آسیب پذیری سلامت جسمی در کودکان بزرگسالان از خانواده های طلاق گرفته و دست نخورده

Abstract Objectives: The current study evaluated family process variables associated with markers of physical health vulnerability. Methods: Retrospective reports of parental caring, conflict, and divorce-specific factors were examined in reference to hostility, somatic symptoms, and illness reports in young adults from divorced (n=253) and intact (n=552) families. Results: Contrary to expectations, participants from divorced and intact families were equivalent on all health-related measures. Within the intact group, parental conflict and low parental caring were associated with hostility, somatic symptoms, and illness reports. Within the divorce group, negative feelings about the divorce were associated with higher hostility, somatic complaints, and illness reports. Conclusions: Results suggest that parental divorce in itself does not increase long-term vulnerability to physical illness; rather it is the negativity of the experience that is associated with vulnerability. Although overall health markers did not differ, the family process variables associated with physical health risk differed for individuals from divorced versus intact families.

Introduction Evidence is mounting that early care-taking experiences can impact long-term physical health [1], physiological stress reactivity, and vulnerability to stress-related illness [2], [3] and [4]. For example, early separation from the primary caregiver has been shown in both animal and human studies to be associated with the development of maladaptive stress responses [4], [5] and [6] and long-term physical health problems [7] and [8]. The quality of parent–child relationships has been associated with long-term neuroendocrine and cardiovascular function [9] and [10]. Perceptions of parental caring have also been associated with long-term psychiatric vulnerability [11], [12] and [13], somatic symptoms [14] and [15], and physical health status in middle-age adults [15]. Exposure to conflict within the family represents another pathway by which family factors may influence long-term physical health vulnerability. High levels of family conflict have been associated in children and adolescents with poorer physical health [16] and [17], exaggerated physiological reactivity [18], slower growth [19], and greater health care utilization [20]. Children of divorce may experience these factors in combination and/or at higher intensity. Accordingly, several lines of evidence suggest that children of divorce may be at higher physical health risk than those from intact families. First, children generally experience a decrease in contact with both parents following the divorce, and many experience minimal or no contact with the noncustodial parent [21]. Second, parental divorce has been shown to be associated in the long-term with lower quality of parent–child relationships [22]. Third, marital conflict tends to both precede and follow parental divorce [21]. The literature on the long-term physical health correlates of parental divorce is sparse, and tends to be inconclusive. There is some evidence to suggest that adult children from divorced families may be at higher risk of physical health problems [23]. Parental divorce has been related to decreased longevity in a long-term follow-up of the Terman Life Cycle Study [24]. However, the Terman study participants were raised in the 1920s, and current social implications of divorce are quite different from those early in the century. Goede and Spruijt [25] reported poorer health in 18–24-year-old females from divorced families relative to intact families, but no differences were found for males. Higher levels of somatic symptoms have been reported in adolescents from divorced relative to intact families [26] and [27]. Maier and Lachman [7] reported significantly higher levels of acute and chronic health problems in middle-aged adults as a function of early parental divorce, an effect that was mediated by current income, education, drug use, and family support. Krause [8] reported that declines in physical health in older adults were related to the combination of early parental separation (by death or divorce) and high levels of current stress. Given the high current divorce rate, potential long-term health correlates of parental divorce remain an important question to be explored. In addition, although specific family process variables have been associated in the broader literature with health vulnerability, it is unclear whether relationships between family process factors and physical health differ between intact and divorced families. The present study examines associations between self-reported parental caring and conflict, physical health, and two markers of long-term physical health vulnerability, somatic symptoms and trait hostility, in young adults from intact and divorced families. The current experience of somatic symptoms (e.g., headaches, nausea, and abdominal pain) can be an indication of physical health vulnerability. Somatic symptoms are often thought to be a response to psychosocial stress, and individuals reporting high levels of somatic symptoms show frequent use of health services [28], higher blood pressure [29], increased disability, and maladaptive illness behavior [30]. A large literature links hostility with heightened risk for cardiovascular and other diseases, and poorer prognosis following cardiac incidents [31] and [32]. Hostility levels assessed in young adults may provide important predictive information concerning long-term risk of heart disease. In a follow-up of college students assessed in the 1960s, higher hostility was predictive of greater coronary risk factors 21–23 years later [33]. Several researchers have reported evidence that trait hostility in children and young adults is associated with exposure to parental conflict [34], [35] and [36] and negative family environment [37]. Increased levels of parental conflict may contribute to the development of higher levels of hostility and increased long-term vulnerability to cardiovascular and other illness. The current study addresses the question of whether parental divorce increases long-term physical health vulnerability, and evaluates the specific family process variables associated with vulnerability. To accomplish this, we evaluated self-reports of physical illness, somatic symptoms, and hostility levels in young adults from divorced and intact families. There were two main objectives of the analyses. First, we were interested in evaluating overall group differences in health vulnerability. We hypothesized that participants from divorced families would report poorer health, as evidenced by higher hostility, somatic symptoms, and illness reports. Second, within-group relationships between health variables and family process variables were evaluated, with the expectation that similar associations between the quality of family relationships and health vulnerability would be found within each group. Specifically, we expected that within each group, low parental caring and high parental conflict would be associated with poor health factors.

