Researchers have found evidence for the contribution of social exchanges to health and well-being. For example, the role of positive social exchanges such as social support has been widely studied (Cantor, 1979; Eckenrode & Hamilton, 2000). Proponents of the stress-buffering effect of social support maintain that support received by an individual from social networks when under stress may intervene between stressors and health by attenuating the individual's stress appraisal response (Cohen & Wills, 1985; House, 1981).
However, taking the perspective that social networks provide an exclusively supportive function, ignores the aspects of relationships that impose a cost, rather than a benefit, on individuals' well-being. Since Rook (1984) pointed out that people in an individual's social networks can be critical and overbearing, several researchers have investigated the effects of negative social exchanges (hereafter called negative exchanges) on individuals' depressive symptoms (Okun & Keith, 1998) and physical health (Edwards, Hershberger, Russell, & Markert, 2001). It has also been argued that negative exchanges may exert a greater effect on individuals' mental health than do positive exchanges (Finch, Okun, Pool, & Ruehlman, 1999; Okun & Keith, 1998).
Although researchers have shown that negative exchanges operate directly to decrease mental health symptoms, less attention has been paid to examining the stress-amplifying effects of negative exchanges, that is, the effects that exacerbate the negative association of stressors with well-being (Shinn, Lehman, & Wong, 1984; Siegel, Raveis, & Karus, 1997). For example, people appear to harbor negative feelings about individuals who are distressed, and to be critical of them for allowing the distressing event to occur in the first place (Krause & Jay, 1991). Social interactions of this kind cause additional stress and, accordingly, result in a worse situation for the individual's mental health.
Adequate empirical support has not been provided in studies concerning the stress-amplifying effects of negative exchanges. Okun, Melichar, and Hill (1990) found that the association between the number of daily negative events and psychological distress was not amplified by negative social exchanges. On the other hand, Vinokur, Price, and Caplan (1996) observed that the financial strain experienced by unemployed persons increased their depressive symptoms by increasing negative exchanges with their partners.
The reason for these inconsistent findings may be, in part, because of the way researchers have conceptualized stressors. Okun et al. (1990) estimated participants' stress levels simply by counting the number of experienced life events. This method is problematic because it is difficult to determine whether the deleterious effects of stress involve a wide range or a few types of events (Krause, 1986). For example, a wife may be critical of her diabetic husband for not following an appropriate diet, because she wants him to engage in beneficial health behaviors (Lewis & Rook, 1999). In this case, negative exchanges with a spouse may result in increasing, rather than decreasing, the well-being of the individual who is suffering from the stress of illness. On the other hand, the aggregation of life event indices would dilute the individual influence of each life event and also obscure the role of negative exchanges that should mediate the association between life event stress and mental health. This idea was illustrated by Vinokur et al. (1996), who found the effect of a specific life event (i.e., financial strain) on depressive symptoms was amplified by negative exchanges of unemployed people with their spouses. However, few empirical investigations of the stress-amplifying effect of negative exchanges in this manner have been conducted.
I have extended the findings recorded in the current literature by examining whether or not negative exchanges increase the effect of either aggregate measures of stressful life events or specific types of life event stressors on depressive symptoms. The following hypotheses were proposed:
Hypothesis 1: The effect of an aggregate measure of life events on depressive symptoms will not be amplified by negative exchanges.
Hypothesis 2: The effect of several specific life events on depressive symptoms will be amplified by negative exchanges.
Students (N = 126) from a women's college (Seitoku University, Matsudo, Japan), enrolled in an introductory psychology class, participated in the research. The group consisted of 65 freshmen (51.6%), 60 sophomores (47.6%), and one senior (0.8%). Two people (one freshman and one senior) were excluded from the sample because they were much older than the other participants (50 and 51 years old, respectively). In addition, three people were excluded because they did not submit a written form giving informed consent for use of the data. Consequently, data from 121 participants were used in the final analysis. The age range was between 18 and 20 (M = 18.9).
Participants completed two scales in the first lecture at the beginning of the semester.
