Introduction

Romantic relationships are typically construed as positive components of individuals’ lives. A meaningful relationship with a significant other may bring companionship, friendship, love, security, and happiness to an individual’s life. It may also bring health benefits. In fact, some research suggests that an absence of significant social relationships may be as detrimental to health as is smoking, high blood pressure, and obesity (House et al. 1988). However, it seems likely that individuals may not always think of social relationships as contributors to their health status, instead focusing on germs, family history, or health behaviors as determinants of their health. The present study will extend our understanding of the links between romantic relationships and health by using qualitative data to examine the extent to which individuals are aware of the impact their significant others have on their health and in what domains. Further, we intend to demonstrate the importance of individuals’ perceptions of their romantic partners’ influence on their health by revealing links between these perceptions and individuals’ reports of their relationship quality and actual health behaviors and outcomes.

The contributions of romantic relationships, particularly marriage, to individuals’ health have been reported in terms of both physical health and psychological well-being. The health of a husband or wife is strongly associated with his or her spouse’s health (Wilson 2001). Additionally, relationship functioning has been found to be associated with cardiovascular, endocrine, and immune functioning (Rankin-Esquer et al. 2000; Robles and Kiecolt-Glaser 2003). Associations have also been found between marital status and specific health outcomes, such that married individuals have better health experiences than nonmarried individuals in terms of: pain and pain-related disability, periodontal disease, rheumatoid arthritis, heart health, neurological disorders, ulcers, self-reports of overall health status, and longevity (Carels et al. 1998; Kiecolt-Glaser and Newton 2001; Levenstein et al. 1995; Marcenes and Sheiham 1996; Medalie et al. 1992; O’Farrell et al. 1998; Rankin-Esquer et al. 2000; Tucker et al. 1996; Turk et al. 1992; Vitaliano et al. 1993; Zautra et al. 1998).

There is also growing evidence for the psychological health benefits of a romantic partner. For example, married individuals have been found to have greater psychological health and report being happier and more satisfied with their lives than their unmarried peers (Bookwala and Schulz 1996; Bradburn 1969; Gove 1979; Gove et al. 1983; Tucker et al. 1996). Young adults who get and stay married report better mental health than those who remain single. Married versus single women report less depression and married versus single men report fewer alcohol problems (Horowitz et al. 1996). In fact, Bloom et al. (1978, p. 869) have gone as far as to assert that, “Of all the social variables whose relationships with the distribution of psychopathology in the population have been studied, none has been more consistently and powerfully associated with this distribution than marital status.”

The positive benefits of romantic relationships are thought to be attributable to both selection and protection effects (Fu and Goldman 1996; Kiecolt-Glaser and Newton 2001). In other words, healthier individuals are more likely to maintain relationships and individuals who are in relationships experience the protective effects they may offer. Specific reasons suggested (and to varying degrees supported by empirical research) for the positive affects of romantic relationships on health include: social support (Robles and Kiecolt-Glaser 2003; Seeman and Syme 1987), economic support (Hahn 1993; Wilson 2001), greater social integration and protection from stress among couples (Hahn 1993; Lewis 1997), an increased sense of meaning and purpose in life that may accompany a romantic relationship (Antonovsky 1979), and romantic partners’ encouragement of positive health beliefs and health behaviors (Joung et al 1995; Markey et al. 2005). In other words, an individual’s physical and psychological experiences are affected by their relationships. For example, married individuals have been found to be more likely to exercise and eat breakfast and less likely to smoke or drink heavily than unmarried individuals (Joung et al. 1995). Further, romantic partners have been found to enhance individuals’ happiness, thus resulting in improvements in health (Cohen 1988; Wickrama et al. 1997).

