1 Behav Genet (2012) 42: DOI /s ORIGINAL RESEARCH Age Differences in Genetic and Environmental Variations in Stress-Coping During Adulthood: A Study of Female Twins Yoon-Mi Hur Alexander J. MacGregor Lynn Cherkas Frances M. K. Williams Tim D. Spector Received: 21 December 2011 / Accepted: 19 April 2012 / Published online: 5 May 2012 Ó Springer Science+Business Media, LLC 2012 Abstract The way people cope with stressors of day to day living has an important influence on health. The aim of the present study was to explore whether genetic and environmental variations in stress-coping differ over time during adulthood. The brief COPE was mailed to a large sample of the UK female twins (N = 4,736) having a wide range of age (20 87 years). Factor analyses of the items of the brief COPE yielded three coping scales: Problem-Solving, Support Seeking, and Avoidance. Monozygotic and dizygotic twin correlations tended to become lower with age for all three scales, suggesting that unique environmental factors may become more important with age during adulthood. Model-fitting results showed that relative influences of unique environmental factors increased from 60 % at age 20 years to 74% at age 87 years for Problem-Solving and 56 % at age 20 years to 76% at age 87 years for Avoidance. During the same age period, genetic factors decreased from 40 to 26 % for Problem-Solving and from 44 to 24 % for Avoidance. For Seeking Support, the magnitude of Edited by Deborah Finkel. Y.-M. Hur Mokpo National University, Jeonnam, South Korea Y.-M. Hur (&) Industry Academic Cooperation, 61 Dorim-ri, Cheonggye-myeon, Muan-gun, Jeonnam, South Korea A. J. MacGregor L. Cherkas F. M. K. Williams T. D. Spector Department of Twin Research, Kings College London, London, UK A. J. MacGregor School of Medicine, University of East Anglia, Norwich, UK genetic and unique environmental factors was not significantly different across the adulthood. For all three scales, shared environmental effects were negligible. Overall, our findings implicate that the effects of environment that stem from idiosyncratic experience of stressful life events accumulate and become increasingly important in adulthood. Keywords Stress-coping Twin Genetics Aging Epigenetics Introduction Coping defined as cognitive and behavioral efforts to manage psychological stress is an important trait related to human adaptation and health (Lazarus 1993). It has been demonstrated that coping is associated with diabetes and cardiovascular disease (Maes Leventhal and Ridder 1996; Snieder et al. 2002) various forms of psychopathology (Sawyer et al. 2009; Vinberg et al. 2010) social functioning (Lazarus 1993) and personality (Watson and Hubbard 1996; Kato and Pedersen 2005; Kendler et al. 1991). For example patients with depression and chronic fatigue syndrome have been shown to use more avoidance coping relative to problem-focused coping strategies than normal controls (Afari et al. 2000; Kendler et al Although it has been argued that coping is a product of social learning and experience (Heszen-Niejodek 1997), several twin studies have shown the importance of genetic influences on variation in coping. Recently, Jang et al. (2007) found in an adult volunteer twin sample that % of the variance for Emotion-Oriented Coping, Task-Oriented Coping, and Social Diversion were due to genetic factors. The results of Jang et al. s study, however, are limited by the small sample size (N = 171 pairs).
