Impact of family communication on self-rated health of couples who visited primary care physicians: A cross-sectional analysis of Family Cohort Study in Primary Care

Introduction Self-rated health (SRH) is a subjective health measurement that predicts mortality and morbidity and reflects mental health and socioeconomic status. Since a couple’s relationship can influence the health status of the individuals involved, poor family communication can negatively influence the health status of its members. The aim of this study was to investigate the factors affecting SRH among married couples in primary care and evaluated the effect of family communication on SRH. Material and methods In this cross-sectional analysis of Family Cohort Study in Primary Care, 469 couples (938 participants) were analyzed to evaluate the relationship between SRH and family communication. Participants answered questionnaires on demographic characteristics and lifestyle factors. The Korean version of the Family Communication Scale of Family Adaptation and Cohesion Evaluation Scale-IV was used to assess family communication, and a 5-point scale of SRH questions was used to assess the SRH status. Multivariate logistic regression analyses were performed in order to evaluate the relationship between family communication and SRH and identify associated factors for good SRH. Results Wives with a high family communication level had higher OR for good SRH. When the husband and wife both reported high family communication levels, the OR for good SRH increased in wives; however, the relationship between family communication and SRH was not significant in husbands. In the multi-adjusted model, the OR for good SRH of husbands increased in those with >12 years of education, moderate drinkers and decreased in current smokers. The OR for good SRH of wives increased in those with age of 60 to 69, those with >12 years of education, and those who participated in vigorous physical activity, and decreased in those with diabetes and depressive mood. Conclusions Our results indicate that improvement in family communication may contribute to better SRH.

Introduction and signed the informed consent, physicians collected data using standardized questionnaires. Patients who had difficulty understanding and writing responses to the questionnaire were excluded. In the end, 520 couples (1,040 participants) were enrolled. The study protocol and the written informed consent form were approved by the Institutional Review Board of Asan Medical Center (IRB number 2009-0347). All participants read and signed the informed consent form before participating in the study.
Among the 1,040 study participants, we excluded 5 participants, who did not provide answers to the SRH questionnaires, 46 participants who did not provide answers to the Korean version of the FCS of the FACE-IV (Family Adaptation and Cohesion Evaluation Scale), and 8 participants who neglected to provide answers for both the SRH and FCS questionnaires. Additionally, we excluded 43 participants whose spouses were excluded for not completing the questionnaires. In total, 51 couples (102 participants) were excluded, and 469 couples (938 participants) were analyzed to investigate the relationship between SRH and family communication.

Data collection and measurement
All of the physicians collected data using standardized questionnaires that included questions regarding demographic characteristics, such as age, sex, educational level, and income, as well as lifestyle factors, such as smoking habits, alcohol consumption, and physical activity. The participants answered the questionnaires individually, which were then reviewed by trained interviewers or their physicians. Educational level was categorized into three groups: <12 years, 12 years and >12 years, and monthly household income was divided into four groups: <2 million won (1,761 USD), 2-4 million won (1,761-3,521 USD), 4-6 million won (3,282 USD), and �6 million won (5,282 USD). Smoking status was determined as nonsmoker, ex-smoker, and current smoker, and alcohol consumption was categorized as non-drinker, moderate drinker, and heavy drinker according to the National Institute on Alcohol Abuse and Alcoholism (NIAAA) definitions. Heavy drinkers were defined as men who drank more than 14 glasses per week and women who drank more than 7 glasses per week [16]. Physical activity was measured using the International Physical Activity Questionnaire Short-Form (IPAQ-SF), which is a scale based on activity intensity and exercise time during the previous 7 days [17]. In addition, information about previous medical history and concomitant medications was collected. Participants also answered Center for Epidemiologic Studies Depression Scale (CES-D) questionnaires, and a depressive mood was defined when participants were diagnosed with depression or had a CES-D score � 21 [18].

