ابراز هیجان و پردازش عمق تروما و ارتباط آنها با بقای طولانی مدت در بیماران مبتلا به ایدز

Abstract Objective: This study examined the relationship between emotional expression and depth processing of trauma and long-term survival of patients living with AIDS. A further purpose was to examine the immune, health behavior and psychosocial correlates of emotional disclosure and depth processing. Method: Subjects wrote essays describing their reactions to past traumas; these were scored for emotional expression and depth processing (positive cognitive appraisal change, experiential involvement, self-esteem enhancement and adaptive coping strategies). Two HIV-seropositive groups were recruited for this study; long-term survivors (LTS; n=46) patients who had survived at least 4 years past a Category C (AIDS defining) symptom prior to starting protease inhibitors and an equivalent HIV-seropositive comparison group (ECOMPLTS; n=89) who had CD4+ cells between 150 and 500, and had no history of Category C symptoms. The groups were equivalent on age, gender, ethnicity, education, employment, income, sexual orientation and route of infection. Results: The group LTS were significantly higher than the ECOMPLTS group on emotional expression and depth processing. Depth processing mediated the relationship between emotional expression and long-term survival status. Depth processing was positively related to CD4+ cell number for women. Emotional expression was also significantly related to viral load (negatively) and to CD4+ cell number (positively) for women only. Interestingly, only depth processing (and not emotional expression) was related to medication adherence and to psychosocial variables (perceived stress and social support). Conclusions: Emotional expression and depth processing were related to long-term survival, however, depth processing was the mediator for this relationship and only depth processing was associated with medication adherence, perceived stress and social support. Our results underscore the importance of depth processing (and not just emotional expression) of traumatic experiences for people living with HIV/AIDS.

Introduction There is considerable variation observed in disease course in those infected with HIV. It is likely that psychological factors are, in part, implicated in this observed variability [1], [2], [3] and [4]. The management of HIV/AIDS can involve responses to a wide variety of disease-related stressors [5] and [6]. It is possible that the emotional disclosure and processing of these stressors may impact the health and immune status of those living with HIV/AIDS, and may represent one of the sources of the observed variation in HIV disease progression [7] and [8]. Emotional disclosure/processing in nonpatient populations Although there is considerable research examining the relationship between emotional disclosure of trauma and health and immune outcomes, much of it has been conducted in nonpatient populations. Emotional disclosure has been related to decreases in health center visits, absences from work and self-reported health problems [9], [10], [11] and [12], although two studies failed to find these effects [13] and [14]. Emotional disclosure of healthy subjects has also been related to favorable changes in immune function, such as greater proliferation to antigens over time [9], increased immune system control over Epstein–Barr viral capsid antigen [15], [16] and [17], higher antibody concentrations in response to hepatitis vaccinations [18] and enhanced natural killer cell function for subjects high in cynical hostility [19]. In addition to emotional disclosure, depth processing of stressful events (e.g., cognitive change, involvement and finding meaning) has been related to beneficial changes in immune function. Comparing three emotional expression conditions (written, verbal and trivial writing), Esterling et al. [16] found that, in addition to group assignment, cognitive change, self-esteem enhancements and seriousness of the trauma predicted greater control of EBV at follow-up. In a follow-up study, the level of experiential involvement was also associated with large decreases in EBV-VCA antibody titers [17]. These studies provide strong support for the relationship between emotional disclosure and immune function but provide only equivocal evidence relating emotional disclosure to favorable health outcomes. These results also establish the connection between depth processing of stressful experiences and improved control over EBV. Emotional disclosure/processing in immune-related disease Several studies have examined the relationship between emotional expression/processing and health outcomes in patients with chronic illnesses related to immune system dysfunction that may provide some evidence of a pathophysiologic link to disease. Smyth et al. [20] reported that written emotional disclosure in patients with asthma or rheumatoid arthritis was related to improved health