STATIN USE IS ASSOCIATED WITH SUBTLE CHANGES IN MUSCLE COMPOSITION AND QUALITY: A QUANTITATIVE IMAGING ANALYSIS USING OAI DATA

Abstract

Subjective statin-associated musculoskeletal symptoms such as myalgia are commonly reported and contribute to statins' non-adherence. Several trials have recently suggested that in the absence of rare myositis, such symptoms are probably a placebo effect. However, no longitudinal study has been conducted to investigate whether statin use can be associated with deterioration in muscle quality. Such an association can be of particular clinical concern in older patients with or at-risk of knee OA, in whom muscle changes can affect the knee OA clinical outcomes. To assess longitudinal changes of validated thigh muscle quality biomarkers associated with statin use in people with or at-risk of knee OA. OAI Statin users at baseline or any of the four annual follow-up visits and non-users were matched for confounding by indication using 1:1 propensity-score (PS) matching. Potential confounders as covariates of the PS-matching included a wide range of demographic variables, comorbid diseases, risk factors, and medications. After PS-matching, 3,772 thighs of 1,910 participants were included (1,886 thighs of statin-users: 1,886 of non-users; age: 62±9 years (average ± SD), range: 45-79; female/male ratio: 1). Participants were further stratified according to baseline radiographic knee OA status (with versus at-risk of knee OA). We trained, tested, and used an automated deep learning model with a 2D U-Net structure to segment thigh muscle MRIs, with similar results to manual segmentation and quantification. Longitudinal muscle quality biomarkers included cross-sectional area (CSA), intramuscular adipose tissue (intra-MAT), contractile percent, and knee extensors and flexors maximum contractile force and specific contractile force (contractile force/CSAs) at baseline, 2nd, and 4th-year follow up. Linear mixed-effects models were used. At baseline, there was no statistically significant imbalance in PS-matching variables or muscle quality biomarkers between PS-matched statin-users and non-users. During four years, statin use was associated with a slight decrease in muscle quality, indicated by decreased knee extensor (quadriceps) maximum and specific contractile forces (mean difference (% of baseline value), 95%CI: -1.85 Newton/year (0.5% of 355.15N at baseline), -3.23 – -0.47, and -0.04 N/cm2/year (0.6% of 6.96N/cm2 at baseline), -0.07 – -0.01), contractile percent of the thigh muscles (-0.03%/year (0.03% of 95.32% at baseline), -0.06 – -0.01), and increased thigh intra-MAT (3.06 mm2/year (0.7% of 461.45mm2 at baseline), 0.53–5.59). However, stratification analysis showed that these results were only present in participants at-risk of knee OA but not those with the existing radiographic knee OA at baseline. Statin use is associated with decreased muscle quality, indicated by reduced contractility of the extensor (quadriceps) muscle and increased intra-muscular fat deposition. However, considering the small size of changes in muscle quality compared to substantial favorable cardiovascular effects of statins in clinical practice, these muscle changes might be considered clinically negligible. NIH NIAMS, under Award Number R01AR079620-01. AG reported receiving funding from Merck Serono, AstraZeneca, Galapagos, Pfizer, Roche, TissueGene (for consultation), and Boston Imaging Core Lab (as the president and stockholder). SD reported that he received funding from Toshiba Medical Systems (for consultation) and grants from GERRAF and Carestream Health (for a clinical trial study). CORRESPONDENCE ADDRESS: sdemehri1@jhmi.edu