Abstract
Objective
To extend the model of effort-reward imbalance at work to close and more general social relationships and test the associations with different measures of health. Lack of reciprocity at work is associated with poorer health in a number of studies. However, few studies have analysed the effect of non-reciprocity in other kinds of social relationships on health.
Methods
The Whitehall II Study is an ongoing prospective study of British civil servants (N=10,308 at baseline in 1985–88). Cross-sectional data from the latest phase (7, N=6,944 in 2002–04) were used in the analyses. The main exposure was a questionnaire measuring non-reciprocal social relations in partnership, parent-children and general trusting relationships. Health measures included the SF-36 mental and physical component Scores, GHQ-30 depression subscale, Jenkins’ Sleep disturbance questionnaire and the Rose Angina questionnaire. Logistic and linear regression models were analysed, adjusted for potential confounders and mediators of the association.
Results
Lack of reciprocity is associated with all measures of poorer health. This association attenuates after adjustment for previous health and additional confounders and mediators, but remains significant in a majority of models. Negative social support from a close person is independently associated with reduced health, but adjusting for this effect does not eliminate the association of non-reciprocity with poor health.
Conclusion
The effort-reward imbalance at work model has been extended to close and more general social relationships. Lack of reciprocity in partnership, parent-children and general trusting relationships is associated with poorer health.
Keywords: Reciprocity, Social Relationships, Effort-Reward Imbalance, SF-36, Rose-Angina
Introduction
The health-adverse impact of severe disappointing life events that disrupt intimate social relationships is well documented (1–3). Yet the adverse health effects of less dramatic but chronically recurring disappointments and frustrations are debatable. This paper applies a theoretical model of non-reciprocal social exchange at work, effort-reward imbalance(4), to less formalized, more intimate role relationships between partners, parents and children, and adults involved in trusting relationships in civic life.
The effort-reward imbalance model is based on the norm of reciprocity in contractual social exchange. According to this norm any action or service provided by person A to person B that has some utility to B, is expected to be returned by person B to A where the exchange expectancy concerns some agreed-upon standard of equivalence(5). If this norm is violated because a service in return is denied or fails to meet an agreed-upon level of equivalence, the social relationship is threatened or, in case of its continuation, strong negative emotions of anger and frustration are elicited among those concerned, resulting in a sense of being treated unfairly and in an unjust way. When experienced recurrently, injustice of exchange in important social role relations may result in adverse effects on health due to chronic stress reactions. Although the approach based on role-related social reciprocity bears some resemblance with the notion of negative social support(6) it is more specific because non-equivalence of return following recurrent investment in a transaction with a close person is considered a crucial stress-eliciting condition. This latter assumption was confirmed in several investigations based on the model of effort-reward imbalance at work(7). This important social role relation concerns the employment contract which defines distinct obligations or tasks to be performed in exchange with equitable rewards. The model of effort-reward imbalance at work claims that failed reciprocity in terms of high ‘costs’ spent and low ‘gains’ received occurs frequently under specific conditions (such as dependency, where employees have no alternative choice in the labour market) and increases the risk of stress-related physical and mental disorders.
In several prospective and cross-sectional epidemiological investigations, significantly elevated relative risks of coronary heart disease, depression, type 2 diabetes, alcohol dependence, and symptoms of ill health were documented in adult working populations exposed to effort-reward imbalance compared to those who were free from this type of chronic stress at work (8–10). Moreover, in ambulatory studies monitoring cardiovascular and hormonal responses over a working day, exposure to, or inappropriate coping with, effort-reward imbalance was related to elevated heart rate activity, reduced heart rate variability and elevated cortisol secretion(11–13).
This model has been extended to include other types of cooperative social relationships, whether characterized by a formal contract or by an informal mutual agreement on shared obligations, such as the marital and parental relationships or trusting relationships in civic life(14). In this approach, negative aspects of close social relationships are analysed as an imbalance experienced between high efforts spent and low (or no) rewards received in turn. Despite the fact that the constraints of demands are often less obvious than in paid employment and that reward transmitters are more often of an emotional than of an economic nature there are sufficient communalities between these different types of social exchange. First, they concern crucial social roles in adult life, in particular the marital and parental or maternal roles, the work role, or civic roles within a community. Social roles are defined by mutual obligations and expectations of behaviour addressed to persons who hold a specific position within a societal structure. As role behaviour is instrumental in meeting important needs of people’s everyday life a violation of role-related norms of reciprocity is likely to elicit strong negative emotions and associated stress reactions. Moreover, social roles such as marital and parental or maternal roles are embedded in the adult life course and may cause recurrent positive or negative experience of social exchange, comparable to the experience of continued exposure to a favourable or unfavourable psychosocial work environment. Similarly, failed reciprocity may be experienced as a chronic condition of unfair treatment or as a critical life event that harms feelings of trust and belonging. Two earlier reports found evidence of an association of failed reciprocity in close social relationships with depressive symptoms and poor self-rated health, but they were restricted to either small samples or older age groups and did not take into account negative affect or prior health status as important confounders(14;15).
We set out to analyse this association in a large study population, the participants of the Whitehall II study of British male and female civil servants. Moreover, to test the robustness of the association, we include five different indicators of health. Within the limitations of a cross-sectional analysis, we test the hypothesis that the experience of failed reciprocity in one of the three types of close social relationships (marital or partnership, parents and children, trusting relationships in civic life) is associated with reduced health. This association is expected to persist after adjustment for important confounders and mediators, in particular negative social support(16–18), and additionally age, gender, employment grade, prior health status, health-related behaviours and social networks. Investigating the potential confounding role of negative social support is of primary importance as non-reciprocal close relations and negative social support may measure distinct but overlapping concepts that may affect health independently of each other. So far, there is limited evidence available for each one of these associations(19–25). Yet, to our knowledge, no previous investigation has explored their separate effects in a single data set.