Negative exchanges were measured with the subscale of the Japanese version (Sumi, 2003) of the Interpersonal Relationship Inventory (IPRI; Tilden, Nelson, & May, 1990). The scale consists of 13 items such as "Some people I care about invade my privacy" and "There is someone in my life who gets mad if we have different opinions". Each item is rated on a 5-point Likert scale using anchors of 1 = strongly disagree and 5 = strongly agree. The items are summed in the total score range of 13 to 65, with a higher score indicating a greater perception of negative exchanges. The reliability of the scale was good (Cronbach's alpha = .88).
Participants' depressive symptoms were measured with the Japanese version (Shima, Shikano, Kitamura, & Asai, 1985) of the Center for Epidemiologic Studies Depression Scale (CES-D; Radloff, 1977). Participants indicated how often during the previous week they had experienced any of the 20 given symptoms. Each item was rated on a 4-point Likert scale ranging from 0 = rarely or none of the time to 3 = most or all of the time. Four positively worded items were reverse scored. Responses were summed to obtain total scores ranging from 0 to 60. Higher scores represented more depressive symptoms weighted by frequency of occurrence. Cronbach's alpha = .91.
Life event stress. I assessed life event experiences as stress indicators, with a 63-item checklist that was developed in Japan (Suzuki, Otsuka, & Kosugi, 2001). Participants were asked if they had experienced any of these listed events during the preceding year. Each event was scored as 1 (experienced) or 0 (not experienced). To avoid conceptual confusion between stressors and negative exchanges, I excluded 15 items that related to interpersonal problems (e.g., "teased about appearance", "jilted", "trouble with a friend"). The remaining 48 items consisted not only of standard events such as "pet died" and "car accident", but also typical events for college students, such as "took an entrance examination", "was asked for a date", and "failure of a presentation made to the class" (see Appendix).
The first part of the analysis included t tests to examine the association of negative exchanges with depressive symptoms. After dividing participants into two groups of high or low perception by mean scores for the negative exchanges, the CES-D scores of the low and high negative exchange groups were compared. Similarly, by performing t tests, the CES-D scores were compared depending on whether or not the participants had experienced an event.
The stress-amplifying effects of negative exchanges were then examined using analysis of covariance (ANCOVA), adjusting for the participants' grade (freshman or sophomore). Initially, general linear models were tested by entering the number of experienced life events, negative exchanges, and their two-way interaction as independent variables. The total CES-D score was used as a dependent variable in this model. The significance of the interaction term was estimated to address the question of whether the effect of life events on depressive symptoms differed depending on the perception of negative exchanges. Next, I investigated the amplifying effects of negative exchanges on an individual life event experience using a similar model, in which each life event, the perception of negative exchanges, and their two-way interaction were entered as independent variables. The significance of the interaction term was investigated as an index of the event-specific amplifying effect.
All statistical analyses were performed using SAS version 9.1.3 software (SAS Institute, 1999). The alpha significance level was set at .05.
Descriptive information for the study variables is shown in Table 1. Although the mean score of the CES-D scale (22.4) was higher than has been recorded for students in other countries or of other ethnicities (Peden, Reed, & Rayens, 2005; Yen, Robins, & Lin, 2000), the tendency was consistent with results reported in a cross-cultural study (Iwata & Buka, 2002).
The mean score of the negative exchanges was nearly equal to the score of Japanese female college students reported in a previous study (Sumi, 2003). Participants were divided into two groups, a low negative exchanges group (n = 62, 51.2%) and a high negative exchanges group (n = 59, 48.9%) for the analyses. The result of a t test showed that the CES-D score of the high negative exchanges group was significantly higher than that of the low negative exchanges group, t(119) = 3.26, p < .01.
On average participants had experienced approximately 10 events during the year. I performed t tests to investigate the association between each event experience and depressive symptoms. Results indicated that the following five events were significantly associated with the higher CES-D scores: parents' financial status deteriorated, t(55.4) = 2.70, p = .01; could not understand how to use a computer, t(118) = 2.05, p < .05; ran out of pocket money, t(119) = 2.46, p < .05; changed residence, t(119) = 3.37, p < .01; and failed a course, t(119) = 2.73, p < .01.
