Published on 19 September 2026
In Schizophrenia, Two Patients in Three Show Sarcopenia in a Turkish Sample
Schizophrenia Research · 2026;289:61-70 · Şişman & Önen
DOI 10.1016/j.schres.2026.01.002
PMID 41548363
Scientific 63
Editorial 72
The essentials
Among 198 adults with schizophrenia consecutively recruited at a hospital in Bursa, Turkey, between October 2023 and March 2024, 66.6% met EWGSOP2 criteria for sarcopenia, most of them at the probable sarcopenia stage, that is, on reduced grip strength alone. Patients with sarcopenia had a higher total PANSS score, 67.94 versus 59.89, lower functioning on the FROGS scale, and lower WHOQOL-BREF quality of life. Total serum antioxidant capacity correlated positively with muscle parameters, but its mean values did not differ between patients with and without sarcopenia, and the multivariable model predicting it explained only 10% of its variance. There was no control group, protein intake, albumin and micronutrient status were not measured, muscle mass was estimated by bioelectrical impedance rather than absorptiometry, and the design was cross-sectional. The prevalence figure is striking, it cannot be transposed as it stands, and no causal relationship can be drawn from these data.
Context
The physical health of patients with schizophrenia is usually approached through its metabolic and cardiovascular side: weight, glucose, lipids, blood pressure. Muscle function is almost never part of that picture. Yet it sits at the intersection of several realities of this population: sedentary behavior, insufficient protein intake, motor side effects of antipsychotics, smoking, and accelerated aging.
Sarcopenia refers to the combined loss of muscle mass and strength. It has mainly been studied in geriatrics, where it is associated with falls, fractures, hospitalizations, functional decline and mortality. Its operational definition, established in 2018 by the European EWGSOP2 working group and published in 2019, combines grip strength, muscle mass and physical performance. This study is one of the few to apply it to a psychiatric population of this size, and, according to its authors, the first to add oxidative stress markers to it.
The study at a glance
| Element | Content |
|---|---|
| Population | 198 adults aged 18 to 65, schizophrenia per DSM-5-TR, recruited consecutively in outpatient clinic, community mental health center and inpatient settings |
| DetailSingle-center recruitment at Bursa Yüksek İhtisas training and research hospital, Turkey, between October 2023 and March 2024. Excluded were patients with intellectual disability, severe neurocognitive or neurological disorder, or inability to consent. Diabetes, hypertension and dyslipidemia were not exclusion criteria. The participation rate is not reported. | |
| Measures | EWGSOP2 criteria: grip dynamometry, bioelectrical impedance analysis, usual walking speed |
| DetailCAMRY EH101 dynamometer, TANITA BC-730 impedance analyzer, walking speed timed over the central 4 meters of an 8-meter corridor. Strength and muscle mass thresholds are those validated in a Turkish population by Bahat et al. 2016: grip strength below 22 kg in women and 32 kg in men, skeletal muscle mass index below 7.4 kg/m² in women and 9.2 kg/m² in men. The walking speed threshold is 0.8 m/s, taken from a meta-analysis in an apparently healthy adult population. Also measured were total antioxidant capacity, total oxidant status and the oxidative stress index, assayed with commercial Rel Assay Diagnostics kits, along with the SARC-F, PANSS, FROGS and WHOQOL-BREF scales. | |
| Comparator | No control group |
| DetailThis is the study’s structuring limitation, acknowledged by the authors. No estimate of excess prevalence relative to a matched general population is possible. | |
| Outcomes | Prevalence and stages of sarcopenia, comparison of physical, clinical, functional and quality-of-life profiles, correlations with oxidative markers, multivariable predictors of these markers |
| DetailPearson or Spearman correlations depending on distribution, chi-square tests, t tests or Mann-Whitney tests. Benjamini-Hochberg correction for multiple comparisons. Multiple linear regressions including age, BMI, number of sarcopenia criteria, total PANSS, total WHOQOL-BREF and the three FROGS subscores, with variable selection at a bivariate threshold of p below 0.10 and variance inflation factors below 3. Analyses performed with SPSS 26. | |