Measurement of family functioning and communication
Each individual answered questionnaires for family functioning and family communication, which were then evaluated independently regardless of one's spouse's answer. Family functioning was evaluated using a Korean version of the FACES-III. Developed by Olson and colleagues to measure the quality of family relations, FACES-III consists of 10 questions assessing family adaptability and 10 questions assessing family cohesion [19]. Family functioning is defined by the combination of the categories of family adaptability and cohesion and is classified into balanced, mid-range, and extreme family types.
Family communication was evaluated using the Korean version of the FCS of the FACE-S-IV [20]. The FCS consists of 10 questions scored on a scale ranging from 1 to 5. According to the sum of the scores, family communication is classified into low , moderate (36)(37)(38)(39), and high (40-50) groups. Higher scores represent better family communication (S1 Appendix).

Measurement of SRH
We asked the following question to assess the SRH status of the participants: "In general, how would you rate your health?" Patients responded using a 5-point scale of excellent, very good, good, fair, and poor [21]. Excellent, very good, and good SRH status were categorized as good SRH status, while fair and poor SRH status were grouped as poor SRH status.

Statistical analyses
The Pearson's chi-square test and Fisher's exact test were used to compare the characteristics of the patients with good SRH status and poor SRH status. Logistic regression models were used to estimate the odds ratios (ORs) and 95% confidence intervals (CIs) of good SRH status in association with each level of family communication. Furthermore, ORs and 95% CIs of the variables associated with good SRH were obtained. We adjusted for age and all covariates that were found to be significant from the Pearson's chi-square test, and analyses were separately performed for husbands and wives. For sensitivity analyses, the analyses were also performed in combination of both sexes. Lastly, we obtained ORs and 95% CIs of good SRH status according to spousal combinations of family communication levels. Further sensitivity analyses were performed after stratifying according to the educational level. We also investigated the relationship between family communication and several health conditions by logistic regression analyses. In this study, p < 0.05 was considered statistically significant, and all data were analyzed using IBM SPSS Statistics for Windows, Version 23.0 (IBM Corp., Armonk, NY, USA). Table 1 lists the basic characteristics of the study participants. All participants included in the analysis were married couples. The mean age (standard deviation) was 57.35(10.12) years, and most participants (68.1%) were between the ages of 50 and 69. Among the participants, 48.7% had >12 years of education, indicating that approximately half of them finished high school. The monthly household income was �6 million won (5,282 USD) for 29.3% of the participants. Of the 12.3% of participants who were current smokers, the majority were men (21.9% of husbands and 1.5% of wives). Similarly, most of the heavy drinkers (16.7% of participants) were men (27.9% of husbands and 5.4% of wives). Vigorous physical activity was observed in 36.1% of the participants. Diabetes was present in 20.8%, while hypertension and dyslipidemia were present in 41.5% and 36.7% of the participants, respectively. Depressive mood was observed in 15.2% of the participants.

Basic characteristics of the study participants
A balanced family type was observed in 26.2% of the participants. Family communication levels were low in 22.0%, moderate in 27.6%, and high in 50.4% of the participants. When assessing one's health status, 49.2% reported his or her health status to be excellent, very good, or good, while 50.8% reported fair or poor health status. Table 2 shows the distribution of SRH status of husbands and wives according to demographic, lifestyle factors and family dynamics. The frequency of good SRH status of husbands was higher in those with higher educational levels (68.1% for >12 years of education) and incomes (33.0% for �6 million won [5,282 USD]) nonsmokers (21.4%) and ex-smokers (60.5%) (p < 0.05). Although family functioning was not significantly different between the two groups (p = 0.660), husbands with higher family communication levels were more likely to be in a  Table).

Association between family communication and SRH
The OR for good SRH increased in wives with high family communication levels (OR 1.91, 95% CI 1.07-3.41); however, the relationship was not significant in husbands (Table 3). This finding was consistent in participants with > 12 years of education (S2 Table). However, there was no significant relationship between family communication and SRH when participants had � 12 years of education. When both sexes were analyzed together, the OR for good SRH increased in those with high family communication levels (OR 2.07, 95% CI 1.33-3.23) (S1 Table). Table 4 presents the ORs for good SRH status according to spousal combinations of family communication levels. The OR for good SRH of wives significantly increased when the husband and wife both reported high family communication levels (OR 2.31, 95% CI 1.32-4.05); however, the relationship was not significant in husbands. This finding was consistent when participants were stratified according to educational level (S3 Table).
S4 Table presents the ORs for health conditions according to family communication levels. The OR for depressive mood increased as family communication aggravated in both husbands and wives. The associations between family communication and hypertension, diabetes mellitus, and dyslipidemia were not significant. Family communication and self-rated health

