Expressive Writing Intervention Improves Quality of Life Among Chinese-American Breast Cancer Survivors: A Randomized Controlled Trial

Expressive Writing Intervention Improves Quality of Life Among Chinese-American Breast Cancer... Abstract Background Expressive writing interventions are shown to confer health benefits for Caucasian cancer survivors. However, few studies reported improved quality of life or studied ethnic minorities. Purpose The study evaluated whether a culturally sensitive expressive writing intervention improved quality of life. Methods Chinese-speaking breast cancer survivors (n = 136) in the USA were randomly assigned to one of three conditions to write three 30-min weekly essays: a cancer-fact condition to write about facts relevant to the cancer experience for three weeks; a self-regulation condition to write about deepest feelings at week 1, stress and coping at week 2, and finding benefits at week 3; or an enhanced self-regulation condition to write about stress and coping at week 1, deepest feelings at week 2, and finding benefits at week 3. Quality of life was assessed by Functional Assessment of Cancer Therapy at baseline, 1-, 3-, and 6-month follow-ups. Results Growth curve models showed that quality of life was increased in the sample from baseline to the 6-month follow-up. The enhanced self-regulation condition had a large and statistically significant effect (d = 0.90, 95% CI [0.02, 1.687]), and the self-regulation condition had a small effect (d = 0.22, 95% CI [−0.79, 1.07]) on quality of life improvement compared with the cancer-fact group. Conclusion Expressive writing is shown to be an effective intervention to improve quality of life for Chinese-American cancer survivors. Future efforts are warranted to disseminate and implement this low-dose and brief intervention in community and clinical settings. Clinical Trial Registration NCT02946619. Expressive writing, Asian, Chinese breast cancer survivors, Randomized controlled trial, Quality of life Currently an estimated 17.3 million U.S. residents identify themselves as Asian [1]. Cancer is the leading cause of death for Asian Americans. Asian-American women have the fastest growing rate of breast cancer in the nation [2]. However, few intervention studies have been conducted on Asian-American cancer survivors, despite concerns about quality of life among this population [3]. In order to reduce this health disparity, we tested the efficacy of a culturally sensitive intervention using an expressive writing paradigm to improve quality of life for Asian-American breast cancer survivors. Expressive writing is a brief intervention designed to promote health by facilitating emotional and cognitive processes. A typical expressive writing experiment randomly assigns participants to either an intervention group to write about their deepest thoughts and feelings related to a traumatic experience or a control group to write about a neutral topic unrelated to deep feelings in three 20-min sessions [4]. Reviews and a meta-analysis with 146 randomized controlled trials among non-Hispanic Whites reveal that expressive writing confers a variety of benefits, including increased physical and psychological well-being [5–7]. Multiple randomized controlled trials also demonstrate that expressive writing delivers promising effects among cancer survivors, such as improving physical symptoms [8, 9], pain [10], sleep [11], depressive symptoms [12], and perceived social support [13]. For example, one study randomly assigned 60 breast cancer patients to one of the three conditions [8]: an emotional disclosure condition to write about their deepest feelings and thoughts, a benefit finding condition to write about positive aspects of their cancer experience, and a control group to write about their cancer facts unrelated to emotions. At a 3-month follow-up, both emotional disclosure and benefit finding reduced medical appointments for cancer-related morbidities. Despite the demonstrated benefits from those studies, a recent meta-analysis reviewed 16 randomized controlled trials and concluded no overall evidence for benefits of writing in psychological, physical, or quality of life outcomes [14]. Conversely, the review did suggest that those who experience low levels of emotional support may benefit from expressive writing. The variable effects of expressive writing among cancer survivors suggest that future research should aim to study populations in need of emotional support in order to optimize the benefits of expressive writing. For example, as previous studies have shown that emotional disclosure or benefit finding alone improves health outcomes, combining emotional disclosure with benefit finding may potentially deliver more benefits. Although the effects of combination writings have rarely been tested among cancer survivors, several studies among undergraduates have examined the effects of writing instructions designed to prompt emotional disclosure in combination with other processes, such as coping, cognitive processing, and positive outlook [15–17]. These studies have demonstrated that the combination writing group improved physical health and psychological well-being compared with a control group. In order to predict whether expressive writing interventions would be beneficial for other ethnic/cultural groups and to guide the design of expressive writing instructions for culturally diverse groups, Lu and Stanton developed a self-regulation moderator model of expressive writing [18]. This model emphasizes the essential role of cognitive reappraisal of stressful events and the facilitative role of emotional disclosure in producing health benefits. Stress and coping theory [19] argues that stress influences health outcomes via appraisal. Along the same line, the self-regulation moderator model postulates cognitive reappraisal as a critical pathway in reducing the impact of stress and producing beneficial effects through expressive writing. Writing that facilitates cognitive reappraisal would bring positive change in the evaluation of stressors and/or the self. Combining emotional disclosure and cognitive reappraisal constitutes a self-regulation process that may promote maximal benefits. A recent study showed that writings combining emotional disclosure with cognitive reappraisal improved positive affect at 1-, 2-, and 4-month follow-ups for both Caucasians and Asians [18]. Writing instructions that facilitate both emotional disclosure and cognitive reappraisal could be beneficial for cancer survivors. Unfortunately, Asian-American breast cancer survivors face culturally specific obstacles that prevent them from seeking emotional support from family and friends. Such obstacles consist of stigma due to cancer and the suppression of emotions in order to bear the burden alone and avoid damaging harmony with others [20]. Expressive writing allows individuals to disclose intrapersonally without damaging harmony or provoking stigma and thus, may be especially suitable to Asian-American communities. In an effort to develop a culturally sensitive intervention, we incorporated both the communities’ input and the previous research among Asians [18]. We created self-regulation writing instructions to facilitate emotional disclosure, cognitive reappraisal, and benefit finding. The self-regulation writing intervention was shown to be culturally sensitive and associated with improved quality of life among Chinese breast cancer survivors in a pilot study [21]. In contrast, however, a recent randomized controlled trial failed to show that the self-regulation writing improved quality of life compared with cancer-fact writing [22]. The study randomly assigned Chinese-speaking breast cancer survivors to three conditions: (a) an emotional disclosure condition to write about deepest thoughts and feelings; (b) a self-regulation condition to write about deepest thoughts and feelings, stress, coping, and benefit finding from their cancer experience; and (c) a cancer-fact writing condition to write about facts related to cancer diagnosis and treatment. Surprisingly, the cancer-fact writing condition reported the highest level of overall quality of life at the 6-month follow-up. Therefore, it is imperative to determine what types of expressive writing can deliver health benefits above and beyond cancer-fact writing for Asian Americans. One explanation for lack of benefits in expressive writing is that people may not benefit from emotional disclosure if they are not ready for it. A focus group interview with Chinese-speaking breast cancer survivors suggested that the participants found it more difficult to write about emotion than writing about stress and its impact (unpublished pilot data). Prior studies found that writings facilitating cognitive reappraisal benefited Asians the most [18, 22]. Engaging in cognitive reappraisal first may be more congruent with Asian culture than engaging in emotional disclosure first because Asian culture values emotional control and restraint over emotional expression [23]. Therefore, we developed new writing instructions in the hope of facilitating cognitive reappraisal first and emotional disclosure second to enhance the benefits of expressive writing. The current study aimed to examine the impact of expressive writing on quality in a randomized controlled trial with three arms: the cancer-fact writing and self-regulation conditions tested in the previous study [22] and a new condition called enhanced self-regulation condition, which used the same writing instructions in the self-regulation condition, but the order of the first and second instructions were reversed. In the enhanced self-regulation condition, participants were asked to write about stress and coping first, then deepest feelings, and finally, positive thoughts and feelings. These writing instructions were designed to facilitate cognitive reappraisal, emotional disclosure, and benefit finding, respectively. The primary hypothesis was that the enhanced self-regulation condition would improve quality of life compared with the cancer-fact condition, and the secondary hypothesis was that the enhanced self-regulation condition would have higher quality of life compared with the self-regulation condition, which would have higher quality of life compared with the cancer-fact condition (shown in Fig. 1). Because Chinese individuals make up the largest Asian subgroup and breast cancer is the most common cancer among Chinese women [24], this study focused on testing the intervention among Chinese-speaking breast cancer survivors in the USA. Cancer survivorship research encompasses the physical and psychosocial sequelae of cancer diagnosis and its treatment, and thus, quality of life is one of the most important outcomes during cancer survivorship. Surprisingly, although expressive writing studies have shown benefits in alleviating psychological and physical distress among cancer survivors, few had reported increased quality of life. If the new writing instructions are shown to improve quality of life, the study will contribute to the general literature not only by applying the paradigm of expressive writing to an understudied population but also by offering new directions to optimize the benefits of expressive writing among cancer survivors. Fig. 1 View largeDownload slide Hypotheses: projection of quality of life by condition across time. Fig. 1 View largeDownload slide Hypotheses: projection of quality of life by condition across time. Methods Participants Participants were recruited from Los Angeles, New York, and Houston metropolitan areas in collaboration with the Herald Cancer Association, a community-based Chinese cancer organization. Potential participants were introduced to the study at cultural events, educational conferences, and support groups and were screened for eligibility. Eligibility criteria were (a) having a breast cancer diagnosis, (b) completing breast cancer surgery within 5 years, and (c) being comfortable writing and speaking Chinese (i.e., Mandarin or Cantonese). Power analysis revealed that 43 participants per group would yield 95% power to test the primary hypothesis with an estimated effect size of d = 0.8 [25] derived from a previous study [18]. To account for 5% estimated attrition, 136 were enrolled in the study (refer to flow chart in Fig. 2). Fig. 2 View largeDownload slide CONSORT chart. Fig. 2 View largeDownload slide CONSORT chart. Procedure The study was approved by several relevant institutional review boards. Participants completed the baseline questionnaire package at home and returned the baseline questionnaire by mail. A simple randomization was conducted using 1:1:1 allocation to assign participants to the three conditions. Randomization was conducted by a well-trained research staff working at the community site. Simple randomization was done by drawing one card randomly from a box and returning the card to the box after drawing. Each of the cards was labeled as 1, 2, or 3. The person who did the randomization did not know what number stood for what condition. The writing instructions were contained in large opaque sealed envelopes, which were labeled as 1, 2, or 3. Participants assigned to the “1” card would be in condition “1” and would receive writing instructions contained in the envelope labeled as “1.” Similarly, participants assigned to the “2” or “3” cards would receive the designated envelopes. Participants were told that the study aimed to understand their experiences through writing and were not made aware of whether they were assigned to an experimental or a control condition. Researchers who had interactions with participants were blinded to participants’ condition assignment. Once randomization was performed, no change in intervention allocation took place. The intervention consisted of three writing sessions starting one week after the baseline assessment. An established protocol was followed [4], with minor modifications based on prior work [18, 22]. One week after returning the baseline questionnaire, each participant received two copies of a calendar and three sealed envelopes. Participants were asked to schedule three weekly writing sessions on the calendar, keep one copy of the calendar as a reminder, and send one copy back so that researchers can remind them of the next writing session. Each envelope was labeled as “week 1”, “week 2”, and “week 3” with writing instructions for the corresponding week. Participants were asked to only open the corresponding envelope at the scheduled writing time and to write continuously for up to 30 min or until they complete one page of writing. For each