2 542 Behav Genet (2012) 42: Somewhat higher estimates of genetic influences were documented by other studies that employed relatively large twin samples. In 446 pairs of reared-together and rearedapart middle- and old-aged twins, Kato and Pedersen (2005) found genetic influences to range from 15 to 51 % on Problem-Solving, Turning to Others, and Avoidance. With the same measure of coping used in the Jang et al. (2007) study, Kozak et al. (2005) found in 612 pairs of adult Polish twins that the contribution of genetic factors ranged from 33 to 39 %. These estimates were similar to those reported by Kendler et al. (1991) who found genetic factors accounted for approximately 30 % of the variation in Problem-Solving and Turning to Others in 827 pairs of young adult female twins in the USA. In summary, previous twin studies of coping indicated that, depending on the measure and sample, genetic factors explained approximately % of individual differences in coping in adulthood. Twin studies have documented evidence of environmental factors in coping as well. However, environmental factors important for stresscoping were those unique to individuals in a family, which explained about % of the variance of stress-coping. The magnitude of shared family environmental influences was zero to very small (Kendler et al. 1991; Kato and Pedersen 2005). While estimates of genetic and environmental contributions to individual differences in coping during adulthood have been relatively well established, very little attention has been given to the issue of whether genetic and environmental variations in coping change over time in adult years. Coping is a dynamic process that depends on situational as well as dispositional factors (Lazarus 1993). A review of 185 gerontological studies has shown that older people are more variable than younger people on a variety of psychological characteristics (Nelson and Dannefer 1992). Research examining coping during adulthood has documented significant age differences in the mean level of coping strategy as well. In general, the younger people tend to use more active, interpersonal, problem-focused forms of coping (e.g., seeking of social support, problem solving) than do the older people, and the older people use more passive, emotion-focused coping styles (e.g., distancing) (Folkman et al. 1987; Lazarus and DeLongis 1983). Across the life span, individuals undergo role changes and consequently, experience different types of environmental stressors. Although genes may predispose individuals to develop certain consistent coping styles to handle various stressors, as a result of accumulation of life experiences including participation in various therapies and intervention programs, individual s coping responses can be altered, which can lead the influence of unique environmental effects in coping responses to increase in late adulthood. Using a cross-sectional twin design, the present study aimed to investigate whether the magnitude of genetic and environmental influences on coping varies during adulthood. It is important to understand the etiological basis of age differences in coping during adulthood because this knowledge will enable us to implement effective prevention and intervention strategies to promote mental health as well as successful adaptation to aging. As far as we understand, to date, only one twin study has explored age differences in genetic and environmental variation in coping. In Kato and Pedersen s (2005) adult twin study, the authors divided the total sample into two age groups cut at the median age (58.19 years) of their sample, but failed to find any significant differences in the magnitude of genetic and environmental variation between the two age groups. However, their results should be interpreted with a caution at least for two reasons. First, the sample size in the study was not sufficiently large (in total, 446 pairs and 447 individual twins) to detect age effects. Unfortunately, it is difficult to figure out the magnitude of the difference in genetic and environmental influences between the two age groups as Kato and Pedersen (2005) did not report the effect size of the difference. Secondly, because they arbitrarily classified their total sample into two subgroups and compared them, the authors may have failed to capture the effects of age which may be manifested across adulthood continuously rather than discretely. In the present study, we treated age as a continuous moderator that can influence genetic and environmental variances in coping during adulthood. As the age range of our sample was broad (20 87 years), this study provided a sensitive test for age differences in heritability and the strength of environmental influences on coping across adulthood. The present study also had a sample larger than any other twin studies of coping published so far, which offered us an increased statistical power to detect effects of age in coping. Methods Sample Subjects, drawn from the TwinsUK register ( ac.uk) (Spector and Williams 2006), consisted of female twins of European ancestry who returned a mail-out survey of self-report questionnaires on a range of topics, mostly unrelated to coping. Recruitment to the TwinsUK register began in 1994 through a national media campaign. Participants were all females because the TwinsUK register was originally developed mainly for studies of adult female bone mineral density. However, twin subjects were not selected for any specific behavioral trait or health status,
