Ancestry, Telomere Length, and Atherosclerosis Risk

Cardiovascular profiles in African Americans and European Americans

African Americans (AfAms) and European Americans (EuAms) display differences in cardiovascular risk factors and mortality from cardiovascular disease (CVD). AfAms show a higher prevalence than EuAms in most cardiovascular risk factors¹. These include high blood pressure^1–3, left ventricular hypertrophy¹, obesity¹ and type 2 diabetes mellitus^1,4. National US epidemiological data indicate that dyslipidemia is less prevalent in AfAms than EuAms¹. However, based on the Multi-Ethnic Study of Atherosclerosis (MESA), AfAms and EuAms have a similar prevalence of dyslipidemia, which is less controlled in AfAms⁵. Other studies also found lower rates of cholesterol testing, dyslipidemia awareness, treatment, and therapeutic goal attainments among AfAms than EuAms^6,7. In some of these studies, ethnic disparities were attenuated by adjustment for social factors and healthcare access⁵, while in other studies they were not⁷. Notably, compared with EuAms and Mexican-Americans, AfAms displayed a higher risk of having one or two of the three major cardiovascular risk factors, i.e., diabetes, dyslipidemia and hypertension; compared with EuAms, AfAms were also more likely to have all three conditions⁸. While these ethnic differences might account for the higher CVD mortality in AfAms than EuAms^9,10, the ethnic differences cannot explain consistent observations of less atherosclerosis in the form of plaques, arterial calcifications and occluded coronary arteries in AfAms than in EuAms (Table 1)^1,11–13. This applies to individuals without or with similar cardiovascular risk factors.^12–18.

Table 1.

Prevalence of cardiovascular risk factors and of coronary atherosclerosis in African Americans and European Americans.

	EuAm Women	AfrAm Women	EuAm Men	AfrAm Men	Ref #
CV Risk Factors (%)
Tobacco	19.8	15.4	22.6	22.8	1
HBP (140/90 or medic)	27.7	42.9	30.1	40.1	1
Diabetes (Medic or Gl>1.26g/l)	7.8	16.9	11.6	18.6	1
Dyslipidemia (>200mg/dl)	45.9	40.7	39.9	37.4	1
Obesity (BMI>30 kg/m2)	33.0	58.0	34.0	38.0	1
Coronary Atherosclerosis (%)
Coronary Artery Calcifications
33–45 years	11.3	4.9	17.6	5.2	11
45–84 years	44.6	36.5	70.4	52.1	12
* Coronary Stenosis
% with Gensini score >10	26.7	14.9			13
% with severe stenosis	34.0	20.0			13

EuAm = European American; AfAm = African American;

^*men and women together

Key Tenet

Here we propose that longer telomeres in AfAms than EuAms^19–21 may attenuate the atherosclerotic risk in AfAms and explain in part the mortality cross-over between elderly AfAms and EuAms. First, however, we provide a brief background on the biology, epidemiology and evolution of telomeres.

Telomere length dynamics

Comprising TTAGGG tandem repeats and their telomere-binding protein complex (shelterin), mammalian telomeres safeguard the ends of the chromosomes²². As somatic cells divide, their telomeres undergo progressive shortening, a process that ultimately leads to cessation of replication, i.e., replicative senescence²³. In humans, embryonic stem cells express the enzyme telomerase, a reverse transcriptase that adds telomere repeats to the ends of the chromosomes, thereby preventing telomere shortening during early intra-uterine life²⁴. However, as telomerase is repressed in somatic tissues during extra-uterine life, age-dependent telomere shortening in these tissues largely reflects the replicative histories of stem cells/progenitor cells on top of the somatic cell hierarchy²⁵. In this way, TL in skeletal muscle, a minimally proliferative tissue, is longer than TL in leukocytes, which represent the highly proliferative hematopoietic system²⁶. Moreover, to accommodate the demands of the growing soma, stem cell/progenitor cell replication is much faster during growth and development. Therefore, the rate of telomere shortening is much faster in childhood than during adulthood^26,27.

Most epidemiological studies have used leukocyte TL (LTL) as a proxy for TL in other somatic tissues, given that TL is ‘proportional’ across somatic tissues within the individual, i.e., a person with short (or long) TL in one somatic tissue has short (or long) TL in other somatic tissues²⁶, regardless of the replicative history of these tissues. This is because at birth and throughout the life course the variation in TL, which is highly heritable²⁸, across individuals is much wider than TL variation across somatic tissues within the individual.

Telomere length evolution across mammals

Why is TL in humans as long as it is? Moreover, is this length arbitrary or is it determined by evolutionary principles? TL is short and telomerase activity in somatic tissues is repressed in long-lived, large mammals compared with short-lived, small mammals²⁹. One potential explanation for these findings is that short telomeres and repressed telomerase diminish cancer risk by curtailing replicative potential^25,30,31. The trade-off might be diminished tissue repair through cell replication and therefore increased propensity to degenerative diseases. Thus, across long-lived, large mammals, telomeres might converge to an optimal length that strikes a balance between cancer and degenerative diseases during the life course of animals from a given species.