| Design | Descriptive cross-sectional study |
| DetailPeer-reviewed publication. The authors declare having received no specific funding and having no competing interests. A corrigendum signed by the same authors was published in the same volume, page 112; its subject is not specified in the version consulted. | |
Quality control
| Point checked | Judgment |
|---|---|
| Validated diagnostic criteria | EWGSOP2 criteria |
| FindingUsing a consensus definition allows comparison with the geriatric literature, which is not always the case in this field. The numeric thresholds applied are, however, Turkish thresholds rather than the reference European ones. | |
| Sample size | Sufficient for a prevalence estimate |
| Finding198 patients give a reasonably precise prevalence estimate, within the limits of recruitment. The a priori calculation set the minimum at 72 participants for a medium effect, power of 0.95 and an alpha risk of 0.05. | |
| Funding | No funding declared |
| FindingThe authors state that they received no specific grant from any public, commercial or nonprofit body, and declare no financial or personal conflict of interest. No industry funding, then, but also no dedicated support. | |
| Control group | Absent |
| FindingWithout controls matched on age, sex and setting, the 66.6% figure cannot be interpreted as an excess attributable to schizophrenia. The authors themselves call for comparisons with matched healthy controls. | |
| Nutritional status | Not measured |
| FindingNeither albumin, nor protein intake, nor micronutrient status. Protein intake is a major determinant of sarcopenia, and the absence of its measurement leaves the most obvious confounder unaddressed. | |
| Physical activity | Collected, not adjusted for |
| Finding71.7% of participants are sedentary, 21.7% engage in low-to-moderate intensity walking and 6.6% in regular exercise. Activity level appears in the correlation tables, with no significant association with sarcopenia indicators, but it does not enter the regression models, and the authors list it among the unaddressed confounders. | |
| Method of measuring muscle mass | Bioelectrical impedance |
| FindingLess precise than dual-energy X-ray absorptiometry, and sensitive to hydration status, which the authors note as a limitation in a population where hydration is variable. The European definition accepts it, but estimate precision suffers. | |
| Cross-sectional design | No causality |
| FindingThe direction of the relationship between symptom severity and sarcopenia cannot be established. Both may stem from a common unmeasured factor. The authors explicitly mention reverse causality as not excluded. | |
| Representativeness | Single-center |
| FindingOne tertiary center, one country, a consecutive rather than random sample, with specific dietary habits and care organization. Transposing the figures to another setting is not warranted. | |
The results
| Result | What the data show |
|---|---|
| Prevalence of sarcopenia | 66.6%, that is 132 of 198 patients, mostly at the probable sarcopenia stage |
| ReadingA high figure, even against the community data the authors cite, 5% to 13% after age 60 and up to 50% after age 80. Mean sample age is 41.3 years, SD 11.1, and does not differ significantly between the two groups, 42.08 versus 39.76 years, p = 0.167. The probable stage rests on grip strength alone, without confirmation from muscle mass. | |
| Total antioxidant capacity | Positively correlated with muscle mass (r = 0.26), basal metabolic rate (r = 0.29), bone mass (r = 0.28) and grip strength (r = 0.21), and negatively with the number of sarcopenia criteria (r = -0.15, p = 0.038) |
| ReadingThese correlations survive Benjamini-Hochberg correction. Mean values of total antioxidant capacity, total oxidant status and the oxidative stress index, however, do not differ between patients with and without sarcopenia (p = 0.089; 0.163; 0.066), and the oxidative stress index correlations do not survive correction. | |
| Multivariable model | R² = 0.10, p = 0.011 for the model predicting total antioxidant capacity |
| ReadingSignificant predictors are BMI (beta = 0.17, p = 0.019), the number of sarcopenia criteria (beta = -0.20, p = 0.014), total PANSS (beta = 0.22, p = 0.024) and total WHOQOL-BREF (beta = 0.19, p = 0.030). This coefficient of determination of 0.10 concerns the variance in antioxidant capacity explained by the model as a whole, not the relationship between antioxidant capacity and muscle mass. The models predicting total oxidant status and the oxidative stress index were not significant. | |