3 Behav Genet (2012) 42: nor did they have prior knowledge of the precise nature of the study. Overall, the distribution of the social class of the twins was very similar to that of the general population in the UK (Spector and Williams 2006). Zygosity was established through a standardized questionnaire and by DNA confirmation in doubtful cases. The study received approvals from King s College London Ethics Committee. In total, 4,736 individual twins consisting of 1,019 pairs of monozygotic (MZ), 981 pairs of dizygotic (DZ), and 736 individual twins were included in data analyses. The mean age of the MZ twins was 52.6 (±13.2, range = 20 87) years and of the DZ twins was 53.5 (±11.9, range = 22 85) years. Measure To measure coping, we carried out a confirmatory factor analysis using 28 items that came from the Brief COPE (Carver 1997) and two additional items tapping behavioral aspects of coping: I eat more than usual to make myself feel better and I do exercise to clear my mind, such as running, walking, swimming, yoga, or going to the gym. The confirmatory factor analysis was performed in a large sample of twins who responded to our 2001 survey (N = 1,720), which is a subsample of the present study. The confirmatory factory analysis yielded three interpretable factors: Problem-Solving (11 items), Seeking Support (7 items) and Avoidance (12 items). Absolute loadings on these three factors ranged from 0.23 to For each of the 30 items, twins were asked to respond on a 4-point Likert-type response format ranging from I usually don t do this at all to I usually do this a lot when experiencing certain stressful events. Responses were summed up to compute the score for each of the three scales. Sample sizes varied across the three coping scales due to incomplete reporting by some subjects. For all three scales, higher scores indicate greater use of the particular coping strategy. Cronbach s alpha coefficients were 0.78 for Problem-Solving, 0.71 for Avoidance, and 0.75 for Seeking Support. The scores of the three scales conformed to a normal distribution (skewness =-0.12 to 0.57; kurtosis =-0.05 to 0.39). The three coping scales in our study are similar to those commonly identified by previous coping research: Problem-Solving refers to responses that seek to eliminate the source of stress by actively dealing with the reality of the situation and is considered positive, practical coping style. Seeking Support represents the coping responses to reduce the stress by seeking informational and emotional supports from others. Avoidance describes a negative, passive coping strategy to engage in another task rather than the task at hand to escape from stressful situations. Problem-Solving was moderately correlated with Seeking Support (r = 0.32), but was independent from Avoidance (r = 0.07). There was a modest correlation between Seeking Support and Avoidance (r = 0.17) in our sample. Overall, this pattern of interscale correlations was consistent with those found in the literature of coping (e.g., Endler and Parker 1990). Statistical analyses There were two steps in data analyses. In the first step, we roughly divided the total sample into two age groups and compared means, variances, twin correlations, and the estimates of genetic and environmental influences for each coping scale across the two age groups. The two age groups were the young/middle-aged adult group (20 60 years) and the old-aged adult group (61 87 years). We combined young adults with middle-aged adults and made the young/middleaged adult group as the number of young adult twins in our sample was very small. This first step served as a preliminary analysis to confirm the results of analyses in the second step where we performed model-fitting analyses incorporating age as a continuous moderator (Purcell 2002). In the age-moderation model, the standard paths a, d, and e, indicating factors attributable to additive genetic (A) and nonadditive genetic (D), and unique environmental (E) influences, respectively, are allowed to vary as a function of a moderator variable (M, age in our analyses). Additive genetic factors (A), i.e., the sum of the average effect of all genes that influence a trait, were set to be correlated at 1.0 and 0.5 for MZ and DZ twins, respectively. Nonadditive genetic factors (D), i.e., those interactive effects of genes were set to be correlated at 1.0 and 0.25 for MZ and DZ twins, respectively. Finally, (E), environmental factors unique to each member of a twin pair and measurement error represented the remaining variance not explained by additive or nonadditve genetic factors. We did not include shared environmental factors in our model because our twin correlations showed no evidence of shared environmental influences (see Table 1). According to our age-moderation model, the phenotypic variance (V p ) can be expressed as: V p =(a+b a M) 2 +(d+b d M) 2 +(e+b e M) 2 In this equation, b a, b d, and b e represent the magnitude of the moderating effects associated with A, D, and E, respectively. If b is significantly different from zero, there is evidence for a moderating effect of age on genetic and environmental influences on coping. That is, influences of additive and nonadditive genetic factors and unique environmental factors vary as a function of age. In addition to the three variance components, we allowed a pathway l? b m M to represent the main effects of age on coping.