Discussion
In this study, we identified the associated factors of SRH among married couples and examined the relationship between family communication on SRH among husbands and wives for the first time. In the analyses of husbands, we found that men with high educational level and men who drank moderately reported good SRH, and current smokers were more likely to have poor SRH. In the analyses of wives, women in their sixties, women with high educational level, and women who participate in vigorous physical activity reported good SRH, while women with diabetes and depressive mood were more likely to be in a poor SRH status. Furthermore, wives with a high level of communication had good SRH, although family communication did not influence husbands' SRH. Additionally, wives were more likely to report good SRH when the husband and the wife both reported high communication levels. SRH is a strong predictor of mortality in both acute and chronic contexts [1,3]. It also influences the incidence of chronic diseases such as coronary heart disease, diabetes, stroke, lung disease, and arthritis [2]. In our crude analyses, husbands with diabetes and depressive mood had poor SRH, while wives with diabetes, hypertension, and depressive mood reported poor SRH. However, after adjusting for potential risk factors, only the association with diabetes mellitus and depressive mood remained significant for wives. Unlike patients with hypertension or dyslipidemia, patients with diabetes have to undertake several self-care activities in order to avoid microvascular and macrovascular complications of diabetes. They need to monitor blood glucose levels, take medications regularly, inject insulins, eat healthy food, and participate in regular physical activity. These activities may cause social stigma to diabetic patients and make them to assess their health status as poor [22]. Wives with depressive mood according to depression diagnosis or CES-D score �21 also demonstrated poor SRH. Since CES-D is a self-reporting questionnaire, those who reported themselves as depressed would also be likely to assess their health status as poor [18]. The relationship between comorbidities and SRH was significant only in wives, which implies that women are psychologically more vulnerable to medical conditions and recognize their medical conditions negatively. It has been reported that male diabetics experience lesser depression and anxiety but more energy and better positive wellbeing [23].
When the relationship between family communication and health conditions was analyzed, depressive mood was the only comorbidity associated with family communication. Emotional Family communication and self-rated health regulation is the essential aspect of the etiology of depression, and family communication is critical in facilitating emotional understanding [24]. Therefore, poor family communication may contribute to depressive symptoms. On the other hand, depressive symptoms can also aggravate family dynamics by influencing the mood of the family members [25]. Although the cross-sectional nature of our analysis does not demonstrate a definite direction between family communication and depressive mood, it certainly illustrates the role of psychological health on family communication and vice versa. Among the lifestyle factors, moderate drinking and vigorous physical activity were associated with good SRH, and current smoking was associated with poor SRH. Men who currently smoke were more likely to be in a poor SRH status, while men who drink moderately reported good SRH. Women, who participate in vigorous physical activity, had good SRH. Harmful effects of cigarette smoking are well recognized nowadays due to various anti-smoking campaigns and actions targeted for general population; thus, current smokers were likely to report their health status as poor. The analysis for women was inconclusive because the number of ex-smokers and current smokers were very few among the wives.
Since the study design is cross-sectional, participants with good health status or who were confident in their own health were likely to be moderate drinkers. This phenomenon was also observed in a previous study that reported an association between good SRH and occasional and moderate drinking [26]. Unlike the study of Valencia-Martin, which included both genders, the relationship between moderate drinking and good SRH was significant only for men in our analysis. This may imply that women, who are confident in their own health, do not engage in drinking behavior as much as healthy men do.
Physical activity is beneficial for both physical fitness and mental health [27]. Previous studies have shown that an increase in moderate-to-vigorous physical activity was associated with improvement in health-related quality of life, and increased physical activity also improved quality of life in patients with comorbidities [28,29]. Likewise, in our study, vigorous physical activity was associated with good SRH after adjusting for comorbidities such as diabetes and hypertension. The relationship between vigorous physical activity and good SRH was observed only for women in our analysis. In a previous study by Morimoto et al, the amount of physical activity had positive effects on quality of life for both men and women; however, women had more preferable effects of maximum intensity on quality of life than men [30]. Further investigation is needed to interpret the gender differences of physical activity on SRH.
