Association Between Temporal Muscle Thickness and Skeletal Muscle Mass, Nutritional Status, and Physical Function in Patients With Post-stroke Hemiparesis: A Cross-Sectional Study

Sarcopenia, the age-related loss of skeletal muscle mass and strength, impairs physical function and worsens recovery and prognosis after stroke. Standard sarcopenia assessments (e.g., handgrip strength, gait speed) can be impractical in stroke patients due to impaired consciousness, bed rest, or paralysis. While DXA offers accurate muscle mass quantification and BIA is accessible, both have limitations (equipment, cost, hydration sensitivity). Temporal muscle thickness (TMT)—visible on routine neuroimaging such as CT/MRI—has been proposed as a surrogate marker for skeletal muscle mass and a predictor of stroke outcomes. Because stroke patients are also at risk of malnutrition (e.g., from dysphagia and inactivity) that can exacerbate sarcopenia, it is clinically important to determine whether TMT reflects not only muscle mass but also nutritional status and physical function. This study investigates the relationships between TMT, skeletal muscle mass/quality, nutritional markers, and lower limb motor function in patients with post-stroke hemiparesis.

Prior work suggests TMT may indicate sarcopenia and predict outcomes in neurological disorders and stroke, leveraging routinely acquired neuroimaging. However, few studies examined its association with nutritional status. Stroke-related factors (dysphagia, decreased activity) elevate malnutrition risk and may accelerate sarcopenia, underscoring the need to evaluate whether TMT captures both muscle and nutrition domains. The study builds on reports linking TMT to prognosis and sarcopenia while addressing gaps regarding nutrition-related indices (e.g., GNRI).

Design: Cross-sectional study at a single center. Consecutive patients with first-ever stroke and hemiparesis admitted to a convalescent rehabilitation ward between October 2021 and October 2022 were screened. Participants: Inclusion—first-ever stroke with hemiparesis, discharged from rehabilitation ward. Exclusions—prior craniotomy; pacemaker/metal implants contraindicating MRI; MRI motion artifacts preventing reliable TMT measurement; inability to complete assessments before discharge. Data collection: Demographics and clinical data (age, sex, BMI, length of stay, SIAS-MLE) were extracted from electronic records. Handgrip strength was measured using a Smedley dynamometer (two trials per hand; highest value used). TMT measurement: T2-weighted brain MRI at admission and discharge on a 3-Tesla scanner (Magnetom Verio; Siemens). Measurements performed using Synapse PACS (Fujifilm). TMT was measured bilaterally at the level of the orbital apex using the supraorbital fissure and orbital apex as landmarks, perpendicular to the muscle’s long axis, following published methods. The average of right and left TMT values was calculated per patient. Two evaluators (a physical therapist with 10 years’ experience and a radiologic technologist with 12 years’ experience) performed measurements; a board-certified radiologist, blinded to clinical data, conducted all measurements used for analysis. Muscle mass and quality: Body composition by InBody S10 in the supine position after rest. Indices: Skeletal Muscle Index (SMI) for muscle mass; extracellular water-to-total body water ratio (ECW/TBW) and phase angle (PhA) as muscle quality markers. Nutritional status: Serum albumin and Geriatric Nutritional Risk Index (GNRI). GNRI = 14.89 × albumin (g/dL) + 41.7 × (actual body weight/ideal body weight), with ideal body weight = height^2 × 22 (kg/m^2). If actual weight > ideal weight, the ratio was set to 1. Statistical analysis: Inter-rater reliability of TMT assessed using ICC (2,1) with 95% CI. Spearman’s rank correlation coefficients (two-tailed) were calculated between TMT and eight variables: age, sex, BMI, SMI, ECW/TBW, GNRI, SIAS-MLE, and handgrip strength. Analyses were performed in R 4.3.1; significance set at p<0.05. Ethics: Approved by the Ethics Committee of the Akita Cerebrospinal and Cardiovascular Center (approval number 21-16; July 27, 2021). Participants were informed of opt-out rights; confidentiality maintained.

- Sample: Of 91 screened stroke patients, 45 met inclusion criteria; exclusions included prior stroke (n=14), pacemaker/metal implants (n=2), craniotomy (n=12), poor MRI quality (n=3), no admission MRI (n=9), and missing data (n=6). - Baseline: Median age 67.8 years (IQR 59.5–75.6). Stroke types: cerebral infarction 52.5%, cerebral hemorrhage 45.0%, subarachnoid hemorrhage 2.5%. Mean GNRI 96.8 ± 7.6. Sex differences: men vs women showed higher handgrip strength (31.6 ± 9.7 vs 18.8 ± 7.8 kg; p<0.001), higher TMT (5.3 ± 1.5 vs 4.9 ± 0.9 mm; p=0.045), and higher SMI (6.8 ± 0.8 vs 5.7 ± 0.7 kg/m²; p<0.001). No sex differences in age, BMI, SIAS-MLE, ECW/TBW, serum albumin, or GNRI. - Reliability: TMT inter-rater reliability was high (ICC (2,1): left 0.812 [95% CI 0.672–0.896]; right 0.796 [95% CI 0.706–0.919]). - Correlations (Spearman’s ρ, p-value): • SIAS-MLE: ρ=0.468, p=0.0023 • SMI: ρ=0.396, p=0.0114 • PhA: ρ=0.528, p<0.001 • GNRI: ρ=0.350, p=0.0267 • ECW/TBW: ρ=−0.460, p=0.0028 • Not significant: age (ρ=−0.004, p=0.9816), BMI (ρ=0.293, p=0.0667), length of stay (ρ=−0.223, p=0.1665), handgrip strength (ρ=0.306, p=0.0544).

TMT measured on routine MRI correlated with skeletal muscle mass (SMI), muscle quality (higher PhA, lower ECW/TBW), nutritional status (higher GNRI), and lower limb motor function (higher SIAS-MLE) in post-stroke hemiparesis, supporting its utility as a practical surrogate marker for sarcopenia and functional status when conventional assessments are limited. High inter-rater reliability between a physical therapist and a radiologic technologist indicates TMT can be consistently measured outside radiology, enhancing feasibility in rehabilitation settings. The association with SIAS-MLE suggests TMT may reflect broader functional capacity beyond cranial musculature, potentially capturing disuse and systemic changes after stroke. Links with GNRI imply TMT may also indicate nutritional status, possibly influenced by dysphagia and reduced masticatory function leading to temporalis atrophy. Differences from some prior studies regarding nutrition associations may relate to timing (measurement at discharge vs stroke onset). While promising, TMT reflects a localized muscle; further work is needed to clarify how well it represents whole-body muscle quality and to define its role in prognostication and care planning.

TMT can be reliably measured by non-specialists and is associated with muscle mass, muscle quality, nutritional status, and lower limb motor function in patients after stroke. It shows promise as a practical tool for sarcopenia screening and comprehensive assessment in rehabilitation settings. Future studies should include larger, multicenter, longitudinal designs and incorporate detailed swallowing assessments to validate clinical applications and clarify causal relationships.

- Single-center study with a small sample size, limiting generalizability. - Cross-sectional design precludes causal inference. - Swallowing function was not evaluated in detail, limiting insights into relationships among dysphagia, nutrition, and TMT.