Of course, romantic relationships are complex and multifaceted and being in a relationship is not inevitably positive. The quality of people’s relationships may be much more consequential for their well-being than is the mere presence or absence of a significant other (Gottman and Notarius 2002). In fact, research suggests that being in an unhappy stable relationship may have a deleterious impact on health (Robles and Kiecolt-Glaser 2003). This may be particularly true if the relationship is stressful and presents a strain on psychological and physical well-being (Burman and Margolin 1992; Gottman and Notarius 2002). McCabe et al. (1996) suggest that the positive benefits of relationships only apply to relationships of at least “medium quality.” Consistent with this notion, single people have somewhat better health outcomes than those in low quality relationships (Glenn and Weaver 1981; McCabe et al. 1996), and unhappy relationships are associated with a decreased chance of individuals making positive health behavior changes, decreased well-being, increased mental disorders, greater health problems, and mortality (Hintikka et al. 1999; Robles and Kiecolt-Glaser 2003; Schafer et al. 2000; Walen and Lachman 2000). Indeed, some researchers have cautioned that finding a romantic partner or getting married is not an antidote to ill health or the key to happiness (Friedman 2000, 2003; Tucker et al. 1996).

Relationship quality is not the only factor contributing to the potential benefits or detriments reaped from a relationship. Gender has been found to moderate the affects of romantic relationships on health. For example, significant gender differences in the protective effect of marriage have been reported: nonmarried men have a 250% greater mortality rate than married men, and nonmarried women have a 50% greater mortality rate than married women (Ross et al. 1990). Further, marital dissolution appears to have a more deleterious influence on men’s health than women’s (Bloom et al. 1978; Tucker et al. 1996) and it seems that getting married may have some negative affects on women’s health (possibly due to the traditional gender role of care-taker that many women assume when they get married; Markey et al. 2005). Thus, some research suggests that men may benefit more from romantic relationships than women do, but the extent to which men and women are uniquely affected by their romantic relationships in specific health domains has yet to be fully explored. For example, it is possible that women positively impact the men they are in relationships with by encouraging them to go to the doctor when they are sick while men may positively impact the women they are in relationships with by encouraging them to participate in physical activities.

The extent to which individuals are thoughtful about the role their romantic relationships may play in their psychological and physical health has yet to be examined. Past research examining links between romantic relationships and health has tended to rely on standardized measures of relationship and health status. Although such an approach allows for a useful examination of the health domains a researcher is interested in, it does not allow for an examination of individuals’ unique health experiences in the context of their relationships. Researchers have not asked individuals to describe (using a more qualitative methodology) how their romantic partners influence their health. A qualitative approach that does not constrain the health issues under investigation (i.e., does not ask about specific health behaviors or outcomes chosen by the investigator) has the potential to reveal health domains that are associated with romantic relationship experiences that have not been previously explored. Further, past research has not investigated whether these perceptions of how individuals believe their romantic partners affect their health are actually related in meaningful ways to their psychological and physical well-being.

In addition to the empirical findings reviewed above, the literature on social control (e.g., Butterfield and Lewis 2002; Tucker et al. 2006) provides a theoretical basis for why individuals’ perceptions of their partners’ influences on their health are important. Social control research suggests that relationships may influence health via indirect and direct processes. The indirect route is via the internalization of relationships in ways that motivate individuals to maintain healthy habits. The direct route entails others providing specific instructions for healthy behaviors (Franks et al. 2006; Umberson 1992). As an extension of these findings, a qualitative approach provides a means of understanding how individuals perceive their romantic partners as attempting to influence their health and the predictive validity of these perceived influences. Understanding individuals’ perceptions of their romantic partners’ influences on their health has implications for health care professionals’ abilities to implement intervention and promotion efforts utilizing romantic partners as sources of support. Once we understand whether or not individuals acknowledge their partners as integral to their decision to participate in certain health behaviors (e.g., eat healthily), we can better determine if a health regimen that involves these partners will be effective.

Aims and Hypotheses

This study will build on past research linking romantic relationships and health by asking participants to list and describe the ways that their significant others influence their health. These qualitative data will then be coded to quantitatively describe potential gender differences in how much individuals think their partners influence their health and the type of influence (positive versus negative) individuals report that their significant other has on their health.

Based on previous research indicating that romantic relationships are usually positive contributors to individuals’ health, our first hypothesis is that both men and women will report experiencing more positive health influences from their partners than negative health influences. Second, because research suggests that men’s health is more affected by their romantic relationships, our second hypothesis is that that men will report being influenced by their partners more than women.