While high educational level was significantly associated with good SRH, we did not find a significant association between income and SRH. Although socioeconomic status can be expressed in terms of income or occupation, we might also suggest that one's level of education reflects his or her fundamental social position because childhood education depends mostly on one's parents' socioeconomic status, and education determines an individual's occupation and income [31]. Educational attainment was previously reported as a predictor of life expectancy, and people with lower educational level had poor SRH and shorter life expectancy in comparison to the more educated [31,32].
In our analyses, wives in their sixties had approximately twice higher odds of good SRH in comparison to the wives whose age is less than 50. In general, the proportions of good SRH decrease according to age since both physical and mental health conditions decline with aging [33]. One possible explanation of increased good SRH in sixties in our study would be that Korean women in sixties often become free from domestic work related to children's education and from housework because this period is when their husbands retire. It has been reported that housework may increase stress and decrease perceived well-being of individuals [34,35]. Age-specific distribution of SRH of a larger Korean population would be an interest of future research.
Wives with a high level of family communication demonstrated good SRH, which implies that communication between couples is an important determinant of health, especially for women. When SRH status was analyzed according to spousal combinations of family communication levels, good SRH significantly increased in wives when the husband and the wife both reported high family communication levels. However, the relationship between SRH and family communication was not significant among the husbands. Family communication has a significant association with a variety of cognitive activities and relational behaviors, as well as individual well-being [36]. Therefore, high family communication facilitates interaction between family members and can positively influence the mental status of family members. Conversely, low family communication can have negative impact on psychological status of family members, which make individuals to report poor SRH. In addition, family communication can be deteriorated when spouses are not healthy. For instance, poor family communication was reported in spouses of patients with Parkinsonism [37]. In our analyses, the relationship between SRH and family communication was observed only in wives. This suggests that psychological aspects of women are more likely to be influenced by environmental factors in comparison to those of men as reported in a previous study [38]. In the subgroup analyses stratified by educational level, the association between family communication and SRH remained significant only in wives with > 12 years of education; whereas, the association between SRH and spousal combinations of family communication levels was significant in wives regardless of their educational level. This finding suggests the possible influence of education on the association between family communication and SRH.
In our study, we evaluated SRH as the only measure of general health. Previous studies have identified SRH as a valid measure of general health although the mechanism is not well understood [7][8][9]. It predicts mortality, morbidity, and even mental health [1][2][3][4][5]. There are social and biological pathways that mediate certain health information from the human body to consciousness, which contribute self-rating of health [7]. We evaluated husbands and wives separately because questionnaires were completed separately although they were couples.
Our study has a few limitations. Since the study design is cross-sectional, we cannot determine whether a high communication level causes good SRH or vice versa. In addition, the study data were obtained with self-reporting questionnaires; thus, the responses could be exaggerated or downscaled. Conversely, this is meaningful in a way that SRH evaluates subjective health status and reflects the quality of life of each participant. We only used SRH as the measure of health; therefore, unmeasured heterogeneity may exist in our analyses. Furthermore, there is a selection bias regarding the following three points. First, the study participants were limited to patients receiving primary care; therefore, our results might not be applicable to the general healthy population. Second, participants who agreed to enroll in the cohort may have different characteristics from those who did not agree to enroll in the cohort. Third, the number of patients varies among the 28 family physicians. However, to the best of our knowledge, this is the first study to investigate the association between family communication and SRH in married couples who sought primary care. We only evaluated family communication between spouses and our analyses did not include family communication between other family members although FCS was initially developed for assessing family communication between parents and children [12]. However, FCS had been used to assess family communication between spouses in previous studies [13]. In this study, we analyzed the relationship between SRH and family communication in several ways, including on an individual basis as well as in terms of spousal combinations. Furthermore, we evaluated gender-specific factors that were associated with good SRH of husbands and wives.