and Behavior 63 (2013) 510–517

Contents lists available at SciVerse ScienceDirect

Hormones and Behavior

journal homepage: www.elsevier.com/locate/yhbeh

Helping hands, healthy body? gene and prosocial behavior interact to buffer the association between stress and physical health

Michael J. Poulin a,⁎, E. Alison Holman b a University at Buffalo, USA b University of California, Irvine, USA article info abstract

Article history: Providing help or support to others buffers the associations between stress and physical health. We examined Received 10 September 2012 the function of the neurohormone oxytocin as a biological mechanism for this stress-buffering phenomenon. Revised 14 January 2013 Participants in a longitudinal study completed a measure of charitable behavior, and over the next two years Accepted 15 January 2013 provided assessments of stressful life events and physician-diagnosed physical ailments. Results indicated Available online 24 January 2013 that charitable behavior buffered the associations between stressful events and new-onset ailments among individuals with the AA/AG genotypes of oxytocin receptor gene (OXTR) variant rs53576, but not among Keywords: fi Oxytocin those with the GG genotype. These results suggest that oxytocin function may signi cantly affect health OXTR and may help explain the associations between prosocial behavior and health. More broadly, these findings Stress are consistent with a role for the caregiving behavioral system in health and well-being. Health © 2013 Elsevier Inc. All rights reserved. Prosocial behavior Volunteering Social behavior Caregiving Genetics

Introduction joint role of charitable behavior, stressful life events, and oxytocin re- ceptor (OXTR) genotype in predicting new-onset physical ailments. Providing help or support to others, while often depicted as burden- some, is robustly associated with better health and well-being. Indeed, volunteering or providing care for close others predicts reduced morbid- Prosocial behavior, caregiving, and oxytocin ity and mortality (e.g., Brown et al., 2003; Konrath et al., 2012; O'Reilly et al., 2008), and increased psychological well being (e.g., Brown et al., Engaging in prosocial behavior may reduce stress in several ways, 2008; Poulin et al., 2010;seePost, 2007, for an overview). such as increasing positive affect or perceived control (Krause, 2006). The benefits of engaging in prosocial behavior may be due to Theoretical models of the caregiving behavioral system further sug- stress-buffering features of helping. Helping or supporting others pre- gest that prosocial behavior may also buffer stress more directly dicts reduced associations between stress and mortality (Krause, 2006; (Brown and Brown, 2006; Goetz et al., 2010). The caregiving system, Okun et al., 2010; Poulin et al., in press)aswellasdepression(Brown which evolved to facilitate parental behavior, motivates efforts to re- et al., 2008). Laboratory research indicates that people respond to acute duce the suffering and/or increase the well-being of any individual stress by increasing their prosocial behavior (von Dawans et al., in thought to be in need (Bell and Richard, 2000; Collins et al., 2010; press), and that helping behavior reduces physiological response to Shaver et al., 2010). Recent models of this system suggest it may do stress (e.g., Floyd et al., 2007), raising the possibility that prosociality so by facilitating social approach behaviors, at least in part via anxiety mayserveasacopingstrategy. reduction (Brown and Brown, 2006; Goetz et al., 2010). We test the prediction that the stress-buffering effects of prosocial Prior research indirectly links the caregiving system with health by behavior are a function of the caregiving behavioral system, and spe- showing that prosocial behavior only promotes health and well-being cifically of the neurohormone oxytocin. To do so, we examine the when accompanied by other-focused motivations (Gillath et al., 2005; Konrath et al., 2012) or when extended towards valued others (Poulin et al., 2010). However, a more direct way to study the role of caregiving ⁎ Corresponding author at: Department of Psychology, Park Hall 206, University at Buffalo, Buffalo, NY 14260, USA. Fax: +1 716 645 3801. in health is to examine a key biological mechanism that supports care- E-mail address: [email protected] (M.J. Poulin). giving: the function of the neurohormone oxytocin. In both humans and