In order to better understand the specific health domains (e.g., eating, physical activity, etc.) in which individuals believe their romantic partners influence their health, we will next examine participants’ qualitative reports of their romantic partners’ health influences to determine the extent to which there are gender differences in the health domains perceived to be influenced by romantic partners, and the extent to which participants feel their partners are a positive or negative influence in different health domains. Analyses determining the domains that are most and least conceptualized as influenced by romantic partners are exploratory.

Next, men’s and women’s reports of their relationship quality will be examined in association with their reports of their partners’ influences on their health. Because previous research indicates that the quality of individuals’ relationships is consequential for their well-being and that low quality relationships may have a deleterious impact on health, our third hypothesis is that participants who report having more loving, understanding, and harmonious relationships will also report that their partners have a relatively more positive influence on their health than do participants who report having less positive relationship experiences.

Finally, to help validate participants’ reports of the impact their significant other has on their health and to determine the predictive utility of these reports, associations between participants’ reports of the impact their significant other has on their health and participants’ health outcomes and behaviors will be analyzed using a variety of health measures collected as part of this study. These analyses will highlight the importance of examining individuals’ perceptions of their partners’ influences on their health by demonstrating relations between these perceptions and individuals actual health behaviors and outcomes. Specifically, we expect that health status (i.e., presence or absence of symptoms) will be related to participants’ perceptions that their significant other encourages them to seek medical help (Hypothesis 4a). Participants’ body mass index (BMI) is expected to be associated with their reports of their partners’ influence on their physical activity (Hypothesis 4b) and eating behaviors (Hypothesis 4c). Participants’ reports of regular physical activity is hypothesized to be related to their perceived significant others’ influence on their physical activity behaviors (Hypothesis 4d). Participants’ smoking (Hypothesis 4e) and drinking behaviors (Hypothesis 4f) are expected to be associated with their reports that their significant others influence their substance use. Finally, participants’ reports of their general stress are hypothesized to be related to their perceptions of their partners’ influences on their stress levels (Hypothesis 4g).

Method

Participants and Procedure

Two-hundred and ten adults residing in the U.S. (105 females, mean age = 23.93 years; 105 males, mean age = 25.93 years) participated in the present study as part of a larger study examining couples’ health. In order to be eligible for this study, participants were required to be involved with their romantic partner in a monogamous dating, cohabitating, or marital relationship for at least one consecutive year. Twenty-eight percent of couples were married, 33% were living with their partners, and 39% were exclusively dating. The mean length of couples’ relationships was 3.82 years. The majority of participants were European-American; 76% of men reported that they were of European-American/ White background (11% were African American, 6% were Asian, and 7% were Hispanic), and 70% of women reported that they were of European-American/White backgrounds (10% were African American, 7% were Asian, 8% were Hispanic, and 5% were of an “other” ethnic background). The socioeconomic backgrounds of participants varied considerably; 54% of men reported incomes under $20,000 per year, 30% reported incomes between $20,000 and $49,000 per year, 14% reported incomes between $50,000 and $75,000, and 2% reported incomes over $75,000 per year. Sixty-nine percent of the women reported incomes under $20,000 per year, 24% reported incomes between $20,000 and $49,000 per year and 7% reported incomes between $50,000 and $75,000.

Participants were recruited from a northeastern university and the surrounding areas using fliers and advertisements. Participants came to a campus laboratory and completed a 1.5 h series of questionnaires (described below) independently from their partner. Participants were each compensated with a check for 25 dollars, or 2 h of introductory psychology research credit (only 14% of the participants elected to be compensated with research credit).

Measure

Romantic Partners Influences on Health

An open-ended questionnaire, created for the purposes of this study, asked participants to specify how they felt their partners influenced their health. The instructions read:

Take a few minutes to think about how your partner influences your health. Your partner may influence everything including: how you feel, behaviors that you participate in that are healthy, and behaviors you participate in that are not healthy. First, list and describe some ways that you think your partner positively influences your health. Try to think of as many things as you can. Second, list and describe some ways that you think your partner negatively influences your health. Try to list as many things as you can.

Blank lines were provided for participants to write out their positive and negative responses. Participants who did not fill all of the lines provided were asked if they had intentionally left blanks (i.e., they could not think of anything else to write) or if they had accidentally left blanks. If participants had accidentally left blanks, they were encouraged to fill in as many comments as they could. Participants provided a broad range of comments describing their perceptions of their partners’ health influences including: “She exercises which motivates me to exercise as well,” “He makes me go to the doctors,” and “She makes me eat healthier.” Additional examples of the types of comments written by participants are included in Table 1.