Results Health vulnerability First, overall group differences between participants from divorced and intact families were assessed. Group means on health and family variables are displayed in Table 1. Contrary to our predictions, participants from divorced families were not significantly different from those from intact families on hostility, somatic symptoms, or illness reports. However, participants from divorced families reported significantly lower parental caring and time spent with either parent, and higher parental conflict than those from intact families. Table 1. Group means Range Divorced group Intact group n Range Overall mean (S.D.) Males mean (S.D.) Females mean (S.D.) n Overall mean (S.D.) Males mean (S.D.) Females mean (S.D.) Hostility 9–65 128 37.4 (8.1) 39.8 (8.1) 35.3 (7.6) 260 37.4 (9.3) 39.9 (8.9) 35.3 (9.1) Somatic symptoms 0–33 128 7.1 (5.9) 5.8 (5.3) 8.3 (6.2) 262 7.1 (5.8) 6.7 (6.1) 7.4 (5.7) Days sick 0–7 128 2.3 (1.9) 1.4 (1.6) 3.1 (1.9) 263 2.3 (1.7) 1.8 (1.7) 2.6 (1.7) Healthcare visits 0–6 128 1.3 (1.6) 1.0 (1.2) 1.7 (1.7) 262 1.2 (1.6) 1.0 (1.6) 1.3 (1.6) Parental conflict* 0–8 127 4.5 (2.3) 4.5 (2.6) 4.5 (2.4) 262 2.3 (1.9) 2.4 (1.8) 2.3 (2.0) Father caring* 0–40 248 23.0 (9.8) 23.3 (9.1) 22.7 (10.4) 551 26.4 (8.4) 26.5 (8.0) 26.3 (8.7) Mother caring* 1–39 253 27.7 (8.4) 28.9 (7.1) 26.9 (9.2) 551 30.9 (5.7) 31.1 (5.3) 30.8 (6.0) Time with father* 0–8 128 3.6 (2.2) 3.5 (2.0) 3.7 (2.4) 262 4.9 (1.9) 4.5 (1.6) 5.2 (2.0) Time with mother* 0–8 128 5.4 (1.9) 5.2 (1.9) 5.6 (1.9) 262 5.9 (1.8) 5.4 (1.6) 6.3 (1.7) * P<.01 for difference in overall means between divorced and intact group. Table options Family process correlates and predictors of health markers Divorce Within the divorce group, feelings about the divorce were significantly correlated with hostility, somatic symptoms, days sick, and healthcare visits, such that those who reported the divorce as a more negative experience were more negative on health measures (see Table 2). In addition, perceived father caring was positively correlated with days sick. The overall MANOVA was significant for feelings about the divorce [F(4,116)=3.7, P<.01], age [F(4,116)=3.5, P=.01], and gender [F(4,116)=12.9, P<.001]. Interactions of feelings about the divorce with age or gender were not significant. Stepwise regression analyses were conducted for each health variable with age, gender, feelings about the divorce, and father caring as potential predictor variables. Results are shown in Table 3. Feelings about the divorce and gender were the only significant variables selected in the final models for somatic symptoms and healthcare visits. The final model for hostility selected gender, age, and feelings about the divorce. Gender, age, and father's caring were selected in the final model for number of days sick in the last semester. Thus, feelings about the divorce emerged as a significant predictor of somatic symptoms, healthcare visits, and hostility. Number of days sick was best predicted by father's caring. Table 2. Correlations between variables within divorce (Div) and intact (Int) groups Parental conflict Father caring Mother caring Time with father Time with mother Living arrangement Feelings about divorce Div Int Div Int Div Int Div Int Div Int Div Int Div Int Somatic symptoms .08 .25*** .03 −.28*** −.11 −.24*** .12 −.10 −.01 −.11 .05 n/a −.20* n/a Hostility .14 .31*** −.12 −.30*** −.06 −.17** −.03 −.24*** .02 −.17** −.06 n/a −.21* n/a Days sick −.03 .05 .23** .04 −.05 −.03 .13 .08 .14 .10 −.06 n/a −.21* n/a Healthcare visits −.03 .16** .07 −.01 −.12 −.07 .00 −.01 −.01 −.08 .04 n/a −.19* n/a Parental conflict −.17 −.40*** −.25** −.24*** −.01 −.30*** −.11 −.17** .00 n/a .01 n/a Father caring .12 .46*** .57*** .62*** .09 .22*** .25** n/a −.17 n/a Mother caring −.03 .33*** .53*** .51*** −.30*** n/a .03 n/a Time with father .08 .45*** .46*** n/a −.10 n/a Time with mother −.28*** n/a −.02 n/a Living arrangement .10 n/a n/a=not applicable/not measured. *** P<.001. * P<.05. ** P<.01. Table options Table 3. Stepwise multiple regression results for divorce groupa Step Variables entered R2 β t Somatic symptoms Model 1 Feelings about divorce .04 −.21 −2.4* Model 2 Feelings about divorce .06 −.20 −2.4* Gender .18 2.1* Hostility Model 1 Gender .06 −.27 −3.1** Model 2 Gender .14 −.30 −3.5** Age −.28 −3.4** Model 3 Gender .18 −.31 −3.8** Age −.28 −3.3** Feelings about divorce −.23 −2.8** Healthcare visits Model 1 Gender .07 .28 4.0** Model 2 Gender .10 .27 3.1** Feelings about divorce −.18 −2.1* Days sick Model 1 Gender .20 .45 5.6** Model 2 Gender .23 .43 5.4** Age −.19 −2.3* Model 3 Gender .26 .42 5.3** Age −.19 −2.4* Father caring .19 2.4* a Variables available for selection in models included age, gender, father’s caring, and