0018-506X/$ – see front matter © 2013 Elsevier Inc. All rights reserved. http://dx.doi.org/10.1016/j.yhbeh.2013.01.004 M.J. Poulin, E.A. Holman / Hormones and Behavior 63 (2013) 510–517 511 non-human mammals, oxytocin plays a prominent role in motivating Method parental care (Campbell, 2010; Carter, 1998; Feldman, 2012). Oxytocin also has stress-buffering effects, moderating neural and behavioral re- Participants and procedure sponses to negative stimuli (Campbell 2010; Kirsch et al., 2005; Poulin et al., 2012) as well as modulating hypothalamic-pituitary-adrenal Data were collected through Internet-based surveys of a large, na- (HPA) axis and cardiovascular stress reactivity (Chen et al., 2011; tionally representative sample (N=2729; Silver et al., 2006) recruited Norman et al., 2012; Rodrigues et al., 2009). Higher blood levels of oxy- using stratified random-digit-dial telephone sampling by Knowledge tocin have even been found to predict faster healing time of lab-induced Networks Inc. (KN). KN panel members are compensated with Internet wounds (Gouin et al., 2010). In humans, oxytocin promotes positive in- access (if needed), points used to obtain merchandise, and cash incen- teractions with close others beyond offspring (Ditzen et al., 2008; tives for completing surveys. Data used for this investigation were col- Feldman, 2012) and facilitates prosociality more generally, including lected at several time points, described below. charitable giving, empathy, and compassion (e.g., Kogan et al., 2011; Poulin et al., 2012; for a review, see Campbell, 2010). Life event and health surveys Together, oxytocin's strong links to caregiving and its known The participants completed a survey on mental/physical health stress-buffering effects make it a plausible mechanism for explaining and lifetime and ongoing stress by September, 2002 (Wave 1; N= the stress-buffering effects of prosocial behavior. However, no prior 1916). They were subsequently invited to participate in two similar research has linked oxytocin's stress-buffering effects to physical follow-up surveys 12 months (Wave 2; N=1571, 82% of the Wave health per se, nor assessed its function as a potential moderator of 1 sample) and 24 months after Wave 1 (Wave 3; N=1771, 92% of the stress-buffering effects of prosocial behavior on health. To some the Wave 1 sample). degree, this may be because the central functions of oxytocin are dif- ficult to observe in humans. That is, oxytocin is both a peripheral hor- Social and political survey mone and central nervous system , with central A subset of those who completed the Wave 1 survey had complet- effects on the amygdala that are believed to motivate prosocial ed a KN-administered survey on social attitudes and involvement, in- behavior by way of threat reduction (Baumgartner et al., 2008; cluding charitable behavior, between April, 2000 and March, 2003. Campbell, 2010). The exact nature and extent of oxytocin's central ef- This sample (n=924) was the base sample for all analyses involving fects remain unclear, but it is possible to broadly examine variability charitable behavior. in oxytocin function by exploring individual differences in oxytocin receptor gene (OXTR) polymorphisms. DNA collection Within the OXTR gene, the single nucleotide polymorphism (SNP) In 2008, KN re-contacted available participants (N =1296) from rs53576 has been shown to predict both stress-buffering and prosocial the larger longitudinal study to request their participation in the behavior, with the “GG” genotype, in particular, appearing to indicate genetic study. Most participants (54.7%; N =711) agreed, and pro- greater inclination for engaging in beneficial psychological and social vided saliva for genotyping using OraGene test kits (http://www. behaviors associated with oxytocin (Bakermans-Kranenburg and van dnagenotek.com/) mailed to their homes. Kits were sent to The Ijzendoorn, 2008; Rodrigues et al., 2009; Tost et al., 2010). Moreover, al- Centre for Applied Genomics (TCAG; http://www.tcag.ca/)for though the exact role of rs53576 “G” and “A” alleles remains unclear, genotyping. There were 704 participants for whom OXTR rs53576 the risk allele (A) has been associated with risk allele-load dependent was successfully genotyped (99% call rate), and of these, 631 had decreases in hypothalamic size and amygdalar activation (Tost et al., completed the Wave 1 survey. This sample (n =631) was the 2010), suggesting that risk for psychosocial dysfunction increases in base sample for all analyses involving OXTR genotype. the presence of an “A” allele. If this is the case, and oxytocin accounts for the stress-buffering effects of prosocial behavior on health, we Final sample would expect to see a significant interaction between prosocial behav- Because not all participants provided data on either charitable ior and the GG genotype that explains the stress-health relationship. behavior or genetics, analyses in the present study are not based This interaction effect could take at least one of two forms. First, it is on a single sample size, but on the largest, most representative sam- possible that prosocial behavior promotes oxytocin function to com- ple providing data for all variables in each analysis. For analyses pensate for the lower sensitivity of the non-GG genotype—i.e., a com- involving just charitable behavior and health, n=924; for analyses pensatory stress-buffering effect specifically among those with the involving just OXTR genotype and health, n =631; and for analyses AA/AG genotype. Second prosocial behavior may combine with involving both charitable behavior and OXTR genotype, n=365. the GG genotype to maximally buffer stress—i.e., a synergistic stress- Across these separate analyses, data from a total of N= 1195 people buffering effect specifically among those with the GG genotype. By con- are examined. trast, the absence of an interaction between prosocial behavior and the GG genotype would cast doubt on the notion that oxytocin helps to ex- Measures plain the stress-buffering effects of prosocial behavior. OXTR The OXTR variant of interest, rs53576, consists of a locus at which The present study either the nucleobase adenine or guanine (denoted by “A” or “G”) can occur. The rs53576 SNP genotyping identified 361 “GG”, 284 “AG”, The present study tests whether oxytocin function, as indicated by and 59 “AA” individuals. It was in Hardy–Weinberg equilibrium (X2 OXTR rs53576 genotype, moderates the stress-buffering effects of [1]=0.1, p>.05) with no genotype-by-sex differences (X2[2]=1.32, prosocial behavior in predicting subsequent longitudinal assessments p=.52) or genotype-by-ethnicity differences (X2[6]=7.22, p=.30). of health status. We predicted that, adjusting for baseline health sta- We followed the same grouping strategy used in prior studies ad- tus and other relevant covariates, greater charitable involvement dressing genotype-related psychological and social phenotypes—GG would predict a lessened association between stress and new-onset respondents were compared to individuals carrying a high risk A al- ailments (i.e., a stress-buffering pattern). In addition, we predicted lele (GG=1, AA/AG=0) (e.g., Poulin et al., 2012; Rodrigues et al., that this stress-buffering effect would be qualified by an interaction 2009; Tost et al., 2010). This approach maintains consistency with (either compensatory or synergistic) between OXTR genotype and prior studies and ensures sufficient statistical power to detect differ- charitable involvement. ences related to the high risk A allele. 512 M.J. Poulin, E.A. Holman / Hormones and Behavior 63 (2013) 510–517

