Interaction of Stress, Lead Burden, and Age on Cognition in Older Men: The VA Normative Aging Study

Junenette L. Peters; Marc G. Weisskopf; Avron Spiro III; Joel Schwartz; David Sparrow; Huiling Nie; Howard Hu; Robert O. Wright; Rosalind J. Wright


Environ Health Perspect. 2010;118(4):505-510. 

In This Article


In this cohort of older men, increased self-report of stress was related to lower cognition. Moreover, an inverse association with blood lead and MMSE was more pronounced among those who reported higher perceived stress using the PSS than among those who reported lower perceived stress. In addition, the combination of perceived stress and lead modified the relationship between age and cognition. This study corroborates laboratory studies and one other human study that indicated lead and stress interact to affect cognitive function (Cory-Slechta et al. 2008; Glass et al. 2009) and further supports the theory that cognitive impairment is not singularly a result of aging but due to risk factors working in concert.

Previous studies have reported that heterogeneity in cognition is especially pronounced in the elderly compared with younger adults (Lupien et al. 2005). Sapolsky et al. (1986) reported that stress exposure over the life course, likely mediated through disrupted stress hormones, significantly affect the aging process. Our group and others have reported a relationship between biomarkers of lead and cognition as well as a negative interaction between lead and age on cognition in older adults (Balbus-Kornfeld et al. 1995; Shih et al. 2007; Weuve et al. 2009). To our knowledge, this is among the first studies to assess the interaction of lead and psychological stress on cognition in older men and the first to investigate the combined association of lead and stress as a modifier of the relationship between age and cognition.

Aging has been associated with an increase in oxidative stress and elevated glucocorticoids (Pardon 2007; Sapolsky et al. 1986). It has also been associated with impaired plasticity of the HPA axis in experimental studies and appears to predict negative effects of stress (Lupien et al. 2007; Pardon 2007). The outcome of chronic stress and aging on brain function shows similarities; however, stress and aging seem to impact cognition via different underlying mechanisms (Pardon 2007). To add to the complexity, there is a paradox in stress–aging interactions: Although some evidence suggests vulnerability to stress can increase with age, other data indicate that the threshold of tolerance to stress may increase with age (Pardon 2007).

As proposed by Cory-Slechta and others (2008), the interactive effect of lead and stress may follow a multihit model. Lead and stress can both work through the activation of the HPA axis, which results in the release of a cascade of hormones such as cortisol. Disruption of the optimal balance of these stress hormones may enhance central nervous system vulnerability if they present insults on the same system of the brain via different mechanisms, overwhelming the ability of the system to maintain homeostasis (Cory-Slechta et al. 2008).

We observed some differences in the relationship between the two stress measures on cognition and their interactive relationship with lead. We found a significant negative relationship between the most stressful life event measure and MMSE score and a marginal negative relationship with PSS. In addition, even though the direction of the association was the same, only PSS showed significant interaction with lead in association with MMSE. The most stressful life event rating assesses a stressful event judged by the respondent to have a negative impact, whereas the perceived stress measures the individual's perception of the current demands exceeding the ability to cope (Cohen et al. 1993). These measures, although correlated, may measure different constructs (i.e., may have independent relationships with disease risk and be mediated by different processes) (Cohen et al. 1993). In a study looking at the stress relationship with the common cold, Cohen et al. (1993) proposed that a) measures such as the life events measure may pick up acute or direct effects, whereas the PSS may be indicative of dispositional affect (i.e., an overall predictable way of responding to situations) and that b) the former may drive the development of symptoms, whereas the latter may be related to increased susceptibility.

We also found a difference between the interactive association of stress and lead among the measures of lead exposure: a significant interaction with blood lead, marginal interaction with patella lead, and no interaction with tibia lead. Of note, there were substantially fewer bone lead measures than blood lead. However, our results for blood lead remained significant after we restricted the blood analyses to those with only bone lead measurements (data not shown). In a previous study in this cohort that looked at the cross-sectional relationship between lead and elevated MMSE, Wright et al. (2003) observed a significant inverse relationship with blood and patella lead but not with tibia lead. These authors also found interactive relationships between blood and patella lead with age predicting elevated MMSE (Wright et al. 2003). The relationship with blood and patella lead is consistent with the theory that bone lead is chronically released into blood, that mobilization rate in aging differs by bone type, and that this mobilized lead contributes to an acceleration in cognitive decline (Wright et al. 2003). Stress has been found to mobilize bone lead stores in animals (Bushnell et al. 1979). In human studies, cortisol decreased mineral density and increased bone loss (Cetin et al. 2001; Dennison et al. 1999; Reynolds et al. 2005). Of interest, the effect of cortisol differs by sex and was observed to affect trabecular (e.g., patella) bone in older men (Dennison et al. 1999; Reynolds et al. 2005). The differential effect of cortisol by sex may partially explain significant results found by Glass et al. (2009) between neighborhood hazards and tibia lead (only tibia lead tested) on cognitive function in a mixed sample of adults. In addition, the contrasting results may reflect a difference in the measure of stress or the limitation of the MMSE used in our study to differentiate between domains of cognitive function that may be associated with different lead exposure measures (Weisskopf et al. 2007).

The combined effect of lead and stress is of particular concern, because HPA axis dysfunction has been linked to myriad disorders, in addition to cognitive impairment, including cardiovascular and metabolic diseases and psychiatric disorders (Cory-Slechta et al. 2008). Indeed, we showed the interactive association of lead and stress on blood pressure and the prospective risk of hypertension in this same cohort (Peters et al. 2007). It also is relevant in the context of low socioeconomic populations where the prevalence of these disorders is high and stress and lead tend to co-occur. Thus, the public health implications may be significant, given the possibility of improved neurobehavioral performance after reducing blood lead, which has been shown in serveral studies (Chuang et al. 2005; Schwartz et al. 2001; Winker et al. 2006), and stress (Bremner et al. 2008).

We note a number of limitations that may be addressed in future research. This study is cross sectional, so temporality cannot be established. It is conceivable that deficits in cognitive function could be a source of stress or produce stressful experiences. As eluded to earlier in the discussion, use of the MMSE may be considered a limited assessment of cognition; however, the strength of the MMSE is that it is a general measure that is widely used and understood. We evaluated relationships using two measures of psychological stress and three measures of lead exposure, raising the issue of multiple comparisons. However, we chose to make these comparisons because of reported differences between the stress and lead exposure measures in their relationship with disease. In addition, although we controlled for a number of risk factors, there is the risk of omitted or inadequately controlled confounders. In addition, this study was not conducted in a low socioeconomic population where there is a greater likelihood of dual exposure to stress and lead. Finally, there are noted differences in the association of stress with cognition and in the interactive effect among males and females (at least in laboratory studies) (Cory-Slechta et al. 2008; Virgolini et al. 2008; Wang et al. 2007). This finding suggests the need to look more closely at sex differences and the relationships found in our study.

In summary, our results show that stress is associated with lower cognition and modifies the relationship of age to cognition among community-dwelling older adult males. Furthermore, stress negatively modifies the relationship of blood lead and cognition, and combined high lead and high stress negatively modify the association of age with cognition.


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.
Post as: