We found a strong association between physical fitness and telomere length in 944 patients with CHD. After adjustment for other patient characteristics, including markers of cardiac disease severity and physical inactivity, participants with low exercise capacity (<5 METS) had a 94% greater odds of having short telomere length than those with high exercise capacity (>7 METS). In addition, participants with low exercise capacity had shorter mean telomere length than those with high exercise capacity (T/S ratio: 0.85 vs. 0.92, p
0.005). This is equivalent to a difference of 169 base pairs (5325 vs. 5494 base pairs). Given that telomeres in this population decrease at an average rate of approximately 42 base pairs/year 
, this can be viewed as equivalent to a 4 year age difference between those with low versus high physical fitness.
Our study demonstrates for the first time a strong relationship between physical fitness and telomere length in a large sample of patients with existing CHD, after controlling for many other patient characteristics, such as severity of CHD. We extend the findings of previous studies on cardiovascular disease and telomere length 
by demonstrating that telomere length varies with fitness level within a population with known cardiovascular disease. The linkage between telomere length and an objective measure of fitness, rather than a self-reported one, is significant because it isolates aerobic fitness, rather than factors that tend to co-vary with greater self-reported physical activity, as being associated with longer telomeres. Many hard-to-measure factors that affect telomere length, such as nutrition and social stressors 
, can differ between those who report exercise and those who do not, potentially limiting the conclusions that can be drawn from self-reported exercise 
. This may explain why previous studies examining the relationship between self-reported physical activity and telomere length have yielded mixed results 
The causal direction of the association between fitness and telomere length cannot be determined by our analyses. Short telomeres could reduce exercise capacity by decreasing the function of the cardiovascular system, an idea supported by the fact that functional telomeres are required for viability of cardiovascular cells in-vitro and that deficient telomeres have been shown to cause cardiovascular disease in mice 
. Alternatively, both low exercise capacity and short telomeres could result from common genetic or environmental factors. A third alternative is that physical inactivity could both reduce exercise capacity and shorten telomeres.
Several studies have suggested mechanisms by which physical inactivity may lead to shorter telomere length. The strongest evidence comes from research by Werner et al. showing that inactivity in mice alters the protein complexes that regulate leukocyte telomere length and structure. In this study, mice were randomized to running or no running wheel conditions for 3 weeks. The sedentary mice then showed lower telomerase activity, lower expression of telomerase reverse transcriptase (TERT), and lower telomere-stabilizing telomere repeat binding factor TRF2 compared with the aerobically conditioned mice 
. A subsequent observational study in humans found that sedentary individuals had decreased leukocyte telomerase and telomere-stabilizing protein compared to individuals with long-term endurance training 
. Other research has suggested indirect mechanisms by which physical inactivity may lead to shorter telomeres. Moderate exercise has been shown to increase anti-oxidant capacity 
, and human studies have shown that higher oxidative stress levels lead to accelerated telomere shortening in leukocytes 
. Furthermore, in-vitro studies have demonstrated that oxidative stress increases telomere attrition 
and decreases telomerase activity 
in numerous cell types 
. Yet another mechanism is that exercise may upregulate anti-inflammatory processes 
, and increased inflammation may then contribute to telomere attrition 
Previous studies have shown that both exercise capacity and telomere length are powerful independent predictors of mortality among men with cardiovascular disease 
. Given the strong association between exercise and telomere length observed in our study, we investigated whether shorter telomere length mediated the association between exercise capacity and mortality. We found, however, that adjusting for telomere length had very little effect on the strong relationship between exercise capacity and mortality. These results suggest that telomere length does not contribute to the association between exercise capacity and mortality in patients with cardiovascular disease.
Among the strengths of the present study is the characterization of numerous clinical, biological, and psychosocial covariates that enable us to exclude possible confounders. However, our study has several limitations that should be considered in the interpretation of our results. First, the association reported in this study is cross-sectional, so no conclusions can be reached regarding causality. Next, no genetic polymorphisms were analyzed in our study, but genetics are known to impact telomere length in people with coronary artery disease 
. Another limitation of our study is that most participants were urban, elderly men, and therefore the results may not generalize to other populations. Finally, our study population consisted entirely of people with stable coronary heart disease, and the results may not apply to either healthy individuals or those immediately post-myocardial infarction.
In summary, our study shows a strong relationship between exercise capacity and telomere length in a population of patients with stable coronary heart disease. The association between self-reported physical activity and telomere length became non-significant after multivariable adjustment. Whether poor physical fitness leads to shorter telomeres, or vice versa, and whether common genetic or other factors may reduce both telomere length and exercise capacity, deserve further study.