Stressful life events Analytic strategy Assessment of lifetime exposure to stressful events was completed by Wave 1 using a count of 37 negative events (e.g., serious illness or Analyses were conducted using Stata 11.0 (Stata Corp. College Sta- injury, natural disaster) a respondent reported experiencing (see tion, TX). Multilevel Poisson regression models were built using Stata's Silver et al., 2002). This measure was modified from the Diagnostic In- xtpois module1 to examine whether OXTR genotype, recent stressors, terview Schedule section on trauma (Robins et al., 1981), expanded and charitable behaviors would predict incidence of physical ailments. using open-ended coding of lifetime traumas reported by a primary These analyses were conducted to predict increased ailment incidence care sample (Holman et al., 2000), and has provided rates of specific at Waves 2 and 3 by controlling for ailments at Wave 1. Because the de- events in this sample comparable to surveys conducted on other com- pendent variable was assessed at more than one point in time, analyses munity samples (e.g., Breslau et al., 1998). The total number of life- took the form of multilevel regression models (mixed effects or hierar- time stressful events reported was used as a measure of lifetime chical linear models; Singer and Willett, 2003), which examine out- adversity. At Waves 2 and 3, respondents were asked which of those comes across multiple waves without increasing Type I error. Since same events had occurred in the past 12 months. The total count of physical ailments was a count variable, these analyses were conducted events reported at each wave—excluding events directly related to as Poisson regressions. Poisson regression models yield regression coef- health, our dependent variable—represented recent stressful events ficients that can be exponentiated as eb and interpreted as incidence experienced between Waves 1 and 2 and between Waves 2 and 3, rate ratios (IRRs). IRRs reflect the ratio by which the predicted count respectively. of the outcome (e.g., number of physical ailments) would change given a one-unit increase in the predictor. IRRs of b1indicatearelative decrease in the outcome's incidence while IRRs of >1 indicate a relative Charitable behavior increase in the outcome's incidence, and IRRs=1 indicate no associa- The social and political survey asked whether participants had en- tion between the predictor and the outcome. gaged in the following activities in the past 12 months specifically as The construction of all models presented involved screening two a way of being socially involved in their communities: donating sets of control variables before entering the variables for recent stress, blood, giving money to or working for a charity, attending PTA meet- OXTR genotype, and charitable behavior: (a) demographics (age, sex, ings, or attending a community group meeting. The total number of ethnicity, education, and income level) and (b) mental health history these activities was used as a measure of involvement in charitable (number of lifetime stressful events, number of physician-diagnosed behaviors. Asking participants about participation in specific, concrete mental health ailments at each wave). activities provides the best possible estimate of long-term charitable involvement (as well as group membership; see below), and has Results been validated for psychological (e.g., Omoto et al., 2010), sociological (Wilson, 2000), and governmental (U.S. Bureau of Labor Statistics, Descriptive statistics 2010) research.

Across the different analyses reported, a total of 1195 people con- Group membership tributed data. Ages ranged from 18 to 89 (M=52.97), females com- The social and political survey also asked the participants wheth- prised 46%, 79% were European-American, 8% were African American, er they participated in any of the following groups, without tying 8% were Latino/a, and 6% were of another ethnicity. At Wave 1, partici- such participation to community service: a club or fraternal organi- pants reported an average of 3.60 physician-diagnosed ailments zation, a veterans group, a religious group, a senior citizen's center (SD=3.24) and those who provided data on charitable activities or group, a women's group, an issue-oriented political organization, reported engaging in an average of 1.31 out of 5 possible charitable be- a non-partisan civic organization, a school club or association, a haviors (SD=1.08). hobby club, a sports team, a youth club, a neighborhood association or community group, or a group representing racial/ethnic issues. Attrition analysis The total number of these group memberships was computed as an index of group membership. Results of a multiple logistic regression predicting participation revealed that individuals who were genotyped for the present study Physical ailments (n=631) did not differ from those not genotyped in the Wave 1 sam- At Waves 1–3, health data were collected using a questionnaire ple in terms of age, sex, ethnicity, income, education, number of psy- modified from the Centers for Disease Control's National Center for chological or physical ailments, or past-year stressful events. Those Health Statistics annual National Health Interview Survey (NHIS; genotyped, however, reported slightly more lifetime adverse events U.S. Department of Health and Human Services, 2000). Respondents (M=9.27) than those not genotyped (M=8.34; OR=1.02, p=.04). were asked, “Has a medical doctor ever diagnosed you as suffering A separate multiple logistic regression examined differences between from any of the following ailments?” with prompts for 33 physical those who provided data on charitable behavior (N=924) versus ailments. Evidence for the validity of this measure comes from a study other Wave 1 participants. Respondents with data on charity were of 25,000 KN panelists (Baker et al., 2003): when compared to the older (M=54.95) than other Wave 1 participants (M=44.31, OR= b 2000 NHIS, prevalence rates for common health outcomes (e.g., 1.05, p .001), less likely to be female (40% of the sample with charity problems, cancer, diabetes, hypertension, ulcer, migraine, and stroke) data versus 62% of the sample without charity data, OR=0.32, b differed in the KN panel by an average of less than 1.5%. From this list p .001) and more likely to be of European-American ethnicity (81% of ailments, we created a count variable of the total number of physical of the sample with charity data versus 69% of the sample without b ailments as the dependent variable in our analyses. charity data, OR=0.63, p .001). Following Wave 1, some individuals did not participate in all subse- quent waves, but this type of missing data is acceptable in multilevel Mental ailments modeling, because individuals contribute to estimation of the model The KN health questionnaire also included questions about doctor- at particular time points even if they cannot do so at all time points diagnosed psychological ailments (anxiety disorder and depression). We created a count variable of the total number of psychological ail- 1 To address possible overdispersion, negative binomial models were also examined; ments as a covariate in our analyses. results were substantively identical. M.J. Poulin, E.A. Holman / Hormones and Behavior 63 (2013) 510–517 513

