S-Index could be estimated with useful internal predictive performance from fixed-pressure inspiratory performance, spirometry, and body composition variables in male patients with COPD or asthma, though external validation is required before clinical implementation.
Peak inspiratory flow rate showed the strongest correlation with S-Index among all candidate predictors.
LassoLars regularized regression numerically achieved the lowest prediction error for S-Index in the full predictor scenario.
Adding spirometry and inspiratory flow variables substantially increased the variance in S-Index explained by prediction models.
Multiple regression model types were evaluated and showed broadly comparable performance for predicting S-Index.
The study population consisted exclusively of male patients with COPD or asthma from a single center.
The S-Index, a measure of dynamic inspiratory muscle strength reflecting pressure generation during flow-dependent inspiration, was assessed using a single-breath test.
The developed prediction models are intended as complementary tools rather than substitutes for direct S-Index measurement.
This research suggests that a measure of breathing muscle strength called the S-Index — which captures how forcefully patients can breathe in when airflow is involved — can be reasonably estimated using simpler, more widely available tests in men with COPD or asthma. The S-Index normally requires specialized equipment, so researchers developed mathematical prediction models using information from lung function tests (spirometry), body composition measurements, and a standardized breathing endurance task where patients breathed against a fixed resistance of 30 cmH₂O until they could no longer continue. Among 123 male patients, how fast a patient could inhale (peak inspiratory flow rate) and how many breathing repetitions they could complete were the most closely linked to the S-Index. The best-performing prediction model (called LassoLars) explained about 76% of the variation in S-Index values across patients, with an average prediction error of roughly 7 cmH₂O. Using only physical characteristics like weight and height explained only about 44–46% of the variation, but adding lung function and breathing flow data pushed this to around 70%, showing that functional breathing tests add meaningful information. Several different types of prediction models — from simple linear equations to more complex machine learning approaches — performed similarly, suggesting the relationship between these variables and S-Index is relatively straightforward. This research suggests that clinics without access to specialized S-Index equipment might be able to estimate this measure from tests they already perform, potentially making it easier to assess respiratory muscle function in COPD and asthma patients. However, the study was conducted in a single center and only in male patients, and the models were only tested on the same data used to build them (internal validation). The authors caution that these models should complement — not replace — direct S-Index measurement, and that testing in new, independent patient groups is needed before these tools could be used in routine clinical care.
Upload your bloodwork. We'll cross-reference your results against this study and 4,700 others.
Upload Your LabsYilmaz Y, Tosun M, Demirkan E, Yalcinkaya M, Efdal A, Boyacı H, et al.. (2026). Prediction models for dynamic respiratory muscle strength in COPD and asthma patients based on fixed-pressure inspiratory muscle performance.. BMC pulmonary medicine. https://doi.org/10.1186/s12890-026-04579-3