Modelling handgrip strength in the presence of confounding variables: results from the Allied Dunbar National Fitness Survey

Abstract

Differences in handgrip strength, caused by risk factors such as physical inactivity, will be influenced by ‘confounding’ variables, e.g. age, body size. The aims of the study were to identify the confounding variables associated with handgrip strength and to assess the benefit that physical activity plays in maintaining grip strength within a population, having adjusted for differences in these confounding variables. The most appropriate linear body size dimension associated with grip strength was height rather than demispan. Non-linear associations with age and body mass were also identified. Handgrip strength peaked in the age group 25 – 34 years for male subjects and in the age group 35 44 years for female subjects. Similarly, handgrip strength increased with body mass until it peaked at a body mass of approximately 100 kg for male and 90 kg for female subjects; thereafter a rapid decline in grip strength was observed. Differences in handgrip strength were found to be significantly associated with levels of physical activity even having controlled for differences in age and body size (height, mass and percentage body fat), but the observed association was not linear. The level of physical activity necessary to maintain an optimal level of handgrip strength was found to be a balance of moderate or vigorous occasions of physical activity.

Keywords:

• Physical Activity Allometric Models Heteroscedastic Errors