(Singer and Willett, 2003). However, patterns of non-response after and females. Testing a three-way interaction among genotype, stress, Wave 1 were examined using Stata's xtgee module (a multilevel appli- and sex (female versus male) revealed no significant differences in cation that allows for binomial distributions), with participation at stress buffering between females (n=321) and males (n=310; p= each wave (yes/no) as the time-varying dependent variable. At each .15). A final follow-up analysis tested whether the OXTR GG×stress subsequent wave, non-participants reported slightly more recent interaction was present in the sample of individuals who had complete stressful events on average (0.56) than did continuing participants data on all variables in our study (n=365). Results indicated that this in- (0.36; OR=1.42, pb.001), but there were no differences with respect teraction was not significant in the smaller sample (IRR=0.98, p=.58). to age, income, ethnicity, education, sex, number of psychological or physical ailments, lifetime adversity, or genotype. Charity and stress buffering A second multilevel Poisson regression tested the interaction be- Charitable behavior, OXTR, stress, and health tween charitable behavior and recent stressors, along with relevant controls, among all participants with data for those variables (n= Our prediction that charitable behavior would interact with OXTR 924; Table 1, Model 2). Results indicated a significant charity×recent genotype to buffer the association between recent stressful events stress interaction (IRR=0.96, pb.001). To examine the simple effects and ailment incidence suggested a three-way interaction among of recent stressors predicting ailment incidence, charitable behavior charitable behavior, OXTR genotype, and recent stressors. However, was recentered at low and high values (1 SD below and above the before testing this hypothesis, we examined the independent mean, respectively). Recent stress predicted increased ailment inci- stress-buffering roles of OXTR genotype and charitable behavior iden- dence more strongly at low levels of charitable behavior (IRR=1.15, tified in prior research. pb.001), than at high levels of charitable behavior (IRR=1.04, p= .02; Fig. 1, Panel B). OXTR and stress buffering To assess whether the stress-buffering role of charitable behavior was A multilevel Poisson regression tested the interaction between OXTR specific to prosocial activities as opposed to general social connection, a genotype and recent stressors, along with relevant controls, among all follow-up analysis included group membership and a group member- participants with data for those variables (n=631; Table 1, Model 1). ship×recent stress interaction term alongside the charitable behavior Results indicated a significant OXTR GG×recent stress interaction variables reported in Table 1, Model 2. Results of this model indicated (IRR=0.90, p=.001). Examining the simple effects of recent stress that group membership did not significantly buffer the association be- for GG versus other genotype by recentering the genotype variable re- tween stress and health (IRR=1.00, p=.64), but the charity×recent vealed that recent stress predicted increased ailment incidence among stress interaction remained significant (IRR=0.96, p=.01). We again AA/AG individuals (IRR=1.09, pb.001), but not among GG individuals examined whether our results differed by sex; a second follow-up anal- (IRR=0.99, p=.51; Fig. 1, Panel A). ysis examined whether the stress-buffering role of charitable behavior As prior research has identified inter-ethnic differences in OXTR differed between males and females. Testing a three-way interaction rs53576, with greater proportions of GG individuals among Cauca- among charitable behavior, stress, and sex (female versus male) revealed sians (Kim et al., 2010), a follow-up analysis tested a three-way inter- no significant differences in stress buffering between females (n=364) action among genotype, stress, and ethnicity (European-American and males (n=560; p=.41). A final follow-up analysis tested whether versus other). This interaction was not significant (p=.20), indicat- the charity×stress interaction was present in the sample of individuals ing no ethnic differences in the stress-buffering role of OXTR who had complete data on all variables our study examined (n=365). rs53576; moreover, the magnitude of the OXTR GG×stress interac- Results indicated that this interaction was not significant in the smaller tion was very similar between Caucasians (IRR=0.91) and those of sample (IRR=0.98, p=.15). other ethnic groups (IRR=0.87). While prior research on OXTR rs53576 has found no sex differences (e.g., Poulin et al., 2012; Charity, OXTR, and stress buffering Rodrigues et al., 2009), female sex was a significant covariate in our Poisson regression tested the three-way interaction among charitable analyses (see Table 1) and oxytocin is associated with many behavior, OXTR genotype, and recent stressors, along with relevant con- female-specific functions such as childbirth and lactation (Carter, trols (Table 1,Model3).Therewasasignificant charity×OXTR GG×stress 1998). Therefore, a second follow-up analysis examined whether interaction (IRR=1.11, p=.002). Examining the simple two-way the stress-buffering role of OXTR rs53576 differed between males (charity×stress) interactions for different genotypes by recoding the ge- notype variable (as GG=0 or AA/AG=0) revealed that charitable behav- Table 1 ior significantly interacted with recent stress to predict ailment incidence Multilevel Poisson regression models of physical health ailment incidence. among AA/AG individuals (IRR=0.94, p=.001), but not among GG indi- Variable IRR (SE) viduals (IRR=1.05, p=.12). Simple effects of recent stressors predicting ailment incidence were Model 1 Model 2 Model 3 n=631 n=924 n=365 examined for both genotypes, with charitable behavior recentered at low and high values (1 SD below and above the mean, respectively). Re- Wave 1 physical ailments 1.19 (0.01)*** 1.20 (0.01)*** 1.20 (0.01)*** Age in years 1.00 (0.00)* 1.00 (0.00) 1.00 (0.00) sults indicated that recent stress did not predict ailment incidence for Female sex 1.09 (0.04)* 1.07 (0.05)* 1.10 (0.05)* GG individuals, regardless of whether they had engaged in low (IRR= Household income 1.01 (0.01)** 1.00 (0.00) 1.01 (0.01) 0.96, p=.42) or high amounts of charitable behavior (IRR=1.06, p= Psychological ailments at each 1.23 (0.03)*** 1.28 (0.05)*** 1.23 (0.04)*** .14).However,forAA/AGindividuals,recentstressspecifically predict- wave ed increased ailment incidence among individuals who engaged in few Recent stressors 1.09 (0.02)*** 1.16 (0.08)*** 1.17 (0.05)** OXTR rs53576 GG genotype 1.00 (0.04) 1.07 (0.09) charitable behaviors (IRR=1.16, p=.001), but not among those who Genotype×recent stressors 0.90 (0.03)** 0.81 (0.06)** engaged in many charitable behaviors (IRR=1.01, p=.80). In sum, Charitable behavior 1.05 (0.09) 1.03 (0.03) charitable behavior buffered the association between recent stressors Charity×recent stressors 0.96 (0.07)* 0.94 (0.02)** and health for AA/AG individuals, but not for GG individuals, among Genotype×charity 0.94 (0.04) Genotype×charity×stressors 1.11 (0.04)** whom there was no stress/health association (Fig. 2). To estimate the magnitude of the contribution of OXTR rs53576, *pb.05, **pb.01, ***pb.001. we computed the effect size of the charity×stress interaction as pre- Note. IRR = incidence rate ratio. Model 1 fit: Wald χ2(8, 1724)=1060.12; Model 2 fit: Wald χ2(8, 2535)=1710.93, pb.001; Model 3 fit: Wald χ2(12, 1028)=772.15. All dicted for AA/AG individuals and for GG individuals. To do so, we psb.001. computed the change in Wald χ2 (the omnibus test of the fitofa 514 M.J. Poulin, E.A. Holman / Hormones and Behavior 63 (2013) 510–517 AB