4 544 Behav Genet (2012) 42: Table 1 Means, variances, twin correlations and genetic and environmental variance components for three coping scales by age Problem-Solving Seeking Support Avoidance Young/middle Old Total Young/middle Old Total Young/middle Old Total Mean Variance r MZ 0.35(0.30, 0.43) 0.26 (0.15, 0.36) 0.33(0.27, 0.38) 0.40(0.34, 0.46) 0.35(0.24, 0.44) 0.38(0.33, 0.43) 0.37(0.30, 0.43) 0.33(0.23, 0.43) 0.36(0.30, 0.41) r DZ 0.07 (-0.01, 0.14) 0.11(-0.01, 0.22) 0.09(0.02, 0.15) 0.16(0.08, 0.23) 0.14(0.02, 0.25) 0.15(0.09, 0.21) 0.15(0.08, 0.22) 0.10(0.01, 0.22) 0.14(0.08, 0.20) A 0(0, 5.8) 3.8(0, 11.0) 0(0, 6.8) 3.8(0, 7.6) 2.9(0, 7.8) 3.6(0, 7.2) 5.7(0, 10.7) 2.7(0, 9.0) 5.0(0, 9.7) D 9.3(3.5, 11.0) 4.6(0, 11.8) 9.2(2.2, 10.8) 3.1(0, 7.8) 3.6(0, 8.3) 3.3(0, 7.4) 3.8(0, 10.9) 4.7(0, 9.6) 3.9(0, 9.9) E 17.3(15.8, 19.1) 22.7(19.5, 26.4) 18.8(17.4, 20.5) 10.1(9.1, 11.2) 11.3(9.6, 13.2) 10.4(9.6, 11.4) 16.0(14.5, 17.7) 15.4(13.3, 17.8) 15.8(14.6, 17.1) A (%) 0(0, 22) 12(0, 34) 0(0, 24) 22(0, 44) 16(0, 43) 21(0, 41) 22(0, 41) 12(0, 39) 20(0, 39) D (%) 35(11, 40) 15(0, 37) 33(7, 38) 18(0, 45) 20(0, 45) 19(0, 43) 15(0, 42) 21(0, 41) 16(0, 40) E (%) 65(60, 72) 73(63, 84) 67(62, 73) 60(54, 66) 64(54, 75) 60(56, 66) 63(57, 69) 67(59, 78) 64(59, 69) 95 % CIs are in parenthesis r MZ = Twin correlation for monozygotic twins; r DZ = Twin correlation for dizygotic twins Young/middle young/middle-aged adults (age range = years), old old-aged adults (age range = years), A additive genetic variance, D nonadditive genetic variance, E unique environmental variance The maximum likelihood raw data option in Mx (Neale et al. 2003) was used in our model-fitting analyses. Mx calculates twice the negative log-likelihood (-2LL) of the data. As the difference in -2LL is Chi-square distributed, when models were nested to each other, the likelihood ratio test (LRT) was used to evaluate alternative models. When models were not nested to each other, Akaike s information criterion (AIC) was used to judge the model. AIC is defined as -2LL for the model minus twice the number of degrees of freedom (AIC =-2LL - 2df). Models having smaller AIC are considered more parsimonious, and are thus preferred (Akaike 1987). To determine the best-fitting, most parsimonious model, we reduced parameters and compared the goodnessof-fit of the reduced models with that of the full model. Results Preliminary analyses MZ twin correlations were consistently greater than DZ twin correlations across all three scales in all age groups, suggesting the importance of genetic influences on coping across adulthood. The presence of nonadditive genetic factors was also indicated as DZ twin correlations were generally lower than half the MZ correlations for all three scales. Most MZ twin correlations fell below 0.40, suggesting that unique environmental factors were substantial for coping. None of the three scales showed evidence for significant shared environmental factors, however. In the total sample, age was very modestly correlated with Problem-Solving (-0.05, p \ 0.01), Seeking Support (-0.04, p \ 0.01), and Avoidance (-0.09, p \ 0.01). These modest correlations suggested that the mean effects of age on coping were minimal in our sample. Table 1 shows that the mean levels for the three coping scales were not significantly different among the three age groups. Of particular interest in the present study is variance. Interestingly, phenotypic variances for Problem-Solving and Seeking Support increased with age, with this trend being more pronounced for Problem-Solving than for Seeking Support. These increases in the phenotypic variance with age were primarily due to the increases of unique environmental influences, which was indicated in the increased estimates of unstandardized variance of unique environmental factors in old-aged adults as compared to those in young/middle-aged adults for Problem-Solving (young/ middle-aged adults = 17.3 and old-aged adults = 22.7) and Seeking Support (young/middle-aged adults = 10.1 and old-aged adults = 11.3) in Table 1. Indeed, the estimates of unstandardized variance of total genetic factors (A? D) for both scales were slightly decreased with age (9.3? 8.4 for Problem-Solving and 6.9? 6.5 for