session, participants were asked to write continuously for up to 30 min. Participants in the self-regulation condition were asked to write about their deepest feelings and thoughts related to their breast cancer experience at week 1, the stress and their coping strategies to deal with the stress at week 2, and positive thoughts and feelings regarding their breast cancer experience at week 3 [22]. Participants in the enhanced self-regulation condition were given the same instructions as the self-regulation condition, but the order was switched between week 1 and week 2. Participants in the cancer-fact condition were asked to write about their cancer diagnosis and treatment as objectively and as detailed as they could for 3 weeks [8, 22]. Participants wrote at home and mailed their essays back in a prepaid envelope after each writing session. They were also given reminders to improve adherence. After the last writing session, they completed manipulation check items described below. At 1, 3, and 6 months after the last writing, participants completed and mailed back follow-up questionnaires assessing quality of life. Measures Functional Assessment of Cancer Therapy general scale: FACT-G This 27-item measure assesses perceived life quality including physical, social, emotional, and functional well-being in the past week [26]. The scale was validated among Chinese breast cancer survivors [27]. The items were rated on a five-point scale from (0) not at all to (4) very much. Total scores were computed by summing all the items. The Cronbach’s alpha of the FACT-G at baseline was 0.94. Manipulation check Participants rated how personal the writings were, how much emotion they revealed in the writings, and how meaningful and valuable the study was on a seven-point scale [22]. Demographic and clinical variables (e.g., age, marital status, treatment received) were self-reported at baseline. Data Analysis Plan Analyses were conducted using Mplus 7.2 [28] as intent-to-treat analyses to include all randomized participants (n = 136). Direct maximum likelihood estimation procedures were used to accommodate missing data. We began by calculating effect sizes with 95% confidence intervals to determine the magnitude of differences in quality of life between the three conditions at each time point. Latent growth curve modeling was then used to examine the impact of the treatment condition on the intraindividual trajectories of quality of life across the four assessments. The latent growth curve was used as the primary test for study hypotheses since we were interested in the effects of the treatment condition on the overall trajectories in quality of life from baseline to the final assessment. As there were no expectations regarding the nature of the trajectories, shape factors were specified such that the loading of the slope factor was fixed at 0 for the baseline assessment, fixed at 1 for the month 6 follow-up assessment, and freely estimated for the intermediate assessments. The mean of the slopes in the latent growth curve models therefore represented the total change in quality of life across the four assessments. Two separate latent growth curves were estimated in which dummy codes representing treatment conditions were specified as predictors of the slope factor of quality of life (the cancer-fact condition was the reference condition in model 1, and the self-regulation condition was the reference condition in model 2). Results for the effects of treatment condition on changes in quality of life were presented as both unstandardized effects and partially standardized effects, which can be interpreted as an effect size akin to Cohen’s d. Model fit for the latent growth curve was evaluated based on commonly used fit indices and criteria [29]: root mean square error of approximation (RMSEA; ≤ 0.08) and the comparative fit index (CFI; ≥ 0.95). Results Sample Characteristics, Randomization, Completion of Intervention, and Attribution Recruitment was from June 2012 to April 2015. The age of the participants ranged from 34 to 84 years old (M = 57.6, SD = 9.5). Among them, 65.4% had an annual income less than $45,000, 65.7 % were married, 85.5% were diagnosed with breast cancer at stages I, II, or III, and the mean time since diagnosis was 26.4 months (SD = 19.2 months; see Table 1 for sample characteristics). Participants were randomly assigned to one of the three conditions: enhanced self-regulation condition (n = 54), self-regulation condition (n = 46), or the cancer-fact condition (n = 36). Participants in the three conditions were compared to determine the success of randomization. ANOVA and chi-square test revealed no significant differences among the conditions in demographic variables (i.e., age, education level, family income, employment, and marital status), clinical variables (i.e., stage and time since cancer diagnosis), and outcome at baseline (i.e., quality of life; p’s > .05), suggesting successful randomization. All statistical tests were two tailed with α = 0.05. Table 1 Participant demographics and clinical characteristics by group Demographic or clinical characteristics  CFC (n = 36)  SRC (n = 46)  ESRC (n = 54)  Overall (n = 136)    N  %  N  %  N  %  N  %  p  Age          .64   Mean (SD)  57.9 (8.2)  58.6 (9.5)  56.9 (9.7)  57.8 (9.2)     Missing  0  0.0  1  2.2  3  5.6  4  2.9    Years in the US                  .20   Mean (SD)  16.6 (9.1)  20.6 (10.6)  17.8(10.8)  18.5 (10.4)     Missing  2  5.6  0  0.0  0  0.0  2  1.5    Marital status:                  .84   Married  22  61.1  29  63.0  37  68.5  88  64.7     Not married  13  36.1  16  34.8  17  31.5  46  33.8     Missing  1  2.8  1  2.2  0  0.0  2  1.5    Highest education                  .99   ≥Some college  22  61.1  28  60.9  33  61.1  83  61.0     No college  14  38.9  18  39.1  20  37.0  52  38.2     Missing  0  0.0  0  0.0  1  1.9  1  0.7    Household income                  .82   ≤$15,000  13  36.1  16  34.8  17  31.5  46  33.8     $15,000–$45,000  10  27.8  13  28.3  20  37.0  43  31.6     $45,000–$75,000  5  13.9  4  8.7  4  7.4  13  9.6     >$75,000  4  11.1  9  19.6  9  16.7  22  16.2     Missing  4  11.1  4  8.7  4  7.4  12  8.8    Employment status                  .53   Working full-time  9  25.0  10  21.7  15  27.8  34  25.0     Working part-time  8  22.2  7  15.2  6  11.1  21  15.4     Retired  6  16.7  9  19.6  5  9.3  20  14.7     Housewife  6  16.7  8  17.4  15  27.8  29  21.3     Unemployed  4  11.1  9  19.6  11  20.4  24  17.6     other  3  8.3  2  4.3  1  1.9  6  4.4     Missing  0  0  1  2.2  1  1.9  2  1.5    Stage                  .10   0  3  8.3  5  10.9  7  13.0  15  11.0     1  17  47.2  10  21.7  16  29.6  43  31.6     2  13  36.1  16  34.8  17  31.5  46  33.8     3  1  2.8  11  23.9  11  20.4  23  16.9     4  0  0.0  1  2.2  3  5.6  4  2.9     Missing  2  5.6  3  6.5  0  0.0  5  3.7    Treatment                     Surgery  34  94.4  43  93.5  50  92.6  127  93.4  .94   Radiation  18  51.4  26  56.5  40  74.1  84  62.2  .06   Chemo  20  55.6  30  65.2  31  58.5  81  60.0  .65  Baseline QOL                  .42   Mean (SD)  70.3 (19.7)  72.5 (22.9)  75.9 (19.0)  73.3 (20.5)     Missing  0  0.0  0  0.0  0  0.0  0  0.0    Demographic or clinical characteristics  CFC (n = 36)  SRC (n = 46)  ESRC (n = 54)  Overall (n = 136)    N  %  N  %  N  %  N  %  p  Age          .64   Mean (SD)  57.9 (8.2)  58.6 (9.5)  56.9 (9.7)  57.8 (9.2)     Missing  0  0.0  1  2.2  3  5.6  4  2.9    Years in the US                  .20   Mean (SD)  16.6 (9.1)  20.6 (10.6)  17.8(10.8)  18.5 (10.4)     Missing  2  5.6  0  0.0  0  0.0  2  1.5    Marital status:                  .84   Married  22  61.1  29  63.0  37  68.5  88  64.7     Not married  13  36.1  16  34.8  17  31.5  46  33.8     Missing  1  2.8  1  2.2  0  0.0  2  1.5    Highest education                  .99   ≥Some college  22  61.1  28  60.9  33  61.1  83  61.0     No college  14  38.9  18  39.1  20  37.0  52  38.2     Missing  0  0.0  0  0.0  1  1.9  1  0.7    Household income                  .82   ≤$15,000  13  36.1  16  34.8  17  31.5  46  33.8     $15,000–$45,000  10  27.8  13  28.3  20  37.0  43  31.6     $45,000–$75,000  5  13.9  4  8.7  4  7.4  13  9.6     >$75,000  4  11.1  9  19.6  9  16.7  22  16.2     Missing  4  11.1  4  8.7  4  7.4  12  8.8    Employment status                  .53   Working full-time  9  25.0  10  21.7  15  27.8  34  25.0     Working part-time  8  22.2  7  15.2  6  11.1  21  15.4     Retired  6  16.7  9  19.6  5  9.3  20  14.7     Housewife  6  16.7  8  17.4  15  27.8  29  21.3     Unemployed  4  11.1  9  19.6  11  20.4  24  17.6     other  3  8.3  2  4.3  1  1.9  6  4.4     Missing  0  0  1  2.2  1  1.9  2  1.5    Stage                  .10   0  3  8.3  5  10.9  7  13.0  15  11.0     1  17  47.2  10  21.7  16  29.6  43  31.6     2  13  36.1  16  34.8  17  31.5  46  33.8     3  1  2.8  11  23.9  11  20.4  23  16.9     4  0  0.0  1  2.2  3  5.6  4  2.9     Missing  2  5.6  3  6.5  0  0.0  5  3.7    Treatment                     Surgery  34  94.4  43  93.5  50  92.6  127  93.4  .94   Radiation  18  51.4  26  56.5  40  74.1  84  62.2  .06   Chemo  20  55.6  30  65.2  31  58.5  81  60.0  .65  Baseline QOL                  .42   Mean (SD)  70.3 (19.7)  72.5 (22.9)  75.9 (19.0)  73.3 (20.5)     Missing  0  0.0  0  0.0  0  0.0  0  0.0    CFC cancer fact condition; SRC self-regulation condition; ESRC enhanced self-regulation condition; QOL Quality of Life. View Large Table 1 Participant demographics and clinical characteristics by group Demographic or clinical characteristics  CFC (n = 36)  SRC (n = 46)  ESRC (n = 54)  Overall (n = 136)    N  %  N  %  N  %  N  %  p  Age          .64   Mean (SD)  57.9 (8.2)  58.6 (9.5)  56.9 (9.7)  57.8 (9.2)     Missing  0  0.0  1  2.2  3  5.6  4  2.9    Years in the US                  .20   Mean (SD)  16.6 (9.1)  20.6 (10.6)  17.8(10.8)  18.5 (10.4)     Missing  2  5.6  0  0.0  0  0.0  2  1.5    Marital status:                  .84   Married  22  61.1  29  63.0  37  68.5  88  64.7     Not married  13  36.1  16  34.8  17  31.5  46  33.8     Missing  1  2.8  1  2.2  0  0.0  2  1.5    Highest education                  .99   ≥Some college  22  61.1  28  60.9  33  61.1  83  61.0     No college  14  38.9  18  39.1  20  37.0  52  38.2     Missing  0  0.0  0  0.0  1  1.9  1  0.7    Household income                  .82   ≤$15,000  13  36.1  16  34.8  17  31.5  46  33.8     $15,000–$45,000  10  27.8  13  28.3  20  37.0  43  31.6     $45,000–$75,000  5  13.9  4  8.7  4  7.4  13  9.6     >$75,000  4  11.1  9  19.6  9  16.7  22  16.2     Missing  4  11.1  4  8.7  4  7.4  12  8.8    Employment status                  .53   Working full-time  9  25.0  10  21.7  15  27.8  34  25.0     Working part-time  8  22.2  7  15.2  6  11.1  21  15.4     Retired  6  16.7  9  19.6  5  9.3  20  14.7     Housewife  6  16.7  8  17.4  15  27.8  29  21.3     Unemployed  4  11.1  9  19.6  11  20.4  24  17.6     other  3  8.3  2  4.3  1  1.9  6  4.4     Missing  0  0  1  2.2  1  1.9  2  1.5    Stage                  .10   0  3  8.3  5  10.9  7  13.0  15  11.0     1  17  47.2  10  21.7  16  29.6  43  31.6     2  13  36.1  16  34.8  17  31.5  46  33.8     3  1  2.8  11  23.9  11  20.4  23  16.9     4  0  0.0  1  2.2  3  5.6  4  2.9     Missing  2  5.6  3  6.5  0  0.0  5  3.7    Treatment                     Surgery  34  94.4  43  93.5  50  92.6  127  93.4  .94   Radiation  18  51.4  26  56.5  40  74.1  84  62.2  .06   Chemo  20  55.6  30  65.2  31  58.5  81  60.0  .65  Baseline QOL                  .42   Mean (SD)  70.3 (19.7)  72.5 (22.9)  75.9 (19.0)  73.3 (20.5)     Missing  0  0.0  0  0.0  0  0.0  0  0.0    Demographic or clinical characteristics  CFC (n = 36)  SRC (n = 46)  ESRC (n = 54)  Overall (n = 136)    N  %  N  %  N  %  N  %  p  Age          .64   Mean (SD)  57.9 (8.2)  58.6 (9.5)  56.9 (9.7)  57.8 (9.2)     Missing  0  0.0  1  2.2  3  5.6  4  2.9    Years in the US                  .20   Mean (SD)  16.6 (9.1)  20.6 (10.6)  17.8(10.8)  18.5 (10.4)     Missing  2  5.6  0  0.0  0  0.0  2  1.5    Marital status:                  .84   Married  22  61.1  29  63.0  37  68.5  88  64.7     Not married  13  36.1  16  34.8  17  31.5  46  33.8     Missing  1  2.8  1  2.2  0  0.0  2  1.5    Highest education                  .99   ≥Some college  22  61.1  28  60.9  33  61.1  83  61.0     No college  14  38.9  18  39.1  20  37.0  52  38.2     Missing  0  0.0  0  0.0  1  1.9  1  0.7    Household income                  .82   ≤$15,000  13  36.1  16  34.8  17  31.5  46  33.8     $15,000–$45,000  10  27.8  13  28.3  20  37.0  43  31.6     $45,000–$75,000  5  13.9  4  8.7  4  7.4  13  9.6     >$75,000  4  11.1  9  19.6  9  16.7  22  16.2     Missing  4  11.1  4  8.7  4  7.4  12  8.8    Employment status                  .53   Working full-time  9  25.0  10  21.7  15  27.8  34  25.0     Working part-time  8  22.2  7  15.2  6  11.1  21  15.4     Retired  6  16.7  9  19.6  5  9.3  20  14.7     Housewife  6  16.7  8  17.4  15  27.8  29  21.3     Unemployed  4  11.1  9  19.6  11  20.4  24  17.6     other  3  8.3  2  4.3  1  1.9  6  4.4     Missing  0  0  1  2.2  1  1.9  2  1.5    Stage                  .10   0  3  8.3  5  10.9  7  13.0  15  11.0     1  17  47.2  10  21.7  16  29.6  43  31.6     2  13  36.1  16  34.8  17  31.5  46  33.8     3  1  2.8  11  23.9  11  20.4  23  16.9     4  0  0.0  1  2.2  3  5.6  4  2.9     Missing  2  5.6  3  6.5  0  0.0  5  3.7    Treatment                     Surgery  34  94.4  43  93.5  50  92.6  127  93.4  .94   Radiation  18  51.4  26  56.5  40  74.1  84  62.2  .06   Chemo  20  55.6  30  65.2  31  58.5  81  60.0  .65  Baseline QOL                  .42   Mean (SD)  70.3 (19.7)  72.5 (22.9)  75.9 (19.0)  73.3 (20.5)     Missing  0  0.0  0  0.0  0  0.0  0  0.0    CFC cancer fact condition; SRC self-regulation condition; ESRC enhanced self-regulation condition; QOL Quality of Life. View Large Figure 2 shows the completion of the writing interventions. Among the 136 participants, 112 (82.4%) completed all three writing sessions, 3 (2.2%) completed the first two writings, and 21 (15.4%) did not complete any writing and withdrew during the writing stage. The completion of writing sessions did not differ by group (χ2 = 2.17, p = .34). Four participants withdrew during the follow-up stage. Completers and noncompleters at each follow-up did not differ in the assigned condition, demographic, medical, and outcome variables at baseline (p’s > .14). Manipulation Check As shown in Table 2, ANOVA revealed that the enhanced self-regulation condition revealed more emotion compared with the self-regulation condition, which revealed more emotion compared with the cancer-fact condition (p’s < .02). The enhanced self-regulation and