Leucocyte telomere length and risk of cardiovascular disease: systematic review and meta-analysis

Abstract

Objective: To assess the association between leucocyte telomere length and risk of cardiovascular disease.

Design: Systematic review and meta-analysis.

Data sources: Studies published up to March 2014 identified through searches of Medline, Web of Science, and Embase.

Eligibility criteria: Prospective and retrospective studies that reported on associations between leucocyte telomere length and coronary heart disease (defined as non-fatal myocardial infarction, coronary heart disease death, or coronary revascularisation) or cerebrovascular disease (defined as non-fatal stroke or death from cerebrovascular disease) and were broadly representative of general populations--that is, they did not select cohort or control participants on the basis of pre-existing cardiovascular disease or diabetes.

Results: Twenty four studies involving 43,725 participants and 8400 patients with cardiovascular disease (5566 with coronary heart disease and 2834 with cerebrovascular disease) were found to be eligible. In a comparison of the shortest versus longest third of leucocyte telomere length, the pooled relative risk for coronary heart disease was 1.54 (95% confidence interval 1.30 to 1.83) in all studies, 1.40 (1.15 to 1.70) in prospective studies, and 1.80 (1.32 to 2.44) in retrospective studies. Heterogeneity between studies was moderate (I(2) = 64%, 41% to 77%, Phet<0.001) and was not significantly explained by mean age of participants (P = 0.23), the proportion of male participants (P = 0.45), or distinction between retrospective versus prospective studies (P = 0.32). Findings for coronary heart disease were similar in meta-analyses restricted to studies that adjusted for conventional vascular risk factors (relative risk 1.42, 95% confidence interval 1.17 to 1.73); studies with ≥ 200 cases (1.44, 1.20 to 1.74); studies with a high quality score (1.53, 1.22 to 1.92); and in analyses that corrected for publication bias (1.34, 1.12 to 1.60). The pooled relative risk for cerebrovascular disease was 1.42 (1.11 to 1.81), with no significant heterogeneity between studies (I(2) = 41%, 0% to 72%, Phet = 0.08). Shorter telomeres were not significantly associated with cerebrovascular disease risk in prospective studies (1.14, 0.85 to 1.54) or in studies with a high quality score (1.21, 0.83 to 1.76).

Conclusion: Available observational data show an inverse association between leucocyte telomere length and risk of coronary heart disease independent of conventional vascular risk factors. The association with cerebrovascular disease is less certain.

Fig 1 Shorter telomere length and risk of coronary heart disease across 20 studies stratified by study design (see table for study acronyms). Study specific estimates were pooled with random effects meta-analysis. Sizes of data markers are proportional to inverse of variance within study. Degree of adjustment: − no adjustment, + adjusted for age and/or sex, ++ adjusted for age, sex, and non-lipid risk factors, +++ adjusted for age, sex, smoking, BMI, diabetes, blood pressure, and lipid markers; ++++ preceding plus adjusted for C reactive protein and physical activity. *Calculated from mean difference in telomere length between cases and controls; †obtained through correspondence. Summary associations for prospective and retrospective studies were not significantly different (P=0.32)

Fig 2 Shorter telomere length and risk of coronary heart disease grouped by recorded study level characteristics. Sizes of data markers are proportional to inverse of variance of relative risk. NR=not reported; NOS=Newcastle-Ottawa scale. *P values for heterogeneity from meta-regression; studies in which characteristic was not reported were not included in calculation of P value; for continuous characteristics, P value reflects linear test of association. †Population source of cohort or controls in case-control studies. Degree of adjustment: −/+ no adjustment or adjusted for age and/or sex, ++adjusted for age, sex, and non-lipid risk factors, +++adjusted for age, sex, BMI, diabetes, smoking, blood pressure, and lipid markers, ++++adjusted for preceding plus C reactive protein and physical activity

Fig 3 Shorter telomere length and cerebrovascular disease risk across 10 studies stratified by study design (see table for study acronyms). Study specific estimates were pooled with random-effects meta-analysis. Sizes of data markers are proportional to inverse of variance within study. Degree of adjustment: +adjusted for age and/or sex, ++adjusted for age, sex, and non-lipid risk factors, +++adjusted for age, sex, smoking, BMI, diabetes, blood pressure, and lipid markers, ++++adjusted for preceding plus C reactive protein and physical activity. *Calculated from mean difference in telomere length between cases and controls; †obtained through correspondence; ‡previously unpublished. Summary associations for prospective and retrospective studies were not significantly different (P=0.15)