feelings about the divorce. Adjusted R2 and standardized β reported. * P<.05. ** P<.01. Table options Because ratings of feelings about the divorce emerged as a strong correlate of health variables for participants from the divorce group, post-hoc analyses were conducted in an attempt to better understand contributors to participants' ratings of the divorce experience. Although a linear relationship did not reach statistical significance, a significant quadratic regression was found for parental conflict [F(2,124)=4.6, P=.01]. Participants reporting either high or low levels of parental conflict were more positive about the divorce experience, while those reporting moderate levels of conflict were more negative. Furthermore, a significant cubic regression was found for father's caring and feelings about the divorce [F(3,122)=2.6, P=.05], in which the high end of father caring was associated with more negative feelings about the divorce and the low end was associated with more positive feelings, while intermediate levels of father caring were not associated with feelings about the divorce. Intact Within the intact group, father caring and mother caring were correlated negatively with hostility and somatic symptoms, time spent with either parent was negatively correlated with hostility, and parental conflict was positively correlated with hostility, somatic symptoms and health care visits (see Table 2). The overall MANOVA was significant for father's caring [F(4,237)=3.3, P=.01], time spent with mother [F(4,237)=2.6, P=.04], parental conflict [F(4,237)=3.8, P<.01], age [F(4,237)=3.0, P=.02], and gender [F(4,237)=4.9, P<.01]. Interactions of family process variables with age or gender were not significant. Results of individual stepwise regressions with health variables are shown in Table 4. The final model predicting somatic symptoms selected father caring, mother caring, and parental conflict. The final model predicting hostility selected parental conflict, gender, and father's caring. Stepwise regression predicting healthcare visits selected age and parental conflict. The final model predicting days sick selected only gender and age. Thus, parental conflict emerged as a significant predictor of somatic symptoms, hostility, and healthcare visits. Parental caring variables were significant predictors of somatic symptoms and hostility. Table 4. Stepwise multiple regression results for intact groupa Variables entered R2 β t Somatic symptoms Model 1 Father caring .07 −.27 −4.4** Model 2 Father caring .10 −.19 −2.8** Mother caring −.18 −2.6** Model 3 Father caring .12 −.14 −1.9 Mother caring −.16 −2.4* Parental conflict .14 2.2* Hostility Model 1 Parental conflict .08 .28 4.7** Model 2 Parental conflict .12 .28 4.7** Gender −.22 −3.7** Model 3 Parental conflict .15 .20 3.1** Gender −.22 3.7** Father caring −.19 −3.0** Healthcare visits Model 1 Age .03 −.17 −2.8** Model 2 Age .05 −.17 −2.7** Parental conflict .16 2.6** Days sick Model 1 Gender .05 .23 3.8** Model 2 Gender .07 .21 3.3** Age −.14 2.3* a Variables available for selection in models included age, gender, mother's caring, father's caring, time with mother, time with father, and parental conflict. Adjusted R2 and standardized β reported. ** P<.01. * P<.05. Table options Family process correlates Also of interest were correlations between family process variables. For both divorce and intact groups, significant negative correlations were found between parental conflict and parental caring, and less time with parents was associated with less caring (see Table 2). Within the intact group, higher parental conflict was also associated with less time with either parent. Within the divorce group, living arrangements following the divorce (higher scores indicating progressively more time living with father) were positively correlated with father's caring and negatively correlated with mother's caring. However, ratings of living arrangements were heavily skewed towards living with the mother (79% lived more with the mother, 11% lived equally with both parents, and 10% lived more with the father). Post-hoc correlational analyses were conducted with the 90% of participants who lived equal or more time with their mothers. Results showed that a positive correlation with father's caring remained (r=.41, P<.001), but the correlation with mother's caring became nonsignificant (r=.10, P=.28), suggesting that perceptions of father's caring increased the more the child lived with the father (up to equal time) without a corresponding decrease in perceptions of mother's caring (see Fig. 1). Young adults' perceptions of parental caring as a function of living ... Fig. 1. Young adults' perceptions of parental caring as a function of living arrangements postdivorce.