cancer-fact groups reported that the study had higher meaning/value compared with the self-regulation (p < .01) group. Table 2 Analyses of variance on essay ratings at the end of writing sessions Dependent variable  CFC (n = 36)  SRC (n = 46)  ESRC (n = 54)  F  p  Essay was personal   Mean  1.83  2.61  2.55  1.33  .2693   SD  1.97  2.12  2.23      Revealed emotions   Mean  2.73ab  3.73a  4.6b  12.51  <.0001   SD  1.66  1.84  1.36      Valuable/meaningful   Mean  5.1a  4.21b  5.04a  4.95  .0088   SD  1.21  1.54  1.18      New perspective   Mean  3.23  3.00  3.64  1.22  .2999   SD  2.18  1.8  1.65      Dependent variable  CFC (n = 36)  SRC (n = 46)  ESRC (n = 54)  F  p  Essay was personal   Mean  1.83  2.61  2.55  1.33  .2693   SD  1.97  2.12  2.23      Revealed emotions   Mean  2.73ab  3.73a  4.6b  12.51  <.0001   SD  1.66  1.84  1.36      Valuable/meaningful   Mean  5.1a  4.21b  5.04a  4.95  .0088   SD  1.21  1.54  1.18      New perspective   Mean  3.23  3.00  3.64  1.22  .2999   SD  2.18  1.8  1.65      Subscripts a, b, and ab indicate significant group differences in the ANOVA. Means with different subscripts in the same row are statistically different with p < .05. CFC cancer fact condition; SRC self-regulation condition; ESRC enhanced self-regulation condition. View Large Table 2 Analyses of variance on essay ratings at the end of writing sessions Dependent variable  CFC (n = 36)  SRC (n = 46)  ESRC (n = 54)  F  p  Essay was personal   Mean  1.83  2.61  2.55  1.33  .2693   SD  1.97  2.12  2.23      Revealed emotions   Mean  2.73ab  3.73a  4.6b  12.51  <.0001   SD  1.66  1.84  1.36      Valuable/meaningful   Mean  5.1a  4.21b  5.04a  4.95  .0088   SD  1.21  1.54  1.18      New perspective   Mean  3.23  3.00  3.64  1.22  .2999   SD  2.18  1.8  1.65      Dependent variable  CFC (n = 36)  SRC (n = 46)  ESRC (n = 54)  F  p  Essay was personal   Mean  1.83  2.61  2.55  1.33  .2693   SD  1.97  2.12  2.23      Revealed emotions   Mean  2.73ab  3.73a  4.6b  12.51  <.0001   SD  1.66  1.84  1.36      Valuable/meaningful   Mean  5.1a  4.21b  5.04a  4.95  .0088   SD  1.21  1.54  1.18      New perspective   Mean  3.23  3.00  3.64  1.22  .2999   SD  2.18  1.8  1.65      Subscripts a, b, and ab indicate significant group differences in the ANOVA. Means with different subscripts in the same row are statistically different with p < .05. CFC cancer fact condition; SRC self-regulation condition; ESRC enhanced self-regulation condition. View Large Preliminary Analysis We first examined mean levels of quality of life by condition across time based on the direct maximum likelihood estimates that accounted for missing data due to attrition. Means, standard deviations, and between condition effect sizes (Cohen’s d) with 95% confidence intervals are presented in Table 3. Baseline differences in quality of life were statistically nonsignificant. An examination of the means by condition across time indicated that quality of life generally increased across time in all three conditions. Based on the confidence intervals of the effect sizes, the self-regulation condition resulted in a higher level of quality of life at the final assessment than the cancer-fact condition that was small in effect size magnitude but statistically nonsignificant. The enhanced self-regulation condition resulted in a higher level of quality of life at the final assessment than the cancer-fact condition that was moderate in effect size magnitude (d = −0.58) and statistically significant. The enhanced self-regulation condition resulted in a higher level of quality of life at the final assessment than the self-regulation condition that was small-to-moderate in effect size magnitude (d = −0.31) but statistically nonsignificant. Thus, effect size analyses suggest that the enhanced self-regulation condition may have had the most robust effects on quality of life. Table 3 Descriptive Statistics and Between Condition Effect Sizes for Quality of Life   CFC (n = 36)  SRC (n = 46)  ESRC (n = 54)    d (95% CI)    Time  M  SD  M  SD  M  SD  CFC vs. SRC  CFC vs. ESRC  SRC vs. ESRC  T1  70.28  19.44  72.48  22.61  75.94  18.82  −0.10 (−0.54: 0.33)  −0.29 (−0.73: 0.14)  −0.17 (−0.56: 0.23)  T2  73.09  20.78  76.21  16.34  78.08  17.18  −0.17 (−0.60: 0.27)  −0.26 (−0.70: 0.17)  −0.11 (−0.50: 0.28)  T3  72.13  15.96  76.85  15.91  80.71  16.34  −0.29 (−0.73: 0.14)  −0.53a (−0.97: −0.08)  −0.24 (−0.63: 0.16)  T4  72.29  18.71  77.21  17.56  82.35  15.90  −0.27 (−0.71: 0.17)  −0.58a (−1.03: −0.14)  −0.31 (−0.70: 0.09)    CFC (n = 36)  SRC (n = 46)  ESRC (n = 54)    d (95% CI)    Time  M  SD  M  SD  M  SD  CFC vs. SRC  CFC vs. ESRC  SRC vs. ESRC  T1  70.28  19.44  72.48  22.61  75.94  18.82  −0.10 (−0.54: 0.33)  −0.29 (−0.73: 0.14)  −0.17 (−0.56: 0.23)  T2  73.09  20.78  76.21  16.34  78.08  17.18  −0.17 (−0.60: 0.27)  −0.26 (−0.70: 0.17)  −0.11 (−0.50: 0.28)  T3  72.13  15.96  76.85  15.91  80.71  16.34  −0.29 (−0.73: 0.14)  −0.53a (−0.97: −0.08)  −0.24 (−0.63: 0.16)  T4  72.29  18.71  77.21  17.56  82.35  15.90  −0.27 (−0.71: 0.17)  −0.58a (−1.03: −0.14)  −0.31 (−0.70: 0.09)  Negative effect sizes indicate that the condition listed first in the comparison had lower levels of quality of life. T1, baseline; T2 1-month follow-up; T3 3-month follow-up; T4 6-month follow-up; CFC cancer fact condition; SRC self-regulation condition; ESRC enhanced self-regulation condition. aThe effect size difference is statistically significant at p < .05 based on the confidence interval of the effect size (i.e., confidence interval does not include 0). View Large Table 3 Descriptive Statistics and Between Condition Effect Sizes for Quality of Life   CFC (n = 36)  SRC (n = 46)  ESRC (n = 54)    d (95% CI)    Time  M  SD  M  SD  M  SD  CFC vs. SRC  CFC vs. ESRC  SRC vs. ESRC  T1  70.28  19.44  72.48  22.61  75.94  18.82  −0.10 (−0.54: 0.33)  −0.29 (−0.73: 0.14)  −0.17 (−0.56: 0.23)  T2  73.09  20.78  76.21  16.34  78.08  17.18  −0.17 (−0.60: 0.27)  −0.26 (−0.70: 0.17)  −0.11 (−0.50: 0.28)  T3  72.13  15.96  76.85  15.91  80.71  16.34  −0.29 (−0.73: 0.14)  −0.53a (−0.97: −0.08)  −0.24 (−0.63: 0.16)  T4  72.29  18.71  77.21  17.56  82.35  15.90  −0.27 (−0.71: 0.17)  −0.58a (−1.03: −0.14)  −0.31 (−0.70: 0.09)    CFC (n = 36)  SRC (n = 46)  ESRC (n = 54)    d (95% CI)    Time  M  SD  M  SD  M  SD  CFC vs. SRC  CFC vs. ESRC  SRC vs. ESRC  T1  70.28  19.44  72.48  22.61  75.94  18.82  −0.10 (−0.54: 0.33)  −0.29 (−0.73: 0.14)  −0.17 (−0.56: 0.23)  T2  73.09  20.78  76.21  16.34  78.08  17.18  −0.17 (−0.60: 0.27)  −0.26 (−0.70: 0.17)  −0.11 (−0.50: 0.28)  T3  72.13  15.96  76.85  15.91  80.71  16.34  −0.29 (−0.73: 0.14)  −0.53a (−0.97: −0.08)  −0.24 (−0.63: 0.16)  T4  72.29  18.71  77.21  17.56  82.35  15.90  −0.27 (−0.71: 0.17)  −0.58a (−1.03: −0.14)  −0.31 (−0.70: 0.09)  Negative effect sizes indicate that the condition listed first in the comparison had lower levels of quality of life. T1, baseline; T2 1-month follow-up; T3 3-month follow-up; T4 6-month follow-up; CFC cancer fact condition; SRC self-regulation condition; ESRC enhanced self-regulation condition. aThe effect size difference is statistically significant at p < .05 based on the confidence interval of the effect size (i.e., confidence interval does not include 0). View Large Treatment Effects: Hypothesis Testing Using Growth Curve Models We next examined the impact of the treatment condition on the longitudinal trajectories of quality of life using latent growth curve (Fig. 3). The model fit was identical and good (χ2 [df = 10, n = 136] = 16.45, RMSEA = 0.07, CFI = 0.98) for the two latent growth curve models. The β parameters from these models represent the mean differences in quality of life trajectories between conditions. In the first model in which the control condition was specified as the reference condition, the enhanced self-regulation condition had a large and statistically significant effect on quality of life trajectories (b = 6.47, 95% CI [1.56, 11.38]; d = 0.90, 95% CI [0.02, 1.687]), and the self-regulation condition had a small but statistically nonsignificant effect on quality of life trajectories (b = 1.58, 95% CI [−0.50, 8.28]; d = 0.22, 95% CI [−0.79, 1.07]). In the second model in which the self-regulation condition was specified as the reference condition, the enhanced self-regulation condition had a moderate but statistically nonsignificant effect on quality of life trajectories (b = 4.89, 95% CI [−1.04, 10.81]; d = 0.68, 95% CI [−0.03, 1.39]). Together, the results of the latent growth curve analyses indicated that quality of life improved across all three treatment conditions (b = 3.96, 95% CI [1.36, 6.57] in the unconditional latent growth curve) but that the enhanced self-regulation condition was associated with the greatest intraindividual changes in quality of life. The improvement in quality of life was larger in the enhanced self-regulation condition than that in the control condition. Fig. 3 View largeDownload slide Growth curve models: quality of life by condition across time. Fig. 3 View largeDownload slide Growth curve models: quality of life by condition across time. Discussion Expressive writing has been shown to improve psychological and physical health outcomes in both healthy and clinical populations [7, 30]. However, few studies have been conducted among ethnic minorities or shown improvement in quality of life among cancer survivors. The present study is the first to show that expressive writing improved quality of life compared with a cancer-fact control among Chinese-American breast cancer survivors using a randomized controlled trial. It suggests that expressive writing is an efficacious, brief intervention for Chinese breast cancer survivors. Confirming our primary hypothesis, the study revealed that the enhanced self-regulation condition improved quality of life compared with the cancer-fact condition. Furthermore, the increase in quality of life from baseline to the 6 month follow-up reached both statistical significance and clinical significance, as a 5–7 points change in Functional Assessment of Cancer Therapy general scale is indicative of clinical significance [31]. These findings are noteworthy because previous examinations of expressive writing interventions among non-Hispanic Whites have not reported improvements in quality of life [32]. Although the study sample lived in the USA for 18 years on average, the baseline quality of life indicated by FACT (M = 73.3) was low, compared with non-Hispanic White female cancer patients (normative value = 82.1) [33] and breast cancer survivors (M = 84.3) [34]. Perhaps, the significant unmet emotional needs among Chinese-speaking breast cancer survivors [35] render this underserved population more likely to experience the benefits of an intervention, suggesting the importance of conducting intervention studies in underserved populations both from scientific and clinical perspectives. Reporting improvement in quality of life for the first time among cancer survivors could also be attributed to the addition of the new writing condition, which is different from the traditional emotional disclosure condition typically used in previous expressive writing studies among non-Hispanic White cancer survivors. The condition that improved quality of life the most was the enhanced self-regulation condition, which was designed to facilitate cognitive reappraisal, emotional disclosure, and benefit finding in that order, and delivered maximum health benefits. In contrast, the self-regulation condition, which was designed to facilitate disclosing emotion, cognitive reappraisal, and benefit finding in that order, only had small and nonsignificant effects compared with the cancer-fact writing. Thus, engaging in cognitive processing first is more beneficial than engaging in emotional disclosure first for the study population. This is in line with prior findings that cognitive reappraisal benefited East Asians the most [18, 22]. Prior work did not find that writing interventions facilitating emotional disclosure alone delivered benefits for Asians [18, 22], and one may conclude that Asians would not benefit from emotional disclosure. This study clarified potential misperception that Asians would not benefit from emotional disclosure, and takes a step further to highlight the importance of ordering effects of writing instructions, specifically, prompting cognitive reappraisal before emotional disclosure for Asians. East Asian culture values emotional control and restraint over emotional expression [23, 36], and thus, Chinese-American cancer survivors may find it unnatural to be asked to dive deeply into their emotions before having the opportunity to reflect on the stressful event that caused their negative emotions. Interestingly, the enhanced self-regulation condition was associated with greater emotional disclosure, suggesting that disclosing stress helped them be better prepared to engage in emotion disclosure. Our findings suggest that starting with “what happened to you and how you were impacted,” as the enhanced self-regulation condition did, rather than “what/how did you feel” is a culturally sensitive approach for addressing distress. Our findings challenge an implicit assumption that expressive writing interventions should always begin with an emotional disclosure. The stimulating findings shed light on the importance of developing and testing culturally based interventions congruent with cultural norms and practices for minority breast cancer survivors. This supports the argument that interventions for minorities need to consider cultural adaptation [37]. Although cultural norms of emotion suppression help to explain the ordering effects, it is premature to conclude that culture is the only explanation. It is also possible that the facilitation of cognitive reappraisal in the enhanced self-regulation condition during the first writing session sets up a positive tone for or gives more purpose to the second session of emotional disclosure. For example, the cognitive reappraisal writing instruction asked people to answer the following questions: “What kind of changes in your life have you made as a result of this experience?” “What things are you able to do better now?” This may prompt a growth mindset, which helps to steer the later session of emotional disclosure toward a more meaningful direction, rather than simply venting. This speculation was supported by the data that the enhanced self-regulation condition was rated more meaningful than the self-regulation condition. Engaging in cognitive reappraisal first and emotional disclosure second may help to form a more comprehensive story than the other way around, as forming a comprehensive