Fig. 1. Number of new ailments per wave, graphed by number of recent stressors (0–2). Note. Panel A represents stress–ailment associations graphed by OXTR rs53576 genotype (GG versus AA/AG). Panel B represents stress-ailment associations graphed by high and low levels of charitable behavior (1 SD above and below the mean, respectively).

Poisson model) when the charity×stress interaction was deleted separately for females and males. The magnitude of the OXTR from the model, and converted this statistic to the corresponding ef- GG×charity×stress interaction was very similar between females fect size estimate, phi (ϕ) by dividing the χ2 by n and taking the (IRR=1.15; n=224) and males (IRR=1.12; n=141). square root of the result. This procedure revealed that the effect size of the charity×stress interaction was ϕ=.34, a “medium” effect size, for AA/AG individuals, but for GG individuals, the same interac- Tests of replicability tion had an effect size of ϕ=.09, a “small” effect size—and one that Fig. 2 indicates is in the opposite direction. While our analyses were conducted in a single sample, we A follow-up analysis examined whether the OXTR GG×charity× conducted further analyses to provide evidence of the replicability stress interaction varied by ethnicity (European-American versus of these findings by testing our models separately in randomly se- other). We did not expect to be able to test a four-way interaction lected halves of our sample. Using random digits from the website (ethnicity×OXTR GG×charity×stress), however, we explored the pos- random.org (http://www.random.org) we randomly designated sibility of ethnic differences by testing the three-way interaction sepa- each participant a member of subsample 1 or subsample 2. All re- rately in Caucasians and those of other ethnicities. As in our analysis sults reported above were replicated and achieved significance inde- of ethnicity and the OXTR GG×stress interaction, the magnitude of the pendently in both subsamples. The OXTR GG stress-buffering effect OXTR GG×charity×stress interaction was very similar between Cauca- (OXTR GG×stress interaction) was present in subsample 1 (n=320, sians (IRR=1.14) and those of other ethnic groups (IRR=1.12) with no IRR=0.91, p=.01) and subsample 2 (n=311, IRR=0.89, p=.02). significant differences between the two groups. Similarly, the charity stress-buffering effect (charity×stress interac- A second follow-up analysis examined whether the OXTR GG× tion) was present in subsample 1 (n=445, IRR=0.96, p=.04) and charity×stress interaction differed between males and females. As subsample 2 (n=479, IRR=0.95, p=.002). Finally, the three-way in- with ethnicity, we did not expect to be able to test a four-way inter- teraction among genotype, charity, and stress was significant in both action (sex×OXTR GG×charity×stress), however, we explored the subsample 1 (n=185, IRR=1.13, p=.01) and subsample 2 (n=180, possibility of sex differences by testing the three-way interaction IRR=1.11, p=.04).

AB

Fig. 2. Number of new ailments per wave, graphed by number of recent stressors (0–2), amount of charitable behavior, and OXTR rs53576 genotype (GG versus AA/AG). Note. High and low values of charitable behavior represent amounts 1 SD above and below the mean, respectively. Separate panels represent the charity×recent stress interaction for the AA/ AG genotype (Panel A) and for the GG genotype (Panel B). M.J. Poulin, E.A. Holman / Hormones and Behavior 63 (2013) 510–517 515

