Independent and Joint Associations of the Planetary Health Diet Index and Triglyceride‐Glucose Index With Mortality and Prevalent Cardiovascular Disease in Adults With Overweight or Obesity: NHANES 2005–2018

The Planetary Health Diet Index (PHDI) reflects adherence to a healthy and sustainable dietary pattern, whereas the triglyceride-glucose (TyG) index is a surrogate marker of insulin resistance. Their independent and joint associations with mortality and prevalent cardiovascular disease (CVD) among adults with overweight or obesity remain insufficiently characterized. To investigate the independent and joint associations of PHDI and the TyG index with all-cause mortality (ACM), cardiovascular mortality (CVM), and prevalent CVD among adults with overweight or obesity. This study included 6443 adults aged ≥ 40 years with overweight or obesity from the National Health and Nutrition Examination Survey (NHANES) 2005-2018. Mortality outcomes were assessed longitudinally using linked mortality records, whereas prevalent CVD was assessed cross-sectionally at baseline. Survey-weighted Cox proportional hazards models were used for ACM and CVM, and survey-weighted logistic regression was used for prevalent CVD. PHDI and TyG were evaluated individually and in four joint groups defined by their weighted medians. In the fully adjusted model, compared with the lowest PHDI quartile, the third and fourth quartiles were associated with lower risks of ACM (HR = 0.76, 95% CI: 0.62-0.94; and HR = 0.74, 95% CI: 0.58-0.95, respectively). The highest PHDI quartile was also associated with lower odds of prevalent CVD (OR = 0.60, 95% CI: 0.45-0.81), whereas the association between PHDI and CVM was less consistent. TyG was not significantly associated with ACM, CVM, or prevalent CVD after full adjustment. In the joint analysis, the High-PHDI and Low-TyG group had a lower risk of CVM than the Low-PHDI and High-TyG group (HR = 0.52, 95% CI: 0.30-0.89), whereas the corresponding associations with ACM and prevalent CVD were not statistically significant. No significant additive or multiplicative interaction was observed, and the joint model did not significantly improve prognostic discrimination compared with the individual indicators. Higher PHDI was associated with lower ACM risk and lower odds of prevalent CVD among adults with overweight or obesity, whereas its association with CVM was less consistent. TyG was not significantly associated with ACM, CVM, or prevalent CVD after full adjustment. Although the High-PHDI and Low-TyG group was associated with lower CVM risk, the findings did not support significant interaction or improved prognostic discrimination from jointly considering PHDI and TyG.