story is proposed to be a key ingredient in producing the benefits of writing [5]. If that is the case, findings of the order effect may not be limited to Chinese Americans. This important ordering effect is the first reported in the expressive writing literature, perhaps due to the lack of a similar study design in previous studies. Expressive writing studies among non-Hispanic White cancer survivors either used emotional disclosure alone [8, 11] or used emotional disclosure prompts first prior to other writing prompts [13]. Similarly, studies using a combination of writing instructions among general populations usually began with giving emotional disclosure prompts first [15–17]. None had begun with cognitive reappraisal and switched to emotional disclosure later. The prevailing design preconception reflects an assumption that understanding emotions needs to happen first to understand stressors. Although the ordering effect has not been tested in the past, it does not preclude the possibility that disclosing stress first enhances the benefits of expressive writing for other ethnic groups including non-Hispanic Whites. A prior study showed that cognitive processing resulting from emotional disclosure writing was linked to improvements in physical symptoms among non-Hispanic Whites [9]. Furthermore, men are found to use less emotional processing and emotional expression than women [38]. Perhaps, men would benefit more from engaging in stress disclosure first than engaging in emotional disclosure first. The ordering effect should be tested for its generalizability in future, and the new writing instructions may prove to be fruitful for other ethnic groups and men as well. Several caveats exist. First, the cancer-fact condition was associated with improved quality of life in a previous study [22]. The self-regulation condition had a small and nonsignificant improvement in quality of life compared with the cancer-fact condition. The study was not powered to detect the difference. It is also possible that the self-regulation condition may have larger benefits compared with a nonwriting assessment only group, and this possibility should be tested in the future. In fact, previous expressive writing studies with cancer survivors suggest that the studies with a nonwriting control group were more likely to find benefits of writing [12, 13]. If that is the case, the effects of the enhanced self-regulation condition are potentially larger than were detected here. Second, the study was subject to self-selection bias and was limited to Chinese-speaking breast cancer survivors. Future studies need to examine other Asian subgroups and other types of cancer. Third, the baseline quality of life score seemed to be higher in the enhanced self-regulation condition; however, it was not statistically different from the other two groups. Nonetheless, it is still possible that those who have a higher quality of life to begin with may benefit more from expressive writing. Future studies examining this possibility require a larger sample. Fourth, the study focused on a sample that had completed cancer treatment, so it remains unclear what kind of writing instructions may benefit those undergoing treatment. It is possible that those who are undergoing treatment may have more negative emotions to reveal and thus benefit from different types of writing instructions. Finally, only self-report measures were used; hence, future studies assessing biomarkers will add to our understanding about the underlying biological mechanism. Despite the limitations, this study has several strengths, including an understudied population, an experimental method, and long-term follow-ups. The study was the first to demonstrate that the expressive writing improved quality of life compared with a fact writing condition using a randomized controlled trial among a subgroup of Asian-American cancer survivors. It shed light onto the oncology literature relevant to an underserved and growing patient population. Expressive writing is shown to be an effective brief intervention for Chinese-American cancer survivors. Specifically, Asians benefited from writing instructions emphasizing cognitive aspects first before emotional disclosure. This culturally tailored intervention challenges the implicit assumption that psychosocial interventions validated among non-Hispanic Whites could be directly generalized to other populations. Future research is warranted to test this intervention and various writing instructions for other ethnic groups in the USA and in other countries. Expressive writing allows people with limited English proficiencies to express and regulate their feelings and thoughts privately in their primary language. It is a relatively low-dose and feasible intervention model that could be easily disseminated and implemented in community and clinical settings. Acknowledgements The study was supported by American Cancer Society MRSGT-10-011-01-CPPB (PI: Q.L.). The authors would like to thank the anonymous reviewers and colleagues (Lara Traeger, William Tsai, Ivan Wu) for their constructive comments and Sidra Deen for proofreading. Compliance with Ethical Standards Authors’ Statement of Conflict of Interest and Adherence to Ethical Standards Authors do not have conflict of interests to disclose. The views expressed in the submitted article are his or her own and not an official position of the institution or funder. Ethical Approval All procedures performed in studies involving human participants were in accordance with the ethical standards of the institutional and/or national research committee and with the 1964 Helsinki declaration and its later amendments or comparable ethical standards. Informed Consent Informed consent was obtained from all individual participants included in the study. References 1. U.S. Census Bureau. Annual estimates of the resident population by sex, race, and hispanic origin for the United States, States, and Counties: April 1, 2010 to July 1, 2014. 2015; Available at http://factfinder.census.gov/faces/tableservices/jsf/pages/productview.xhtml?src=bkmk. Accessibility verified January 12, 2017 2. Gomez SL, Quach T, Horn-Ross PLet al.   Hidden breast cancer disparities in Asian women: Disaggregating incidence rates by ethnicity and migrant status. Am j Public Health . 2010; 100( Suppl 1): S125– S131. Google Scholar CrossRef Search ADS PubMed  3. Wen KY, Fang CY, Ma GX. Breast cancer experience and survivorship among Asian Americans: A systematic review. j Cancer Surviv . 2014; 8( 1): 94– 107. Google Scholar CrossRef Search ADS PubMed  4. Pennebaker JW, Beall SK. Confronting a traumatic event: Toward an understanding of inhibition and disease. j Abnorm Psychol . 1986; 95( 3): 274– 281. Google Scholar CrossRef Search ADS PubMed  5. Pennebaker JW. Writing about emotional experiences as a therapeutic process. Psychol Sci . 1997; 8( 3): 162– 166. Google Scholar CrossRef Search ADS   6. Smyth JM. Written emotional expression: Effect sizes, outcome types, and moderating variables. j Consult Clin Psychol . 1998; 66( 1): 174– 184. Google Scholar CrossRef Search ADS PubMed  7. Frisina PG, Borod JC, Lepore SJ. A meta-analysis of the effects of written emotional disclosure on the health outcomes of clinical populations. j Nerv Ment Dis . 2004; 192( 9): 629– 634. Google Scholar CrossRef Search ADS PubMed  8. Stanton AL, Danoff-Burg S, Sworowski LAet al.   Randomized, controlled trial of written emotional expression and benefit finding in breast cancer patients. j Clin Oncol . 2002; 20( 20): 4160– 4168. Google Scholar CrossRef Search ADS PubMed  9. Milbury K, Spelman A, Wood Cet al.   Randomized controlled trial of expressive writing for patients with renal cell carcinoma. j Clin Oncol . 2014; 32( 7): 663– 670. Google Scholar CrossRef Search ADS PubMed  10. Rosenberg HJ, Rosenberg SD, Ernstoff MSet al.   Expressive disclosure and health outcomes in a prostate cancer population. Int j Psychiatry Med . 2002; 32( 1): 37– 53. Google Scholar CrossRef Search ADS PubMed  11. de Moor C, Sterner J, Hall Met al.   A pilot study of the effects of expressive writing on psychological and behavioral adjustment in patients enrolled in a Phase II trial of vaccine therapy for metastatic renal cell carcinoma. Health Psychol . 2002; 21( 6): 615– 619. Google Scholar CrossRef Search ADS PubMed  12. Henry EA, Schlegel RJ, Talley AE, Molix LA, Bettencourt BA. The feasibility and effectiveness of expressive writing for rural and urban breast cancer survivors. Oncol Nurs Forum . 2010; 37( 6): 749– 757. Google Scholar CrossRef Search ADS PubMed  13. Gellaitry G, Peters K, Bloomfield D, Horne R. Narrowing the gap: The effects of an expressive writing intervention on perceptions of actual and ideal emotional support in women who have completed treatment for early stage breast cancer. Psychooncology . 2010; 19( 1): 77– 84. Google Scholar CrossRef Search ADS PubMed  14. Zachariae R, O’Toole MS. The effect of expressive writing intervention on psychological and physical health outcomes in cancer patients–a systematic review and meta-analysis. Psychooncology . 2015; 24( 11): 1349– 1359. Google Scholar CrossRef Search ADS PubMed  15. Cameron LD, Nicholls G. Expression of stressful experiences through writing: Effects of a self-regulation manipulation for pessimists and optimists. Health Psychol . 1998; 17( 1): 84– 92. Google Scholar CrossRef Search ADS PubMed  16. Lepore SJ, Greenberg MA. Mending broken hearts: Effects of expressive writing on mood, cognitive processing, social adjustment and health following a relationship breakup. Psychol Health . 2002; 17( 5): 547– 560. Google Scholar CrossRef Search ADS   17. Ullrich PA, Lutgendorf SL. Journaling about stressful events: Effects of cognitive processing and emotional expression. Ann Behav Med . 2002; 24( 3): 244– 250. Google Scholar CrossRef Search ADS PubMed  18. Lu Q, Stanton AL. How benefits of expressive writing vary as a function of writing instructions, ethnicity and ambivalence over emotional expression. Psychol Health . 2010; 25( 6): 669– 684. Google Scholar CrossRef Search ADS PubMed  19. Lazarus RS, Folkman S. Stress, Appraisal, and Coping . New York, NY: Springer Publishing Company; 1984. 20. Warmoth K, Cheung B, You J, Yeung NCY, Lu Q. Exploring the social needs and challenges of chinese american immigrant breast cancer survivors: A qualitative study using an expressive writing approach. Int j Behav Med . 2017; 24( 6): 827– 835. Google Scholar CrossRef Search ADS PubMed  21. Lu Q, Zheng D, Young L, Kagawa-Singer M, Loh A. A pilot study of expressive writing intervention among Chinese-speaking breast cancer survivors. Health Psychol . 2012; 31( 5): 548– 551. Google Scholar CrossRef Search ADS PubMed  22. Lu Q, Wong C, Gallagher M, Tou R, Young L, Loh A. Expressive writing among Chinese American breast cancer survivors: A randomized controlled trial. Health Psychol . 2016. 23. Gross JJ, John OP. Individual differences in two emotion regulation processes: Implications for affect, relationships, and well-being. j Pers Soc Psychol . 2003; 85( 2): 348– 362. Google Scholar CrossRef Search ADS PubMed  24. Kwong SL, Chen MSJr, Snipes KP, Bal DG, Wright WE. Asian subgroups and cancer incidence and mortality rates in California. Cancer . 2005; 104( 12 Suppl): 2975– 2981. Google Scholar CrossRef Search ADS PubMed  25. Cohen J. Statistical Power Analysis for the Behavioral Sciences . 2nd ed. Hillsdale, NJ: Erlbaum; 1988. 26. Cella DF, Tulsky DS. Quality of life in cancer: Definition, purpose, and method of measurement. Cancer Invest . 1993; 11( 3): 327– 336. Google Scholar CrossRef Search ADS PubMed  27. Wan C, Zhang D, Yang Zet al.   Validation of the simplified Chinese version of the FACT-B for measuring quality of life for patients with breast cancer. Breast Cancer Res Treat . 2007; 106( 3): 413– 418. Google Scholar CrossRef Search ADS PubMed  28. Muthén LK, Muthén BO. Mplus User’s Guide . 7th ed. Los Angeles, CA: Muthén & Muthén; 1998–2015. 29. Hu LT, Bentler PM. Cutoff criteria for fit indexes in covariance structure analysis: Conventional criteria versus new alternatives. Struct Equ Model . 1999; 6( 1): 1– 55. Google Scholar CrossRef Search ADS   30. Frattaroli J. Experimental disclosure and its moderators: A meta-analysis. Psychol Bull . 2006; 132( 6): 823– 865. Google Scholar CrossRef Search ADS PubMed  31. Cella D, Eton DT, Fairclough DLet al.   What is a clinically meaningful change on the Functional Assessment of Cancer Therapy-Lung (FACT-L) Questionnaire? Results from Eastern Cooperative Oncology Group (ECOG) Study 5592. j Clin Epidemiol . 2002; 55( 3): 285– 295. Google Scholar CrossRef Search ADS PubMed  32. Merz EL, Fox RS, Malcarne VL. Expressive writing interventions in cancer patients: A systematic review. Health Psychol Rev . 2014; 8( 3): 339– 361. Google Scholar CrossRef Search ADS PubMed  33. Brucker PS, Yost K, Cashy J, Webster K, Cella D. General population and cancer patient norms for the Functional Assessment of Cancer Therapy-General (FACT-G). Eval Health Prof . 2005; 28( 2): 192– 211. Google Scholar CrossRef Search ADS PubMed  34. Rogers LQ, Courneya KS, Anton PMet al.   Effects of the BEAT Cancer physical activity behavior change intervention on physical activity, aerobic fitness, and quality of life in breast cancer survivors: A multicenter randomized controlled trial. Breast Cancer Res Treat . 2015; 149( 1): 109– 119. Google Scholar CrossRef Search ADS PubMed  35. Lu Q, Yeung NC, You J, Dai J. Using expressive writing to explore thoughts and beliefs about cancer and treatment among Chinese American immigrant breast cancer survivors. Psychooncology . 2016; 25( 11): 1371– 1374. Google Scholar CrossRef Search ADS PubMed  36. Wei M, Su JC, Carrera S, Lin SP, Yi F. Suppression and interpersonal harmony: A cross-cultural comparison between Chinese and European Americans. j Couns Psychol . 2013; 60( 4): 625– 633. Google Scholar CrossRef Search ADS PubMed  37. Hall GC, Ibaraki AY, Huang ER, Marti CN, Stice E. A Meta-analysis of cultural adaptations of psychological interventions. Behav Ther . 2016; 47( 6): 993– 1014. Google Scholar CrossRef Search ADS PubMed  38. Stanton AL, Kirk SB, Cameron CL, Danoff-Burg S. Coping through emotional approach: Scale construction and validation. j Pers Soc Psychol . 2000; 78( 6): 1150– 1169. Google Scholar CrossRef Search ADS PubMed  © The Society of Behavioral Medicine 2018 http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Annals of Behavioral Medicine Oxford University Press

Expressive Writing Intervention Improves Quality of Life Among Chinese-American Breast Cancer Survivors: A Randomized Controlled Trial

Loading next page...