outcomes at 4-month follow-up that were assessed for some patients to be clinically significant. The effects of a verbal emotional expression intervention on several affective and disease-related outcomes were also examined in patients with rheumatoid arthritis [21]. Patients were randomly assigned to talk about stressful or trivial events. At 3-month follow-up, patients in the stressful disclosure condition reported less affective disturbance and better physical functioning. Also, comparing two groups of patients with HIV, higher levels of emotional disclosure and depth processing (cognitive change, self-esteem enhancements, adaptive coping strategies and involvement) of traumatic experiences were associated with asymptomatic health status in patients with AIDS who had very low CD4+ cell number compared with a HIV-seropositive comparison group [7]. In addition, depth processing was positively related to natural killer cell number, which mediated the relationship between depth processing and healthy survival with AIDS. Bower et al. [8] rated interviews conducted with 40 HIV-seropositive men who had recently experienced an AIDS-related bereavement and reported that men who found meaning had a less rapid decline in CD4+ cells and had lower rates of AIDS-related mortality at 2–3-year follow-up. These results provide some initial evidence relating emotional disclosure to clinically relevant outcomes in immune-related illnesses and some preliminary results underscoring the relevance of depth processing of trauma to immune and health status in patients living with HIV. The research reviewed above also raises the question as to the relative contributions of emotional expression and depth processing to related immune system and health benefits. Pennebaker's theory of inhibition [10] and [11] suggests that emotional disclosure confers health and immunologic benefit by reducing the inhibition associated with denying or avoiding upsetting emotional experiences associated with traumatic or stressful events. More recently, this theory has been modified to include consideration of processing variables such as cognitive change [22] and [23]. Others have hypothesized that resolution of traumatic experiences is based upon emotional disclosure followed by cognitive reappraisal rather than affective discharge alone [13] and [14]. In a written emotional expression study [7], it was reported that depth processing mediated the relationship between emotional disclosure and asymptomatic status in patients with AIDS. These results provide some preliminary evidence suggesting that the relationship between emotional disclosure and protected health status, in this patient population, may be governed by the extent to which written emotional disclosure facilitates depth processing of the traumatic experience. A central purpose of our study was to assess the relative contributions of disclosure and/or processing on long-term survival in people living with HIV/AIDS and to determine whether depth processing was the pathway by which disclosure might have its effect on long-term survival. Purpose The primary purpose of this study was to determine whether emotional disclosure (positive and negative expression) and processing (cognitive change, self-esteem enhancement, adaptive coping strategies and involvement) were related to long-term survival with AIDS. A second purpose was to determine whether depth processing mediated the relationship between emotional disclosure and long-term survival with AIDS. A third exploratory purpose of this study was to examine the relationship between emotional expression/processing and disease progression markers in HIV (CD4+ and viral load).

Results Demographic characteristics The sample ranged in age, gender, ethnicity, education, employment, income, sexual orientation and stressful life events as indicated in Table 1. The groups (LTS vs. ECOMPLTS) did not differ significantly on gender [χ2(1)=1.00, P=.32] or age [t(133)=−1.51, P=.13] with women representing approximately one quarter of the LTS group and one third of the ECOMPLTS group and an average age of 39 and 37 years, respectively. The sample across both subject groups was ethnically diverse with Caucasians, African Americans and Hispanics similarly represented in both groups [χ2(3)=4.54, P=.21]. The groups did not differ significantly on educational experience [χ2(3)=6.24, P=.12] with approximately two thirds of the sample having some educational experience beyond high school (some college or trade school or college graduates). Neither did the groups differ on employment status [χ2(4)=7.33, P=.12] or income [χ2(3)=1.82, P=.52] with approximately 75% of our sample on disability earning less than US$10,000 a year. Approximately half of our sample were heterosexual [χ2(2)=1.35, P=.42]. Medically relevant variables Most of our sample reported acquiring HIV through sexual contact with no significant difference between the groups whether they ascribed it to