Methods
Data
The Whitehall II study is an ongoing longitudinal study of 10,308 male and female civil servants (initially aged 35 to 55) based in London, originally set up in 1985(26). Repeated data collection was carried out in 1989 (phase 2), 1991–1993 (phase 3) 1995 (phase 4), 1997–1999 (phase 5), 2001 (phase 6) and 2002–04 (phase7). In this paper, cross sectional data from the seventh phase (N=6914) were analysed, for the most part, as that was when the main exposure (non-reciprocal social relations) was measured.
Measures
Main health indicators (phases 5 and 7 of the study)
Two measures of self-reported physical and mental functioning are measured, based on the UK standard version of the Short Form 36 Health Survey (SF-36)(27). This is a 36 item questionnaire covering issues relating to physical, psychological and social functioning. Detailed information on the use of the SF-36 in the Whitehall study has been reported elsewhere(28). The original eight scales of this questionnaire can be summarized into physical and mental functioning components by a method based on factor analysis(29). Each score ranges from 0 to 100, with low scores implying low functioning. A mean of 50 is observed in the general US population. We defined scores in the lowest quartile of the Physical Component Score (PCS) and of the Mental Component Score (MCS) to represent reduced health in terms of poor physical and mental functioning.
As a third indicator of reduced health we assessed sleep problems, applying the scale described by Jenkins et al.(30) This scale consists of 4 items, asking respondents how often in the past month they had woken up several times in the night, had trouble staying asleep, or woke up too early in the morning. There were 6 response options, ranging from 0 (not at all) to 5 (22–31 days). The total amount of sleep problems was computed with scores ranging from 0 (no problems) to 20 (severe problems). The Cronbach’s alpha in this sample was 0.77. Being in the highest quartile of the sleep problems scale represents the group with the greatest sleep problems.
A four item depression subscale scored 0–12 (Cronbach’s alpha=0.88) was identified from the 30 item General Health Questionnaire on the basis of factor analysis and comparison with the items of the depression subscale of the 28 item General Health Questionnaire(6). The depression subscale was dichotomised so that the highest scoring quartile (scores >3) represented the highest level of depressive symptoms which was contrasted to the remaining group. As a fifth indicator of health we included the seven item Rose angina questionnaire(31). This widely used measure defines angina pectoris as pain located over the sternum or in both the left chest and left arm that is precipitated by exertion, causes the person to stop, and goes away in 10 minutes or less.
Main exposure (phase 7)
Non-reciprocal social relations were measured using an 9 item questionnaire from which three scales were extracted by factor analysis: a scale containing items relating to marital (or partnership) life (3 items); a scale measuring critical parent-children exchange (4 items); and a scale measuring failed reciprocity in unspecified trusting relationships (2 items) (14) (see Table 1 and Appendix). All items are answered in two steps. Firstly, subjects agree or disagree on whether or not the item content describes a typical experience of their social relationships. Subsequently, respondents who agree are asked to evaluate to what extent, they usually feel distressed (“bothered”) by this experience. Ratings are given on a four-point Likert-scale ranging from not at all to very much. Both answers are combined by creating a scale ranging from 0 (disagree) to 4 (agree, very much distressed). For each factor, means and standard deviation are calculated, and tertiles of the scores are defined (the lower two tertiles of the scale for children were combined as there were too few participants in the lowest tertile). The validity and reliability of these three factors have been repeatedly established(14;15).
Table 1.
Factor Analysis of items on Lack of Reciprocity from Partners, Children and General Trust: Whitehall II, Phase 7
| Factor 1 Partner | Factor 2 Children | Factor 3 General Trust | |
|---|---|---|---|
| Good relationship with partner ahead of own needs | −0.72 | 0.04 | 0.06 |
| Mutual understanding in relationship to partner | −0.85 | −0.01 | 0.02 |
| Balance of give and take in relationship to partner | −0.86 | 0.00 | 0.00 |
| Well respected by children | 0.00 | 0.67 | 0.03 |
| Sufficient support from children | −0.11 | 0.85 | −0.05 |
| Childred educated up to the level you hoped for | 0.18 | 0.63 | 0.07 |
| Appreciation for providing help in family | −0.18 | 0.77 | −0.05 |
| Disappointment | −0.14 | 0.05 | 0.76 |
| Injustice | 0.06 | −0.03 | 0.90 |
| Eigenvalue | 3.12 | 1.54 | 1.08 |
| Explained variance | 34.70 | 17.09 | 12.03 |
| Cronbach’s alpha | 0.78 | 0.69 | 0.62 |
Other covariates (phase 7)
Health behaviours such as smoking, physical activity (measured in METS scores), excessive alcohol intake (greater than 28 units/week for men, 21 units/week for women) and fruit/vegetable consumption were measured from the phase 7 questionnaire. Negative social support and social networks were also measured from the phase 7 questionnaire using four items included in the Close Person’s Questionnaire (32) for the former and questions on frequency of contact with relative and friends(32) drawn from Berkman and Syme (33) for the latter.
Analysis
We first analysed the factorial structure of the non-reciprocity items and whether separate scales for each type of close relationship could be constructed (Table 1). After exploration of bivariate associations of all health measures and main socio-demographic variables with non-reciprocity scores (Table 2) the main hypothesis is tested using linear and logistic regression models (Table 3). The baseline model adjusts for age and sex. Further nested models were analysed to examine the effect of confounders and mediators such as negative social support (model 2), previous health status (assessed at phase 5), employment grade, health- related behaviours and social networks (model 3). Previous health status was specific to the health outcome analysed, so for example the analysis of phase 7 sleep problems was adjusted for sleep problems at phase 5. Health related behaviours included smoking, alcohol consumption, physical activity and fruit/vegetable consumption.