 
/lp/ou_press/expressive-writing-intervention-improves-quality-of-life-among-chinese-yukUs0Spx9
Copyright
© The Society of Behavioral Medicine 2018
ISSN
0883-6612
eISSN
1532-4796
D.O.I.
10.1093/abm/kax067
Publisher site
See Article on Publisher Site

Abstract

Abstract Background Expressive writing interventions are shown to confer health benefits for Caucasian cancer survivors. However, few studies reported improved quality of life or studied ethnic minorities. Purpose The study evaluated whether a culturally sensitive expressive writing intervention improved quality of life. Methods Chinese-speaking breast cancer survivors (n = 136) in the USA were randomly assigned to one of three conditions to write three 30-min weekly essays: a cancer-fact condition to write about facts relevant to the cancer experience for three weeks; a self-regulation condition to write about deepest feelings at week 1, stress and coping at week 2, and finding benefits at week 3; or an enhanced self-regulation condition to write about stress and coping at week 1, deepest feelings at week 2, and finding benefits at week 3. Quality of life was assessed by Functional Assessment of Cancer Therapy at baseline, 1-, 3-, and 6-month follow-ups. Results Growth curve models showed that quality of life was increased in the sample from baseline to the 6-month follow-up. The enhanced self-regulation condition had a large and statistically significant effect (d = 0.90, 95% CI [0.02, 1.687]), and the self-regulation condition had a small effect (d = 0.22, 95% CI [−0.79, 1.07]) on quality of life improvement compared with the cancer-fact group. Conclusion Expressive writing is shown to be an effective intervention to improve quality of life for Chinese-American cancer survivors. Future efforts are warranted to disseminate and implement this low-dose and brief intervention in community and clinical settings. Clinical Trial Registration NCT02946619. Expressive writing, Asian, Chinese breast cancer survivors, Randomized controlled trial, Quality of life Currently an estimated 17.3 million U.S. residents identify themselves as Asian [1]. Cancer is the leading cause of death for Asian Americans. Asian-American women have the fastest growing rate of breast cancer in the nation [2]. However, few intervention studies have been conducted on Asian-American cancer survivors, despite concerns about quality of life among this population [3]. In order to reduce this health disparity, we tested the efficacy of a culturally sensitive intervention using an expressive writing paradigm to improve quality of life for Asian-American breast cancer survivors. Expressive writing is a brief intervention designed to promote health by facilitating emotional and cognitive processes. A typical expressive writing experiment randomly assigns participants to either an intervention group to write about their deepest thoughts and feelings related to a traumatic experience or a control group to write about a neutral topic unrelated to deep feelings in three 20-min sessions [4]. Reviews and a meta-analysis with 146 randomized controlled trials among non-Hispanic Whites reveal that expressive writing confers a variety of benefits, including increased physical and psychological well-being [5–7]. Multiple randomized controlled trials also demonstrate that expressive writing delivers promising effects among cancer survivors, such as improving physical symptoms [8, 9], pain [10], sleep [11], depressive symptoms [12], and perceived social support [13]. For example, one study randomly assigned 60 breast cancer patients to one of the three conditions [8]: an emotional disclosure condition to write about their deepest feelings and thoughts, a benefit finding condition to write about positive aspects of their cancer experience, and a control group to write about their cancer facts unrelated to emotions. At a 3-month follow-up, both emotional disclosure and benefit finding reduced medical appointments for cancer-related morbidities. Despite the demonstrated benefits from those studies, a recent meta-analysis reviewed 16 randomized controlled trials and concluded no overall evidence for benefits of writing in psychological, physical, or quality of life outcomes [14]. Conversely, the review did suggest that those who experience low levels of emotional support may benefit from expressive writing. The variable effects of expressive writing among cancer survivors suggest that future research should aim to study populations in need of emotional support in order to optimize the benefits of expressive writing. For example, as previous studies have shown that emotional disclosure or benefit finding alone improves health outcomes, combining emotional disclosure with benefit finding may potentially deliver more benefits. Although the effects of combination writings have rarely been tested among cancer survivors, several studies among undergraduates have examined the effects of writing instructions designed to prompt emotional disclosure in combination with other processes, such as coping, cognitive processing, and positive outlook [15–17]. These studies have demonstrated that the combination writing group improved physical health and psychological well-being compared with a control group. In order to predict whether expressive writing interventions would be beneficial for other ethnic/cultural groups and to guide the design of expressive writing instructions for culturally diverse groups, Lu and Stanton developed a self-regulation moderator model of expressive writing [18]. This model emphasizes the essential role of cognitive reappraisal of stressful events and the facilitative role of emotional disclosure in producing health benefits. Stress and coping theory [19] argues that stress influences health outcomes via appraisal. Along the same line, the self-regulation moderator model postulates cognitive reappraisal as a critical pathway in reducing the impact of stress and producing beneficial effects through expressive writing. Writing that facilitates cognitive reappraisal would bring positive change in the evaluation of stressors and/or the self. Combining emotional disclosure and cognitive reappraisal constitutes a self-regulation process that may promote maximal benefits. A recent study showed that writings combining emotional disclosure with cognitive reappraisal improved positive affect at 1-, 2-, and 4-month follow-ups for both Caucasians and Asians [18]. Writing instructions that facilitate both emotional disclosure and cognitive reappraisal could be beneficial for cancer survivors. Unfortunately, Asian-American breast cancer survivors face culturally specific obstacles that prevent them from seeking emotional support from family and friends. Such obstacles consist of stigma due to cancer and the suppression of emotions in order to bear the burden alone and avoid damaging harmony with others [20]. Expressive writing allows individuals to disclose intrapersonally without damaging harmony or provoking stigma and thus, may be especially suitable to Asian-American communities. In an effort to develop a culturally sensitive intervention, we incorporated both the communities’ input and the previous research among Asians [18]. We created self-regulation writing instructions to facilitate emotional disclosure, cognitive reappraisal, and benefit finding. The self-regulation writing intervention was shown to be culturally sensitive and associated with improved quality of life among Chinese breast cancer survivors in a pilot study [21]. In contrast, however, a recent randomized controlled trial failed to show that the self-regulation writing improved quality of life compared with cancer-fact writing [22]. The study randomly assigned Chinese-speaking breast cancer survivors to three conditions: (a) an emotional disclosure condition to write about deepest thoughts and feelings; (b) a self-regulation condition to write about deepest thoughts and feelings, stress, coping, and benefit finding from their cancer experience; and (c) a cancer-fact writing condition to write about facts related to cancer diagnosis and treatment. Surprisingly, the cancer-fact writing condition reported the highest level of overall quality of life at the 6-month follow-up. Therefore, it is imperative to determine what types of expressive writing can deliver health benefits above and beyond cancer-fact writing for Asian Americans. One explanation for lack of benefits in expressive writing is that people may not benefit from emotional disclosure if they are not ready for it. A focus group interview with Chinese-speaking breast cancer survivors suggested that the participants found it more difficult to write about emotion than writing about stress and its impact (unpublished pilot data). Prior studies found that writings facilitating cognitive reappraisal benefited Asians the most [18, 22]. Engaging in cognitive reappraisal first may be more congruent with Asian culture than engaging in emotional disclosure first because Asian culture values emotional control and restraint over emotional expression [23]. Therefore, we developed new writing instructions in the hope of facilitating cognitive reappraisal first and emotional disclosure second to enhance the benefits of expressive writing. The current study aimed to examine the impact of expressive writing on quality in a randomized controlled trial with three arms: the cancer-fact writing and self-regulation conditions tested in the previous study [22] and a new condition called enhanced self-regulation condition, which used the same writing instructions in the self-regulation condition, but the order of the first and second instructions were reversed. In the enhanced self-regulation condition, participants were asked to write about stress and coping first, then deepest feelings, and finally, positive thoughts and feelings. These writing instructions were designed to facilitate cognitive reappraisal, emotional disclosure, and benefit finding, respectively. The primary hypothesis was that the enhanced self-regulation condition would improve quality of life compared with the cancer-fact condition, and the secondary hypothesis was that the enhanced self-regulation condition would have higher quality of life compared with the self-regulation condition, which would have higher quality of life compared with the cancer-fact condition (shown in Fig. 1). Because Chinese individuals make up the largest Asian subgroup and breast cancer is the most common cancer among Chinese women [24], this study focused on testing the intervention among Chinese-speaking breast cancer survivors in the USA. Cancer survivorship research encompasses the physical and psychosocial sequelae of cancer diagnosis and its treatment, and thus, quality of life is one of the most important outcomes during cancer survivorship. Surprisingly, although expressive writing studies have shown benefits in alleviating psychological and physical distress among cancer survivors, few had reported increased quality of life. If the new writing instructions are shown to improve quality of life, the study will contribute to the general literature not only by applying the paradigm of expressive writing to an understudied population but also by offering new directions to optimize the benefits of expressive writing among cancer survivors. Fig. 1 View largeDownload slide Hypotheses: projection of quality of life by condition across time. Fig. 1 View largeDownload slide Hypotheses: projection of quality of life by condition across time. Methods Participants Participants were recruited from Los Angeles, New York, and Houston metropolitan areas in collaboration with the Herald Cancer Association, a community-based Chinese cancer organization. Potential participants were introduced to the study at cultural events, educational conferences, and support groups and were screened for eligibility. Eligibility criteria were (a) having a breast cancer diagnosis, (b) completing breast cancer surgery within 5 years, and (c) being comfortable writing and speaking Chinese (i.e., Mandarin or Cantonese). Power analysis revealed that 43 participants per group would yield 95% power to test the primary hypothesis with an estimated effect size of d = 0.8 [25] derived from a previous study [18]. To account for 5% estimated attrition, 136 were enrolled in the study (refer to flow chart in Fig. 2). Fig. 2 View largeDownload slide CONSORT chart. Fig. 2 View largeDownload slide CONSORT chart. Procedure The study was approved by several relevant institutional review boards. Participants completed the baseline questionnaire package at home and returned the baseline questionnaire by mail. A simple randomization was conducted using 1:1:1 allocation to assign participants to the three conditions. Randomization was conducted by a well-trained research staff working at the community site. Simple randomization was done by drawing one card randomly from a box and returning the card to the box after drawing. Each of the cards was labeled as 1, 2, or 3. The person who did the randomization did not know what number stood for what condition. The writing instructions were contained in large opaque sealed envelopes, which were labeled as 1, 2, or 3. Participants assigned to the “1” card would be in condition “1” and would receive writing instructions contained in the envelope labeled as “1.” Similarly, participants assigned to the “2” or “3” cards would receive the designated envelopes. Participants were told that the study aimed to understand their experiences through writing and were not made aware of whether they were assigned to an experimental or a control condition. Researchers