heterosexual or homosexual contact [χ2(2)=4.11, P=.11]. The groups did not differ significantly with respect to either CD4+ cell number [t(133)=−0.158, P=.52] or HIV-1 viral loadlog10 [t(133)=−1.07, P=.55]. The groups were also equivalent on antiretroviral medication regimen at study entry with more than three quarters of each group receiving some form of combination therapy [χ2(3)=3.52, P=.25]. The groups did not differ on past STDs [t(133)=0.897, P=.94], adherence to antiretroviral medications [t(77)=−0.977, P=.91] or stressful life events in the past 6 months [t(133)=0.265, P=.64]. An analysis of the topics discussed in the subject essays revealed that approximately half of the subjects in each group wrote about their reactions to receiving an HIV-positive diagnosis. Within the LTS group, the next most frequently discussed issues were HIV-related bereavement (26%), HIV symptom onset (15%), stress associated with lifestyle change (e.g., job loss, medication side effects, relationship issues and modifying life goals) (9%) and disclosure of HIV status (4%). Within the ECOMPLTS group, the next most commonly discussed essay topics were stress associated with lifestyle change (19%), HIV-related bereavement (15%), disclosure of HIV status (10%) and symptom onset (8%). Gender differences Although the genders were equivalently represented in both groups men and women differed significantly on relevant demographic and medically relevant variables. The average age for men in the sample (mean=40) was significantly higher than for women (mean=36) [t(133)=2.88, P<.01]. Although the majority of both men and women had some college experience, more women failed to complete high school and more men had completed a college degree [χ2(3)=20.11, P<.01]. The genders also differed significantly with respect to ethnicity [χ2(3)=38.59, P<.01]: 70% of women and only 20% of men were African–American. Although men had significantly higher incomes than women [χ2(3)=11.83, P<.01], men and women did not differ significantly on employment status [χ2(4)=2.45, P>.1]. There were substantial differences between the genders on sexual orientation [χ2(2)=52.40, P<.01], with over 90% of the women and only 27% of men identifying themselves as heterosexual. Men and women did not differ with respect to HIV medication regimens [χ2(3)=3.19, P>.1] or CD4+ number [t(133)=−.640, P>.1]. However, women had significantly higher viral loadlog10 than men [t(133)=−2.62, P=.01]. Men and women did not differ on measures of emotional disclosure or depth processing with all P's>.1. Reliability One third of the essays were scored by a second rater to provide a measure of interrater reliability for the emotional disclosure and processing variables. Interrater reliability for the emotional expression measures ranged from r=.95 to .99 with all Pearson correlations significant at P<.01. Reliability for the processing variables ranged from r=.83 to .90 also with all Pearson correlations significant at P<.01. Table 2 provides the mean score for each rater on each of the variables. The absence of substantial mean differences provides additional evidence of the reliability of the measures. Essays were also rated for seriousness of the stressful experience. Interrater reliability for this variable was .74 (P<.01). Table 2. Interrater reliability for emotional expression and depth processing measures from 43 essays Variable Rater 1 Rater 2 r Mean (S.D.) Mean (S.D.) Emotional expression measures Total word count 147.16 (131.46) 146.26 (128.02) .98** Positive emotion words 1.26 (1.79) 1.21 (1.73) .96** Negative emotion words 2.56 (2.41) 2.58 (2.46) .95** Depth processing measures Involvement 4.84 (1.60) 4.72 (1.74) .83** Cognitive change appraisal 3.88 (1.82) 4.00 (2.17) .88** Self-esteem enhancement 3.61 (1.83) 3.42 (1.99) .86** Adaptive coping strategies 3.91 (1.94) 3.74 (2.11) .90** ** Indicates P<.001. Table options Group differences on emotional disclosure and depth processing: omnibus test In order to test the differences between the LTS and ECOMPLTS groups while controlling for overall Type I error rate, the word count, total emotional expression and the depth processing variables were subjected to an omnibus Hotelling's trace test. The multivariate test was significant [F(6,128)=2.77, P=.02]. The results of follow-up univariate t-tests are shown in Table 3. Table 3. Means, standard deviations and t-tests for the LTS (n=46) and ECOMPLTS (n=89) groups on essays rated for emotional expression and depth processing variables Variable LTS ECOMPLTS t P Mean (S.D.) Mean (S.D.) Disclosure variables Total words 190.71 (116.66) 136.51 (91.84) 2.94 .006 (Total emotional words) 5.78 (4.48) 4.35 (3.33) 2.08 .035 Positive emotion words 1.96 (2.29) 1.38 (1.77) 1.60 .140 Negative emotion words 3.76 (3.01) 2.91 (2.48) 1.71 .084 Depth processing variables Cognitive change appraisal 4.22 (1.83) 3.23 (1.59) 3.25 .001 Involvement 5.18 (1.57) 4.18 (1.80) 3.16 .001 Self-esteem enhancement 4.24 (2.00) 