Table 1 Representative examples of the positive and negative comments participants provided regarding how their significant others influence their health, how the responses were coded, and the total number of responses in each code.
Full size table

All of participants’ responses were later typed exactly as they were written. Then, four researchers read through all of the responses and created lists of possible categories that would encompass all or nearly all of the participants’ responses. The nine categories that consistently emerged across the four researchers’ lists and proved to encompass nearly all of participants’ responses were: Eating, Physical Activity, Medical Help/Treatment, Sexual Activity, Substance Use, Sleep, Self-Esteem, Personality traits/Characteristics that contribute to health (i.e., personality qualities or characteristics of a person’s romantic partner—e.g., optimism, conscientiousness about health issues—that are perceived as influencing health), and Stress. Additionally, a category “other” was created for the responses that could not be coded into one of the nine content categories (e.g., “She buys soap”; see Table 1 for example items from each category). The 1,143 responses were then coded independently by three trained researchers into the ten categories. If participants’ responses contained two separate health themes, e.g., “My partner helps me eat healthy, which makes me feel good about myself,” then these responses were coded into the two relevant health categories (e.g., eating and self-esteem). The average two judge agreement for these codes was fairly high (M agreement = 89%, M Kappa = .86). When there were disagreements about the codes, raters discussed their codes until a consensus was reached. The number of responses coded into each category are presented in Table 1. As seen in this table, 90% of the responses could be coded into one of the nine health categories.

Relationship Measures

Fifteen items from the Marital Interaction Scale (Braiker and Kelley 1979) were used to assess the level of love and conflict in couples’ relationships. The 10-item love scale assesses an individual’s sense of belonging, love, and attachment to a romantic partner (e.g. “How close do you feel to your partner?”). The Cronbach’s alphas for the love scale in the present study were .85 for women and .84 for men. The 5-item conflict scale assesses behavioral conflict and communication of negative affect within a relationship (e.g. “How often do you feel angry or resentful to your partner?”). The Cronbach’s alphas for the conflict scale were .81 for women and .72 for men. Since the majority of the participants in this study were not married, the measures were amended to read “significant other,” instead of “spouse.”

Each participant’s perception of understanding of their romantic partner was assessed by utilizing the “Perspective Taking” scale of the Marital Experiences Questionnaire (Stets 1993). This 4 item scale was designed to assess an individual’s sense of understanding their romantic partner and includes items such as: “I understand my significant other” and “I understand my significant others’ feelings quite well.” The Cronbach’s alphas for the “Understanding of Significant Other” scale for women and men in this study were .81 and .67, respectively. Additionally, each participant’s belief that their romantic partner understands them was assessed by replacing the pronoun “I” with “he/she” in each of the original four items (e.g., “I understand my significant other” was changed to “He/she understands me”). The Cronbach’s alphas for this “Understanding from Significant Other” scale was .85 for women and .81 for men. Again, since the majority of the participants in this study were not married, the measures were amended to read “significant other,” instead of “spouse.”

Health Measures

Men’s and women’s weight status was assessed as a potential correlate of their reported significant others’ influences on their health (i.e., eating behaviors and physical activity behaviors influenced by partners). Weight status was operationalized using Body Mass Index scores (BMI; (weight(kg)/height2(m)). Based on the recommendations of Lohman et al. (1988) three height and weight measurements were collected for each participant by trained research assistants and individuals’ average weight and height measurements were used to calculate their BMI.

Participants’ reports of their participation in physical activities were examined as a possible correlate of the physical activity behaviors that they believed were influenced by their partners. Physical activity was assessed using 6 items created for the purposes of this study that were based on items included in measures designed by Tucker and colleagues (see Tucker et al. 1997). Participants indicated whether or not they almost never, sometimes, often, or almost all of the time participated in physical activities including exercising (e.g., jogging) and team sports. In this sample, the Cronbach’s alphas of this measure for women was .68 and for men it was .70.