Table 2.
Unadjusted Mean (N) of Lack of Reciprocity from Partners, Children and General Trust by health and sociodemographic variables: Whitehall II, Phase 7
| Partner | Child | Trust | |
|---|---|---|---|
| Sleep Problems | |||
| Lowest quartile (fewest problems) | 0.58 (1284) | 0.15 (1054) | 0.70 (1489) |
| Quartile 2 | 0.72 (1391) | 0.22 (1143) | 0.81(1581) |
| Quartile 3 | 0.85 (1458) | 0.25 (1203) | 0.93 (1659) |
| Highest Quartile (highest problems) | 1.12 (1307) | 0.39 (1071) | 1.23 (1564) |
| p for trend | 0.00 | 0.00 | 0.00 |
| GHQ-Depression | |||
| No | 0.76 (4815) | 0.22 (3960) | 0.85 (5554) |
| Case | 1.28 (620) | 0.52 (507) | 1.45 (732) |
| p for trend | 0.00 | 0.00 | 0.00 |
| SF_36 PCS | |||
| Lowest quartile (worst health) | 0.94 (1233) | 0.32 (1081) | 1.12 (1526) |
| Quartile 2 | 0.81 (1358) | 0.26 (1106) | 0.92 (1537) |
| Quartile 3 | 0.73 (1378) | 0.21 (1137) | 0.78 (1560) |
| Highest Quartile (best health) | 0.81 (1354) | 0.23 (1053) | 0.83 (1533) |
| p for trend | 0.00 | 0.00 | 0.00 |
| SF_36 MCS | |||
| Lowest quartile (worst health) | 1.30 (1238) | 0.47 (1016) | 1.40 (1511) |
| Quartile 2 | 0.85 (1337) | 0.25 (1056) | 0.88 (1539) |
| Quartile 3 | 0.62 (1377) | 0.19 (1128) | 0.74 (1553) |
| Highest Quartile (best health) | 0.55 (1371) | 0.13 (1177) | 0.65 (1553) |
| p for trend | 0.00 | 0.00 | 0.00 |
| Rose Angina | |||
| No | 0.81 (5241) | 0.25 (4294) | 0.90 (6051) |
| Yes | 1.11 (197) | 0.41 (177) | 1.19 (238) |
| p for trend | 0.00 | 0.00 | 0.00 |
| Agegroup | |||
| 50–55 | 0.95 (1018) | 0.34 (710) | 0.99 (1154) |
| 55–60 | 0.87 (1664) | 0.27 (1278) | 0.99 (1893) |
| 60–65 | 0.82 (1172) | 0.27 (996) | 0.92 (1355) |
| 65–74 | 0.69 (1603) | 0.19 (1504) | 0.79 (1913) |
| p for trend | 0.00 | 0.00 | 0.00 |
| Employment Grade | |||
| High | 0.82 (2738) | 0.24 (2238) | 0.85 (2966) |
| Intermediate | 0.83 (2258) | 0.28 (1767) | 0.97 (2689) |
| Low | 0.75 (435) | 0.23 (459) | 1.00 (629) |
| p for trend | 0.24 | 0.04 | 0.00 |
| Sex | |||
| Men | 0.77 (4160) | 0.25 (3471) | 0.84 (4543) |
| Women | 0.97 (1297) | 0.27 (1017) | 1.11 (1772) |
| p for trend | 0.00 | 0.26 | 0.00 |
| Marital Status | |||
| Married/Cohabit/Civil Partnership | 0.79 (5014) | 0.24 (3916) | 0.81 (4942) |
| Widowed | 1.05 (133) | 0.77 (24) | 1.12 (633) |
| Divorced | 1.47 (191) | 0.42 (312) | 1.76 (453) |
| Separated | 0.84 (109) | 0.25 (227) | 0.91 (274) |
| p for trend | 0.00 | 0.00 | 0.00 |
Table 3.
Linear and Logistic Regression coefficients (95% CI) of poor health by Lack of Reciprocity and Negative Social Support- continues on next page
| Sleep Problems | Model 1 | Model 2 | Model 3 | N |
|---|---|---|---|---|
| Lack of Reciprocity:Partner 1 | 4813 | |||
| Intermediate3 | 0.09 (0.02, 0.17) | 0.04 (−0.04, 0.11) | 0.03 (−0.04, 0.09) | 1976 |
| High3 | 0.44 (0.35, 0.52) | 0.30 (0.21, 0.38) | 0.12 (0.05, 0.19) | 1551 |
| Negative social support 1 | 4813 | |||
| Intermediate3 | 0.22 (0.14, 0.30) | 0.13 (0.06, 0.20) | 981 | |
| High3 | 0.33 (0.26, 0.41) | 0.19 (0.13, 0.25) | 1766 | |
| Lack of Reciprocity:Children 2 | 3933 | |||
| High3 | 0.27 (0.19, 0.35) | 0.22 (0.14, 0.29) | 0.08 (0.01, 0.14) | 1017 |
| Negative social support 1 | 3933 | |||
| Intermediate3 | 0.19 (0.10, 0.28) | 0.11 (0.03, 0.18) | 772 | |
| High3 | 0.35 (0.27, 0.42) | 0.17 (0.11, 0.23) | 1400 | |
| Lack of Reciprocity: Trust 1 | 5525 | |||
| Intermediate3 | 0.07 (−0.01, 0.16) | 0.05 (−0.04, 0.13) | 0.02 (−0.04, 0.09) | 825 |
| High3 | 0.37 (0.30, 0.43) | 0.31 (0.24, 0.38) | 0.13 (0.08, 0.19) | 1640 |
| Negative social support 1 | 5525 | |||
| Intermediate3 | 0.24 (0.16, 0.32) | 0.14 (0.07, 0.20) | 1109 | |
| High3 | 0.39 (0.32, 0.45) | 0.22 (0.16, 0.27) | 1955 | |
| General Health Questionnaire-30 Depression | ||||
| Lack of Reciprocity:Partner 1 | 4750 | |||
| Intermediate3 | 0.06 (−0.06, 0.18) | −0.06 (−0.18, 0.06) | −0.02 (−0.12, 0.09) | 1953 |
| High3 | 0.67 (0.54, 0.80) | 0.34 (0.21, 0.47) | 0.16 (0.04, 0.28) | 1529 |
| Negative social support 1 | 4750 | |||
| Intermediate3 | 0.24 (0.11, 0.37) | 0.12 (0.01, 0.23) | 973 | |
| High3 | 0.79 (0.67, 0.90) | 0.46 (0.36, 0.56) | 1738 | |
| Lack of Reciprocity:Children 2 | 3877 | |||
| High3 | 0.63 (0.50, 0.75) | 0.50 (0.37, 0.62) | 0.27 (0.17, 0.38) | 999 |