Testing the Stress-amplifying Effects of Negative Exchanges
An ANCOVA adjusted for the participants' grade was performed, based on the number of life events experienced, with the perception of negative exchanges and their interaction term included as independent variables. Neither the number of events nor the interaction of these with the perception of negative exchanges had a statistically significant effect on depressive symptoms, F(1, 115) = 2.80, ns; F(1, 115) = 2.18, ns, respectively.
The event-specific ANCOVA models were tested consecutively, that is, each event experience, the perception of negative exchanges, and their interaction term were entered in the models as independent variables, adjusting for the participants' grade. The analyses were performed on five events that indicated a significant association with depressive symptoms in preliminary analyses, that is, parents' financial status deteriorated, could not understand how to use a computer, ran out of pocket money, changed residence, and failed a course. Of these two (could not understand how to use a computer, and changed residence) indicated a significant main effect, F(1, 115) = 4.10, p < .05, and F(1, 116) = 7.10, p < .01 respectively, whereas they did not significantly interact with the perception of negative exchanges, F(1, 115) = 0.27, ns; F(1, 116) = 0.17, ns, respectively. The results suggest that those event experiences were associated with more depressive symptoms regardless of the perception of negative exchanges.
The three remaining events (parents' financial status deteriorated, ran out of pocket money, and failed a course) also indicated a significant main effect to increase depressive symptoms, F(1, 115) = 7.78, p < .01; F(1, 116) = 6.13, p < .05; F(1, 116) = 5.11, p < .05, respectively. Analyses also indicated that these events significantly interacted with the perception of negative exchanges, F(1, 115) = 7.32, p < .01; F(1, 116) = 5.21, p < .05; F(1, 116) = 4.90, p < .05, respectively. The results suggest that the effect of the three events on depressive symptoms was moderated by the perception of negative exchanges.
Results in Table 2 show the nature of these interactions by displaying the mean CES-D scores based on the event experience and perception of negative exchanges. Further tests were conducted to examine the simple main effects of the event experience on each negative exchanges group using the Tukey-Kramer adjustment method. After controlling for the participants' grade, results indicated that among the high negative exchanges group, a high CES-D score was significantly associated with three events, that is, parents' financial status deteriorated, t(118) = 3.69, p < .01; ran out of pocket money, t(119) = 3.15, p < .01; and failed a course, t(119) = 3.10, p < .01. For the low negative exchanges group, however, there were no significant differences in the CES-D score regardless of whether or not the participants had experienced an event.
As research on stressful life events has evolved, researchers have devised increasingly complex models of the stress-health relationship. Models in which the direct effect of stress on well-being only has been examined have been replaced by elaborate conceptual schemas in which the effect of stress is moderated by factors such as socioeconomic status, living arrangements, personality, and social relationships. In this study, the intricacies of the stress-health relationship were further illuminated by an examination of negative exchanges as a contributor to the effect of life events on depressive symptoms.
The first hypothesis was that the effect of an aggregate measure of life events on depressive symptoms would not be amplified by negative exchanges. Data analysis supported the hypothesis because the two-way interaction between the number of experienced life events and negative exchanges was not statistically significant. This is consistent with previous findings, in which the amplifying effect of negative exchanges was investigated by using the number of life events as a stress indicator (Okun et al., 1990). In this study, only 5 out of a list of 48 events were significantly associated with depressive symptoms, although, on average, the participants reported having experienced 10 life events. It may be that some events participants had experienced were not so stressful as to cause depressive symptoms. The aggregate measure of life stress generated by simply counting the number of experienced events would mask the effect of events that were actually stressful and, therefore, make the role of negative exchanges as an amplifier less clear.
Results of analyses examining the effect of each listed life event on depressive symptoms suggest that some events, by themselves, are deleterious to individuals' well-being, and that other events become stressful when combined with negative exchanges. The findings in this study, therefore, also supported the second hypothesis that the effect of several specific life events on depressive symptoms would be amplified by negative exchanges.