The symptom checklist (Sherbourne et al. 1992) was used to assess participants’ general health status (i.e., their potential need for medical care, which was expected to be associated with their partners’ encouragement to seek medical help). This 13-item measure asks participants to rate their experience of several common health problems (e.g., backaches, upset stomach, cold, allergies) in the past month using a Likert scale (1 = never experienced the symptom to 6 = every day). In this sample, the Cronbach’s alphas of this measure for women was .70 and for men it was .71.

Men’s and women’s participation in alcohol consumption and smoking behaviors was assessed using two items: “During the past 30 days, on how many days did you smoke cigarettes?” and “During the past 30 days, on how many days did you have at least one drink of alcohol?” Possible responses ranged from “0 days” to “All 30 days.” These items are similar to substance use items used in national studies of health behaviors (e.g., National Longitudinal Study of Adolescent Health, see Section 28 of the survey; Harris et al. 2003), and were examined in this study as potential correlates of significant others’ perceived influences on substance use behaviors.

Participants’ self-reported general stress was assessed as a potential correlate of their perceptions of their significant others’ impact on their stress levels. General experiences of stress were assessed using the Perceived Stress Scale (Cohen and Williamson 1987). This is a 14-item measure that asks participants to indicate, using a Likert scale (1 = never to 5 = very often), the extent to which they have experienced stress in the past month. Sample items include, “In the last month, how often have you felt nervous and ‘stressed’?” and “In the last month, how often have you felt difficulties were piling up so high that you could not overcome them?” In this sample, the Cronbach’s alphas of this measure for women was .87 and for men it was .86.

Results

In the following analyses, the results obtained were similar regardless of the type of relationship (i.e., single, cohabitating, married) examined; thus, all participants are included together in the reported results. In total, participants provided 1,143 statements regarding the ways they perceived their significant others’ influenced their health. Analyses were first conducted to examine Hypothesis 1 (i.e., both men and women would report experiencing more positive health influences from their partners than negative health influences) and Hypothesis 2 (i.e., men would report being influenced by their partners more than women). In order to examine these hypotheses, the total number of positive health influence comments and the total number of negative health influence comments that participants provided were summed for each individual and this sum score was used as the outcome in the following analyses. Using the romantic dyad as the level of analysis, a 2 (male vs. female) × 2 (positive vs. negative) repeated measures ANOVA was conducted. Because the dyad is the level of analysis, both gender and positiveness were treated as within-dyad variables. Consistent with Hypothesis 1, across both genders, the mean number of positive health influences (M = 3.57) was significantly higher than the mean number of negative health influences (M = 1.88; F (1,104) = 145.83, p < .0001). However, contrary to Hypothesis 2, the mean number of health influences provided by women (M = 3.01) was significantly higher than the mean number of health influences provided by men (M = 2.45; F (1, 104) = 9.52, p < .01). The interaction between gender and positiveness was not significant (F (1, 104) = .83, p > .05).

In order to explore exactly how participants thought their partners influenced their health, the ten health categories used by judges to code participants’ comments were examined. To determine any significant differences that might exist between these health categories a 2 (male vs. female) × 2 (positive vs. negative) × 10 (health category) repeated measures ANOVA was conducted. Similar to the previous ANOVA, because the dyad is the level of analysis, gender, positiveness, and health category were all treated as within-dyad variables. This analysis replicated the main effects of gender (F (1,104) = 9.52, p < .01) and positiveness (F ( 1,104) = 145.83, p < .0001) found in the earlier ANOVA analysis. Additionally, the main effect of health category was significant (F (9 , 936) = 55.99, p < .001). In other words, participants believed their romantic partners influenced their health more in some health categories than other health categories. The mean number of health influences provided by participants in each category was, from most influential to least influential: eating (M = .80), physical activity (M = .48), self-esteem (M = .29), personality traits/characteristics that contribute to health (M = .26), stress (M = .22), substance use (M = .20), medical help/treatment (M = .08), sleep (M = .06), and sex (M = .03). However, the results from the ANOVA also revealed significant interactions of gender × category (F (9,936) = 2.71, p < .01), positiveness × health category (F (9,936) = 9.61, p < .01), and the three-way interaction between gender × positiveness × category (F (9,936) = 2.77, p < .01).