5 Behav Genet (2012) 42: Table 2 Model-fitting results for the three coping scales Model Problem-Solving Seeking Support Avoidance -2LL AIC df DX 2 p -2LL AIC df DX 2 p -2LL AIC df DX 2 p 1 ADEade AEae AEe AEa AE ADEde ADEae ADEad ADEe ADEd ADEa ADE The best-fitting model is in bold type A additive genetic effects, D nonadditive genetic effects, E unique environmental effects including measurement error. a the age-moderator effects on additive genetic factors, d the age-moderator effects on nonadditive genetic factors, e the age-moderator effects on unique environmental factors, -2LL -2 log likelihood Seeking Support ), confirming that the increase of the phenotypic variance for these two scales occurred due to the increase of unique environmental factors rather than genetic factors. For Avoidance, the phenotypic variance diminished with age. Although unstandardized variances of both unique environmental factors and total genetic factors were lowered with age, the decrease was more pronounced in the total genetic variance (9.5? 7.4) than in the unique environmental variance (16.0? 15.4). We therefore concluded that the decrease in the phenotypic variance in Avoidance could be largely due to a decrease in genetic variance. We tested these observations by fitting the agemoderation model below. Model-fitting analyses incorporating moderating effects of age Table 2 summarizes the results of fitting models with agemoderators. Figures 1, 2, and 3 present graphical descriptions of standardized variance components in the best-fitting model for the three coping scales. It should be noted that we did not attempt to remove A from the full model as it has been argued that an extreme amount of nonadditive genetic effects with no effects of additive genes is unlikely to explain genetic variance for complex traits (Eaves 1988). However, as our sample did not have sufficient statistical power to distinguish between A and D, all genetic variance was absorbed into D for the Problem- Solving scale, resulting in the estimate of A to be zero. For this reason, we summed up A and D to represent broad heritability, i.e., total genetic variance in Figs. 1, 2, and 3. Among the models tested for Problem-Solving, Model 9 was the best by AIC and LRT. In Model 9, the parameters, A, D, E, and the age-moderator on E(e) alone were maintained. The unstandardized unique environmental variance for Problem-Solving increased from 13.4 (95 % CI: ) at age 20 years to 25.6 (95 % CI: ) at age 87 years, whereas the total unstandardized genetic variance was the same [9.0 (95 % CI: )] from age 20 to 87 years in Model 9. Consistent with observations Standardized variance component 100% 90% 80% 70% 60% 50% 40% 30% 20% 10% 0% A + D Age in years Fig. 1 Standardized variance components of total genetic (A? D) and unique environmental factors including measurement error (E) for Problem-Solving from age 20 to 87 years in the best-fitting model E
6 546 Behav Genet (2012) 42: Standardized variance component 100% 90% 80% 70% 60% 50% 40% 30% 20% 10% 0% A+D Age in years Fig. 2 Standardized variance components of total genetic (A? D) and unique environmental factors including measurement error (E) for Seeking Support from age 20 to 87 years in the best-fitting model Standardized variance component 100% 90% 80% 70% 60% 50% 40% 30% 20% 10% 0% made from Table 1, these results suggested that the increase of the phenotypic variance with age found for Problem-Solving was attributable to the increase of unique environmental variance. In terms of the standardized variance component (Fig. 1), total genetic factors decreased from 40 % (95 % CI: %) at age 20 years to 26 % (95 % CI: %) at age 87 years, whereas unique environmental factors increased from 60 % (95 % CI: %) at age 20 years to 74 % (95 % CI: %) at age 87 years in Model 9. For Seeking Support, Model 5 was the best one, where none of the age-moderators was significant, with only A and E parameters being statistically significant. These results suggested that total genetic and unique environmental variances were constant across adulthood in E A+D Age in years Fig. 3 Standardized variance components of total genetic (A? D) and unique environmental factors including measurement error (E) for