| Negative social support 1 | 3877 | |||
| Intermediate3 | 0.26 (0.12, 0.40) | 0.11 (−0.01, 0.23) | 764 | |
| High3 | 0.82 (0.71, 0.94) | 0.48 (0.37, 0.58) | 1378 | |
| Lack of Reciprocity: Trust 1 | 5453 | |||
| Intermediate3 | 0.05 (−0.08, 0.18) | 0.00 (−0.13, 0.13) | −0.01 (−0.13, 0.10) | 816 |
| High3 | 0.62 (0.52, 0.73) | 0.50 (0.40, 0.61) | 0.25 (0.16, 0.34) | 1610 |
| Negative social support 1 | 5453 | |||
| Intermediate3 | 0.29 (0.17, 0.41) | 0.16 (0.06, 0.26) | 1098 | |
| High3 | 0.84 (0.74, 0.94) | 0.49 (0.40, 0.58) | 1924 | |
| SF-36 Physical Component Score | ||||
| Lack of Reciprocity:Partner 1 | 4633 | |||
| Intermediate3 | −0.38 (−0.97, 0.22) | −0.15 (−0.75, 0.45) | −0.16 (−0.66, 0.34) | 1926 |
| High3 | −1.24 (−1.87, −0.61) | −0.63 (−1.30, 0.04) | −0.23 (−0.79, 0.33) | 1484 |
| Negative social support 1 | 4633 | |||
| Intermediate3 | −0.51 (−1.15, 0.14) | −0.09 (−0.63, 0.45) | 936 | |
| High3 | −1.45 (−2.03, −0.87) | −0.50 (−0.98, −0.02) | 1695 | |
| Lack of Reciprocity:Children 2 | 3783 | |||
| High3 | −1.02 (−1.64, −0.41) | −0.79 (−1.41, −0.16) | −0.36 (−0.88, 0.16) | 978 |
| Negative social support 1 | 3783 | |||
| Intermediate3 | −0.77 (−1.49, −0.04) | −0.27 (−0.87, 0.34) | 731 | |
| High3 | −1.46 (−2.07, −0.84) | −0.45 (−0.96, 0.05) | 1340 | |
| Lack of Reciprocity: Trust 1 | 5314 | |||
| Intermediate3 | −0.74 (−1.41, −0.07) | −0.65 (−1.32, 0.02) | −0.07 (−0.62, 0.47) | 793 |
| High3 | −1.42 (−1.94, −0.90) | −1.19 (−1.72, −0.66) | −0.57 (−1.00, −0.14) | 1565 |
| Negative social support 1 | 5314 | |||
| Intermediate3 | −0.77 (−1.39, −0.16) | −0.23 (−0.73, 0.27) | 1059 | |
| High3 | −1.54 (−2.06, −1.02) | −0.57 (−0.99, −0.14) | 1873 | |
| SF_36 Mental Component Score | ||||
| Lack of Reciprocity:Partner 1 | 4633 | |||
| Intermediate3 | −0.87 (−1.46, −0.28) | −0.25 (−0.83, 0.33) | −0.43 (−0.95, 0.10) | 1926 |
| High3 | −4.83 (−5.45, −4.20) | −3.15 (−3.80, −2.49) | −1.84 (−2.43, −1.25) | 1484 |
| Negative social support 1 | 4633 | |||
| Intermediate3 | −1.60 (−2.23, −0.97) | −1.04 (−1.60, −0.48) | 936 | |
| High3 | −3.97 (−4.53, −3.41) | −2.54 (−3.05, −2.04) | 1695 | |
| Lack of Reciprocity:Children 2 | 3783 | |||
| High3 | −3.31 (−3.92, −2.71) | −2.54 (−3.14, −1.95) | −1.23 (−1.76, −0.70) | 978 |
| Negative social support 1 | 3783 | |||
| Intermediate3 | −1.89 (−2.59, −1.20) | −1.15 (−1.76, −0.54) | 731 | |
| High3 | −4.67 (−5.26, −4.09) | −2.93 (−3.45, −2.41) | 1340 | |
| Lack of Reciprocity: Trust 1 | 5314 | |||
| Intermediate3 | −0.36 (−1.02, 0.30) | −0.10 (−0.74, 0.55) | 0.17 (−0.40, 0.73) | 793 |
| High3 | −3.95 (−4.46, −3.43) | −3.25 (−3.76, −2.75) | −1.70 (−2.16, −1.24) | 1565 |
| Negative social support 1 | 5314 | |||
| Intermediate3 | −1.90 (−2.49, −1.30) | −1.29 (−1.82, −0.77) | 1059 | |
| High3 | −4.63 (−5.13, −4.13) | −3.01 (−3.46, −2.56) | 1873 | |
| Rose Angina | ||||
| Lack of Reciprocity:Partner 1 | 5227 | |||
| Intermediate4 | 1.37 (0.91, 2.05) | 1.25 (0.83, 1.89) | 1.34 (0.87, 2.06) | 2142 |
| High4 | 1.99 (1.33, 2.96) | 1.56 (1.02, 2.39) | 1.51 (0.96, 2.36) | 1695 |
| Negative social support 1 | 5227 | |||
| Intermediate4 | 1.01 (0.65, 1.59) | 0.91 (0.57, 1.46) | 1056 | |
| High4 | 1.75 (1.23, 2.49) | 1.47 (1.01, 2.13) | 1938 | |
| Lack of Reciprocity:Children 2 | 4277 | |||
| High4 | 1.46 (1.04, 2.04) | 1.32 (0.94, 1.86) | 1.19 (0.83, 1.70) | 1120 |
| Negative social support 1 | 4277 | |||
| Intermediate4 | 1.35 (0.86, 2.10) | 1.12 (0.70, 1.79) | 836 | |
| High4 | 1.84 (1.29, 2.63) | 1.54 (1.06, 2.25) | 1539 | |
| Lack of Reciprocity: Trust 1 | 6002 | |||
| Intermediate4 | 1.09 (0.73, 1.62) | 1.04 (0.70, 1.56) | 1.06 (0.70, 1.62) | 897 |
| High4 | 1.53 (1.15, 2.06) | 1.39 (1.03, 1.87) | 1.15 (0.84, 1.57) | 1801 |
| Negative social support 1 | 6002 | |||
| Intermediate4 | 1.18 (0.79, 1.75) | 1.01 (0.67, 1.52) | 1197 | |
| High4 | 1.89 (1.40, 2.56) | 1.60 (1.16, 2.21) | 2142 | |
Reference category is the lowest tertile
Reference category is the lowest and middle tertile
Linear Regression beta coefficients
Logistic Regression odds ratio coefficients
Model 1- adjusted for age and sex
Model 2- adjusted for Model 1, age, sex and negative social support
Model 3- adjusted for Model 2, prior health status, grade, health behaviours, social networks
Results