The interactions found in the ANOVA analysis suggest that the main effect of health category was moderated by gender and positiveness. In order to further probe these omnibus interactions, ten separate 2 (male vs. female) × 2 (positive vs. negative) repeated measure ANOVAs were conducted within each health category. Of these ten ANOVAs, five produced at least one significant effect. These significant ANOVAs suggested that women were more likely than men to think their significant others had an impact on their health in the domains of physical activity (M women = .55, M men = .41, F (1,104) = 6.33, p < .05) and self-esteem (M women = .40; M men = .17, F (1,104) = 12.34, p < .01). Across gender, participants tended to think their partners had a more positive than negative impact on their health by encouraging healthy eating habits (M positive = .87, M negative = .77, F (1,104) = 5.51, p < .05), physical activity (M positive = .80; M negative = .29, F (1,104) = 60.11, p < .01) , influencing them to seek medical help (M positive = .17, M negative = .01, F (1,104) = 22.77, p < .01), by having personality traits/ characteristics that enhanced their sense of health and well-being (M positive = .45, M negative = .17, F (1,104) = 8.44, p < .01), and by increasing their self-esteem (M = .70, M = .10, F (1,104) = 40.92, p < .01). Furthermore, for self-esteem, a significant 2-way interaction was found between gender and positiveness (F (1,104) = 11.41, p < .01) suggesting that, although men thought their romantic partners’ influence on their self-esteem was slightly more positive (M = .27) than negative (M = .07), women thought their romantic partners’ influence on their self-esteem was considerably more positive (M = .70) than negative (M = .09).

Next, analyses were used to examine Hypothesis 3 (i.e., participants who report having more loving, understanding, and harmonious relationships will also report that their partners have a relatively more positive influence on their health than do participants who report having less positive relationship experiences). To investigate this hypothesis, the number of ways a participant thought their significant other positively influenced their health was divided by the total number of ways a participant thought their partner influenced their health. The resulting percentage could range from 0% (i.e., all of the ways the participant thought their partner influenced their heath was negative) to 100% (i.e., all of the ways the participant thought their partner influenced their health was positive). As shown in Table 2, for both men and women, the percentage of positive health influence was significantly related to their sense of love, conflict, understanding of their significant other, and sense of understanding from their significant other. Specifically, consistent with Hypothesis 3, men and women who believed their partners had a relatively positive impact on their health tended to be in relationships that were more loving, understanding, and harmonious than participants who thought their partners had a relatively negative impact on their health.

Table 2 Correlations between participants’ reports of their relationship experiences and their perceived partners’ influences on their health.
Full size table

Finally, in order to examine Hypotheses 4a–4g, participants’ reports of their significant others’ influences on their health were examined in relation to their actual health behaviors and outcomes. To determine relations among individuals’ perceived health influences and actual health, six separate multiple regression analyses were conducted within each health category for each gender. In these analyses, the number of times a participant reported that their partner influenced their health in a positive and negative manner was used to predict their BMI, physical activity, medical symptoms, alcohol consumption, smoking behavior, and stress. In other words, in each analysis an outcome (e.g., BMI, physical activity, medial symptoms, etc.) was regressed on the number of positive health influences in a given health category and the number of negative health influences in a given health category. Such analyses provide a parsimonious means of determining if perceived positive and negative health influences in a particular health category are related to actual health behaviors and outcomes.

Table 3 presents the multiple R values obtained from these analyses and the constructs that were hypothesized to be related to each other (see the values in bold). As shown in this table, for men, six of the seven hypotheses were confirmed. Specifically, men’s BMIs were related to their perceived partners’ influence on their eating (Hypothesis 4c) and physical activity (Hypothesis 4b), their physical activity behaviors were related to their perceived partners’ influence on physical activity (Hypothesis 4d), the number of symptoms they reported was related to their perceived partners’ influence on their medical help-seeking (Hypothesis 4a), and their smoking (Hypothesis 4e) and drinking (Hypothesis 4f) were related to their perceived partners’ influence on their substance use. For women, four of the seven hypothesized relations were confirmed. In particular, women’s BMIs were related to their perceived partners’ influence on their eating (Hypothesis 4c), their physical activity behaviors were related to their perceived partners’ influence on physical activity (Hypothesis 4d), and their smoking (Hypothesis 4e) and drinking (Hypothesis 4f) were related to their perceived partners’ influence on substance use. What is also noteworthy is that only three of the 94 (i.e., 3%) regression analyses that were not predicted to be significant produced a significant result. Taken as a whole, and consistent with our hypotheses that participants’ perceptions of their partners health influences would be associated with their health, these analyses provide support for the validity of the coding scheme utilized and the relations between participants’ perceptions of their partners’ health influences and their actual health behaviors and outcomes.

