Effect of Inspiratory Muscle Training on Diaphragm and Abdominal Wall Muscle Thickness with Fatty Liver Density in Elderly Women: A Randomized Controlled Trial

Abstract

Background and Objectives: Post-menopausal estrogen decline is considered a contributing factor to sarcopenia, and inspiratory muscle training (IMT) may provide benefits in this demographic. This study examined the impact of a four-week IMT program on diaphragm thickness, abdominal wall muscle thickness (AWMT; transversus abdominis, internal oblique, and external oblique), and liver fat percentage in healthy elderly women. Materials and Methods: Twenty-six women aged 60–80 years were randomly assigned to an IMT group (n = 13) or a control group (n = 13). The IMT group used the PowerBreathe® Classic device at 40% of maximal inspiratory pressure (MIP), with weekly increments of 10%. Training was performed twice daily, five days per week, with 30 breathing cycles per session (60 per day). The control group maintained their usual routines. AWMT, diaphragm thickness (DT), and fatty liver density (FLD) were measured by a radiologist before and after the intervention. Results: After four weeks, the IMT group showed significant improvements in all parameters compared to controls. Mid-diaphragm thickness (MDT) increased by 11.44% (effect size (ES) = 0.358, p < 0.001) versus 0.76% in controls (p = 0.271). Posterior diaphragm thickness (PDT) improved by 7.48% (ES = 0.282, p < 0.001) versus 0.38% (p = 0.564). Right AWMT increased by 12.7% (ES = 0.492, p < 0.001) compared to 0.10% (p = 0.872), and left AWMT increased by 9.93% (ES = 0.395, p < 0.001) versus 2.64% (p = 0.014). FLD improved by 11.79% (ES = 0.959, p < 0.001) in the IMT group, while the control group showed no meaningful change (−0.13%, p = 0.847). Conclusions: A short-term IMT protocol significantly enhanced diaphragm and AWMT and reduced liver fat in elderly women. These findings support the use of IMT as a simple, non-invasive intervention to preserve musculoskeletal and metabolic health in aging populations. © 2025 by the authors.