Gait speed (GS) is an important predictor of disability, falls, institutionalization and death among elderly people. Our aim was to assess which factors are associated with higher GS in a sample of physically active elderly.

A cross-sectional study was performed in a sample of 43 self-sufficient and active elderly (12 men and 31 women) aged 65–82 years. Anthropometric features (weight and height), mobility (Tinetti test), physical and mental health (Short Form 12 - SF12 questionnaire), physical activity (Physical Activity Scale for Elderly - PASE questionnaire), strength and power of lower limbs (Myotest Pro accelerometer) and GS were measured. A multivariable linear regression model was built in order to identify which variables were associated with higher GS.

The final multivariable linear regression model included gender, fall in the previous year, hypertension, age, BMI and Mental Health score. Furthermore, it explained nearly 2/3 of the variability in GS (R²=0.64). Male sex, hypertension and a higher Mental Health score were associated with higher GS, whereas fall in the previous year and higher values of both age and BMI were associated with lower GS. A further model which included an interaction between sex and BMI (R²=0.68) revealed that the negative association between GS and BMI was found among women but not among men.

In addition to the well-known associations between GS and demographic and anthropometric characteristics, we reported a positive association between mental health and GS among generally healthy and physically active community-dwelling elderly.