Commentary flags statistical and design limits of Sjögren nutrition study

A new commentary in Advances in Rheumatology raises methodological concerns about a cross-sectional study that linked nutritional status to extraglandular manifestations in patients with Sjögren’s syndrome. The commentary, authored by Dong Zheyu and Shi Jieyi and published on 2 November 2026, addresses the original study by Ösken and colleagues, which was published on 9 June 2026 and combined two nutritional screening tools, the Controlling Nutritional Status score and the Prognostic Nutritional Index, to explore how malnutrition relates to extraglandular manifestations of the disease.

The commentators’ first concern centers on a logical inconsistency in how the Prognostic Nutritional Index cutoff was defined. The original study used the formula PNI = (10 × serum albumin [g/dL]) + (0.005 × total lymphocyte count [/mm³]), in which higher values reflect better nutritional status. Consistent with that definition, the original study reported that patients with extraglandular manifestations had significantly lower PNI levels than those without them, at 49.3 versus 54.5. Yet the same study’s receiver operating characteristic analysis defined the predictive cutoff as PNI > 51.625 for predicting extraglandular manifestations. According to the commentary, a more reasonable cutoff should be revised to PNI ≤ 51.625, so that lower PNI values indicate a higher risk of developing extraglandular manifestations rather than the opposite.

The second statistical issue concerns potential multicollinearity in the multivariate regression model. The original study simultaneously incorporated the ESSDAI disease activity score, the CONUT score, and the PNI as independent variables. The commentators point out that both CONUT and PNI are calculated from serum albumin and peripheral lymphocyte counts, while the ESSDAI score also contains hematological parameters. This high overlap in variable composition, they argue, inevitably produces strong internal correlation among the predictors, which can yield unstable odds ratio values and biased statistical results. They recommend retaining only one nutritional index, either CONUT or PNI, to improve the reliability and robustness of the regression model.

Beyond the statistical concerns, the commentary highlights a fundamental limitation of the cross-sectional design itself. Because the original study captured a single point in time, it cannot clarify the causal direction between malnutrition and extraglandular manifestations. The commentators note that it remains unclear whether malnutrition contributes to the development of extraglandular manifestations, or whether systemic inflammation related to those manifestations leads to subsequent malnutrition. This ambiguity, they argue, affects the credibility of any causal inference drawn from the data.

A further design limitation involves unadjusted confounding factors, particularly immunosuppressive treatment. The commentary states that the original study failed to collect and adjust for immunosuppressive treatment data. Glucocorticoids, which are commonly used in the clinical management of Sjögren’s syndrome, can significantly alter nutritional and hematological indicators by inducing hyperglycemia, reducing serum albumin levels, and redistributing peripheral lymphocytes. Because these are precisely the components from which CONUT and PNI are calculated, the lack of adjustment for glucocorticoid exposure and dosage may confound the true association between nutritional status and extraglandular manifestations, introducing what the commentators describe as unavoidable selection bias.

The third major point addresses what the composite indices can and cannot measure. CONUT and PNI are composite prognostic markers that reflect both systemic inflammation and nutritional status simultaneously. The commentators argue that relying solely on these two indicators cannot effectively distinguish protein-energy malnutrition from inflammation-induced catabolism, two pathological processes with different underlying mechanisms. To separate them, they recommend comprehensive collection of dynamic body mass index changes and quantitative dysphagia severity data, which would help differentiate the two processes and further clarify the mechanism of nutritional disorders in Sjögren’s syndrome patients with extraglandular manifestations.

Sjögren’s syndrome is a systemic autoimmune condition in which extraglandular manifestations extend beyond the hallmark dryness of the eyes and mouth, and nutritional compromise in affected patients has practical clinical significance. The original study’s approach of pairing the CONUT score with the PNI was described by the commentators as an innovative combination for exploring the relationship between nutritional status and these manifestations. The commentary nonetheless frames its critique as constructive, stating that the suggestions are intended to help refine the statistical analysis and conclusions of the study rather than to dismiss its contribution.

The methodological points raised have broader relevance for research using composite nutritional indices in inflammatory disease. Because CONUT and PNI both derive from albumin and lymphocyte measurements, they inherently blend nutritional and inflammatory signals, which complicates interpretation when disease activity measures also include hematological parameters. The commentary’s recommendation to avoid entering overlapping composite variables into the same regression model reflects a standard concern about multicollinearity, where correlated predictors destabilize coefficient estimates and make individual odds ratios difficult to interpret.

The commentators close by commending the original authors for their innovative and rigorous research and expressing hope that the suggestions will improve the study’s statistical analysis and conclusions. They also state that they look forward to more high-quality research in this field. The exchange appears in the Matters Arising category of Advances in Rheumatology, an open-access journal, and is available under open access terms. The commentary does not present new patient data; its scope is limited to reanalysis of the methods and interpretation of the original cross-sectional study, and its recommendations concern cutoff direction, variable selection, confounder adjustment, and the addition of anthropometric and dysphagia measures in future work.

Subject of Research: Methodological commentary on a cross-sectional study of nutritional screening and extraglandular manifestations in Sjögren's syndrome

Article Title: Commentary on “Nutritional screening and extraglandular manifestations in Sjögren disease: a cross sectional study using the controlling nutritional status score and prognostic nutritional index”

Article References: Zheyu, D., & Jieyi, S. (2026). Commentary on “Nutritional screening and extraglandular manifestations in Sjögren disease: a cross sectional study using the controlling nutritional status score and prognostic nutritional index”. Advances in Rheumatology, 66(1), Article 63. https://doi.org/10.1186/s42358-026-00589-1

Image Credits: AI Generated

DOI: 10.1186/s42358-026-00589-1

Keywords: Sjögren's syndrome, CONUT score, Prognostic Nutritional Index, extraglandular manifestations, malnutrition, multicollinearity, cross-sectional study, glucocorticoids, ESSDAI, systemic inflammation, Advances in Rheumatology, Matters Arising