Two events (parents' financial status deteriorated and ran out of pocket money) could be categorized as financial stressors. Vinokur and van Ryn (1993) found that there was a significant amplifying effect of negative exchanges on the relationship between financial stressors and psychological well-being. My findings confirm these results. People in financial difficulty may actually receive less assistance from their social network members, because this type of stressor has been found to be especially likely to be a source of embarrassment for them (Krause & Jay, 1991). For example, asking friends or acquaintances for money may bother them and may even result in provoking their anger (Fukukawa, 2007). In addition, financial strain may give rise to growing frustration that may trigger and sustain a variety of destructive interaction patterns among family members (Berkowitz, 1989). Indeed, Wadsworth and Compas (2002) found that family economic hardship was related to anxiety/depression and aggressive behavior of adolescent children partly caused by family conflict that was activated by the stressor. Thus, while certain life events, such as a recent death, tend to foster great support from significant network members, financial strain is a type of stressor that tends to foster intensely negative exchanges that amplify its deleterious effect on individuals.
The other event for which the effect was amplified by negative exchanges to increase depressive symptoms was failing a course. Because all participants were freshmen and sophomores, of itself the event would not be so stressful as to make them feel in danger of not being able to graduate. Therefore, as is the case with financial stressors, the result in this study requires an explanation of how this event becomes stressful via the mechanism of activating negative exchanges.
One possible explanation may be that this event is strongly associated with an individual's identity as a student. According to social identity theory (Burke, 1991; Thoits, 1991), stressful events arising in salient social roles exert an especially harmful effect on well-being, because they have the capacity to undermine the identity associated with those roles. A social role represents a position within a group and is associated with normative expectations that provide a basis for evaluating the adequacy of role performance (Krause, 1999). Thus, events such as failing a course could undermine the participant's identity as a student by activating problematic interactions among network members, for example, by disappointing one's parents or eliciting feelings of inferiority with friends. Unfortunately, relationships between identity salience and stress have rarely been evaluated empirically, so that my findings should be viewed with caution.
There are limitations in this study. First, because all the participants were female, the results cannot be generalized. For example, it has been found that the stress-buffering effect of social support is stronger for females than males (Olstad, Sexton, & S0gaard, 2001). Sex differences should be investigated in terms of the stress-amplifying effects of negative exchanges as well. Second, the results were based on cross-sectional data. A more strictly designed study using longitudinal data is needed to support the present findings, especially regarding etiological links between study variables. Third, the combination between stress and negative exchanges should be investigated in a stricter manner. Similar to unpleasant but wholesome advice from a dietitian, negative exchanges could potentially contribute to improving individuals' physical well-being, while temporarily decreasing psychological well-being. The effect of negative exchanges on health cannot be measured precisely by using analyses based only on depressive symptoms as an outcome. The present findings would be modified and extended if more positive outcomes were measured.
Despite these limitations, the findings supplement studies in which the relationship between social exchanges and mental health is examined. For example, the findings about a combination between specific life events and negative exchanges may help in understanding how negative exchanges act as modifiers in the association between stress and its outcomes. Researchers' findings on the stress-buffering effect of social support have contributed to enhancing the predictability of life event stress on health outcomes. Theoretically, the present findings that negative exchanges have a stress-amplifying effect could also make a positive contribution to stress research. Moreover, a practical implication of the results is that intervention programs for people under stress should be designed to help them avoid some types of social exchanges. This may be a groundbreaking insight for some mental health practitioners, because intervention for reconstructing social networks has been previously aimed mainly at increasing the availability of social exchanges. For this reason, studies in which the focus is on negative, as well as positive, aspects of social relationships have the potential to contribute to stress research and also to individuals' mental health.