Avoidance from age 20 to 87 years in the best-fitting model E Seeking Support. Although the pattern of the phenotypic variance in two age groups observed in Table 1 implied possible increase in unique environmental variance with age, the effects were not large enough to attain statistical significance. Total standardized genetic variance and unique environmental variance for Seeking Support in Model 5 were, respectively, 38 % (95 % CI: %) and 62 % (95 % CI: %) (Fig. 2). For Avoidance, Model 4 was the best by AIC and LRT. The parameters A, E, and the age-moderator on A (a) were maintained in Model 4. The unstandardized total genetic variance for Avoidance decreased from 12.9 (95 % CI: ) at age 20 years to 5.0 (95 % CI: ) at age 87 years, whereas the unstandardized unique environmental variance was stable [16.1 (95 %CI: )]. Therefore, the decrease in the phenotypic variance from age 20 to 87 years shown in Table 1 was attributed to the decrease in genetic variance during adulthood. In terms of the standardized variance component (Fig. 3), genetic factors decreased from 44 % (95 % CI: 30 55%) at age 20 years to 24 % (95 % CI: 0 50 %) at age 87 years, whereas unique environmental factors increased from 56 % (95 % CI: %) at age 20 years to 76 % (95 % CI: %) at age 87 years in Model 4. Discussion Our study suggests that phenotypic, genetic, and unique environmental variations in human coping responses tend to vary over time during adulthood. In terms of unstandardized variance, the phenotypic variance of Problem- Solving increased with age due to an increase of unique environmental variance with age. In contrast, the phenotypic variance of Avoidance decreased with age due to a decrease of genetic variance with age. For Seeking Support, the phenotypic, genetic, and unique environmental variations were constant during adulthood, although the parameter estimates in Table 1 showed a pattern of age difference similar to that found for Problem-Solving. It is interesting to note that unstandardized variance of genetic and unique environmental factors for Problem- Solving and Avoidance changed with age in different directions. Problem-Solving is considered adaptive coping behavior, whereas Avoidance, maladaptive coping style (Kendler et al. 1991). A recent study by Vinberg et al. (2010) demonstrated that among individuals at a high risk for depression, the use of Problem-Solving strategy in the face of stressful life events played a role of protective factors, whereas frequent use of Avoidance coping strategies exacerbated the health outcome. Our results showed that unstandardized genetic variance for Avoidance was significantly reduced later in life. Given the
7 Behav Genet (2012) 42: previous finding (e.g., Lieberman and Tobin 1983) indicating that the ways the elderly coped with stress made a significant difference in survival and well-being, we speculate that those who were predisposed to have extremely high level of Avoidance coping strategies were not present in our sample due to health and adaptation problems, leading to a smaller genetic variance at older ages. The lower mean level of Avoidance in the old-aged group as compared to the young/middle-aged group supported this speculation. In terms of standardized variance, whereas unique environmental factors increased, genetic factors decreased over time for both Problem-Solving and Avoidance. Charles and Almeide (2007) found unique environmental influences on the severity of perceived stress to increase with age during adulthood. Given that coping style is related to perception of stressful situations (Lazarus 1993), our results are consistent with those reported by Charles and Almeide (2007). Why do unique environmental influences on stress-coping increase during adulthood? Our findings suggest that the effects of idiosyncratic experiences of stressful events accumulate over the life course. Thus, unique environmental influences increasingly contribute to individual difference in coping patterns during adulthood. Folkman et al. (1987) found that sources of stress were different between younger and older adults: the younger adults reported finance, work, parenting, and family issues as salient life events, whereas the older adults reported