Discussion (e.g., Bell and Richard, 2000; Brown and Brown, 2006; Sober and Wilson, 1998). While prior research has demonstrated positive associations be- One implication of a need to invest in others is that seeking oppor- tween prosocial behavior and health, including the stress-buffering tunities to help or provide support to others might be as natural and role of prosocial behavior, the mechanisms for these associations have important during times of adversity as seeking support from others. not been clear. Drawing on research and theory on the caregiving be- Recent research supports this view: under acute stress, individuals havioral system, we predicted that the stress-buffering effects of can become more, not less, prosocial (von Dawans et al., in press). prosocial behavior would be explained by the function of the neurohor- Paradoxically, however, seeking to support others for the express mone oxytocin—specifically that oxytocin receptor (OXTR)genotype purpose of benefiting the self may be self-defeating, given that prior would interact with prosocial behavior to buffer the association be- research has found that such motivations actually undermine the tween stress and health. Our findings support this hypothesis: engaging benefits of prosocial behavior (Gillath et al., 2005; Konrath et al., in charitable behaviors and OXTR rs53576 GG genotype both buffered 2012). These prior findings indicate that it may be impractical to ad- the association between stress and health. However, examining them vise individuals to cope with stress by engaging in prosocial behavior. in combination revealed that charity only buffered the association be- Such advice might deserve further qualification as our findings sug- tween stress and health for respondents with the OXTR rs53576 AA/ gest that the benefits of engaging in prosocial behavior may be limit- AG genotypes, which are thought to render individuals less sensitive ed to OXTR rs53576 AA/AG individuals. to oxytocin than the GG genotype—a compensatory effect. These find- ings provide the first evidence of a link between oxytocin receptor ge- Contributions, limitations, and future directions notype (rs53576) and human health outcomes, and are consistent with the possibility that oxytocin accounts for the stress-buffering role The present study is the first study using prospective data from a of prosocial behavior in predicting health. diverse real-world sample to (a) demonstrate an association between the oxytocin system and human health, and (b) provide evidence suggesting that oxytocin is the mechanism linking prosocial behavior Helping, oxytocin, and health with health. We also extend prior research (e.g., Poulin et al., 2012) indicating that the effects of OXTR genotype depend on social context. Our findings add to a growing literature that indicates that prosocial Despite these strengths, we acknowledge the limitations of our study. behavior bolsters physical health, and that it specifically does so by buff- First, this study featured a relatively small sample of individuals with ering the association between stress and health (Brown et al., 2008; all of the necessary data to test interactions among charitable behav- Krause, 2006; Okun et al., 2010; Poulin et al., 2012). However, our find- ior, OXTR genotype, and stress. Even though we found a significant ings also extend prior research by suggesting a plausible mechanism for three-way (OXTR GG×charity×stress interaction among these vari- the effects of prosocial behavior: the function of the neurohormone ables in this small sample, we did not detect significant OXTR GG× oxytocin. Notably, our findings suggest that individual differences in stress and charity×stress two-way interactions in individuals with the OXTR gene can predict stress-related physical health problems. For all data, but only in larger groups of individuals who had just the rel- individuals with AA/AG genotypes, stressful life events significantly evant data for those analyses. However, the presence of the three- predicted new-onset ailments. For individuals with the GG genotype, way (OXTR GG×charity×stress) interaction result may help explain there was no such association. This finding alone, while a result of one this phenomenon. Because both of these two-way interactions are qual- study only, is worth future investigation. ified by a third variable—i.e., there is a three-way OXTR GG×charity× An even more relevant finding for the oxytocin-as-mechanism hy- stress interaction—any simple two-way effect combines individuals pothesis is that charitable behavior buffered the association between for whom the two-way effect is present with individuals for whom stress and health among those with AA/AG genotypes. This effect is con- the effect is absent. This lowers the effect size of the two-way interac- sistent with the possibility that engaging in charitable behavior provides tion, making it probabilistically more difficult to detect in a smaller asufficient boost in oxytocin to compensate for the relative lack of oxyto- versus a larger sample. cin sensitivity thought to characterize these genotypes. This compensato- Another implication of our relatively small sample is that our find- ry effect is made plausible by the fact that prosociality has been linked to ings about the OXTR rs53576 genotype should be considered prelim- endogenous release of oxytocin in the past (e.g., Feldman, 2012). Future inary, and in need of exact replication (Sullivan, 2007). However, the research should determine whether endogenous oxytocin or other effects fact that our results replicated exactly in randomly-selected subsam- of prosocial behavior (e.g., positive affect, an increased sense of control) ples of our data boosts our confidence in the replicability of these explain why charitable behavior is stress-buffering specifically among findings. AA/AG respondents. Next, given that this was a non-experimental study, it is important to emphasize that the causal effects of OXTR and charity on health re- Why do people need people? main unproven. The longitudinal design of our study, including controls for baseline health that allow for prediction of new-onset ailments, pro- It is broadly recognized that people have a “need to belong,” vide some evidence that the associations we found were not solely due (Baumeister and Leary, 1995), and that the human need for social to health status predicting levels of charitable