Table 3. Smoking impacting on DNA methylation-based telomere length.
| DNAmTL | LTL | ||||||
| Cohort | N | Variable | T | P | T | P | |
| FHS1 | 878 | Pack years | -2.08 | 3.81E-2 | 0.54 | 0.59 | |
| WHI BA231 | 97 | Pack years | 0.45 | 6.53E-1 | 1.30 | 0.2 | |
| JHS1 | 100 | Smoker | -3.00 | 3.38E-3 | -1.87 | 6.40E-2 | |
| LBC 19212 | 404 | Pack years | -3.53 | 4.59E-4 | -0.92 | 0.36 | |
| LBC 19362 | 796 | Pack years | -3.55 | 4.04E-4 | 0.55 | 0.58 | |
| UK Twin2 | 792 | Smoker | -2.79 | 5.23E-3 | -3.23 | 1.23E-3 | |
| BHS1 | 831 | Smoker | -6.83 | 1.70E-11 | -1.83 | 6.77E-2 | |
| ALL | 4039 | Smoking impact | -8.55 | 1.21E-17 | -2.19 | 0.029 | |
The Smoker variable was coded as a binary variable for never versus ever smokers, where estimate was referred to ever smokers.
1Telomere length was based on terminal restriction fraction measurement by Southern blotting.
2Telomere length was measured using quantitative real-time polymerase chain reaction (qPCR).
The table summarized the meta-analysis for the smoking impact on plasma DNA methylation based telomere length (DNAm TL) versus the impact on qPCR or southern blot-based leukocyte TL. The association tests were performed at each study cohort respectively then combined by the Stouffer’s method weighted by square root of sample size. The smoking variable was based on a binary variable for former/current smokers versus never or pack years provided the variable was available.