Data regarding the impact of different lipid measures on cardiovascular diseases (CVD) and mortality events is not consistent. We aimed to evaluate the relationship between different lipid parameters and incident CVD and mortality events in an Iranian population over a median follow-up of 11.9 years.
The study was conducted on 2532 men and 2986 women aged ≥ 40 years. Multivariate adjusted hazard ratios (HRs), using age as time scale, were calculated for every 1 standard deviation (SD) increase in total cholesterol (TC), logarithm-transformed triglycerides (ln-TGs), low density lipoprotein-cholesterol (LDL-C), high density lipoprotein-cholesterol (HDL-C), non-HDL-C, TC/HDL-C and ln-TGs/HDL-C. Covariates included gender (female as reference), body mass index, education status, low physical activity, smoking, blood pressure status (normotension, prehypertension and hypertension), glucose tolerance status (normal glucose tolerance, prediabetes and diabetes) and lipid lowering drugs. The same analyses were also repeated for tertiles of all lipid measures. Considering the absence of interaction between gender and lipid parameters, we used a sex-adjusted analysis. For analyses of mortality events, prevalent CVD was adjusted as well (All p for interactions > 0.1).
A total of 789 new CVD events, 279 cardiovascular (CV) and 270 non-CV deaths occurred. In multivariate analysis, all lipid measures except HDL-C showed significant risk for new CVD events with HRs ranged from 1.14 to 1.27 for ln-TGs/HDL-C and LDL-C, respectively (all p-values ≤ 0.001). Considering CV mortality, there were significant positive associations between TC, LDL-C, non-HDL-C, TC/HDL-C and CV mortality events in sex-adjusted analysis; however after multivariate analysis, these associations attenuated and reached to null. Applying lipid measures as categorical variables, only TC displayed a positive association with CV mortality in multivariate analysis [TC ≥ 6.14 mmol/L: HR 1.43 (1.04–1.98)].
In multivariate analysis, there were negative significant associations between all lipid measures except HDL-C and non-CV mortality; every 1-SD increase in TC, LDL-C, non-HDL-C, ln-TGs ,TC/HDL-C and ln-TGs/HDL-C was associated with 24, 25, 27, 19, 23 and 17 % decreased risk in non-CV mortality (all p-values ≤ 0.01).
These findings indicate divergent associations of TC, LDL-C, non-HDL-C, TC/HDL-C, TGs and TGs/HDL-C with CVD vs non-CV mortality, demonstrating a higher risk for the former and lower risk for the latter.
Electronic supplementary material
The online version of this article (doi:10.1186/s12986-016-0102-1) contains supplementary material, which is available to authorized users.