Development of a Multivariable Prognostic PREdiction Model for 1-year Risk of FALLing in Community-dwelling Older Adults in a Non-clinical Setting

Falls in community-dwelling older adults is a frequent problem with an incidence of 30 % in over-65s and 50 % in over-80s. Incidences are expected to increase significantly in the future due to population aging. For instance, as of 2017, the global population older than 65 years is estimated to be 962 million and will increase to 1.4 and 2.1 billion in 2030 and 2050 respectively. In Denmark, falls are the most common accidents among older adults with around 36,000 fall accidents seen annually by the Danish health services and approximately 680 deaths yearly. This high frequency of fall accidents may also support the fact that falls in Denmark are the fourth most common reason for years lived with disability, thereby giving rise to reduced quality of life. Also, falls are associated with elevated morbidity, mortality, poorer physical functioning and early admission to long-term care facilities. Thus, this frequent and escalating problem of fall accidents is of major concern. Fall prevention is therefore highly relevant. It is recognised that fall-preventive strategies should take on a multifaceted approach due the multifactorial aetiology of falls. This is substantiated by more than 400 risk factors of falling that have now been identified. These spread across different domains including socio-demographics, medical conditions (e.g. atrial fibrillation), medication, physical performance (e.g. reduced lower extremity strength or reaction time), psychology (e.g. depression or fear of falling) and cognition (e.g. global cognitive impairment or reduced executive functioning). In order to aid health care professionals in targeting fall-preventive interventions, individual assessments of fall risk are imperative. In Denmark, municipalities are obliged to perform preventive initiatives to preserve the physical, mental and social health along with the functional capacity and quality of life of their older adults. The aim of these initiatives is to enable the older adults to live an independent and meaningful life for as long as possible. Recently, The Danish Health Authority released an updated manual to support this work. This emphasised the need for development of a validated prediction model to be used in a municipally environment to identify older adults at risk of falling. This is due to the abovementioned consequences of falls. To the knowledge of the authors, this is in line with literature being sparse on prognostic prediction models on falls in community-dwelling older adults with data collected outside a clinical environment (i.e. hospitals, GPs and screening or assessment centres). Objectives: Primary: To develop and internally validate a multifactorial prognostic prediction model on fall risk in community-dwelling older adults in a non-clinical setting. The intended use of the model is, for municipalities, to identify and refer citizens with high risk of falls to fall-preventive interventions. Secondary:

1. 1.To estimate time-consumption for the final prediction model.
2. 2.To describe the prevalence of arrhythmias in community-dwelling older adults.