The relationship between adult socio-economic position and white blood cell telomere length in 1552 female twins reported by Cherkas et al. (2006) has aroused much public interest and comment, including news coverage on the BBC and other networks (http://news.bbc.co.uk/go/pr/fr/-/1/hi/healthy/5188742.stm). There seems to be a fascination in being able to encapsulate the effects of social inequalities on aging in a single measure of white blood cell telomere length, translate differences between nonmanual and manual workers into a single figure (seven biological years), however imprecise, and attribute the cause to psychosocial stress, in particular work stress, despite the lack of evidence for such a mechanistic pathway in the sample under study. As social class was assigned on the basis of the highest occupational class of the female twin or their partner, the effects of work stress would have to act indirectly in many cases, through effects on partners. We should also be cautious of the findings on 17 twin pairs most discordant for socio-economic position that suggest telomere length was shorter in the more disadvantaged twin, as the sample size was small and no results were given for concordant twins.
The paper is of interest because it is the first report of a relationship between telomere length and socio-economic position. However, as Thomas von Zglinicki noted in a comment quoted in the recent BBC news story the finding is ‘astonishingly weak’ given the strong effect of socio-economic position on chances of health and survival, and according to Dr von Zglinicki seems not to have been supported in other studies, still unpublished.
Many previous studies from the 1970s and 1980s documented, but did not explain, social class differences in health (Blaxter, 1981; Feinstein, 1993); we can only hope that this paper by Cherkas et al. does not usher in a similar mass of papers, based on similarly weak cross-sectional methods, of social class differences in various biomarkers of aging. (If such studies are published, editors should be alert for reports of both positive and null findings.) Longitudinal studies are needed to test the hypothesis that it is exposure to the disadvantages (psychosocial or material) of being in the manual social class in adult life that increases the risk of accelerated aging, by investigating whether adult social class is related to changes in telomere length during later life and, if so, whether this mediates the effects of social class on functional decline and risk of age-related diseases.
White blood cell telomere length is highly heritable and highly variable from birth and throughout life. A one-off measure reflects the combined effects of initial length and the rate of attrition. The lifetime trajectory in telomere attrition apparently consists of a rapid loss until around age 5, then a period of stability until early adulthood, followed by gradual decline thereafter (Frenck et al., 1998; Zeichner et al., 1999). Thus, aging (in terms of white blood cell telomere length) is occurring at a faster pace at the beginning of life; indeed the dichotomy between growth and aging becomes increasing blurred, and should probably be viewed as part of the same continuum (Aviv et al., 2003). From a life course aging perspective (Kuh & Ben-Shlomo, 2004), it clearly would be interesting to know whether poor childhood social conditions or early growth patterns characterized by low birth weight followed by accelerated postnatal weight gain (both associated with an increased risk of cardiovascular and other age-related chronic diseases) are associated with a greater rate of attrition in telomere length than better childhood conditions or more favourable patterns of weight gain (Demerath et al., 2004). The rate of telomere attrition with height growth would also be of interest given the debate over whether shorter or taller people have a shorter lifespan or a higher risk of cardiovascular disease (McCarron et al., 2002; Samaras et al., 2003), and that shorter telomere length is related to increased mortality, particularly from cardiovascular disease (Cawthon et al., 2003). There is plenty of potential confounding to unravel, and distinctions to be made between factors that predict individual and population differences. Cherkas and colleagues would probably say that the lack of a relationship between education and telomere length, and the findings of the twin pair analysis that controls for genetic and early life effects, suggest that the cross-sectional relationship between adult socio-economic position and telomere length is unlikely to be explained by early life experience. However, life course research that investigates the lifetime trajectory of telomere length in relation to growth, lifetime socio-economic conditions and later functional decline and disease has the potential to be of greater relevance for understanding the aging process, and inequalities in that process, than a raft of cross-sectional studies that link adult social class to telomere length.