Table 3 Multiple R values from regression analyses predicting health outcomes from positive and negative perceived partners’ health influences.
Full size table

Discussion

In this study, we aimed to extend our understanding of the associations between romantic relationships and health by examining individuals’ perceptions of the influence their significant others have on their health. Individuals’ perceptions of their partners’ influences on their health were analyzed for potential gender differences, differences in terms of the perceived negative versus positive influence of their partners, and differences in terms of the health domains influenced. Further, we examined relations between couples’ relationship quality and their perceptions of their romantic partners’ influences on their health. Finally, we investigated whether or not associations between individuals’ perceptions of their romantic partners’ influence on their health and their actual health existed.

Perceptions of Romantic Partners’ Health Influences

Consistent with our first hypothesis, couples in this study tended to perceive their partners as relatively positive influences on their health. This finding concurs with past research employing standardized assessments that has tended to find that romantic relationships are positively associated with individuals’ physical and psychological health (House et al. 1988; Robles and Kiecolt-Glaser 2003). However, it is worth noting that, prior to this study, participants have not been provided a means of indicating specific ways in which their partners have a positive or negative influence on their health. The present findings indicate that, across gender, participants tended to think their significant others had a more positive than negative impact on their health by encouraging them to maintain healthy eating habits, participate in physical activities, seek medical help, by increasing their self-esteem (especially for women), and by maintaining personality traits/ characteristics that enhanced their health and well-being. It is interesting that participants were aware of both physical and psychological health benefits provided by their partners, and that many participants made reference to personality traits/characteristics of their partners that they felt enhanced their health in some way (e.g., “He’s patient when I nag or am moody.”). Further, it is noteworthy that participants’ perceptions of their partners’ influences on their eating and physical activity behaviors were commented on frequently, given that very little research (for an exception see, Markey et al. 2001) has examined romantic partners’ influences in these specific health domains.

Contrary to our second hypothesis, women tended to think their romantic partners had more of an influence on their health than men did. In particular, women were more likely to think that their significant others had an impact on their health in the domains of physical activity and especially in terms of enhancing their self-esteem. Given that past research suggests that men’s health may be more susceptible to influences from a romantic partner (especially positive influences) than is women’s health, this finding was not expected (Ross et al. 1990). However, there are a number of possible explanations for this finding. First, most past research using standardized survey assessments has not been able to examine specific psychological and physical health domains such as those examined in this study (e.g., eating behaviors, self-esteem, etc.). Health outcomes examined in these studies have usually been more objective medical indicators of health (e.g., presence or absence of heart disease). Thus, the methodology employed in this study—asking participants to generate their own ideas about how specifically their significant other influences their health—may explain this unexpected gender difference by allowing for the investigation of more diverse health issues. It is also possible that women may spend more time than men thinking about their health and may be more aware of the significance of their relationships to their health. Some research suggests that men are less likely than women to perceive themselves as being at risk for health problems, are less likely to seek support for health problems than are women, and traditionally have been conceptualized as “stronger,” “healthier,” and less in need of medical care than women (see Courtenay 2000). Further, women are more likely than men to experience socialization throughout their lives that encourages “care-taking” and nurturing behaviors, potentially leading them to be more concerned with health issues than men and with the role of health issues in their relationships (Courtenay 2000; Eagly 1987).