| Symptom severity | Total PANSS 67.94 in patients with sarcopenia versus 59.89, p < 0.001, d = 0.59 |
| ReadingAll three subscores are higher: negative (p < 0.001), general psychopathology (p = 0.022) and positive (p = 0.036). Correlations with sarcopenia indicators are stronger for negative symptoms (r = 0.235 to 0.349) than for positive symptoms (r = 0.074 to 0.191). This is a cross-sectional association: severity may reduce physical activity, sarcopenia may worsen functional impact, and a third factor may drive both. | |
| Functioning and quality of life | Total FROGS 46.87 versus 50.50 (p = 0.005) and total WHOQOL-BREF 62.44 versus 66.47 (p = 0.005) |
| ReadingFunctioning is measured with the FROGS functional remission scale, not a global functioning scale. Since both scales partly correlate with symptomatology, these results are not independent of the previous one. The largest effect sizes concern grip strength (d = 1.88) and walking speed (d = 0.88), with the effect sizes for functional and quality-of-life scores reported inconsistently at two points in the authors’ text. | |
The 66.6% figure is the one everyone will remember, and it is precisely the one that calls for the most caution. It does not mean that schizophrenia multiplies the risk of sarcopenia, for lack of controls. Nor does it mean that two-thirds of patients in any other population would be affected, for lack of representativeness: the sample is Turkish, single-center and gathered without controls. It also rests, for the most part, on cases classified as probable, that is, on reduced grip strength without confirmation from muscle mass. What it does say, and that is not nothing, is that measurable muscle impairment is very common in a population where no one looks for it.
Critical appraisal
| Domain | Judgment |
|---|---|
| Originality of the question | Genuine |
| FindingMuscle function is a blind spot in somatic monitoring in psychiatry. Asking the question with a validated instrument has value in itself, and adding redox markers is novel in this population. | |
| Internal validity | Weak |
| FindingNo controls, no nutritional indicators, muscle mass measured by bioelectrical impedance, cross-sectional design. These four limitations are acknowledged by the authors and compound each other. | |
| Interpretation of the oxidative pathway | Not to be overinterpreted |
| FindingMean values of the three redox markers do not differ between patients with and without sarcopenia, the multivariable model explains 10% of the variance in antioxidant capacity, and the oxidative stress index correlations do not survive correction for multiple comparisons. On a single serum measurement, this opens a question, it does not resolve it. | |
| Role of treatment | Documented, little analyzed |
| FindingAntipsychotic class, use of a long-acting injectable formulation and adherence were compared between groups with no significant difference (p = 0.632; 0.314; 0.209). Mean treatment duration is 9.36 years, SD 7.69, and correlates weakly with the SARC-F score (r = 0.182, p = 0.010). Neither the specific molecule, nor the dose, nor cumulative exposure were analyzed, which the authors acknowledge. | |
| Authors’ conclusion | More assertive than the design supports |
| FindingThe authors recommend integrating systematic sarcopenia screening and early intervention into routine psychiatric care. A single-center cross-sectional study without controls and without nutritional data describes an association; it does not establish a basis for a screening strategy. | |
| Generalization | Not warranted |
| FindingOne center, one country, consecutive rather than random recruitment. The strength and muscle mass thresholds used are those validated in a Turkish population, which serves the local reading but complicates comparison with work using European or Asian thresholds. | |
Level of evidence
Confidence is moderate that frequent muscle impairment exists in this population. It is low on the magnitude of the phenomenon outside this center, and very low on any mechanistic explanation. The study’s main contribution is not a result, it is a shift in the question being asked.