3.21 (1.72) 3.10 .002 Adaptive coping strategies 4.49 (1.94) 3.46 (1.71) 3.14 .001 Table options Group differences on emotional disclosure The LTS group wrote significantly more words than the ECOMPLTS group [t(133)=2.94, P=.006]. Although the LTS group wrote more total emotional words that the ECOMPLTS group [t(133)=2.08, P=.035], the differences between the groups for positive [t(133)=1.60, P=.14] and negative emotion words [t(134)=−1.71, P=.08] were not significant. Group differences on depth processing The LTS group displayed significantly more cognitive appraisal change [t(133)=3.25, P=.001], experiential involvement [t(133)=3.16, P=.001], self-esteem enhancements [t(133)=3.10, P=.01] and adaptive coping strategies [t(133)=3.14, P=.001] than the ECOMPLTS group. The means and standard deviations for the two groups for each of the processing variables are presented in Table 3 It is possible that the experience of living with HIV for a longer period of time and/or experiencing more serious trauma (e.g., experiencing a potentially life threatening AIDS infection or neoplasm) may be causing the higher levels of emotional expression and depth processing observed in the LTS group. In order to increase our confidence that the differences between the groups in emotional expression and depth processing were not a result of long-survival, we examined the relationship between both time since HIV diagnosis and seriousness of the stressful event written about, with group membership. Although the groups did not differ significantly on seriousness of the traumatic event [t(133)=−1.58, P=.12], the LTS group had significantly longer time since their HIV diagnosis than the ECOMPLTS group [t(133)=3.139, P=.001]. In order to control for this difference, group membership was regressed on each of the emotional expression and processing variables controlling for time since diagnosis. A similar pattern of results emerged. The LTS group scored significantly higher than the ECOMPLTS group on total words [t(132)=2.50, P=.01], total emotional words [t(132)=2.10, P=.04], cognitive appraisal change [t(132)=3.311, P=.001], experiential involvement [t(132)=2.92, P=.004], self-esteem enhancement [t(132)=2.88, P=.005] and adaptive coping strategies [t(132)=3.12, P=.002]. The groups did not differ significantly on positive emotional words [t(132)=1.40, P=.17] or negative emotional words [t(132)=1.88, P=.06]. Depth processing as a mediator of emotional expression and long-term survival We hypothesized that depth processing would mediate the relationship between emotional expression and long-term survival with AIDS. The processing variables (cognitive appraisal change, experiential involvement, self-esteem enhancement and adaptive coping strategies) were combined in one depth processing factor. The mediational effect was tested by path analysis in line with established procedures [37]. The results of these analyses are presented below and depicted in Fig. 1. Depth processing as a mediator of the relationship between emotional expression ... Fig. 1. Depth processing as a mediator of the relationship between emotional expression and long-term survival with AIDS (controlling for time since diagnosis). Path diagram model for testing direct and indirect effects of emotional expression on long-term survival with AIDS with exploration of depth processing as a possible mediator. (The numbers in parentheses are partial correlations describing the direct relationships in the model controlling for time since diagnosis, the numbers outside parentheses are standardized β coefficients in the model with long-term survival as the dependent variable and time since diagnosis, emotional expression and depth processing as the independent variables.) Figure options In order to establish significant relationships between the variables (controlling for time since HIV diagnosis on which the groups differed significantly), long-term survival was regressed on total emotional expression [β=.17, t(132)=2.097, P=.04]. Depth processing was significantly regressed on time since diagnosis and emotional expression [β=.55, t(132)=7.68, P<.001]. Then, long-term survival was regressed on time since diagnosis and depth processing [β=.28, t(132)=3.37, P=.001]. These direct relationships are specified in parentheses in Fig. 1. Finally, long-term survival was regressed on time since diagnosis, total emotional expression and depth processing Depth processing was identified as a mediator of the relationship between total emotional expression and long-term survival because the beta weight associated with depth processing in this second regression equation was significant [β=.26, t(131)=2.63, P=.01] and the β weight associated with total emotional expression (which had been significant in the first regression equation) was no longer significant [β=.03, t(131)=0.29, P=.77]. Emotional expression/processing and immune relationships In order to examine the relationship between emotional