retirement, death of a loved one, physical debilitation, and health issues to be important sources of stress. Stressful experiences are not limited to major life events, but also include the ongoing difficult conditions of daily life. As the number of stressful life events increases with age and as new sets of life events come into effects during adulthood, relatively small environmental influences on coping style in early adulthood can become increasingly important in late adulthood. There is a growing body of evidence that epigenetic factors can have a profound impact on behavioral phenotypes (Petronis 2006). Epigenetics refers to phenotypic changes caused by mechanisms that are unrelated to changes in the underlying DNA sequence, most notably driven by histone modifications and DNA methylation (Petronis 2006). From the epigenetic point of view, phenotypic differences in MZ twins could result, in part, from their epigenetic differences that change over the lifetime. Fraga et al. (2005) reported that DNA methylation differences between MZ twins increase with age, with divergence in time spent together, and with different lifestyles, implicating that the impact of genetic and unique environmental factors in psychological traits can change with age. In our study, MZ twin correlations were lower in the old-aged group than in the young/middle-aged group for all three coping scales (Table 1), suggesting a possibility that differences in epigenetic markings may have occurred in coping responses. Evidence for the role of epigenetic changes in determining individual differences in stresscoping comes from rodent studies. For example, Weaver et al. (2006) found that decreased licking, grooming, and nursing by rat mothers reduced DNA methylation and histone acetylation at a glucocorticoid receptor gene promoter in the hippocampus of the offspring, resulting in increased stress responses in later life, suggesting that environmental factors, by regulating gene expression, can lead to change in stress-coping responses later in life. Another example can be found in the Murgatroyd et al. (2009) study where the authors found that early life stress in mice can dynamically control DNA methylation in postmitotic neurons to generate stable changes in arginine vasopressin (AVP) gene expression that trigger alteration in stress-coping behavior. It would be of interest in future research to explore epigenetic mechanisms in human stress-coping behavior. The present study has several limitations that deserve attention. First, our results were on the basis of female twins, and therefore, cannot be generalized to adult males because sex differences in genetic and environmental variances in coping have been reported (Kato and Pedersen 2005). Secondly, although we attempted to use information on age in our sample maximally by treating age as a continuous variable, our data were cross-sectional in nature. People of different ages may vary in coping because they grew up under different historical conditions in which people perceive and deal with problems of living differently. Replication of our findings using longitudinal data will disentangle these cohort effects from the aging processes. Third, in the present paper we examined how age moderated genetic and environmental influences on coping. However, it is possible that other variables such as marital status and socioeconomic status may have an impact on genetic and environmental factors in coping during adulthood. Future studies should examine how these variables moderate genetic and environmental influences on coping. Finally, our twins are adult volunteers recruited through media campaigns. Although our sample has been shown to be representative of a singleton population ascertained from General Practice registers (Spector and Williams 2006), as with most adult volunteer samples, it is possible that those with better health and more adaptive twins have participated in our study. Acknowledgments The first author (Y.M. Hur) is supported by the Pioneer Fund, USA and Korea Research Foundation Grants. TwinsUK register is supported by the Wellcome Trust. We would like to thank twins who participated in our study and Irina Gillham-Nasenya for the data management.
