In 1,387 NHANES participants aged 60 and older, MASLD (hepatic steatosis index ≥36 plus a cardiometabolic risk factor) showed no association with global cognition after full adjustment (β = 0.027; p = 0.526). Cognition worsened in a graded way with the number of metabolic risk factors: two or more predicted lower global cognition (β = -0.23) and memory (β = -0.33), while processing speed fell only at the highest burden. High FIB-4 fibrosis scores were linked to lower CERAD memory (β = -0.319) but not to other measures.
The practical message is that the liver may be a marker of risk, not the culprit. Fatty liver clusters with hypertension, dyslipidemia, diabetes and obesity, and the findings suggest those conditions, not hepatic fat itself, account for most of the cognitive signal. That fits broader evidence that midlife vascular and metabolic control protects the aging brain. Several caveats limit confidence. The design is cross-sectional, so reverse causation and temporal ordering cannot be assessed. Steatosis was defined with a hepatic steatosis index, not imaging or biopsy, which misclassifies some participants. The fibrosis finding rests on a single borderline signal among many tests, with no multiplicity correction visible. Survivor bias in older NHANES samples may also mute any liver effect. Treat this as confirmatory, not paradigm-shifting: it argues for managing blood pressure, lipids, glucose and weight as the cognitive-protection strategy, while longitudinal cohorts with imaging-defined MASLD must settle whether fatty liver adds independent risk.