In Table 1, the factorial structure of items measuring nonreciprocal social exchange outside work is displayed. The factor loadings suggest three factors may be distinguished, corresponding to the expected domains of non-reciprocity in partnership, children and general trust. The amount of variance explained and the Cronbach’s α values are satisfactory. In addition, further analysis was carried out to test if these factors were distinct from each other and the measure of negative social support. Factor analysis using oblique rotation allowing for correlated factors revealed moderate to weak correlations (<0.45) between the factors. There was a moderate correlation (0.43) between negative social support and non-reciprocal partnership relationships, however factor analysis of the items comprising these scales revealed distinct and non-overlapping factors.
Table 2 represents the mean of the non-reciprocal relationships (for partners, children and general trust) by the health measures and key covariates. Almost all differences are statistically significant with higher means of non-reciprocity among those with reduced health. In addition, means are higher in younger compared to older participants, in women compared to men, and in divorced/widowed/separated versus married participants. A social gradient was found for general trust, but not for relationships with partners or children.
Table 3 represents the results of the linear and logistic regression analyses testing the main hypothesis with respect to four continuous measures of health, (1) sleep problems, (2) GHQ-depression, (3) SF-36 Physical (PCS) and (4) SF-36 Mental (MCS) functioning, and and the binary health measure of (5) Rose angina. Model 1 adjusts for age, and gender Model 2 is nested within Model 1 and further adjusts for negative social support. Model 3 is nested within Model 2 and further adjusts for previous health status, employment grade, health behaviours (smoking, alcohol, physical activity and diet) and social networks. Insert Table 3 about here
The results in all the analyses in Model 1 column of Table 3 show a consistent pattern of association of poor health in the group scoring highest on measures of failed reciprocity compared to the lowest scoring group. Higher levels of non-reciprocity were associated with higher levels of sleep problems, GHQ-30 depression and lower levels of SF_36 physical and metal health. Health differences are particularly pronounced in case of depression and poor mental functioning. Elevated odds of Rose-angina are also observed with increasing non-reciprocity. Some reductions in the size of these coefficients are observed when negative social support is taken into account (Model 2), although nearly all the associations remain statistically significant. There is a further reduction in these associations when previous health, employment grade, health behaviours and social networks are taken into account (Model 3). The confidence intervals in 10 out of 15 model estimates for the highest level of non-reciprocity indicate statistically significant associations.
In addition to the results displayed in Table 3, we examined 2-way interaction effects between non-reciprocity and gender, age, marital status and occupational grade, but found little evidence of effect modification of non-reciprocity by gender, age, marital status and occupational grade.
Discussion
This study found consistent associations of reports on failed reciprocity in social relationships with five measures of reduced health in the Whitehall II study population. Effects remained significant after adjusting for negative social support, an important condition associated with poorer health(32). Findings support the significance for health of a theoretical model, the effort-reward imbalance model, that focuses on experienced violation of the norm of reciprocity in core social roles in adult life beyond the work role, namely in marital and parental roles and civic roles(4;34). The findings of this study are in line with those from two previous investigations(14;15)., but are more robust as they are based on a large sample, include five different health measures and are adjusted for the effects of important confounders and mediators.