Relationship Quality and Perceived Partners’ Health Influences

Consistent with our third hypothesis, participants’ perceptions of their partners’ influences on their health was positively associated with reports of their relationship quality. For both men and women, the percentage of positive health influences they reported was related to their sense of love, conflict, understanding of their significant other, and sense of understanding from their significant other. Specifically, men and women who believed their romantic partners had a relatively positive impact on their health tended to be in relationships that were more loving, understanding, and harmonious than participants who thought their romantic partner had a relatively negative impact on their health. This finding extends past research that has focused almost exclusively on long-term martial relationships. Couples in this study were relatively young and had maintained relationships for an average of 3 years. Thus, it appears that romantic relationship correlates of health may not be confined to married partners who have maintained extensive relationships.

Of course, this study does not provide a means of determining whether participants’ positive relationship experiences influenced their health or whether their positive health experiences influenced their relationships. Consistent with past research examining married couples (e.g., Kiecolt-Glaser and Newton 2001), it is likely that a bidirectional relation exists between couples’ romantic relationship experiences and health experiences. Participants in this study who were healthier may have been more apt to maintain high quality relationships and individuals in high quality relationships are likely to experience the protective effects they offer. Future research could use the methodology used in this study to longitudinally examine the relations between romantic relationships and health and more conclusively determine the direction of effects among these constructs.

Relations Among Perceived Partners’ Health Influences and Health Outcomes and Behaviors

Although women in this study reported that their romantic partners had a more significant influence on their health than men did, consistent with our fourth hypothesis, both men’s and women’s perceptions of their significant others’ influences on their health tended to be associated with their actual health outcomes. Specifically, men’s reports of their significant others’ influences on their health were associated with their BMIs, physical activity, medical help-seeking, and their drinking and smoking behaviors. These findings are consistent with past research suggesting that women may be positive influences on their significant others’ health by caring for them in domains traditionally conceptualized as feminine—making healthy meals, scheduling doctor appointments etc. (Courtenay 2000). Thus, future research may be able to extend these findings using longitudinal data to confirm that men not only perceive their partners as important health influences, but that this awareness may translate into actual changes in their health.

Women’s reports of their significant others’ influences on their health were associated with their BMIs, physical activity, and their drinking and smoking behaviors. This finding suggests that men may have a more integral role in women’s health behaviors than has been documented in previous research (Ross et al. 1990). These findings may suggest that as gender roles are changing, and young couples are more likely to maintain egalitarian and less “traditional” relationships than in the past, men are increasingly involved in the health of the women they are in relationships with.

Limitations

The current study attempted to link various perceived health influences (e.g., eating, physical activity, etc.) to six different health outcomes. Because all of the data examined in this study were collected at the same time point (i.e., before the health categories were created) it was not possible to predict the wide array of perceived health influences that participants reported. Future studies could extend this research by examining romantic partners’ health influences in domains including sleeping patterns, sexual behaviors, self-esteem, and their partners’ personality traits.

This study is also somewhat limited by its use of primarily self-report measures of health behaviors and outcomes. These self-report measures could be corroborated with reliable measures of objective health outcomes and physician-reports of participants’ health in future research. Further, longitudinal assessments of participants’ perceptions of their partners’ influences on their health and participants’ health outcomes could lend insight into the causal direction of effects among these constructs.

Conclusions and Implications

This research approaches understanding associations between romantic relationships and health from a perspective unique from past research. Instead of looking for links between relationship status and health outcomes using standardized questionnaires, we asked participants to generate their own ideas regarding how their romantic partners influenced their health. Findings from this study suggest that romantic partners are not only aware of the impact that their significant other has on their health, but that individuals recognize both the psychological and physical health benefits and detriments provided by their significant others. Further, individuals’ perceptions of the role that their significant others have on their health is associated with both their relationship quality and their health behaviors and outcomes.

At this point in the scientific examination of influences on health, it seems undeniable that the social context of health is important. For the majority of adults, their social experiences revolve around the lives they share with their significant other. Thus, understanding how individuals’ view their romantic partners’ role in their health, and the predictive validity of these perceived influences, may have implications for health care professionals’ abilities to implement intervention and promotion efforts utilizing romantic partners as sources of support. Interestingly, in the current health care climate, adult family members are routinely involved in attempts to improve children’s health (see Minuchin et al. 1979; Tinsley et al. 2002), yet significant others are rarely considered as agents of health behavior change for each other. The present study suggests that including significant others in intervention and prevention efforts may have potential health benefits—from increasing individuals’ self-esteem to improving their eating behaviors.