What is established: in this sample, two patients in three meet EWGSOP2 criteria, most often at the probable stage. What is suggested: a link between symptom severity, functioning and muscle impairment. What remains hypothesis: the role of oxidative stress. What is not measured, and probably matters most: nutrition in all its forms, protein intake, albumin and micronutrients. Physical activity, for its part, is indeed collected, on a self-reported basis, and shows massive sedentary behavior, but it enters no adjustment model.
The colleague test
What an experienced colleague would say if shown this study in two minutes, between two consultations.
Two-thirds is enormous, but without controls I do not know what to make of it. What strikes me more is that I never looked for it. I shake my patients’ hands every day, and I never thought about what that handshake was telling me.
Translation for practice: the authors argue for systematic screening, and the study design cannot support that. What it does make visible is a comorbidity that standard follow-up does not look at. Asking about exertional fatigue, difficulty rising from a chair, or difficulty carrying groceries costs nothing and can prompt a referral for nutritional assessment or rehabilitation care.
What you can do with this
- Add two or three functional questions to extended somatic follow-up: climbing stairs, rising from a chair without using the hands, carrying a bag of groceries.
- When strength loss or reduced mobility appears, explore nutritional status and activity level, which standard psychiatric follow-up does not ask about, before attributing everything to negative symptoms.
- Refer for nutritional assessment or physical and rehabilitation medicine when functional impact is clear, without waiting for advanced age.
- Do not treat the 66.6% figure as applying to any other clinical population. It describes a consecutive sample from one Turkish center, without controls, and rests mainly on probable cases.
- Keep the question rather than the figure: muscle function is absent from our monitoring checklists, and nothing justifies it staying that way.
- The course of action is set out in the NICE decision tree for schizophrenia in adults.
Frequently asked questions
Do two-thirds of patients with schizophrenia have sarcopenia?
In this Turkish sample of 198 patients, 66.6% met EWGSOP2 criteria, most at the probable stage, which rests on grip strength alone. Without a control group or representative recruitment, this figure cannot be extended to other populations. It should be presented as a local observation, not as a reference prevalence.
Are antipsychotics to blame?
The study does not answer this question. Antipsychotic class, use of a long-acting injectable formulation and adherence did not differ between patients with and without sarcopenia, and mean treatment duration, close to ten years, shows only a weak correlation with sarcopenia indicators. Neither the specific molecule, nor the dose, nor cumulative exposure were analyzed. Several mechanisms remain plausible, including sedation, motor side effects and weight gain with relative lean mass loss, but no data presented here can distinguish or quantify them.
Should sarcopenia screening be implemented?
The authors recommend systematic screening in psychiatry and propose grip strength measurement as a simple tool. However, no sarcopenia screening protocol is validated in outpatient psychiatry, and a single-center cross-sectional study without controls is not enough to establish one. The reasonable approach remains clinical: ask about strength and mobility, and refer when the impact is real.
Does oxidative stress explain the muscle loss?
The data show positive correlations between serum antioxidant capacity and muscle parameters, in the range of r = 0.21 to 0.29. But mean redox marker values do not differ between patients with and without sarcopenia, and the multivariable model predicting antioxidant capacity explains only 10% of its variance. This is a lead, not a demonstrated mechanism.
Annotated bibliography
Source study. Şişman Ö, Önen S. Prevalence and multidimensional impact of sarcopenia in schizophrenia: Associations with oxidative stress, functioning, and symptom severity. Schizophrenia Research. 2026;289:61-70. DOI 10.1016/j.schres.2026.01.002 · PMID 41548363. A corrigendum signed by the same authors was published in Schizophrenia Research, 2026;289:112; its subject is not specified in the version consulted.
Editorial collections
Tags
Verified on August 12, 2026 against the full text of the publication and its supplementary material where available. This analysis underwent an independent double reading. The English version was checked for conformity on September 19, 2026, against the figures of the French version and against the source. How we verify what we publish
This analysis is intended for healthcare professionals. It does not constitute a prescribing recommendation and does not replace individual clinical judgment.
Analysis from Psychiatry Evidence Base, evidence-based psychiatry, explained with rigor.