who had interactions with participants were blinded to participants’ condition assignment. Once randomization was performed, no change in intervention allocation took place. The intervention consisted of three writing sessions starting one week after the baseline assessment. An established protocol was followed [4], with minor modifications based on prior work [18, 22]. One week after returning the baseline questionnaire, each participant received two copies of a calendar and three sealed envelopes. Participants were asked to schedule three weekly writing sessions on the calendar, keep one copy of the calendar as a reminder, and send one copy back so that researchers can remind them of the next writing session. Each envelope was labeled as “week 1”, “week 2”, and “week 3” with writing instructions for the corresponding week. Participants were asked to only open the corresponding envelope at the scheduled writing time and to write continuously for up to 30 min or until they complete one page of writing. For each session, participants were asked to write continuously for up to 30 min. Participants in the self-regulation condition were asked to write about their deepest feelings and thoughts related to their breast cancer experience at week 1, the stress and their coping strategies to deal with the stress at week 2, and positive thoughts and feelings regarding their breast cancer experience at week 3 [22]. Participants in the enhanced self-regulation condition were given the same instructions as the self-regulation condition, but the order was switched between week 1 and week 2. Participants in the cancer-fact condition were asked to write about their cancer diagnosis and treatment as objectively and as detailed as they could for 3 weeks [8, 22]. Participants wrote at home and mailed their essays back in a prepaid envelope after each writing session. They were also given reminders to improve adherence. After the last writing session, they completed manipulation check items described below. At 1, 3, and 6 months after the last writing, participants completed and mailed back follow-up questionnaires assessing quality of life. Measures Functional Assessment of Cancer Therapy general scale: FACT-G This 27-item measure assesses perceived life quality including physical, social, emotional, and functional well-being in the past week [26]. The scale was validated among Chinese breast cancer survivors [27]. The items were rated on a five-point scale from (0) not at all to (4) very much. Total scores were computed by summing all the items. The Cronbach’s alpha of the FACT-G at baseline was 0.94. Manipulation check Participants rated how personal the writings were, how much emotion they revealed in the writings, and how meaningful and valuable the study was on a seven-point scale [22]. Demographic and clinical variables (e.g., age, marital status, treatment received) were self-reported at baseline. Data Analysis Plan Analyses were conducted using Mplus 7.2 [28] as intent-to-treat analyses to include all randomized participants (n = 136). Direct maximum likelihood estimation procedures were used to accommodate missing data. We began by calculating effect sizes with 95% confidence intervals to determine the magnitude of differences in quality of life between the three conditions at each time point. Latent growth curve modeling was then used to examine the impact of the treatment condition on the intraindividual trajectories of quality of life across the four assessments. The latent growth curve was used as the primary test for study hypotheses since we were interested in the effects of the treatment condition on the overall trajectories in quality of life from baseline to the final assessment. As there were no expectations regarding the nature of the trajectories, shape factors were specified such that the loading of the slope factor was fixed at 0 for the baseline assessment, fixed at 1 for the month 6 follow-up assessment, and freely estimated for the intermediate assessments. The mean of the slopes in the latent growth curve models therefore represented the total change in quality of life across the four assessments. Two separate latent growth curves were estimated in which dummy codes representing treatment conditions were specified as predictors of the slope factor of quality of life (the cancer-fact condition was the reference condition in model 1, and the self-regulation condition was the reference condition in model 2). Results for the effects of treatment condition on changes in quality of life were presented as both unstandardized effects and partially standardized effects, which can be interpreted as an effect size akin to Cohen’s d. Model fit for the latent growth curve was evaluated based on commonly used fit indices and criteria [29]: root mean square error of approximation (RMSEA; ≤ 0.08) and the comparative fit index (CFI; ≥ 0.95). Results Sample Characteristics, Randomization, Completion of Intervention, and Attribution Recruitment was from June 2012 to April 2015. The age of the participants ranged from 34 to 84 years old (M = 57.6, SD = 9.5). Among them, 65.4% had an annual income less than $45,000, 65.7 % were married, 85.5% were diagnosed with breast cancer at stages I, II, or III, and the mean time since diagnosis was 26.4 months (SD = 19.2 months; see Table 1 for sample characteristics). Participants were randomly assigned to one of the three conditions: enhanced self-regulation condition (n = 54), self-regulation condition (n = 46), or the cancer-fact condition (n = 36). Participants in the three conditions were compared to determine the success of randomization. ANOVA and chi-square test revealed no significant differences among the conditions in demographic variables (i.e., age, education level, family income, employment, and marital status), clinical variables (i.e., stage and time since cancer diagnosis), and outcome at baseline (i.e., quality of life; p’s > .05), suggesting successful randomization. All statistical tests were two tailed with α = 0.05. Table 1 Participant demographics and clinical characteristics by group Demographic or clinical characteristics  CFC (n = 36)  SRC (n = 46)  ESRC (n = 54)  Overall (n = 136)    N  %  N  %  N  %  N  %  p  Age          .64   Mean (SD)  57.9 (8.2)  58.6 (9.5)  56.9 (9.7)  57.8 (9.2)     Missing  0  0.0  1  2.2  3  5.6  4  2.9    Years in the US                  .20   Mean (SD)  16.6 (9.1)  20.6 (10.6)  17.8(10.8)  18.5 (10.4)     Missing  2  5.6  0  0.0  0  0.0  2  1.5    Marital status:                  .84   Married  22  61.1  29  63.0  37  68.5  88  64.7     Not married  13  36.1  16  34.8  17  31.5  46  33.8     Missing  1  2.8  1  2.2  0  0.0  2  1.5    Highest education                  .99   ≥Some college  22  61.1  28  60.9  33  61.1  83  61.0     No college  14  38.9  18  39.1  20  37.0  52  38.2     Missing  0  0.0  0  0.0  1  1.9  1  0.7    Household income                  .82   ≤$15,000  13  36.1  16  34.8  17  31.5  46  33.8     $15,000–$45,000  10  27.8  13  28.3  20  37.0  43  31.6     $45,000–$75,000  5  13.9  4  8.7  4  7.4  13  9.6     >$75,000  4  11.1  9  19.6  9  16.7  22  16.2     Missing  4  11.1  4  8.7  4  7.4  12  8.8    Employment status                  .53   Working full-time  9  25.0  10  21.7  15  27.8  34  25.0     Working part-time  8  22.2  7  15.2  6  11.1  21  15.4     Retired  6  16.7  9  19.6  5  9.3  20  14.7     Housewife  6  16.7  8  17.4  15  27.8  29  21.3     Unemployed  4  11.1  9  19.6  11  20.4  24  17.6     other  3  8.3  2  4.3  1  1.9  6  4.4     Missing  0  0  1  2.2  1  1.9  2  1.5    Stage                  .10   0  3  8.3  5  10.9  7  13.0  15  11.0     1  17  47.2  10  21.7  16  29.6  43  31.6     2  13  36.1  16  34.8  17  31.5  46  33.8     3  1  2.8  11  23.9  11  20.4  23  16.9     4  0  0.0  1  2.2  3  5.6  4  2.9     Missing  2  5.6  3  6.5  0  0.0  5  3.7    Treatment                     Surgery  34  94.4  43  93.5  50  92.6  127  93.4  .94   Radiation  18  51.4  26  56.5  40  74.1  84  62.2  .06   Chemo  20  55.6  30  65.2  31  58.5  81  60.0  .65  Baseline QOL                  .42   Mean (SD)  70.3 (19.7)  72.5 (22.9)  75.9 (19.0)  73.3 (20.5)     Missing  0  0.0  0  0.0  0  0.0  0  0.0    Demographic or clinical characteristics  CFC (n = 36)  SRC (n = 46)  ESRC (n = 54)  Overall (n = 136)    N  %  N  %  N  %  N  %  p  Age          .64   Mean (SD)  57.9 (8.2)  58.6 (9.5)  56.9 (9.7)  57.8 (9.2)     Missing  0  0.0  1  2.2  3  5.6  4  2.9    Years in the US                  .20   Mean (SD)  16.6 (9.1)  20.6 (10.6)  17.8(10.8)  18.5 (10.4)     Missing  2  5.6  0  0.0  0  0.0  2  1.5    Marital status:                  .84   Married  22  61.1  29  63.0  37  68.5  88  64.7     Not married  13  36.1  16  34.8  17  31.5  46  33.8     Missing  1  2.8  1  2.2  0  0.0  2  1.5    Highest education                  .99   ≥Some college  22  61.1  28  60.9  33  61.1  83  61.0     No college  14  38.9  18  39.1  20  37.0  52  38.2     Missing  0  0.0  0  0.0  1  1.9  1  0.7    Household income                  .82   ≤$15,000  13  36.1  16  34.8  17  31.5  46  33.8     $15,000–$45,000  10  27.8  13  28.3  20  37.0  43  31.6     $45,000–$75,000  5  13.9  4  8.7  4  7.4  13  9.6     >$75,000  4  11.1  9  19.6  9  16.7  22  16.2     Missing  4  11.1  4  8.7  4  7.4  12  8.8    Employment status                  .53   Working full-time  9  25.0  10  21.7  15  27.8  34  25.0     Working part-time  8  22.2  7  15.2  6  11.1  21  15.4     Retired  6  16.7  9  19.6  5  9.3  20  14.7     Housewife  6  16.7  8  17.4  15  27.8  29  21.3     Unemployed  4  11.1  9  19.6  11  20.4  24  17.6     other  3  8.3  2  4.3  1  1.9  6  4.4     Missing  0  0  1  2.2  1  1.9  2  1.5    Stage                  .10   0  3  8.3  5  10.9  7  13.0  15  11.0     1  17  47.2  10  21.7  16  29.6  43  31.6     2  13  36.1  16  34.8  17  31.5  46  33.8     3  1  2.8  11  23.9  11  20.4  23  16.9     4  0  0.0  1  2.2  3  5.6  4  2.9     Missing  2  5.6  3  6.5  0  0.0  5  3.7    Treatment                     Surgery  34  94.4  43  93.5  50  92.6  127  93.4  .94   Radiation  18  51.4  26  56.5  40  74.1  84  62.2  .06   Chemo  20  55.6  30  65.2  31  58.5  81  60.0  .65  Baseline QOL                  .42   Mean (SD)  70.3 (19.7)  72.5 (22.9)  75.9 (19.0)  73.3 (20.5)     Missing  0  0.0  0  0.0  0  0.0  0  0.0    CFC cancer fact condition; SRC self-regulation condition; ESRC enhanced self-regulation condition; QOL Quality of Life. View Large Table 1 Participant demographics and clinical characteristics by group Demographic or clinical characteristics  CFC (n = 36)  SRC (n = 46)  ESRC (n = 54)  Overall (n = 136)    N  %  N  %  N  %  N  %  p  Age          .64   Mean (SD)  57.9 (8.2)  58.6 (9.5)  56.9 (9.7)  57.8 (9.2)     Missing  0  0.0  1  2.2  3  5.6  4  2.9    Years in the US                  .20   Mean (SD)  16.6 (9.1)  20.6 (10.6)  17.8(10.8)  18.5 (10.4)     Missing  2  5.6  0  0.0  0  0.0  2  1.5    Marital status:                  .84   Married  22  61.1  29  63.0  37  68.5  88  64.7     Not married  13  36.1  16  34.8  17  31.5  46  33.8     Missing  1  2.8  1  2.2  0  0.0  2  1.5    Highest education                  .99   ≥Some college  22  61.1  28  60.9  33  61.1  83  61.0     No college  14  38.9  18  39.1  20  37.0  52  38.2     Missing  0  0.0  0  0.0  1  1.9  1  0.7    Household income                  .82   ≤$15,000  13  36.1  16  34.8  17  31.5  46  33.8     $15,000–$45,000  10  27.8  13  28.3  20  37.0  43  31.6     $45,000–$75,000  5  13.9  4  8.7  4  7.4  13  9.6     >$75,000  4  11.1  9  19.6  9  16.7  22  16.2     Missing  4  11.1  4  8.7  4  7.4  12  8.8    Employment status                  .53   Working full-time  9  25.0  10  21.7  15  27.8  34  25.0     Working part-time  8  22.2  7  15.2  6  11.1  21  15.4     Retired  6  16.7  9  19.6  5  9.3  20  14.7     Housewife  6  16.7  8  17.4  15  27.8  29  21.3     Unemployed  4  11.1  9  19.6  11  20.4  24  17.6     other  3  8.3  2  4.3  1  1.9  6  4.4     Missing  0  0  1  2.2  1  1.9  2  1.5    Stage                  .10   0  3  8.3  5  10.9  7  13.0  15  11.0     1  17  47.2  10  21.7  16  29.6  43  31.6     2  13  36.1  16  34.8  17  31.5  46  33.8     3  1  2.8  11  23.9  11  20.4  23  16.9     4  0  0.0  1  2.2  3  5.6  4  2.9     Missing  2  5.6  3  6.5  0  0.0  5  3.7    Treatment                     Surgery  34  94.4  43  93.5  50  92.6  127  93.4  .94   Radiation  18  51.4  26  56.5  40  74.1  84  62.2  .06   Chemo  20  55.6  30  65.2  31  58.5  81  60.0  .65  Baseline QOL                  .42   Mean (SD)  70.3 (19.7)  72.5 (22.9)  75.9 (19.0)  73.3 (20.5)     Missing  0  0.0  0  0.0  0  0.0  0  0.0    Demographic or clinical characteristics  CFC (n = 36)  SRC (n = 46)  ESRC (n = 54)  Overall (n = 136)    N  %  N  %  N  %  N  %  p  Age          .64   Mean (SD)  57.9 (8.2)  58.6 (9.5)  56.9 (9.7)  57.8 (9.2)     Missing  0  0.0  1  2.2  3  5.6  4  2.9    Years in the US                  .20   Mean (SD)  16.6 (9.1)  20.6 (10.6)  17.8(10.8)  18.5 (10.4)     Missing  2  5.6  0  0.0  0  0.0  2  1.5    Marital status:                  .84   Married  22  61.1  29  63.0  37  68.5  88  64.7     Not married  13  36.1  16  34.8  17  31.5  46  33.8     Missing  1  2.8  1  2.2  0  0.0  2  1.5    Highest education                  .99   ≥Some college  22  61.1  28  60.9  33  61.1  83  61.0     No college  14  38.9  18  39.1  20  37.0  52  38.2     Missing  0  0.0  0  0.0  1  1.9  1  0.7    Household income                  .82   ≤$15,000  13  36.1  16  34.8  17  31.5  46  33.8     $15,000–$45,000  10  27.8  13  28.3  20  37.0  43  31.6     $45,000–$75,000  5  13.9  4  8.7  4  7.4  13  9.6     >$75,000  4  11.1  9  19.6  9  16.7  22  16.2     