Life Events List
1. Involved in trouble in club activities
2. Beginning of a new school term
3. Parents' financial status deteriorated
4. Took a certification examination
5. Retired from a club
6. Started a part-time job
7. Parents divorced
8. Pet died
9. Parents quarreled
10. Took an entrance examination
11. Involved in trouble in the workplace
12. Became a representative of a club
13. Periodic examination
14. Failed a certification examination
15. Could not understand how to use a computer
16. Personal injury or illness
17. Car accident
18. Was asked for a date
19. Involved in trouble during a trip
20. Failed an employment examination
21. Unsuccessful club activities
22. Involved in trouble in dormitory life
23. Death of a family member
24. Ran out of pocket money
25. Death of a close friend
26. Failure of a grade in school
27. Entered a university
28. Caused trouble in a club
29. Changed residence
30. Suffered from a disaster
32. Started job hunting
33. Serious illness of a family member
34. Started a diet
35. Damaged data file
36. Listened to a friend's problems
37. Attended a driving school
38. Not accepted at a college of choice
39. Rejected after a part-time job interview
40. Failure of a presentation made to the class
41. Began to write a graduation thesis
42. Started living alone
43. Lost possessions
44. Ill-prepared for a presentation
45. Caused trouble in the workplace
46. Increased club activities
47. Failed a course
48. Required to write an assignment
Berkowitz, L. (1989). Frustration-aggression hypothesis: Examination and reformulation. Psychological Bulletin, 106, 59-73. http://doi.org/b4j
Burke, P. J. (1991). Identity processes and social stress. American Sociological Review, 56, 836-849. http://doi.org/d7rz4p
Cantor, M. H. (1979). Neighbors and friends: An overlooked resource in the informal support system. Research on Aging, 1, 434-463. http://doi.org/dkd5fh
Cohen, S., & Wills, T. A. (1985). Stress, social support, and the buffering hypothesis. Psychological Bulletin, 98, 310-357. http://doi.org/b4k
Eckenrode, J., & Hamilton, S. (2000). One-to-one support intervention. In S. Cohen, L. G. Underwood, & B. H. Gottlieb (Eds.), Social support
measurement and intervention: A guide for health and social scientists (pp. 246-277). New York: Oxford University Press.
Edwards, K. J., Hershberger, P. J., Russell, R. K., & Markert, R. J. (2001). Stress, negative social exchange, and health symptoms in university students. Journal of American College Health, 50, 75-79. http://doi.org/b4m
Finch, J. F., Okun, M. A., Pool, G. J., & Ruehlman, L. S. (1999). A comparison of the influence of conflictual and supportive social interactions on psychological distress. Journal of Personality, 67, 581-621. http://doi.org/b4n
Fukukawa, Y. (2007). Rouka to sutoresu no shinrigaku: Taijin-kankeironteki apuro-chi [Stress, social relationships, and aging]. Tokyo: Koubundou.
House, J. S. (1981). Work stress and social support. Reading, MA: Addison-Wesley.
Iwata, N., & Buka, S. (2002). Race/ethnicity and depressive symptoms: A cross-cultural/ethnic comparison among university students in East Asia, North and South America. Social Science and Medicine, 55, 2243-2252. http://doi.org/b4p
Krause, N. (1986). Social support, stress, and well-being among older adults. Journal of Gerontology, 41, 512-519.
Krause, N. (1999). Stress and the devaluation of highly salient roles in late life. The Journals of Gerontology, Series B: Psychological Sciences and Social Sciences, 54B, 99-108. http://doi. org/b4q
Krause, N., & Jay, G. (1991). Stress, social support, and negative interaction in later life. Research on Aging, 13, 333-363. http://doi.org/b4r
Lewis, M. A., & Rook, K. S. (1999). Social control in personal relationships: Impact on health behaviors and psychological distress. Health Psychology, 18, 63-71. http://doi.org/b4s
Okun, M. A., & Keith, V. M. (1998). Effects of positive and negative social exchanges with various sources on depressive symptoms in younger and older adults. The Journals of Gerontology, Series B: Psychological Sciences and Social Sciences, 53B, 4-20. http://doi.org/b4t
Okun, M. A., Melichar, J. F., & Hill, M. D. (1990). Negative daily events, positive and negative social ties, and psychological distress among older adults. The Gerontologist, 30, 193-199. http://doi. org/b4v
Olstad, R., Sexton, H., & S0gaard, A. J. (2001). The Finnmark Study. A prospective population study of the social support buffer hypothesis, specific stressors and mental distress. Social Psychiatry and Psychiatric Epidemiology, 36, 582-589. http://doi.org/b4w
Peden, A. R., Reed, D. B., & Rayens, M. K. (2005). Depressive symptoms in adolescents living in rural America. The Journal of Rural Health, 21, 310-316. http://doi.org/b4x
Radloff, L. S. (1977). The CES-D Scale: A self-report depression scale for research in the general population. Applied Psychological Measurement, 1, 385-401. http://doi.org/b4z
Rook, K. S. (1984). The negative side of social interaction: Impact on psychological well-being. Journal of Personality and Social Psychology, 46, 1097-1108. http://doi.org/b42 SAS Institute. (2006). Base SAS 9.1.3 procedures guide. Cary, NC: Author.