Failed reciprocity may be linked with poor health through unhealthy behaviours (addictive behaviours may increase in non-reciprocal relationships), although there was little evidence for this from this study. Failed reciprocity may also affect health by eliciting strong negative emotions and stress reactions, which may be either chronic (experience of continued unjust treatment) or acute (experience of a critical life event of deceived trust), but in all cases it is the mismatch between efforts spent and rewards received in turn that matters most. Adjusting for negative social support from a close person reduced the associations between the different health outcomes and non-reciprocity, although both remained independently associated with health. Non-reciprocal partnerships were moderately correlated with negative social support from a close person which suggests that some of the association between non-reciprocity and health may be mediated through negative social support(35). Although mean scores of failed reciprocity from partner and general trust were higher in women than in men we did not observe gender-specific differences in the strength of association with health. Previous reports found gender differences with regard to health effects of providing and receiving support (23–25). For instance, in a Finnish prospective study giving more support than receiving was beneficial for health among women, whereas in men, the reversed effect was found(23). However, in this study, providing and receiving support was not weighted against efforts spent, as in the current approach, and the two remaining types of social relationships, parental and civic roles, were not included.
Despite these merits several limitations must be considered. First, given the cross-sectional design reverse causation cannot be excluded, especially so in case of depression and poor mental functioning. Depressed people may perceive their close social relationships in a more negative way, they may be more vulnerable even to minor experiences of unfair treatment, or their behaviour may evoke non-reciprocal responses in their partners. This interpretation is less likely in case of angina pectoris where similar associations with reduced health were observed. The analyses were adjusted for previous health status, reducing the likelihood of reverse causation. As all the health measures and all exposure measures are based on self-reported data the problem of common method variance needs to be addressed(36). Unfortunately, negative affect was not measured concurrently with the variables analysed, although adjusting for negative affect measured around fifteen years earlier did not change the results. It should also be noted that in previous studies on associations of effort-reward imbalance with self-reported health effects in general remained significant after adjusting for negative affect(37),(38).
A further limitation concerns the measurement of experienced failed reciprocity which is exclusively based on participants’ self assessment. We were not able to validate information by triangulating data with those received from partners. Obviously, more detailed observational or experimental studies of microsocial exchange(19;20) are needed to supplement epidemiological findings. Moreover, for reasons explained above, a quantitative estimate of a mismatch between effort spent and reward received, as applied to the measurement of effort-reward imbalance at work(7), was not feasible in the frame of assessing non-reciprocity in close social relationships. Given the constraints of economic measures in large-scale epidemiological investigations, the few items of this questionnaire cannot adequately represent the rich dynamics of imbalanced interpersonal exchange. Important additional aspects of non-reciprocity (e.g., unpaid domestic work, unequal role distribution) including exposure time have not been assessed. On the other hand, our measure comprises condensed information on quality of exchange in three important domains of close social relationships, and these domains are replicated as independent scales at the measurement level(14;15).
Although we included five different measures of health that were only moderately interrelated, no clinically validated health data were included in this analysis. Further investigations are needed to link failed social reciprocity prospectively with incident disease, as was done previously in case of effort-reward-imbalance at work (reviews (8),(9),(10). A final limitation is related to the sample under study. Despite the fact that several previous findings from the Whitehall II study were replicated in other cohort studies(18;39) we cannot exclude the possibility that the results reflect experiences that are typical for male and female civil servants or for a middle-aged to early old age population in North Western Europe. However, similar results were reported from different age groups in two other countries, the United States and Germany, thus reducing the probability of this limitation(14;15).
In conclusion, we found a consistent association of failed reciprocity in three domains of close social relationships with reduced health in a large cohort of middle-aged to early old age men and women, the Whitehall II cohort. Findings support the core notion of a theoretical model, the effort-reward imbalance model, which has mainly been analyzed in the frame of work and health. If corroborated by additional evidence, this approach may have direct implications for counseling activities in the prevention and treatment of disorders with a more proximal stress-related background.
Acknowledgments
The Whitehall II study has been supported by grants from the Medical Research Council; British Heart Foundation; Health and Safety Executive; Department of Health; National Heart Lung and Blood Institute (HL36310), US, NIH: National Institute on Aging (AG13196), US, NIH; Agency for Health Care Policy Research (HS06516); and the John D and Catherine T MacArthur Foundation Research Networks on Successful Midlife Development and Socio-economic Status and Health. We thank all participating Civil Service departments and their welfare, personnel, and establishment officers; the Occupational Health and Safety Agency; the Council of Civil Service Unions; all participating civil servants in the Whitehall II study; all members of the Whitehall II study team.
We thank all members of the Whitehall II Study team. We also thank all participating civil servants in the Whitehall II Study, the participating civil service departments, their welfare, personnel, and establishment officers, the Occupational Health and Safety Agency, London, and the Council of Civil Service Unions, London.
Appendix
Items from the Non-Reciprocal Social Relations Questionnaire
The following questions are related to your personal life situation, your children, your spouse or partner and other people you deal with. Please, give a response that fits best with your typical experiences.
Have you ever had to put the relationship with your partner ahead of your own needs in order to maintain a good relationship?
Has the relationship with your partner always been characterised by mutual understanding?
Have you always been satisfied with the balance between what you have given your partner and what you have received in return?
Compared to your expectations, do you think that you are generally well respected by your children?
Have you always received sufficient support from your children in difficult situations?
Has/Have your child/children been educated up to the level you had hoped for?
Do you feel that you have always received adequate appreciation for providing help in your family (such as taking care of your grand-children)?
Has any person you gave your trust seriously disappointed or hurt you?