Missing  4  11.1  4  8.7  4  7.4  12  8.8    Employment status                  .53   Working full-time  9  25.0  10  21.7  15  27.8  34  25.0     Working part-time  8  22.2  7  15.2  6  11.1  21  15.4     Retired  6  16.7  9  19.6  5  9.3  20  14.7     Housewife  6  16.7  8  17.4  15  27.8  29  21.3     Unemployed  4  11.1  9  19.6  11  20.4  24  17.6     other  3  8.3  2  4.3  1  1.9  6  4.4     Missing  0  0  1  2.2  1  1.9  2  1.5    Stage                  .10   0  3  8.3  5  10.9  7  13.0  15  11.0     1  17  47.2  10  21.7  16  29.6  43  31.6     2  13  36.1  16  34.8  17  31.5  46  33.8     3  1  2.8  11  23.9  11  20.4  23  16.9     4  0  0.0  1  2.2  3  5.6  4  2.9     Missing  2  5.6  3  6.5  0  0.0  5  3.7    Treatment                     Surgery  34  94.4  43  93.5  50  92.6  127  93.4  .94   Radiation  18  51.4  26  56.5  40  74.1  84  62.2  .06   Chemo  20  55.6  30  65.2  31  58.5  81  60.0  .65  Baseline QOL                  .42   Mean (SD)  70.3 (19.7)  72.5 (22.9)  75.9 (19.0)  73.3 (20.5)     Missing  0  0.0  0  0.0  0  0.0  0  0.0    CFC cancer fact condition; SRC self-regulation condition; ESRC enhanced self-regulation condition; QOL Quality of Life. View Large Figure 2 shows the completion of the writing interventions. Among the 136 participants, 112 (82.4%) completed all three writing sessions, 3 (2.2%) completed the first two writings, and 21 (15.4%) did not complete any writing and withdrew during the writing stage. The completion of writing sessions did not differ by group (χ2 = 2.17, p = .34). Four participants withdrew during the follow-up stage. Completers and noncompleters at each follow-up did not differ in the assigned condition, demographic, medical, and outcome variables at baseline (p’s > .14). Manipulation Check As shown in Table 2, ANOVA revealed that the enhanced self-regulation condition revealed more emotion compared with the self-regulation condition, which revealed more emotion compared with the cancer-fact condition (p’s < .02). The enhanced self-regulation and cancer-fact groups reported that the study had higher meaning/value compared with the self-regulation (p < .01) group. Table 2 Analyses of variance on essay ratings at the end of writing sessions Dependent variable  CFC (n = 36)  SRC (n = 46)  ESRC (n = 54)  F  p  Essay was personal   Mean  1.83  2.61  2.55  1.33  .2693   SD  1.97  2.12  2.23      Revealed emotions   Mean  2.73ab  3.73a  4.6b  12.51  <.0001   SD  1.66  1.84  1.36      Valuable/meaningful   Mean  5.1a  4.21b  5.04a  4.95  .0088   SD  1.21  1.54  1.18      New perspective   Mean  3.23  3.00  3.64  1.22  .2999   SD  2.18  1.8  1.65      Dependent variable  CFC (n = 36)  SRC (n = 46)  ESRC (n = 54)  F  p  Essay was personal   Mean  1.83  2.61  2.55  1.33  .2693   SD  1.97  2.12  2.23      Revealed emotions   Mean  2.73ab  3.73a  4.6b  12.51  <.0001   SD  1.66  1.84  1.36      Valuable/meaningful   Mean  5.1a  4.21b  5.04a  4.95  .0088   SD  1.21  1.54  1.18      New perspective   Mean  3.23  3.00  3.64  1.22  .2999   SD  2.18  1.8  1.65      Subscripts a, b, and ab indicate significant group differences in the ANOVA. Means with different subscripts in the same row are statistically different with p < .05. CFC cancer fact condition; SRC self-regulation condition; ESRC enhanced self-regulation condition. View Large Table 2 Analyses of variance on essay ratings at the end of writing sessions Dependent variable  CFC (n = 36)  SRC (n = 46)  ESRC (n = 54)  F  p  Essay was personal   Mean  1.83  2.61  2.55  1.33  .2693   SD  1.97  2.12  2.23      Revealed emotions   Mean  2.73ab  3.73a  4.6b  12.51  <.0001   SD  1.66  1.84  1.36      Valuable/meaningful   Mean  5.1a  4.21b  5.04a  4.95  .0088   SD  1.21  1.54  1.18      New perspective   Mean  3.23  3.00  3.64  1.22  .2999   SD  2.18  1.8  1.65      Dependent variable  CFC (n = 36)  SRC (n = 46)  ESRC (n = 54)  F  p  Essay was personal   Mean  1.83  2.61  2.55  1.33  .2693   SD  1.97  2.12  2.23      Revealed emotions   Mean  2.73ab  3.73a  4.6b  12.51  <.0001   SD  1.66  1.84  1.36      Valuable/meaningful   Mean  5.1a  4.21b  5.04a  4.95  .0088   SD  1.21  1.54  1.18      New perspective   Mean  3.23  3.00  3.64  1.22  .2999   SD  2.18  1.8  1.65      Subscripts a, b, and ab indicate significant group differences in the ANOVA. Means with different subscripts in the same row are statistically different with p < .05. CFC cancer fact condition; SRC self-regulation condition; ESRC enhanced self-regulation condition. View Large Preliminary Analysis We first examined mean levels of quality of life by condition across time based on the direct maximum likelihood estimates that accounted for missing data due to attrition. Means, standard deviations, and between condition effect sizes (Cohen’s d) with 95% confidence intervals are presented in Table 3. Baseline differences in quality of life were statistically nonsignificant. An examination of the means by condition across time indicated that quality of life generally increased across time in all three conditions. Based on the confidence intervals of the effect sizes, the self-regulation condition resulted in a higher level of quality of life at the final assessment than the cancer-fact condition that was small in effect size magnitude but statistically nonsignificant. The enhanced self-regulation condition resulted in a higher level of quality of life at the final assessment than the cancer-fact condition that was moderate in effect size magnitude (d = −0.58) and statistically significant. The enhanced self-regulation condition resulted in a higher level of quality of life at the final assessment than the self-regulation condition that was small-to-moderate in effect size magnitude (d = −0.31) but statistically nonsignificant. Thus, effect size analyses suggest that the enhanced self-regulation condition may have had the most robust effects on quality of life. Table 3 Descriptive Statistics and Between Condition Effect Sizes for Quality of Life   CFC (n = 36)  SRC (n = 46)  ESRC (n = 54)    d (95% CI)    Time  M  SD  M  SD  M  SD  CFC vs. SRC  CFC vs. ESRC  SRC vs. ESRC  T1  70.28  19.44  72.48  22.61  75.94  18.82  −0.10 (−0.54: 0.33)  −0.29 (−0.73: 0.14)  −0.17 (−0.56: 0.23)  T2  73.09  20.78  76.21  16.34  78.08  17.18  −0.17 (−0.60: 0.27)  −0.26 (−0.70: 0.17)  −0.11 (−0.50: 0.28)  T3  72.13  15.96  76.85  15.91  80.71  16.34  −0.29 (−0.73: 0.14)  −0.53a (−0.97: −0.08)  −0.24 (−0.63: 0.16)  T4  72.29  18.71  77.21  17.56  82.35  15.90  −0.27 (−0.71: 0.17)  −0.58a (−1.03: −0.14)  −0.31 (−0.70: 0.09)    CFC (n = 36)  SRC (n = 46)  ESRC (n = 54)    d (95% CI)    Time  M  SD  M  SD  M  SD  CFC vs. SRC  CFC vs. ESRC  SRC vs. ESRC  T1  70.28  19.44  72.48  22.61  75.94  18.82  −0.10 (−0.54: 0.33)  −0.29 (−0.73: 0.14)  −0.17 (−0.56: 0.23)  T2  73.09  20.78  76.21  16.34  78.08  17.18  −0.17 (−0.60: 0.27)  −0.26 (−0.70: 0.17)  −0.11 (−0.50: 0.28)  T3  72.13  15.96  76.85  15.91  80.71  16.34  −0.29 (−0.73: 0.14)  −0.53a (−0.97: −0.08)  −0.24 (−0.63: 0.16)  T4  72.29  18.71  77.21  17.56  82.35  15.90  −0.27 (−0.71: 0.17)  −0.58a (−1.03: −0.14)  −0.31 (−0.70: 0.09)  Negative effect sizes indicate that the condition listed first in the comparison had lower levels of quality of life. T1, baseline; T2 1-month follow-up; T3 3-month follow-up; T4 6-month follow-up; CFC cancer fact condition; SRC self-regulation condition; ESRC enhanced self-regulation condition. aThe effect size difference is statistically significant at p < .05 based on the confidence interval of the effect size (i.e., confidence interval does not include 0). View Large Table 3 Descriptive Statistics and Between Condition Effect Sizes for Quality of Life   CFC (n = 36)  SRC (n = 46)  ESRC (n = 54)    d (95% CI)    Time  M  SD  M  SD  M  SD  CFC vs. SRC  CFC vs. ESRC  SRC vs. ESRC  T1  70.28  19.44  72.48  22.61  75.94  18.82  −0.10 (−0.54: 0.33)  −0.29 (−0.73: 0.14)  −0.17 (−0.56: 0.23)  T2  73.09  20.78  76.21  16.34  78.08  17.18  −0.17 (−0.60: 0.27)  −0.26 (−0.70: 0.17)  −0.11 (−0.50: 0.28)  T3  72.13  15.96  76.85  15.91  80.71  16.34  −0.29 (−0.73: 0.14)  −0.53a (−0.97: −0.08)  −0.24 (−0.63: 0.16)  T4  72.29  18.71  77.21  17.56  82.35  15.90  −0.27 (−0.71: 0.17)  −0.58a (−1.03: −0.14)  −0.31 (−0.70: 0.09)    CFC (n = 36)  SRC (n = 46)  ESRC (n = 54)    d (95% CI)    Time  M  SD  M  SD  M  SD  CFC vs. SRC  CFC vs. ESRC  SRC vs. ESRC  T1  70.28  19.44  72.48  22.61  75.94  18.82  −0.10 (−0.54: 0.33)  −0.29 (−0.73: 0.14)  −0.17 (−0.56: 0.23)  T2  73.09  20.78  76.21  16.34  78.08  17.18  −0.17 (−0.60: 0.27)  −0.26 (−0.70: 0.17)  −0.11 (−0.50: 0.28)  T3  72.13  15.96  76.85  15.91  80.71  16.34  −0.29 (−0.73: 0.14)  −0.53a (−0.97: −0.08)  −0.24 (−0.63: 0.16)  T4  72.29  18.71  77.21  17.56  82.35  15.90  −0.27 (−0.71: 0.17)  −0.58a (−1.03: −0.14)  −0.31 (−0.70: 0.09)  Negative effect sizes indicate that the condition listed first in the comparison had lower levels of quality of life. T1, baseline; T2 1-month follow-up; T3 3-month follow-up; T4 6-month follow-up; CFC cancer fact condition; SRC self-regulation condition; ESRC enhanced self-regulation condition. aThe effect size difference is statistically significant at p < .05 based on the confidence interval of the effect size (i.e., confidence interval does not include 0). View Large Treatment Effects: Hypothesis Testing Using Growth Curve Models We next examined the impact of the treatment condition on the longitudinal trajectories of quality of life using latent growth curve (Fig. 3). The model fit was identical and good (χ2 [df = 10, n = 136] = 16.45, RMSEA = 0.07, CFI = 0.98) for the two latent growth curve models. The β parameters from these models represent the mean differences in quality of life trajectories between conditions. In the first model in which the control condition was specified as the reference condition, the enhanced self-regulation condition had a large and statistically significant effect on quality of life trajectories (b = 6.47, 95% CI [1.56, 11.38]; d = 0.90, 95% CI [0.02, 1.687]), and the self-regulation condition had a small but statistically nonsignificant effect on quality of life trajectories (b = 1.58, 95% CI [−0.50, 8.28]; d = 0.22, 95% CI [−0.79, 1.07]). In the second model in which the self-regulation condition was specified as the reference condition, the enhanced self-regulation condition had a moderate but statistically nonsignificant effect on quality of life trajectories (b = 4.89, 95% CI [−1.04, 10.81]; d = 0.68, 95% CI [−0.03, 1.39]). Together, the results of the latent growth curve analyses indicated that quality of life improved across all three treatment conditions (b = 3.96, 95% CI [1.36, 6.57] in the unconditional latent growth curve) but that the enhanced self-regulation condition was associated with the greatest intraindividual changes in quality of life. The improvement in quality of life was larger in the enhanced self-regulation condition than that in the control condition. Fig. 3 View largeDownload slide Growth curve models: quality of life by condition across time. Fig. 3 View largeDownload slide Growth curve models: quality of life by condition across time. Discussion Expressive writing has been shown to improve psychological and physical health outcomes in both healthy and clinical populations [7, 30]. However, few studies have been conducted among ethnic minorities or shown improvement in quality of life among cancer survivors. The present study is the first to show that expressive writing improved quality of life compared with a cancer-fact control among Chinese-American breast cancer survivors using a randomized controlled trial. It suggests that expressive writing is an efficacious, brief intervention for Chinese breast cancer survivors. Confirming our primary hypothesis, the study revealed that the enhanced self-regulation condition improved quality of life compared with the cancer-fact condition. Furthermore, the increase in quality of life from baseline to the 6 month follow-up reached both statistical significance and clinical significance, as a 5–7 points change in Functional Assessment of Cancer Therapy general scale is indicative of clinical significance [31]. These findings are noteworthy because previous examinations of expressive writing interventions among non-Hispanic Whites have not reported improvements in quality of life [32]. Although the study sample lived in the USA for 18 years on average, the baseline quality of life indicated by FACT (M = 73.3) was low, compared with non-Hispanic White female cancer patients (normative value = 82.1) [33] and breast cancer survivors (M = 84.3) [34]. Perhaps, the significant unmet emotional needs among Chinese-speaking breast cancer survivors [35] render this underserved population more likely to experience the benefits of an intervention, suggesting the importance of conducting intervention studies in underserved populations both from scientific and clinical perspectives. Reporting improvement in quality of life for the first time among cancer survivors could also be attributed to the addition of the new writing condition, which is different from the traditional emotional disclosure condition typically used in previous expressive writing studies among non-Hispanic White cancer survivors. The condition that improved quality of life the most was the enhanced self-regulation condition, which was designed to facilitate cognitive reappraisal, emotional disclosure, and benefit finding in that order, and delivered maximum health benefits. In contrast, the self-regulation condition, which was designed to facilitate disclosing emotion, cognitive reappraisal, and benefit finding in that order, only had small and nonsignificant