Shima, S., Shikano, T., Kitamura, T., & Asai, M. (1985). Atarasii yokuutsu-sei jiko-hyouka syakudo ni tsuite [New self-rating scale for depression]. Clinical Psychiatry, 27, 717-723.
Shinn, M., Lehman, S., & Wong, N. W. (1984). Social interaction and social support. Journal of Social Issues, 40, 55-76. http://doi.org/b43
Siegel, K., Raveis, V. H., & Karus, D. (1997). Illness-related support and negative network interactions: Effects on HIV-infected men's depressive symptomatology. American Journal of Community Psychology, 25, 395-420. http://doi.org/b44
Sumi, K. (2003). Reliability and construct validity of the Japanese version of the Interpersonal Relationship Inventory. Perceptual and Motor Skills, 97, 135-140. http://doi.org/fst7pm
Suzuki, A., Otsuka, Y., & Kosugi, S. (2001, November). Daigakusei wo taisyou to shita raifu-ibento syakudo no sakusei no kokoromi [A pilot study for developing a life event scale for college students]. Poster presentation at the 65th annual convention of the Japanese Psychological Association, Tsukuba, Japan.
Thoits, P. A. (1991). On merging identity theory and stress research. Social Psychology Quarterly, 54, 101-112. http://doi.org/b52mb6
Tilden, V. P., Nelson, C. A., & May, B. A. (1990). The IPR inventory: Development and psychometric characteristics. Nursing Research, 39, 337-343. http://doi.org/b45
Vinokur, A. D., & van Ryn, M. (1993). Social support and undermining in close relationships: Their independent effects on the mental health of unemployed persons. Journal of Personality and Social Psychology, 65, 350-359. http://doi.org/b46
Vinokur, A. D., Price, R. H., & Caplan, R. D. (1996). Hard times and hurtful partners: How financial strain affects depression and relationship satisfaction of unemployed persons and their spouses. Journal of Personality and Social Psychology, 71, 166-179. http://doi.org/b47
Wadsworth, M. E., & Compas, B. E. (2002). Coping with family conflict and economic strain: The adolescent perspective. Journal of Research on Adolescence, 12, 243-274. http://doi.org/b48
Yen, S., Robins, C. J., & Lin, N. (2000). A cross-cultural comparison of depressive symptom manifestation: China and the United States. Journal of Consulting and Clinical Psychology, 68, 993-999. http://doi.org/b49
Yasuyuki Fukukawa, Department of Psychology, Waseda University.
Correspondence concerning this article should be addressed to: Yasuyuki Fukukawa, Department of Psychology, School of Humanities and Social Sciences, Waseda University, 1-24-1 Toyama Shinjuku-ku, Tokyo 162-8644, Japan. Email: email@example.com
Table 1. Descriptive Information for Study Variables Variable M SD CES-D 22.4 11.6 Negative social exchanges 34.5 10.5 Number of life events experienced 10.2 3.8 Notes: (N = 121). Score range of age, negative social exchanges, CES-D, and life events are 18-20, 13-65, 0-60, and 0-48, respectively. Table 2. Mean CES-D Scores and Standard Errors of the Amplified Events by Number of Event Experiences and Perception of Negative Exchanges Life event Negative Not experienced exchanges M SE Parents' financial status Low 19.2 1.5 deteriorated High 22.2 1.8 Ran out of pocket money Low 19.0 1.5 High 22.3 1.9 Failed a course Low 19.1 1.4 High 23.3 1.7 Life event Negative Experienced t exchanges M SE Parents' financial status Low 19.1 2.4 .03 deteriorated High 33.4 2.4 3.69 ** Ran out of pocket money Low 19.5 2.2 .22 High 32.1 2.5 3.15 ** Failed a course Low 19.3 3.1 .05 High 34.3 3.1 3.10 ** Note: ** p < .01.…