Has anyone ever committed an injustice against you or betrayed you without you being compensated for it?
| Yes | No | Not Applicable |
If Yes/No: How much does this bother you?
| Not At All | A Little | Somewhat | Greatly |
Items of Negative Social Support from the Close Person’s Questionnaire
Thinking about the person you are closest to, please tell us how you would rate the practical and emotional support they have provided for you IN THE LAST 12 MONTHS.
| Not at all | A little | Quite a lot | A great deal | |
|---|---|---|---|---|
| e) How much in the last 12 months did this person give you worries, problems and stress? | ||||
| i) How much in the last 12 months would you have liked to have confided more in this person? | ||||
| j) How much in the last 12 months did talking to this person make things worse? | ||||
| n) How much in the last 12 months would you have liked more practical help with major things from this person? |
Footnotes
Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
Reference List
- 1.Brown GW. Life Events and Illness. London: Guilford; 1989. [Google Scholar]
- 2.Berkman LF, Glass T. Social integration, social networks, social support and health. In: berkman l, Kawachi I., editors. Social Epidemiology. Oxford: OUP; 2000. pp. 137–73. [Google Scholar]
- 3.Cacioppo JT, Hawkley LC, Crawford LE, Ernst JM, Burleson MH, Kowalewski RB, et al. Loneliness and health: potential mechanisms. Psychosom Med. 2002 May;64(3):407–17. doi: 10.1097/00006842-200205000-00005. [DOI] [PubMed] [Google Scholar]
- 4.Siegrist J. Adverse health effects of high-effort/low-reward conditions. J Occup Health Psychol. 1996 Jan;1(1):27–41. doi: 10.1037//1076-8998.1.1.27. [DOI] [PubMed] [Google Scholar]
- 5.Gouldner AW. The norm of social reciprocity. A preliminary statement. Am Soc Rev. 1960;25:161–79. [Google Scholar]
- 6.Stansfeld SA, North FM, White I, Marmot MG. Work characteristics and psychiatric disorder in civil servants in London. J Epidemiol Community Health. 1995 Feb;49(1):48–53. doi: 10.1136/jech.49.1.48. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Siegrist J, Starke D, Chandola T, Godin I, Marmot M, Niedhammer I, et al. The measurement of effort-reward imbalance at work: European comparisons. Soc Sci Med. 2004 Apr;58(8):1483–99. doi: 10.1016/S0277-9536(03)00351-4. [DOI] [PubMed] [Google Scholar]
- 8.Siegrist J. Social reciprocity and health: new scientific evidence and policy implications. Psychoneuroendocrinology. 2005 Nov;30(10):1033–8. doi: 10.1016/j.psyneuen.2005.03.017. [DOI] [PubMed] [Google Scholar]
- 9.Tsutsumi A, Kawakami N. A review of empirical studies on the model of effort-reward imbalance at work: reducing occupational stress by implementing a new theory. Soc Sci Med. 2004 Dec;59(11):2335–59. doi: 10.1016/j.socscimed.2004.03.030. [DOI] [PubMed] [Google Scholar]
- 10.van Vegchel N, de Jonge J, Bosma H, Schaufeli W. Reviewing the effort-reward imbalance model: drawing up the balance of 45 empirical studies. Soc Sci Med. 2005 Mar;60(5):1117–31. doi: 10.1016/j.socscimed.2004.06.043. [DOI] [PubMed] [Google Scholar]
- 11.Vrijkotte TG, van Doornen LJ, de Geus EJ. Effects of work stress on ambulatory blood pressure, heart rate, and heart rate variability. Hypertension. 2000 Apr;35(4):880–6. doi: 10.1161/01.hyp.35.4.880. [DOI] [PubMed] [Google Scholar]
- 12.Steptoe A, Siegrist J, Kirschbaum C, Marmot M. Effort-reward imbalance, overcommitment, and measures of cortisol and blood pressure over the working day. Psychosom Med. 2004 May;66(3):323–9. doi: 10.1097/01.psy.0000126198.67070.72. [DOI] [PubMed] [Google Scholar]
- 13.Eller NH, Netterstrom B, Hansen AM. Psychosocial factors at home and at work and levels of salivary cortisol. Biol Psychol. 2006 Oct;73(3):280–7. doi: 10.1016/j.biopsycho.2006.05.003. [DOI] [PubMed] [Google Scholar]
- 14.von dem Knesebeck O, Siegrist J. Reported nonreciprocity of social exchange and depressive symptoms. Extending the model of effort-reward imbalance beyond work. J Psychosom Res. 2003 Sep;55(3):209–14. doi: 10.1016/s0022-3999(02)00514-7. [DOI] [PubMed] [Google Scholar]
- 15.von dem Knesebeck O, Siegrist J. Non-reciprocity in close social relationships, depressive symptoms, and poor self-rated health. Soz Praventivmed. 2004;49(5):336–43. doi: 10.1007/s00038-004-3071-5. [DOI] [PubMed] [Google Scholar]
- 16.Stansfeld SA, Rael EG, Head J, Shipley M, Marmot M. Social support and psychiatric sickness absence: a prospective study of British civil servants. Psychol Med. 1997 Jan;27(1):35–48. doi: 10.1017/s0033291796004254. [DOI] [PubMed] [Google Scholar]
- 17.Rael EG, Stansfeld SA, Shipley M, Head J, Feeney A, Marmot M. Sickness absence in the Whitehall II study, London: the role of social support and material problems. J Epidemiol Community Health. 1995 Oct;49(5):474–81. doi: 10.1136/jech.49.5.474. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Chandola T, Martikainen P, Bartley M, Lahelma E, Marmot M, Michikazu S, et al. Does conflict between home and work explain the effect of multiple roles on mental health? A comparative study of Finland, Japan, and the UK. Int J Epidemiol. 2004 Aug;33(4):884–93. doi: 10.1093/ije/dyh155. [DOI] [PubMed] [Google Scholar]
