Quadriceps strength, quadriceps power, and gait speed in older U.S. adults with diabetes mellitus: results from the National Health and Nutrition Examination Survey, 1999-2002

Abstract

Objectives: To examine the independent association between diabetes mellitus (and its duration and severity) and quadriceps strength, quadriceps power, and gait speed in a national population of older adults.

Design: Cross-sectional nationally representative survey.

Setting: United States.

Participants: Two thousand five hundred seventy-three adults aged 50 and older in the National Health and Nutrition Examination Survey 1999-2002 who had assessment of quadriceps strength.

Methods: Diabetes mellitus was ascertained according to questionnaire. Measurement of isokinetic knee extensor (quadriceps) strength was performed at 60º/s. Gait speed was assessed using a 20-foot walk test. Multiple linear regression analyses were used to assess the association between diabetes mellitus status and outcomes, adjusting for potential confounders or mediators.

Results: Older U.S. adults with diabetes mellitus had significantly slower gait speed (0.96 ± 0.02 m/s) than those without (1.08 ± 0.01 m/s; P < .001). After adjusting for demographic characteristics, weight, and height, diabetes mellitus was also associated with significantly lower quadriceps strength (-4.6 ± 1.9 Nm; P = .02) and power (-4.9 ± 2.0 W; P = .02) and slower gait speed (-0.05 ± 0.02 m/s; P = .002). Associations remained significant after adjusting for physical activity and C-reactive protein. After accounting for comorbidities (cardiovascular disease, peripheral neuropathy, amputation, cancer, arthritis, fracture, chronic obstructive pulmonary disease), diabetes mellitus was independently associated only with gait speed (-0.04 ± 0.02 m/s; P = .02). Diabetes mellitus duration in men and women was negatively associated with age-adjusted quadriceps strength (-5.7 and -3.5 Nm/decade of diabetes mellitus, respectively) and power (-6.1 and -3.8 W/decade of diabetes mellitus, respectively) (all P ≤ .001, no significant interactions according to sex). Glycosylated hemoglobin was not associated with outcomes after accounting for body weight.

Conclusion: Older U.S. adults with diabetes mellitus have lower quadriceps strength and quadriceps power that is related to the presence of comorbidities and walk slower than those without diabetes mellitus. Future studies should investigate the relationship between hyperglycemia and subsequent declines in leg muscle function.

Figure 1. Association of diabetes duration with age-adjusted quadriceps muscle strength and power and gait speed in older U.S. adults

The association of self-reported diabetes duration (years) with age-adjusted quadriceps strength, quadriceps power, and gait speed in a nationally representative population of older U.S. men (Figures 1A, 1C, 1E; n=180) and women (Figures 1B, 1D, 1F; n=138) is plotted for each participant with diabetes and represented by an open circle. A fitted regression line is shown in each figure with 95% confidence intervals represented as grey shading around the line. The negative association of years living with diabetes and age-adjusted quadriceps strength (p≤0.001) and quadriceps power (p≤0.001) are significant for both men and women.

Figure 1. Association of diabetes duration with age-adjusted quadriceps muscle strength and power and gait speed in older U.S. adults

The association of self-reported diabetes duration (years) with age-adjusted quadriceps strength, quadriceps power, and gait speed in a nationally representative population of older U.S. men (Figures 1A, 1C, 1E; n=180) and women (Figures 1B, 1D, 1F; n=138) is plotted for each participant with diabetes and represented by an open circle. A fitted regression line is shown in each figure with 95% confidence intervals represented as grey shading around the line. The negative association of years living with diabetes and age-adjusted quadriceps strength (p≤0.001) and quadriceps power (p≤0.001) are significant for both men and women.