Molecular predictors of hand strength

Abstract

Purpose Grip strength predicts disability and mortality, yet few biomarkers of hand strength have been identified. We surveyed a wide array of serum cytokines, metabolites and joint tissue markers to identify novel independent predictors of grip and pinch strength. Methods Subjects were originally recruited for a longitudinal study of radiographic knee osteoarthritis (OA) progression. At three year follow up, measurement of grip strength (Jamar Hydraulic Hand Dynamometer) and key pinch strength (Jamar Hydraulic Pinch Gauge) were performed twice on each hand, and the greatest value was used for analyses. To control for hand OA, bony enlargements of the four distal and five proximal interphalangeal joints, the five metacarpophalangeal joints, and the carpometacarpal joint were summed (range 0-15) for the stronger hand. Blood (2 hour postprandial) was collected, and serum stored at -80oC. Cytokines and inflammatory markers (adiponectin; brain-derived neurotrophic factor; d dimer; granulocyte colony stimulating factor; interleukins 1, 6, and 8; regulated upon activation normal T cell expressed and secreted; tumor necrosis factor alpha, TNFα; TNFα receptors 1 and 2; TNF related apoptosis-inducing ligand, TRAIL; soluble vascular cell adhesion molecule 1; and vascular endothelial growth factor) and joint tissue markers (C2C; CPII; cartilage oligomeric matrix protein; hyaluronic acid, HA; and matrix metalloproteinase 3) were measured by immunoassay. Acylcarnitines (ACs, n=45) and amino acids (AAs, n=15) were measured using a targeted mass-spectrometry platform. Additional metabolites (glucose; α-hydroxybutyrate; total ketones; lactate; non-esterified fatty acids; and uric acid) were measured with a Hitachi 911 analyzer. Computed variables included global arginine bio-availability ratio (GABR, arginine/(ornithine + citrulline) and C2C:CPII. All variables except age, HA, MCP-1, and TRAIL were log-transformed because of non-normal distributions. Principal components analysis, with varimax rotation, reduced separately the AAs and ACs each to a 1 factor solution. Multivariable correlations were performed between grip or pinch strength and serum markers, controlling for age, sex, and height. Candidate predictors of grip or pinch strength were analyzed by linear regression, controlling for age, sex, height and sum of bony enlargements followed by multivariable modeling of significant variables by stepwise selection (p<0.05). Results The subject population (n=138) was 73% female, mean (SD) age of 65.8 (11.5) years, BMI 31.4(6.9) kg/m2, grip strength of 27.4 (10.0) kg, and pinch strength of 5.9 (1.9) kg. Grip and pinch strength correlated with age, sex, and height (p<0.05), but not weight, BMI, or sum of bony enlargements. Multivariable modeling identified sex, age, height, C2C, and the AA Factor as significant, independent predictors of grip strength; the model predicted 55% of the variability of grip strength and was significant at p<0.0001. For pinch strength, multivariable modeling identified sex, age, height, the ratio C2C:CPII, GABR, and the AC Factor as significant, independent predictors; the model predicted 49% of the variability of pinch strength and was significant at p<0.0001. Conclusion This exploratory analysis identified several novel, independent predictors of both grip and pinch strength. As seen previously, age, sex, and height were the strongest predictors of hand strength. AA factor predicted grip strength, perhaps reflecting sarcopenia status. The association of C2C or the ratio of C2C and CPII, markers of collagen breakdown and synthesis, respectively, with grip and pinch strength suggests the involvement of joint tissue turnover with hand function. The significant associations of GABR and the AC factor implicate metabolic pathways in pinch strength. Acknowledgements: This study was supported by Claude D Pepper OAIC P30 AG028716 and P01 AR050245.