Joint associations of low-density lipoprotein cholesterol and prognostic nutritional index with mortality: a prospective cohort study in UK Biobank
L Li, J Li, H Liu, G Aa, H Sun, P Sun, X Zhou, Q YangAbstract
Background
Low-density lipoprotein cholesterol (LDL-C) is a key cardiovascular (CV) risk marker, but its association with mortality in primary prevention may be influenced by the “cholesterol paradox.” The prognostic nutritional index (PNI) is an established indicator of systemic nutritional and immunologic status. Prior studies suggest that the higher mortality observed at low LDL-C levels may reflect underlying malnutrition and inflammatory burden rather than a direct causal effect of LDL-C. However, evidence remains limited regarding the extent to which nutritional-immune status modifies the association between LDL-C levels and cause-specific mortality.
Purpose
To examine joint associations of LDL-C and PNI with all-cause, CV, and non-CV mortality in a primary prevention population.
Methods
We analyzed data from the UK Biobank, including participants without baseline CV disease. Outcomes were all-cause death, CV death, and non-CV death. Restricted cubic spline (RCS) analyses were used to assess potential nonlinear dose-response associations of LDL-C and the PNI with mortality outcomes. Heatmaps were generated to visualize risk across joint categories of LDL-C and PNI. Participants were stratified into LDL-C and PNI tertiles to compare mortality risks across subgroups.
Results
Among 310,160 participants (median age 56.1 years) followed for a median of 13.5 years, there were 21,899 deaths, including 3,831 CV and 18,068 non-CV deaths. RCS analyses showed that the lower PNI was associated with higher risks of all outcomes, whereas LDL-C showed cause-specific: lower LDL-C levels were associated with higher risks of all-cause and non-CV mortality, while higher LDL-C were linearly associated with increased CV mortality. In Cox models, the LDL-C×PNI interaction was statistically significant (P = 0.03). Heatmaps showed that all-cause mortality risk was greatest among participants with both low PNI and low LDL-C, and the excess risk associated with low LDL-C was attenuated at higher PNI levels (HR, 1.04; 95% CI, 0.98-1.12 for LDL-C tertile 1 vs tertile 2 in the highest PNI tertile). A similar pattern was observed for non-CV mortality. By contrast, CV mortality risk was greatest in the low PNI-high LDL-C region. Among participants in the middle PNI tertile, lower LDL-C was associated with a lower risk of CV mortality (HR, 0.77; 95% CI, 0.66-0.90 for LDL-C tertile 1 vs tertile 2).
Conclusions
In this primary prevention population, LDL-C and the PNI showed joint associations with mortality risk, with evidence of effect modification. The excess risk associated with low LDL-C levels appeared to be partly attributable to nutritional-immune status as captured by PNI. Notably, the risk patterns for CV vs non-CV mortality occupied distinct regions across the joint LDL-C-PNI spectrum. These findings suggest lipid-related risk should be interpreted alongside nutritional-immune status to contextualize CV vs non-CV mortality risk.LDL-C and PNI with mortalityJoint effects