Antipsychotic Use and the Risk of Hip Fracture Among Community-Dwelling Persons With Alzheimer's Disease

Objective: To study whether antipsychotic use is associated with a risk of hip fracture among individuals with Alzheimer's disease and to compare the risk according to the duration of use and the 2 most frequently used antipsychotics. Methods: The MEDALZ (Medication and Alzheimer's disease) cohort consisted of community-dwelling Finnish persons with clinically verified diagnoses of Alzheimer's disease, including 70,718 persons newly diagnosed according to NINCDS-ADRDA and DSM-IV criteria between 2005 and 2011. Antipsychotic use was modeled from prescription register data, and hip fractures (ICD-10 S72.0-72.2) were identified from the Hospital Discharge Register. The incidence of hip fractures was compared between new users and nonusers of antipsychotics, among various time durations of antipsychotic use, and between quetiapine users and risperidone users. Results: Antipsychotic use versus nonuse was associated with an increased risk of hip fractures (adjusted hazard ratio [HR] = 1.54; 95% CI, 1.39-1.70). The risk was increased from the first days of use and remained increased thereafter. Quetiapine was associated with a similar risk of hip fracture as risperidone for the first 2.7 years of use (adjusted HR = 0.98; 95% CI, 0.79-1.21). Compared with low-dose (<= 0.5 mg) risperidone use, higher risperidone doses (> 0.5 mg) were associated with a higher risk of hip fracture (adjusted HR = 1.72; 95% CI, 1.32-2.24). Conclusions: Since the risk of hip fracture was increased from the first days of use, our results confirm the need for setting a high threshold for initiating antipsychotic use among persons with Alzheimer's disease to avoid serious adverse events. If antipsychotic use is initiated, the duration of use should be limited, as the risk of hip fracture does not attenuate with long-term use.