Serum 25-hydroxyvitamin D, liver enzymes and FIB-4 fibrosis risk: a statistically robust but clinically negligible association from 31,607 adult laboratory records

Abstract Low serum 25-hydroxyvitamin D is repeatedly reported alongside abnormal liver chemistry and higher fibrosis scores, but its size, and how much survives adjustment for general health, are rarely quantified. We analysed 31,607 adult records from 22,103 patients with co-ordered 25-hydroxyvitamin D and a complete liver panel at a single Slovenian laboratory, 2014–2023. Correlations with every hepatobiliary analyte were negative but small: gamma-glutamyltransferase was strongest (rho ‑0.156, 95% CI ‑0.167 to ‑0.145; 2.43% of rank variance), FIB-4 0.65% (rho ‑0.081, ‑0.092 to ‑0.069). Each 25 nmol/L higher 25-hydroxyvitamin D was crudely associated with 5.21% lower FIB-4 and 23% lower odds of FIB-4 above 2.67 (odds ratio 0.765, 0.743–0.787). Adjustment for age, sex, albumin, season, era and kidney function removed 80% of the continuous signal (‑1.06%, ‑1.62 to ‑0.49) and 75% of the log-odds (0.935, 0.906–0.966). The one-record-per-patient adjusted association was null (0.967, 0.933–1.002). Five negative-control analytes vitamin D does not regulate, identically adjusted (model 4), showed residuals of similar size (median |OR‑1| 0.050). The association is statistically unambiguous, clinically negligible, and would bring fewer than 0.5% of records within reach of the high-risk threshold. We present crude and adjusted estimates as bounds, not causal effects.