Average telomere length decreases with aging, and is commonly measured in immune cells taken from a blood sample. Telomeres are a part of the mechanism that limits the number of times most cells can divide. Tissues are made up almost entirely of such limited cells, each losing a little telomere length during each division. When telomeres become short, the cell self-destructs or becomes senescent and ceases to divide. An associated stem cell population supports the tissue by delivering a continual supply of new daughter cells with long telomeres to replace the losses. Thus average telomere length is a function of cell division rates and cell replacement rates. Since stem cell activity declines with aging, it shouldn't be surprising to see that average telomere length does as well. In fact average telomere length in immune cells is highly variable between individuals, and even with circumstances for the same individual, and the rate of decline is small. It thus makes a pretty terrible measure of aging, a point reinforced by the numbers in this open access paper.
In the current study, we first examined the cross-sectional associations between leukocyte telomere length (LTL) and age, and, like previous reports, we found an inverse relationship with increasing age. Second, using up to five measurements across 20 years, we found that LTL decreases with age in a two-slope model with a small acceleration of decline after 69.3 years of age. Men have shorter telomere lengths than women, and genetic variation has an additional influence on overall LTL.
Several earlier studies have reported an inverse association between age and telomere length, as did we, and we further demonstrated that women have longer LTL, which is in line with earlier research. Taking our results and prior literature together, shorter telomeres in men could result from very small but consistent attrition throughout adulthood rather than a steeper decline compared to women in old age. Moreover, previous literature from cross-sectional and longitudinal studies has suggested a linear relationship between telomere length and age. We found both the one-slope and the two-slope models to be significant, with a substantially better fit of the latter. While the overall average trend was linear, there was systematic variability around the average trend, better described in a two-slope model accounting for more individual differences. The magnitude of this age-related decline was small overall, and with slight acceleration in the old-old. This observation is in line with earlier research in the field where faster decline in LTL is believed to take place in childhood and old age.
The two-slope trajectory analyses supported both familial and non-familial influences on LTL, with equal contributions to average LTL level (at age 69) and non-familial sources featuring more prominently in the change before age 69 than after age 69. This suggests that in young-old age, individual-specific lifestyle factors may prove more relevant to accelerated LTL shortening above and beyond familial and environmental contributions to overall LTL; however, in old-old age, familial factors may become increasingly salient to accelerated LTL shortening. Moreover, we note that the variation in rate of change was larger in young-old age; hence, evaluating variation in trajectories beyond the assumption of simple linearity and average trends is important for understanding etiological underpinnings.