Time trend and seasonal pattern of fall incidence among Chinese older population in Hong Kong 2005-2014

Abstract

Background and Objectives: It remains unclear about the time trend in fall incidence among older population in Hong Kong. The seasonal pattern of fall reported in existing local studies need to be further confirmed. To examine time trend and seasonal pattern of fall from 2005 to 2014 among Hong Kong community-dwelling Chinese adults aged 65 years and over (seniors). Methods: A large cohort of Hong Kong seniors first applying for long-term care services from 2005 to 2014 was obtained for secondary data analysis. The incidence of fall was calculated based on the number of falls reported by seniors within past 90 days. Age-sex-standardized fall incidence was calculated, and logistic regression model was used to examine both time trend and seasonal variation in fall occurrence, while controlling for gender, age and all two-way interaction terms. Results: About 32% of this cohort of 89,100 seniors experienced at least one fall in 90 days prior to interview. The age-sex-standardized incidence of fall (per person-years) among the seniors in Hong Kong decreased from 2.40 in 2005 to 1.97 in 2014. In the logistic regression model, there was an annual reduction of 2.3% (95% CI 1.8%-2.8%) in 90-day fall risk, and the seasonal variation was indicated by the highest odds ratio (1.35, 95% CI 1.25-1.45) within 90-days prior to February (i.e. falls during November to February), compared with lowest one within 90-days prior to July. There was no significant age or gender difference in annual declined rate and seasonality of fall occurrence. Conclusions: There was a downward trend in fall incidence among older Hong Kong Chinese from 2005 to 2014. For seasonal variation, the fall risk peaked during winter months. Further interventions need to address potentially modifiable factors, such as implementing educational publicity and sending reminder to community-dwelling seniors in the fall season to prevent falls.