Long-term insulin resistance is associated with quantitative increases in CT-based measures of interstitial lung disease (ILD), according to findings from a multicenter cohort study presented at the 2026 American Thoracic Society (ATS) International Conference in Orlando, Florida.1
Insulin Resistance and Interstitial Lung Disease: Study Overview
The analysis, drawn from the Multicenter AIDS Cohort Study (MACS), included 517 men with or at risk for HIV who had both serial insulin resistance measurements and cardiac CT imaging available. Investigators at Johns Hopkins calculated a time-weighted homeostatic model assessment of insulin resistance (HOMA-IR) using up to 5 years of annual measurements prior to CT, capturing cumulative rather than point-in-time exposure.1
Insulin Resistance and Interstitial Lung Disease: Key Data Points
- The study used up to 5 years of serial HOMA-IR measurements to calculate cumulative insulin resistance exposure via time-weighted average
- Each doubling in long-term insulin resistance associated with 1.58% increase in quantitative interstitial lung disease on computed tomography
- Ground glass opacity—a potential marker of early pulmonary inflammation—drove the CT association; lung fibrosis link did not persist after adjustment
- Findings held regardless of HIV or diabetes status
The mean cohort age was 57.5 years, mean BMI was 26.6 kg/m², and mean pack-years was 12.5. Approximately 59.8% of participants were living with HIV, and 8.7% had diabetes. The mean quantitative ILD percentage was 10.6% (SD 9.33).1
In adjusted models accounting for age, BMI, smoking history, HIV status, race, and medically treated diabetes, each doubling in time-weighted HOMA-IR was associated with a 1.58% greater percentage of lungs with any ILD feature (95% CI, 0.38 to 2.78). The association appeared driven primarily by ground glass opacity (adjusted mean difference 1.43%; 95% CI, 0.44 to 2.43). The relationship with lung fibrosis observed in unadjusted models did not reach statistical significance after adjustment (adjusted mean difference 0.15%; 95% CI, −0.10 to 0.40). No effect modification was found by HIV or diabetes status.1
In an interview at ATS, investigator Sarath Raju, MD, from Johns Hopkins University, emphasized that cumulative insulin resistance exposure was central to the study’s design.
“Within COPD, we think about someone's pack-years, how long they've smoked over time. We don't just think about how much they smoked that day," Raju said. "We wanted to understand their cumulative exposure to insulin resistance. The time-weighted average gave us a way to think of someone's long-term insulin resistance because we know you don't develop lung disease overnight.”