factors for hip fractures and could be effectively used to screen the patients with high fracture risk in the hospital. Preventive intervention can be given to these patients.