TL is short in humans compared to most mammals²⁹. Still, TL variance across the general population is considerable, amounting to 3–4 kilobases (kb)^27,32. Detrimental mutations in telomere maintenance genes can result in specific diseases that are the outcome of extremely short TL²⁹, but these diseases are relatively rare. The question then is whether in the general population, individuals with constitutively short telomeres have a higher risk for degenerative diseases. This might be the case for atherosclerosis, which is perhaps the main degenerative disease that afflicts contemporary humans.

Telomere length and atherosclerosis

Two meta-analyses concluded that short LTL is associated with CVD, based on data from individuals of principally European ancestry^33,34. In general, short LTL is associated with arterial aging, as expressed in arterial stiffness³⁵, coronary artery calcifications^36,37, and severity of atherosclerotic plaques in the coronary arteries^38,39 and the carotid arteries^40–43.

What might be the explanation for these associations? For years, the dominant thinking in telomere epidemiology has been that LTL is a passive biomarker of the cumulative burden of inflammation and oxidative stress during adult life. Thus, individuals with a higher burden of inflammation and oxidative stress, including obese individuals, smokers, patients with diabetes and those with atherosclerotic CVD, were presumed to have a faster age-dependent LTL shortening during adult life⁴⁴.

However, the variation in LTL across adults of the same age are largely the outcome of the inter-individual variation in LTL during the first two decades of life^27,32 rather than inter-individual variation in the rate of LTL attrition during adulthood⁴⁵. Thus, short LTL might precede by several decades the manifestations of atherosclerotic CVD and related metabolic risks^46,47. For instance, short LTL is observed long before the onset of carotid lesions and predicts their progression over time⁴⁰. Short LTL also precedes the clinical manifestation of insulin resistance^47,48. Moreover, not only short LTL, but also single nucleotide polymorphisms (SNPs) associated with short LTL^39,49 are associated with increased propensity for atherosclerotic CVD. These findings principally exclude reverse causality. i.e., atherosclerotic CVD engenders short LTL, supporting the thesis that short LTL antecedes the clinical manifestations of atherosclerotic CVD.

Longer leukocyte telomere length in individuals of recent African ancestry than in individuals of recent European ancestry

In light of the short LTL-CVD association, it is noteworthy that compared with EuAms, AfAms display not only a longer LTL^19–21,50, but also a higher prevalence than EuAMs of SNPs associated with a longer TL⁵¹. Although the underlying reasons are not well understood, findings that LTL of sub-Saharan Africans is even longer than that of AfAms suggest that the northbound migration from equatorial Africa might have resulted in TL shortening in Europeans²⁰. One potential cause of telomere shortening in Europeans might be melanoma. Previous investigators had excluded melanoma as an evolutionary force⁵². However, given that not only longer LTL but also alleles associated with longer LTL are over-represented in patients with melanoma^53,54, without evolutionary-driven telomere shortening, whites of European ancestry would have been more susceptible to melanoma than they are at present²⁰.

Telomere length and “mortality crossover” between AfAms and EuAms

Overall, AfAms have a shorter life expectancy than EuAms. However, after the age of 80, AfAms display a longer life expectancy than EuAms– a phenomenon referred to as the “mortality crossover”^55,56. Corti et al.⁵⁵ reported that after the age of 80 AfAms had a 25% lower risk of all-cause mortality than EuAms, a difference that was attributed to a 55% lower atherosclerotic coronary heart disease mortality in AfAms. Other studies suggested that socio-economic status and other environmental factors explain the mortality crossover between AfAms and EuAms^56,57. However, in most of these studies the mortality crossover remained significant after adjustments for environmental factors⁵⁵. Thus, the mortality crossover might be related to different trajectories of coronary heart disease, which could be partially explained by the longer LTL in AfAms vs. EuAms.

Conclusions and perspectives

Since the inception of the discipline of telomere epidemiology, LTL has been regarded as a passive “biomarker” of aging and atherosclerotic CVD. However, an emerging body of research points to an active role of TL in atherosclerosis, which is reason enough to seek further insights into the potential role of TL in the different propensities of AfAms and EuAms to atherosclerotic CVD. The longer LTL in AfAms than EuAms might partially explain the less susceptibility of AfAms than EuAms to atherosclerosis and their longer life expectancy after the age of eighty years. Perhaps, the first step towards gaining further insight into these intriguing ethnic differences is to perform large-scale genome-wide association studies of LTL in multiple ethnic groups. Identifying new LTL- associated SNPs might help construct a comprehensive genetic map leading to mechanistic insights into the role of TL, as expressed in LTL, in a host of human diseases..