involvement. However, contact has implications for physical health. While this need is often we acknowledge the need for converging evidence from experimental interpreted as a need for support, or for the mere presence of others, and field studies. our findings, along with those from previous studies (e.g., Brown et In addition, while we assessed several common types of prosocial in- al., 2003; Schwartz and Sendor, 1999), indicate that giving help or volvement, our charitable behaviors assessment was not comprehen- support to others is at least as good a predictor of health and sive. Other approaches may identify different patterns of associations well-being, if not a better one. A role for oxytocin in these salutary ef- among charitable behaviors, OXTR genotype, and stress. For example, fects of prosocial behavior would indicate that behaving prosocially it is possible that charitable behaviors were not stress-buffering for taps into the behavioral caregiving system, and would suggest that GG individuals because they engaged in more day-to-day supportive humans not only have a need to belong, but may have a need to invest activities that buffered their stress (cf. Kogan et al., 2011). Moreover, in or care for others. While speculative, this notion is consistent with while we examined OXTR genotypes as a way of making inferences evolutionary accounts of human motivation in which maximizing about the effects of oxytocin, it is possible that, over the course of a the well-being of relatives or group members is a selected-for trait life span, individual differences in OXTR genotype may also shape 516 M.J. Poulin, E.A. Holman / Hormones and Behavior 63 (2013) 510–517 aspects of people's personalities and social environments, potentially Campbell, A., 2010. Oxytocin and human social behavior. Pers. Soc. Psychol. Rev. 14, 281–295. confounding the association between OXTR and health. Ideally, future Carter, C.S., 1998. Neuroendocrine perspectives on social attachment and love. work could examine proximal as well as distal indicators of oxytocin Psychoneuroendocrinology 23, 779–818. function and health, not all of which share this limitation, such as pe- Chen, F.S., Kumsta, R., von Dawans, B., Monakhov, M., Ebstein, R.P., Heinrichs, M., 2011. Common oxytocin receptor gene (OXTR) polymorphism and social support interact ripheral oxytocin assessment (e.g., Gouin et al., 2010). Finally, it was be- to reduce stress in humans. Proc. Natl. Acad. Sci. U. S. A. 108, 19937–19942. yond the scope of the present research to examine the physiological Collins, N.L., Ford, M.B., Guichard, A.C., Kane, H.S., Feeney, B.C., 2010. Responding to mechanisms (e.g., modulated HPA axis and cardiovascular stress reac- need in intimate relationships: social support and caregiving processes in couples. tivity) by which charitable behavior and oxytocin might buffer the im- In: Mikulincer, M., Shaver, P.R. (Eds.), Prosocial Motives, Emotions, and Behavior: the Better Angels of Our Nature. American Psychological Association, Washington, pact of stressful events (Chen et al., 2011; Norman et al., 2012; DC, pp. 367–389. Rodrigues et al., 2009). Future research should examine whether the Ditzen, B., Schaer, M., Gabriel, B., Bodenmann, G., Ehlert, U., Heinrichs, M., 2008. Intra- OXTR rs53576 SNP and prosocial behavior interact in a compensatory nasal oxytocin increases positive communication and reduces levels dur- ing couple conflict. Biol. Psychiatry 65, 728–731. manner to buffer physiologic markers of stress, and if so, whether Feldman, R., 2012. Oxytocin and social affiliation in humans. Horm. Behav. 61, these effects can predict health outcomes. 380–391. Floyd, K., Mikkelson, A.C., Tafoya, M.A., Farinelli, L., La Valley, A.G., Judd, J., Haynes, M.T., Davis, K.L., Wilson, J., 2007. Affectionate communication accelerates neuroendo- Conclusion crine stress recovery. Health Commun. 22, 123–132. Gillath, O., Shaver, P.R., Mikulincer, M., Nitzberg, R.E., Erez, A., Van Ijzendoorn, M.H., 2005. Attachment, caregiving, and volunteering: placing volunteerism in an The results of this study provide novel insights into the links be- attachment-theoretical framework. Pers. Relatsh. 12, 425–446. tween social behavior and health, identifying significant roles of char- Goetz, J.L., Keltner, D., Simon-Thomas, E., 2010. Compassion: an evolutionary analysis fi and empirical review. Psychol. Bull. 136, 351–374. itable behavior, oxytocin, and stress. Importantly, these ndings shed Gouin, J.P., Carter, C.S., Pournajafi-Nazarloo, H., Glaser, R., Malarkey, W.B., Loving, T.J., light on how the evolutionarily-preserved caregiving system, by way Stowell, J., Kiecolt-Glaser, J.K., 2010. Marital behavior, oxytocin, , and of the neurohormone oxytocin, may contribute to the biological un- wound healing. Psychoneuroendocrinology 35, 1082–1090. derpinnings of stress-related illness. We hope that these findings en- Holman, E.A., Silver, R.C., Waitzkin, H., 2000. Traumatic life events in primary care patients: a study in an ethnically-diverse sample. Arch. Fam. Med. 9, 802–810. courage future research not only on the mechanisms linking stress Kim, H.S., Sherman, D.K., Sasaki, J.Y., Xu, J., Chu, T.Q., Ryu, C., Suh, E.M., Graham, K., and health, but on novel therapeutic approaches to help individuals Taylor, S.E., 2010. Culture, distress, and oxytocin receptor polymorphism (OXTR) fl adjust to adversity. interact to in uence emotional support seeking. Proc. Natl. Acad. Sci. U. S. A. 107, 15717–15721. Kirsch, P., Esslinger, C., Chen, Q., Mier, D., Lis, S., Siddhanti, S., Gruppe, H., Mattay, V.S., Gallhofer, B., Meyer-Lindenberg, A., 2005. Oxytocin modulates neural circuitry for Acknowledgments social cognition and fear in humans. J. Neurosci. 25, 11489–11493. Kogan, A., Saslow, L.R., Impett, E.A., Oveis, C., Keltner, D., Saturn, S.R., 2011. Thin-slicing Project funding for the original study was provided by the Josiah study of the oxytocin receptor (OXTR) gene and the evaluation and expression of the prosocial disposition. Proc. Natl. Acad. Sci. U. S. A. 108, 19189–19192. Macy Jr. Foundation grant SF03-09 to E. Alison Holman, and the US Konrath, S., Fuhrel-Forbis, A., Lou, A., Brown, S.L., 2012. Motives for volunteering are as- National Science Foundation grants BCS-9910223, BCS-0211039, and sociated with mortality risk. Health Psychol. 