- 19.Kiecolt-Glaser JK, Glaser R, Cacioppo JT, MacCallum RC, Snydersmith M, Kim C, et al. Marital conflict in older adults: endocrinological and immunological correlates. Psychosom Med. 1997 Jul;59(4):339–49. doi: 10.1097/00006842-199707000-00001. [DOI] [PubMed] [Google Scholar]
- 20.Kiecolt-Glaser JK, Loving TJ, Stowell JR, Malarkey WB, Lemeshow S, Dickinson SL, et al. Hostile marital interactions, proinflammatory cytokine production, and wound healing. Arch Gen Psychiatry. 2005 Dec;62(12):1377–84. doi: 10.1001/archpsyc.62.12.1377. [DOI] [PubMed] [Google Scholar]
- 21.Medalie JH, Goldbourt U. Angina pectoris among 10,000 men. II. Psychosocial and other risk factors as evidenced by a multivariate analysis of a five year incidence study. Am J Med. 1976 May 31;60(6):910–21. doi: 10.1016/0002-9343(76)90921-9. [DOI] [PubMed] [Google Scholar]
- 22.Mayou R. Prediction of emotional and social outcome after a heart attack. J Psychosom Res. 1984;28(1):17–25. doi: 10.1016/0022-3999(84)90036-9. [DOI] [PubMed] [Google Scholar]
- 23.Vaananen A, Buunk BP, Kivimaki M, Pentti J, Vahtera J. When it is better to give than to receive: long-term health effects of perceived reciprocity in support exchange. J Pers Soc Psychol. 2005 Aug;89(2):176–93. doi: 10.1037/0022-3514.89.2.176. [DOI] [PubMed] [Google Scholar]
- 24.Takizawa T, Kondo T, Sakihara S, Ariizumi M, Watanabe N, Oyama H. Stress buffering effects of social support on depressive symptoms in middle age: reciprocity and community mental health. Psychiatry Clin Neurosci. 2006 Dec;60(6):652–61. doi: 10.1111/j.1440-1819.2006.01579.x. [DOI] [PubMed] [Google Scholar]
- 25.Jou YH, Fukada H. Stress, health, and reciprocity and sufficiency of social support: the case of university students in Japan. J Soc Psychol. 2002 Jun;142(3):353–70. doi: 10.1080/00224540209603904. [DOI] [PubMed] [Google Scholar]
- 26.Marmot MG, Smith GD, Stansfeld S, Patel C, North F, Head J, et al. Health inequalities among British civil servants: the Whitehall II study. Lancet. 1991 Jun 8;337(8754):1387–93. doi: 10.1016/0140-6736(91)93068-k. [DOI] [PubMed] [Google Scholar]
- 27.Ware JE, Jr, Sherbourne CD. The MOS 36-item short-form health survey (SF-36). I. Conceptual framework and item selection. Med Care. 1992 Jun;30(6):473–83. [PubMed] [Google Scholar]
- 28.Hemingway H, Nicholson A, Stafford M, Roberts R, Marmot M. The impact of socioeconomic status on health functioning as assessed by the SF-36 questionnaire: the Whitehall II Study. Am J Public Health. 1997 Sep;87(9):1484–90. doi: 10.2105/ajph.87.9.1484. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29.Ware JE, Kosinski M, Keller SD. SF-36 physical and mental health summary scales: a user’s manual. Boston, M.A: The Health Institute; 1994. [Google Scholar]
- 30.Jenkins CD, Stanton BA, Niemcryk SJ, Rose RM. A scale for the estimation of sleep problems in clinical research. J Clin Epidemiol. 1988;41(4):313–21. doi: 10.1016/0895-4356(88)90138-2. [DOI] [PubMed] [Google Scholar]
- 31.Rose G, Hamilton PS, Keen H, Reid DD, McCartney P, Jarrett RJ. Myocardial ischaemia, risk factors and death from coronary heart-disease. Lancet. 1977 Jan 15;1(8003):105–9. doi: 10.1016/s0140-6736(77)91701-9. [DOI] [PubMed] [Google Scholar]
- 32.Stansfeld S, Marmot M. Deriving a survey measure of social support: the reliability and validity of the Close Persons Questionnaire. Soc Sci Med. 1992 Oct;35(8):1027–35. doi: 10.1016/0277-9536(92)90242-i. [DOI] [PubMed] [Google Scholar]
- 33.Berkman LF, Syme SL. Social networks, host resistance, and mortality: a nine-year follow-up study of Alameda County residents. Am J Epidemiol. 1979 Feb;109(2):186–204. doi: 10.1093/oxfordjournals.aje.a112674. [DOI] [PubMed] [Google Scholar]
- 34.Siegrist J. Reciprocity in basic social exchange and health: can we reconcile person-based with population-based psychosomatic research? J Psychosom Res. 1998 Aug;45(2):99–105. doi: 10.1016/s0022-3999(98)00019-1. [DOI] [PubMed] [Google Scholar]
- 35.Stansfeld S. Social support and social cohesion. In: Marmot M, Wilkinson R, editors. Social determinants of health. 2. Oxford: OUP; 2006. [Google Scholar]
- 36.Chen PY, Spector PE. Negative affectivity as the underlying cause of correlations between stressors and strains. J Appl Psychol. 1991 Jun;76(3):398–407. doi: 10.1037/0021-9010.76.3.398. [DOI] [PubMed] [Google Scholar]
- 37.Bosma H, Peter R, Siegrist J, Marmot M. Two alternative job stress models and the risk of coronary heart disease. Am J Public Health. 1998 Jan;88(1):68–74. doi: 10.2105/ajph.88.1.68. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 38.Ostry AS, Kelly S, Demers PA, Mustard C, Hertzman C. A comparison between the effort-reward imbalance and demand control models. BMC Public Health. 2003 Feb 27;3:10. doi: 10.1186/1471-2458-3-10. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 39.Fuhrer R, Shipley MJ, Chastang JF, Schmaus A, Niedhammer I, Stansfeld SA, et al. Socioeconomic position, health, and possible explanations: a tale of two cohorts. Am J Public Health. 2002 Aug;92(8):1290–4. doi: 10.2105/ajph.92.8.1290. [DOI] [PMC free article] [PubMed] [Google Scholar]