expression/depth processing and HIV-1 viral load and CD4 cell number, control variables (age, gender, ethnicity, education, income and antiretroviral medications) were first entered as a block regressed separately for CD4 and HIV-1 viral load. Only age was significantly related to CD4 cell number and both age and antiretroviral medication were related to HIV-1 viral load. Therefore, in the subsequent analyses, both age and antiretroviral medication were entered as control variables. It has been suggested that there are gender-specific differences in viral load associated with HIV disease status [38]. Consequently, we tested the interaction of gender and each of the emotional expression/processing variables when examining the relationship with HIV-1 viral load and CD4 number. HIV-1 viral load As a central marker of disease progression, we examined the relationship between emotional expression/depth processing and HIV-1 viral load. Controlling for age and antiretroviral medication, we tested the significance of the interaction term for gender and the specific emotional expression/processing term. The interactions with gender were significant for positive emotions [t(129)=−2.00, P=.048] and approaching significance for total emotional expression [t(129)=−1.751, P=.08]. The interactions with gender were not significant for depth processing [t(129)=−1.31, P=.19] or negative emotional expression [t(129)=−.001, P=.99]. We then examined the relationships separately for men and women (still controlling for age and antiretroviral medication). The results of these analyses are presented in Table 4. Although, as indicated above, some of the interaction terms were not significant, the gender-specific analyses are presented for each disclosure and processing variables to allow for comparison of the effects. Table 4. The relationship between depth processing and emotional expression variables and immune (HIV-1 viral load, CD4+ cell number), health behavior (medication adherence) and psychosocial (perceived stress, social support) measures Correlatesa Depth processing Total emotion words Positive words Negative words Immune measures Viral load women (n=41) −.23, P=.16 −.30, P=.06 −.46, P=.002 .06, P=.73 Men (n=94) .07, P=.46 .13, P=.19 .03, P=.78 .13, P=.19 CD4+ number women (n=41) .49, P=.001 .29, P=.08 .48, P=.001 −.10, P=.54 Men (n=94) −.12, P=.23 −.10, P=.37 −.18, P=.10 .00, P=.99 Health behaviors Med adherence (% missed doses) −.30, P=.007 −.11, P=.35 −.07, P=.73 −.11, P=.34 Psychosocial measures Perceived stress −.23, P=.008 .04, P=.80 −.08, P=.36 .09, P=.31 Social supportb −.24, P=.003 .05, P=.55 −.03, P=.79 .07, P=.32 a The statistics reported for the immune measures are standardized β coefficients regressing viral load (CD4) on control variables (age, antiretroviral medication) and the specific emotional expression/processing variable separately for men and women. The statistic reported for health behaviors and psychosocial measures are Pearson correlations for the whole sample. b A high score on the RULS reflects lower social support. Table options Significant negative relationships emerged, for the women in our sample (n=41), between HIV-1 viral loadlog10 and positive emotional expression (β=−.46, P=.002). No significant relationships emerged for men between emotional expression/processing and HIV-1 viral load. CD4 cell number The relationship between emotional expression/depth processing and CD4 cell number was examined using the same methodology described above for HIV-1 viral load, whereby we examined the significance of the relationship between the Gender×Disclosure Variable interaction and CD4. Significant interactions were observed for depth processing [t(129)=3.107, P=.002], total [t(129)=2.00, P=.048] and positive [t(129)=3.55, P=.001] emotional expression but not for negative emotional expression [t(132)=−0.355, P=.72]. The results of this analysis are presented in Table 4. Significant relationships were also observed, for women, between CD4 cell count and depth processing (β=.49, P=.001), and positive emotional expression (β=.48, P=.001). In summary, for women, positive emotional expression was related to both viral load and CD4 cell number, and depth processing was related to CD4 cell number only. No significant relationships emerged for men. Putative mediators of depth processing/expression and long-term survival Depth processing was significantly and positively related to adherence to anti-HIV medication regimens (r=−.30, P=.007) and to social support (r=−.25, P=.003) and negatively related to perceived stress (r=−.23, P=.008). These variables were considered as possible mediators of the relationship between depth processing and long-term survival. Although they were all significantly related to depth processing, they did not meet the initial requirements for mediation and the analysis was suspended. Emotional expression measures were not significantly related to any of these measures with all P's>.30.