31, 87–96. BCS-0215937 to Roxane Cohen Silver. The supplemental genetic study Krause, N., 2006. Church-based social support and mortality. J. Gerontol. B Psychol. Sci. Soc. Sci. 61, S140–S146. was supported by the Robert Wood Johnson Foundation Nurse Faculty Norman, G.J., Hawkley, L., Luhmann, M., Ball, A.B., Cole, S.W., Berntson, G.G., Cacioppo, Scholars grant #68046 and an award from the UC Irvine School of J.T., 2012. Variation in the oxytocin receptor gene influences neurocardiac reactiv- Medicine's Committee on Research and Graduate Academic Programs ity to social stress and HPA function: a population based study. Horm. Behav. 61, – both given to E. Alison Holman. We thank our colleagues Drs. Roxane 134 139. Okun, M.A., August, K.J., Rook, K.S., Newsom, J.T., 2010. Does volunteering moderate the Silver, Daniel McIntosh, Virginia Gil-Rivas, and Judith Andersen for relation between functional limitations and mortality? Soc. Sci. Med. 71, 1662–1668. their integral role in the design and implementation of the original Omoto, A.M., Snyder, M., Hackett, J.D., 2010. Personality and motivational antecedents – study. We thank Knowledge Networks Government, Academic, and of activism and civic engagement. J. Pers. 78, 1703 1734. O'Reilly, D., Connolly, S., Rosato, M., Patterson, C., 2008. Is caring associated with an in- Non-profit Research team of J. Michael Dennis and Rick Li for their assis- creased risk of mortality? A longitudinal study. Soc. Sci. Med. 67, 1282–1290. tance with collecting saliva samples; and Drs. Richard Wintle and Tara Post, S.G. (Ed.), 2007. Altruism and Health: Perspectives From Empirical Research. Paton of the Center for Applied Genomics for their expert advice and as- Oxford University Press, New York. Poulin, M.J., Brown, S.L., Ubel, P.A., Smith, D.M., Jankovic, A., Langa, K.M., 2010. Does a sistance with genotyping. helping hand mean a heavy heart? Helping behavior and well-being among spouse caregivers. Psychol. Aging 25, 108–117. Poulin, M.J., Holman, E.A., Buffone, A., 2012. The neurogenetics of nice: oxytocin and References vasopressin receptor genes and prosocial behavior. Psychol. Sci. 23, 44–452. Poulin, M.J., Brown, S.L., Dillard, A., Smith, D.M., in press. Stress does not predict in- Baker, L.C., Bundorf, M.K., Singer, S., Wagner, T.H., 2003. Validity of the survey of health creased mortality among those who give to others. Am. J. Public Health. and Internet and knowledge network's panel and sampling. Stanford University, Robins, L.N., Helzer, J.E., Croughan, J.L., Williams, J.B.W., Spitzer, R.L., 1981. Diagnostic Palo Alto, CA. Interview Schedule: Version III. National Institute of Mental Health, Rockville, MD. Bakermans-Kranenburg, M.J., van IJzendoorn, M.H., 2008. Oxytocin receptor (OXTR) Rodrigues, S.M., Saslow, L.R., Garcia, N., John, O.P., Keltner, D., 2009. Oxytocin receptor and transporter (5-HTT) genes associated with observed parenting. genetic variation relates to empathy and stress reactivity in humans. Proc. Natl. Soc. Cogn. Affect. Neurosci. 3, 128–134. Acad. Sci. U. S. A. 106, 21437–21441. Baumeister, R.F., Leary, M.R., 1995. The need to belong: desire for interpersonal attach- Schwartz, C.E., Sendor, M., 1999. Helping others helps oneself: response shift effects in ments as a fundamental human motivation. Psychol. Bull. 117, 497–529. peer support. Soc. Sci. Med. 48, 1563–1575. Baumgartner, T., Heinrichs, M., Vonlanthen, A., Fischbacher, U., Fehr, E., 2008. Oxytocin Shaver, P.R., Mikulincer, M., Shemesh-Iron, M., 2010. A behavioral-systems perspective shapes the neural circuitry of trust and trust adaptation in humans. 58, on prosocial behavior. In: Mikulincer, M., Shaver, P.R. (Eds.), Prosocial Motives, 639–650. Emotions, and Behavior: the Better Angels of Our Nature. American Psychological Bell, D.C., Richard, A.J., 2000. Caregiving: the forgotten element in attachment. Psychol. Association, Washington, DC, pp. 72–91. Inq. 11, 69–83. Silver, R.C., Holman, E.A., McIntosh, D.N., Poulin, M.J., Gil-Rivas, V., 2002. Nationwide Breslau, N., Kessler, R.C., Chilcoat, H.D., Schultz, L.R., Davis, G.C., Andreski, P., 1998. longitudinal study of psychological responses to September 11. JAMA 288, Trauma and posttraumatic stress disorder in the community: the 1996 Detroit 1235–1244. Area Survey of Trauma. Arch. Gen. Psychiatry 55, 626–632. Silver, R.C., Holman, E.A., McIntosh, D.N., Poulin, M., Gil-Rivas, V., Pizarro, J., 2006. Cop- Brown, S.L., Brown, R.M., 2006. Target article: selective investment theory: recasting ing with a national trauma: a nationwide longitudinal study of responses to the the functional significance of close relationships. Psychol. Inq. 17, 1–29. terrorist attacks of September 11. In: Neria, Y., Gross, R., Marshall, R., Susser, E. Brown, S.L., Nesse, R.M., Vinokur, A.D., Smith, D.M., 2003. Providing social support may (Eds.), 9/11: Mental Health in the Wake of Terrorist Attacks. Cambridge University be more beneficial than receiving it: results from a prospective study of mortality. Press, New York, pp. 45–70. Psychol. Sci. 14, 320–327. Singer, J.D., Willett, J.B., 2003. Applied Longitudinal Data Analysis: Modeling Change Brown, S.L., Brown, R.M., House, J.S., Smith, D.M., 2008. Coping with spousal loss: and Event Occurrence. Oxford University Press, New York. potential buffering effects of self-reported helping behavior. Pers. Soc. Psychol. Sober, E., Wilson, D.S., 1998. Unto Others: the Evolution and Psychology of Unselfish Bull. 34, 849–861. Behavior. Harvard University Press, Cambridge, MA. M.J. Poulin, E.A. Holman / Hormones and Behavior 63 (2013) 510–517 517

Sullivan, P.F., 2007. Spurious genetic associations. Biol. Psychiatry 61, 1121–1126. U.S. Department of Health and Human Services, 2000. National Health Interview Tost, H., Kolachana, B., Hakimi, S., Lemaitre, H., Verchinski, B.A., Mattay, V.S., Meyer- Survey Questionnaire 2000. National Center for Health Statistics, Hyattsville, Lindenberg, A., 2010. A common allele in the oxytocin receptor gene (OXTR)impacts MD. prosocial temperament and human hypothalamic–limbic structure and function. von Dawans, B., Fischbacher, U., Kirschbaum, C., Fehr, E., Heinrichs, M., 2012. The social Proc. Natl. Acad. Sci. U. S. A. 107, 13936–13941. dimension of stress reactivity: acute stress increases prosocial behavior in humans. U.S. Bureau of Labor Statistics, 2010. Volunteering in the United States. Retrieved July 8, Psychol. Sci 23, 651–660. 2011, from http://www.bls.gov/news.release/volun.nr0.htm. Wilson, J., 2000. Volunteering. Annu. Rev. Sociol. 26, 215–240.