Association of leptin levels and relative leptin deficiency with steatotic liver disease in the general population

Leptin replacement therapy has shown promising results in treating metabolic steatohepatitis (MASH) patients with leptin levels below the 25th percentile for their sex and BMI category (“relative leptin deficiency”, RLD). However, the clinical utility of the RLD definition for identifying high-risk individuals for liver disease in general screening populations remains unclear. We analyzed the association between leptin levels and steatotic liver disease (SLD) in 636 participants from the screening population of the SAKKOPI registry (2007–2020) who underwent metabolic phenotyping, leptin measurement, and liver assessment including ultrasound, controlled attenuation parameter (CAP), and transient elastography (kPa). SLD was categorized as MASLD, MetALD, ALD, cryptogenic SLD, or no SLD. RLD was defined using NHANES III-based sex- and BMI-specific cutoffs. Leptin was analyzed categorically (RLD vs. non-RLD) and continuously, including leptin-to-BMI ratios (ln[leptin]/ln[BMI]) to assess leptin as a marker independent of body mass. Each two-fold increase in leptin was independently associated with MASLD (adjusted RRR 1.58, 95% CI 1.20–2.09, p = 0.001) and MetALD (RRR 1.74, 95% CI 1.02–2.98, p = 0.043). Leptin-to-BMI ratio analysis confirmed this relationship, with the highest quintile showing 83.7 dB/m higher CAP values compared to the lowest quintile (p < 0.001). Individuals classified as having RLD (n = 112, 18%) demonstrated significantly better metabolic health and lower SLD prevalence (27% vs. 48%, p < 0.001). The leptin/BMI ratio showed specificity for steatosis parameters (CAP, fatty liver index) but no association with liver stiffness (kPa, p = 0.55) or APRI. However, FIB-4 scores were higher in the lowest leptin/BMI quintile (1.27 vs. 1.11, p = 0.011), an association that persisted after age and sex adjustment. In this screening cohort of the general population, NHANES III-derived RLD thresholds may not effectively identify high-risk individuals for SLD in this population, but rather appear to select metabolically healthier participants. While the persistent FIB-4 association deserves consideration, the absence of corresponding associations with direct fibrosis measurements suggests caution in interpretation. These findings indicate that current RLD definitions appear to have limited utility for risk stratification in this specific screening population, though their applicability in high-risk clinical settings requires further investigation.