8 548 Behav Genet (2012) 42: References Afari N, Karen B, Schmaling RH, Hartman S, Goldberg J, Buchwald DS (2000) Coping strategies in twins with chronic fatigue and chronic fatigue syndrome. J Psychosom Res 48: Akaike H (1987) Factor analysis and AIC. Psychometrica 52: Carver CS (1997) You want to measure coping but your protocol s too long: consider the brief COPE. Int J Behav Med 4: Charles ST, Almeide DM (2007) Genetic and environmental effects on daily life stressors: more evidence for greater variation in later life. Psychol Aging 22: Eaves LJ (1988) Dominance alone is not enough. Behav Genet 18(1):27 33 Endler NS, Parker JDA (1990) Multidimensional assessment of coping: a critical evaluation. J Pers Soc Psychol 58(5): Folkman S, Lazarus RS, Pimley S, Novacek J (1987) Age differences in stress and coping processes. Psychol Aging 2: Fraga MF, Ballestar E, Paz MF, Ropero S, Setien F, Ballestar ML, Heine-Suner D, Cigudosa JC, Urioste M, Benitez J, Boix- Chornet M, Sanchez-Aguilera A, Ling C, Carlsson E, Poulsen P, Vaag A, Stephen Z, Spector TD, Wu Y-Z, Plass C, Esteller M (2005) Epigenetic differences arise during the lifetime of monozygotic twins. PNAS 102: Heszen-Niejodek I (1997) Coping style and its role in coping with stressful encounters. Eur Psychol 2: Jang KL, Thordarson DS, Stein MB, Cohan SL, Taylor S (2007) Coping styles and personality: a biometric analysis. Anxiety Stress Coping 20:17 24 Kato K, Pedersen NL (2005) Personality and coping: a study of twins reared apart and twins reared together. Behav Genet 35: Kendler KS, Kessler RC, Heath AC, Neale MC, Eaves LJ (1991) Coping: a genetic epidemiological investigation. Psychol Med 21: Kim-Cohen J, Caspi A, Taylor A, Williams B, Newcombe R, Craig IW, Moffitt TE (2006) MAOA, maltreatment, and gene environment interaction predicting children s mental health: new evidence and a meta-analysis. Mol Psychiatry 11(10): Kozak B, Strelau J, Miles JNV (2005) Genetic determinants of individual differences in coping styles. Anxiety Stress Coping 18:1 15 Lazarus RS (1993) Coping theory and research: past, present, and future. Psychosom Med 55(3): Lazarus RS, Delongis A (1983) Psychological stress and coping in aging. Am Psychol 38: Lieberman MA, Tobin SS (1983) The experience of old age: stress, coping, and survival. Basic Books, New York Maes S, Leventhal H, Ridder TD (1996) Coping with chronic diseases. In: Zeidner M, Endler NS (eds) Handbook of coping: theory, research, applications. Wiley, New York Murgatroyd C, Patchev AV, Wu Y, Micale V, Bockmühl Y, Fischer D, Holsboer F, Wotjak CT, Almeida OF, Spengler D (2009) Dynamic DNA methylation programs persistent adverse effects of early-life stress. Nat Neurosci 12: Neale M, Boker SM, Xie G, Maes H (2003). Mx: statistical modeling. Department of Psychiatry, VCU, Richmond Nelson EA, Dannefer D (1992) Aged heterogeneity: fact or fiction? The fate of diversity in gerontological research. Gerontologist 32(1):17 23 Petronis A (2006) Epigenetics and twins: three variations on the theme. Trends Genet 22(7): Purcell S (2002) Variance components models for gene environment interaction in twin analysis. Twin Research 5: Sawyer MG, Pfeiffer S, Spence SH (2009) Life events, coping and depressive symptoms among young adolescents. a one-year prospective study. J Affect Disord 117:48 54 Snieder H, Harshfield GA, Barbeau P, Pollock DM, Pollock JS, Treiber FA (2002) Dissecting the genetic architecture of the cardiovascular and renal stress response. Biol Psychol 61:73 95 Spector TD, Williams FM (2006) The UK adult Twin Registry (TwinsUK). Twin Res Hum Genetics 9: Vinberg M, Froekjaer VG, Kessing LV (2010) Coping styles in healthy individuals at risk of affective disorder. J Nerv Ment Dis. 198(39 44):2010 Watson D, Hubbard B (1996) Adaptational style and dispositional structure: coping in the context of the five-factor model. J Pers 64: Weaver IC, Cervoni N, Champagne FA, D Alessio AC, Sharma S, Seckl JR, Dymov S, Szyf M, Meaney MJ (2006) Epigenetic programming by maternal behavior. Nat Neurosci 7: