Associations between superoxide dismutase, malondialdehyde and all-cause mortality in older adults: a community-based cohort study

Aging is the accumulation process of diverse deleterious changes in the cells and tissues with advancing age, leading to increased risk of disease and death [1]. The free radical theory proposes that aging is the cumulative result of oxidative damage to the cells and tissues of the body that arises primarily as a result of aerobic metabolism [2]. Substantial correlative evidence suggests a link of oxidative stress, an imbalance between the production and degradation of reactive oxygen species (ROS), to aging [3]. Severe oxidative stress produces excessive ROS, such as superoxide, hydroxyl, and hydrogen peroxide, which may result in damage to DNA, proteins and other cell injuries [4, 5]. The body encloses a complex antioxidant defense grid that collectively acts against free radicals. The first line defense antioxidants included superoxide dismutase (SOD), catalase and glutathione peroxidase, which play an indispensable in the entire defense strategy of antioxidants [6]. SOD is the first detoxification enzyme and most powerful antioxidant in the cell [6]. In mammals, there are three forms of SOD: SOD1 or the Cu/Zn-SOD, SOD2 or the Mu-SOD, and SOD3 or the extracellular SOD [7]. In Saccharomyces cerevisiae [8, 9] and Drosophila models [9, 10], SOD may protect against oxidative damage and extend life span. In addition to ROS, oxidative stress may also induce uncontrolled lipid peroxidation, which in turn, can result in cell injuries via DNA damage and directly inhibit proteins [11, 12]. Malondialdehyde (MDA) is a stable end product of lipid peroxidation and therefore can be used as an indirect measure of the cumulative lipid peroxidation. Both SOD and MDA have been considered as biological markers of oxidative stress [13].

Although in vitro data and studies in animal models have suggested that SOD and MDA may be associated with aging, epidemiologic evidence regarding their associations with mortality remains sparse. A cohort of 507 healthy residents indicated that high serum SOD activity was associated with protective effects against mortality from cancer [14]. This finding was consistent with a nested case-control study of 3653 participants [15]. In addition, blood MDA level has been shown to be associated with mortality in patients with HIV [16], chronic heart failure [17], and breast cancer [18]. However, epidemiologic studies investigating the associations of blood SOD activity, or MDA with all-cause mortality in community-dwelling older adults are still lacking. In this study, we prospectively evaluated these associations using the datasets from the Chinese Longitudinal Healthy Longevity Survey (CLHLS) in longevity areas.

In this prospective cohort study of community-dwelling older adults, we observed a dose-response relationship between blood SOD activity and all-cause mortality in women, but not in men. The association was unlikely to be modified by age, residence, education, smoking, alcohol drinking, vegetable, fruit consumption, and BMI. We did not observe a significant association between MDA level and mortality.

The free radical theory positions an increased production of ROS at the center of age-related cellular injuries [3, 30]. Because SOD enzymes play a key role in regulating cellular ROS levels, decreased SOD activity may increase the vulnerability of cells to oxidative stress and lead to premature death [7]. In Saccharomyces cerevisiae [8, 9] and Drosophila models [9], over-expression of SOD protects against oxidative damage and extend life span. Feeding SOD mimetic drugs, such as Euk-8 and MitoQ, increase lifespan in Drosophila melanogaster [31]. In humans, the association between SOD activity with life span remains unclear although some observational studies have suggested that high SOD activity is associated with decreased mortality from cancer [14, 15]. In a cohort of 507 healthy Japanese participants aged 40 years or above with an 18-year follow-up, there was no significant association between blood SOD activity and all-cause mortality (HR for the lowest quartile vs. the highest quartile 1.35, 95% CI 0.64 to 2.97) [14]; this failure to detect an association may be a result of the small sample size and low death rate (12.6%) that would have provided inadequate study power to detect a difference. The negative association between SOD activity and mortality agreed with previous studies investigating SOD genes. In a Danish cohort of 1650 older adults (aged 92 to 93 years), individuals with the MnSOD rs4880 C alleles, known to be associated with high SOD2 enzyme activity [32], were associated with decreased mortality (HR for CC vs. CT 0.91, 95% CI 0.86 to 0.97) [33]. A prospective cohort study of 2799 subjects aged > 40 years suggested that the homozygous T allele of rs1041740 was associated with all-cause mortality (HR for TT vs. CC 1.53, 95% CI 1.01 to 2.30) [34].

Perhaps one of the most striking findings of this study was the sex-specific difference in SOD. Consistent with previous studies [35, 36], our results suggested that the average SOD activity was higher in women than in men. A sex-specific difference in the association of SOD activity with mortality was also shown in our analyses. In a Drosophila model, higher SOD activity was far more likely to increase lifespan in females than in males [10], and feeding exogenous antioxidants increased the lifespan of only female flies [31]. The mechanism for the sex-specific differences of SOD on longevity remains unclear. Possible explanations include 1) differences in metabolic rates may lead to different levels of oxidative damage between males and females [37], 2) endocrinological differences by sex may result in different sensitivity to oxidative stress [38], and 3) interactions between the SOD gene with the sex chromosomes [39].

Though in vitro and in vivo studies have suggested that MDA is a stable end product of lipid peroxidation that is associated with aging [12], we did not observe a significant association between MDA level and mortality in older adults. Additional analysis of the relationship between MDA level and SOD activity did not show any significant association. Although the pathway of MDA production by enzymatic processes is well known, its biological functions and its possible dose-dependent dual role have not been fully studied [40].

To the best of our knowledge, this is the first prospective cohort study to evaluate the association of SOD activity, MDA level with mortality in community-dwelling older adults. The strength of this study included the large sample size, careful adjustment for established and potential risk factors, investigation of potential interactions, and robust results of sensitivity analysis.

Our study has several limitations. First, we take single SOD activity test as exposure, which may be unable to present the overall SOD activity level for individual participant because oxidative stress is a dynamic process. However, we focused on the population average SOD level as this study is a population-based observational study. Such method has been commonly used to evaluate the association between SOD activity with other endpoints including cognitive decline, diabetes, hypertension, and cancer mortality. Future studies may consider repeated measure of SOD activity to minimize the potential influence of the dynamic change in SOD though such bias can hardly be fully controlled. Second, despite our best efforts to adjust for established and potential confounders, residual confounding by other unmeasured or unknown factors remains possible. Third, we cannot undertake additional analyses to investigate the type of death because such data were not collected in CLHLS. Fourth, a total of 369 participants (16.6%) were lost to follow-up in this study. However, the potential influence would be minor as the proportions of lost to follow-up were generally balanced among groups and there were no major differences in the main characteristics between those lost to follow-up and others (age: P = 0.09, sex: P = 0.18, residence: P = 0.052; education: P = 0.03 (17.8% vs. 22.8%); smoking: P = 0.34, alcohol drinking: P = 0.41, BMI: P = 0.32). Last, this study only explored baseline SOD activity and MDA level, the potential changes in these markers overtime were not adjusted.

Overall, our analyses indicated that higher baseline plasma SOD activity level was associated with lower overall mortality in older women but not in men. SOD activity may be considered as a prognostic biomarker for mortality in older women. Our findings suggested the importance of integrative approach in the application of different parameters of oxidative stress as potential prognostic biomarkers. Though a causal effect cannot be confirmed due to the observational study design, this study added epidemiological evidence to the free radical theory of aging. Our findings also lend support for interventions to increase the SOD activity, such as exercise and energy-restricted diet [41], to increase longevity.