|Home | About | Journals | Submit | Contact Us | Français|
Triglyceride levels and their role in ischemic stroke are the topic of intense research (Leonards et al., 2010). Large-scale epidemiological studies with more than 10,000 participants each were able to detect an association between triglycerides and stroke risk (Bansal et al., 2007; Freiberg et al., 2008). Specifically, such studies have revealed a relationship between non-fasting triglyceride levels and ischemic stroke risk. However, the results of studies in smaller cohorts remain inconclusive (Leonards et al., 2010). Are large studies necessary to detect an effect of non-fasting triglycerides because triglycerides exhibit a relatively weak influence on stroke risk or are there other factors involved?
Even in a primary care cohort of 6,621 unselected patients, taken from the prospective cohort DETECT study (diabetes cardiovascular risk-evaluation: targets and essential data for commitment of treatment), we struggled to replicate findings from the larger studies (Bansal et al., 2007; Freiberg et al., 2008). Details of the DETECT study design can be found elsewhere (Pieper et al., 2005; Wittchen et al., 2005; Schneider et al., 2008). Briefly, however, triglyceride levels were measured in venous blood samples taken at time of patient enrolment using a Roche Modular automatic analyzer (Roche Diagnostics Scandinavian, Bromma, Sweden). Fasting was defined as having eaten no earlier than 8h prior to phlebotomy. The endpoint chosen for this analysis was incident ischemic stroke as reported by physicians and/or patients occurring in the 4–5years follow-up period.
Consistent with Freiberg et al. (2008), triglyceride levels were stratified into quintiles (<89, 89–176, 177–265, 266–353, and ≥354mg/dL). Hazard ratios (HRs) were then calculated by Cox proportional hazard model for individual strata of triglycerides and verified by Schoenfeld residuals. Additionally, HRs were calculated for dimensional triglyceride levels (increase of 1 SD) and scaled to an increase of 89mg/dL for interpretation to follow (Freiberg et al., 2008). Also for reasons of comparison, age, total cholesterol, alcohol consumption, smoking, hypertension, atrial fibrillation, lipid lowering therapy, diabetes, BMI, and HDL were included in triglyceride analyses (Freiberg et al., 2008). Because of the low incidence base rate of 97 strokes (23 in women), we applied a forward stepping algorithm to minimize covariates and avoided over-adjustment (Triglyceride Cononary Disease Consortium and Emerging Risk Factors Collaboration, 2010) using a minimum triglyceride HR change of 5% as an inclusion criterion of a covariate into the model. All statistical analyses were conducted with the software package STATA 10.2 (Stata Corp.).
We found in the unadjusted model and models adjusted for age only, a significant association of fasting triglyceride levels with ischemic stroke in women (HR 1.2, 95% CI 1.2–1.4, p<0.01) but not in men (HR 1.0, 95% CI 0.8–1.1, p>0.5). This finding concurred with two long term cohort studies from Japan (Iso et al., 2007) and Norway (Njolstad et al., 1996) suggesting a possible gender difference in triglycerides’ predictive role for stroke. No significant association between fasting or non-fasting triglycerides and stroke was found however after multiple adjustments (Table (Table11).
Our study may have been underpowered and the low number of events was a major limitation. A post hoc power analysis revealed a 13.1% chance of incorrectly rejecting the null hypothesis at a 95% confidence interval. Based on these findings, one may conclude that if any the relationship between triglycerides and stroke is relatively weak and possibly limited to females.
Overall, however, the association of fasting and non-fasting triglyceride levels with stroke remains inconclusive. Despite several proposed mechanisms describing how triglycerides might possibly increase the risk of stroke (Papagianni et al., 2004; Karepov et al., 2008), the observed associations between triglycerides and vascular events do not provide much support of a causal relationship (Third Report of the National Cholesterol Education, 2002). On the other hand, a generalized lack of standardization could explain the discrepancies between our results and other studies. Neither time points of blood draws after the last meal nor the content of that meal have been standardized in most of the studies discussed here. We therefore advocate a more standardized approach to future research in an effort to garner meaningful outcomes without needing to include tens of thousands of participants. A prospective study (Ebinger et al., 2010), using a novel oral triglyceride tolerance test (blood draws after a 10h fast and 3h after intake of 250mL of 32% fat cream) to test this theory and thereby circumvent the possible problems associated with a lack of standardization is currently under way.
Diabetes cardiovascular risk-evaluation: targets and essential data for commitment of treatment (DETECT) is a cross-sectional and prospective-longitudinal, nationwide clinical epidemiological study. DETECT is supported by an unrestricted educational grant of Pfizer GmbH, Karlsruhe, Germany. Principal investigator: Prof. Dr. H.-U. Wittchen; Staff members: Dr. H. Glaesmer, E. Katze, Dipl.-Math. J. Klotsche, Dipl.-Psych. L. Pieper, Dipl.-Psych. A. Bayer, Dipl.-Psych. A. Neumann. Steering Committee: Prof. Dr. H. Lehnert (Magdeburg, Coventry), Prof. Dr. G.K. Stalla (Munich), Prof. Dr. M.A. Zeiher (Frankfurt), Professor Dr. M. Wehling (Mannheim); Advisory Board: Prof. Dr. W. März (Graz/Heidelberg), Prof. Dr. S. Silber (Munich), Prof. Dr. Dr. U. Koch (Hamburg), PD Dr. D. Pittrow (Munich, Dresden). The research leading to these results has received funding from the Federal Ministry of Education and Research via the grant Center for Stroke Research Berlin (01 EO 0801), the Volkswagen Foundation (Lichtenberg program to Matthias Endres), DFG (NeuroCure), and European Union (European Stroke Network).