effects compared with the cancer-fact writing. Thus, engaging in cognitive processing first is more beneficial than engaging in emotional disclosure first for the study population. This is in line with prior findings that cognitive reappraisal benefited East Asians the most [18, 22]. Prior work did not find that writing interventions facilitating emotional disclosure alone delivered benefits for Asians [18, 22], and one may conclude that Asians would not benefit from emotional disclosure. This study clarified potential misperception that Asians would not benefit from emotional disclosure, and takes a step further to highlight the importance of ordering effects of writing instructions, specifically, prompting cognitive reappraisal before emotional disclosure for Asians. East Asian culture values emotional control and restraint over emotional expression [23, 36], and thus, Chinese-American cancer survivors may find it unnatural to be asked to dive deeply into their emotions before having the opportunity to reflect on the stressful event that caused their negative emotions. Interestingly, the enhanced self-regulation condition was associated with greater emotional disclosure, suggesting that disclosing stress helped them be better prepared to engage in emotion disclosure. Our findings suggest that starting with “what happened to you and how you were impacted,” as the enhanced self-regulation condition did, rather than “what/how did you feel” is a culturally sensitive approach for addressing distress. Our findings challenge an implicit assumption that expressive writing interventions should always begin with an emotional disclosure. The stimulating findings shed light on the importance of developing and testing culturally based interventions congruent with cultural norms and practices for minority breast cancer survivors. This supports the argument that interventions for minorities need to consider cultural adaptation [37]. Although cultural norms of emotion suppression help to explain the ordering effects, it is premature to conclude that culture is the only explanation. It is also possible that the facilitation of cognitive reappraisal in the enhanced self-regulation condition during the first writing session sets up a positive tone for or gives more purpose to the second session of emotional disclosure. For example, the cognitive reappraisal writing instruction asked people to answer the following questions: “What kind of changes in your life have you made as a result of this experience?” “What things are you able to do better now?” This may prompt a growth mindset, which helps to steer the later session of emotional disclosure toward a more meaningful direction, rather than simply venting. This speculation was supported by the data that the enhanced self-regulation condition was rated more meaningful than the self-regulation condition. Engaging in cognitive reappraisal first and emotional disclosure second may help to form a more comprehensive story than the other way around, as forming a comprehensive story is proposed to be a key ingredient in producing the benefits of writing [5]. If that is the case, findings of the order effect may not be limited to Chinese Americans. This important ordering effect is the first reported in the expressive writing literature, perhaps due to the lack of a similar study design in previous studies. Expressive writing studies among non-Hispanic White cancer survivors either used emotional disclosure alone [8, 11] or used emotional disclosure prompts first prior to other writing prompts [13]. Similarly, studies using a combination of writing instructions among general populations usually began with giving emotional disclosure prompts first [15–17]. None had begun with cognitive reappraisal and switched to emotional disclosure later. The prevailing design preconception reflects an assumption that understanding emotions needs to happen first to understand stressors. Although the ordering effect has not been tested in the past, it does not preclude the possibility that disclosing stress first enhances the benefits of expressive writing for other ethnic groups including non-Hispanic Whites. A prior study showed that cognitive processing resulting from emotional disclosure writing was linked to improvements in physical symptoms among non-Hispanic Whites [9]. Furthermore, men are found to use less emotional processing and emotional expression than women [38]. Perhaps, men would benefit more from engaging in stress disclosure first than engaging in emotional disclosure first. The ordering effect should be tested for its generalizability in future, and the new writing instructions may prove to be fruitful for other ethnic groups and men as well. Several caveats exist. First, the cancer-fact condition was associated with improved quality of life in a previous study [22]. The self-regulation condition had a small and nonsignificant improvement in quality of life compared with the cancer-fact condition. The study was not powered to detect the difference. It is also possible that the self-regulation condition may have larger benefits compared with a nonwriting assessment only group, and this possibility should be tested in the future. In fact, previous expressive writing studies with cancer survivors suggest that the studies with a nonwriting control group were more likely to find benefits of writing [12, 13]. If that is the case, the effects of the enhanced self-regulation condition are potentially larger than were detected here. Second, the study was subject to self-selection bias and was limited to Chinese-speaking breast cancer survivors. Future studies need to examine other Asian subgroups and other types of cancer. Third, the baseline quality of life score seemed to be higher in the enhanced self-regulation condition; however, it was not statistically different from the other two groups. Nonetheless, it is still possible that those who have a higher quality of life to begin with may benefit more from expressive writing. Future studies examining this possibility require a larger sample. Fourth, the study focused on a sample that had completed cancer treatment, so it remains unclear what kind of writing instructions may benefit those undergoing treatment. It is possible that those who are undergoing treatment may have more negative emotions to reveal and thus benefit from different types of writing instructions. Finally, only self-report measures were used; hence, future studies assessing biomarkers will add to our understanding about the underlying biological mechanism. Despite the limitations, this study has several strengths, including an understudied population, an experimental method, and long-term follow-ups. The study was the first to demonstrate that the expressive writing improved quality of life compared with a fact writing condition using a randomized controlled trial among a subgroup of Asian-American cancer survivors. It shed light onto the oncology literature relevant to an underserved and growing patient population. Expressive writing is shown to be an effective brief intervention for Chinese-American cancer survivors. Specifically, Asians benefited from writing instructions emphasizing cognitive aspects first before emotional disclosure. This culturally tailored intervention challenges the implicit assumption that psychosocial interventions validated among non-Hispanic Whites could be directly generalized to other populations. Future research is warranted to test this intervention and various writing instructions for other ethnic groups in the USA and in other countries. Expressive writing allows people with limited English proficiencies to express and regulate their feelings and thoughts privately in their primary language. It is a relatively low-dose and feasible intervention model that could be easily disseminated and implemented in community and clinical settings. Acknowledgements The study was supported by American Cancer Society MRSGT-10-011-01-CPPB (PI: Q.L.). The authors would like to thank the anonymous reviewers and colleagues (Lara Traeger, William Tsai, Ivan Wu) for their constructive comments and Sidra Deen for proofreading. Compliance with Ethical Standards Authors’ Statement of Conflict of Interest and Adherence to Ethical Standards Authors do not have conflict of interests to disclose. The views expressed in the submitted article are his or her own and not an official position of the institution or funder. Ethical Approval All procedures performed in studies involving human participants were in accordance with the ethical standards of the institutional and/or national research committee and with the 1964 Helsinki declaration and its later amendments or comparable ethical standards. Informed Consent Informed consent was obtained from all individual participants included in the study. References 1. U.S. Census Bureau. Annual estimates of the resident population by sex, race, and hispanic origin for the United States, States, and Counties: April 1, 2010 to July 1, 2014. 2015; Available at http://factfinder.census.gov/faces/tableservices/jsf/pages/productview.xhtml?src=bkmk. Accessibility verified January 12, 2017 2. Gomez SL, Quach T, Horn-Ross PLet al.   Hidden breast cancer disparities in Asian women: Disaggregating incidence rates by ethnicity and migrant status. Am j Public Health . 2010; 100( Suppl 1): S125– S131. Google Scholar CrossRef Search ADS PubMed  3. Wen KY, Fang CY, Ma GX. Breast cancer experience and survivorship among Asian Americans: A systematic review. j Cancer Surviv . 2014; 8( 1): 94– 107. Google Scholar CrossRef Search ADS PubMed  4. Pennebaker JW, Beall SK. Confronting a traumatic event: Toward an understanding of inhibition and disease. j Abnorm Psychol . 1986; 95( 3): 274– 281. Google Scholar CrossRef Search ADS PubMed  5. Pennebaker JW. Writing about emotional experiences as a therapeutic process. Psychol Sci . 1997; 8( 3): 162– 166. Google Scholar CrossRef Search ADS   6. Smyth JM. Written emotional expression: Effect sizes, outcome types, and moderating variables. j Consult Clin Psychol . 1998; 66( 1): 174– 184. Google Scholar CrossRef Search ADS PubMed  7. Frisina PG, Borod JC, Lepore SJ. A meta-analysis of the effects of written emotional disclosure on the health outcomes of clinical populations. j Nerv Ment Dis . 2004; 192( 9): 629– 634. Google Scholar CrossRef Search ADS PubMed  8. Stanton AL, Danoff-Burg S, Sworowski LAet al.   Randomized, controlled trial of written emotional expression and benefit finding in breast cancer patients. j Clin Oncol . 2002; 20( 20): 4160– 4168. Google Scholar CrossRef Search ADS PubMed  9. Milbury K, Spelman A, Wood Cet al.   Randomized controlled trial of expressive writing for patients with renal cell carcinoma. j Clin Oncol . 2014; 32( 7): 663– 670. Google Scholar CrossRef Search ADS PubMed  10. Rosenberg HJ, Rosenberg SD, Ernstoff MSet al.   Expressive disclosure and health outcomes in a prostate cancer population. Int j Psychiatry Med . 2002; 32( 1): 37– 53. Google Scholar CrossRef Search ADS PubMed  11. de Moor C, Sterner J, Hall Met al.   A pilot study of the effects of expressive writing on psychological and behavioral adjustment in patients enrolled in a Phase II trial of vaccine therapy for metastatic renal cell carcinoma. Health Psychol . 2002; 21( 6): 615– 619. Google Scholar CrossRef Search ADS PubMed  12. Henry EA, Schlegel RJ, Talley AE, Molix LA, Bettencourt BA. The feasibility and effectiveness of expressive writing for rural and urban breast cancer survivors. Oncol Nurs Forum . 2010; 37( 6): 749– 757. Google Scholar CrossRef Search ADS PubMed  13. Gellaitry G, Peters K, Bloomfield D, Horne R. Narrowing the gap: The effects of an expressive writing intervention on perceptions of actual and ideal emotional support in women who have completed treatment for early stage breast cancer. Psychooncology . 2010; 19( 1): 77– 84. Google Scholar CrossRef Search ADS PubMed  14. Zachariae R, O’Toole MS. The effect of expressive writing intervention on psychological and physical health outcomes in cancer patients–a systematic review and meta-analysis. Psychooncology . 2015; 24( 11): 1349– 1359. Google Scholar CrossRef Search ADS PubMed  15. Cameron LD, Nicholls G. Expression of stressful experiences through writing: Effects of a self-regulation manipulation for pessimists and optimists. Health Psychol . 1998; 17( 1): 84– 92. Google Scholar CrossRef Search ADS PubMed  16. Lepore SJ, Greenberg MA. Mending broken hearts: Effects of expressive writing on mood, cognitive processing, social adjustment and health following a relationship breakup. Psychol Health . 2002; 17( 5): 547– 560. Google Scholar CrossRef Search ADS   17. Ullrich PA, Lutgendorf SL. Journaling about stressful events: Effects of cognitive processing and emotional expression. Ann Behav Med . 2002; 24( 3): 244– 250. Google Scholar CrossRef Search ADS PubMed  18. Lu Q, Stanton AL. How benefits of expressive writing vary as a function of writing instructions, ethnicity and ambivalence over emotional expression. Psychol Health . 2010; 25( 6): 669– 684. Google Scholar CrossRef Search ADS PubMed  19. Lazarus RS, Folkman S. Stress, Appraisal, and Coping . New York, NY: Springer Publishing Company; 1984. 20. Warmoth K, Cheung B, You J, Yeung NCY, Lu Q. Exploring the social needs and challenges of chinese american immigrant breast cancer survivors: A qualitative study using an expressive writing approach. Int j Behav Med . 2017; 24( 6): 827– 835. Google Scholar CrossRef Search ADS PubMed  21. Lu Q, Zheng D, Young L, Kagawa-Singer M, Loh A. A pilot study of expressive writing intervention among Chinese-speaking breast cancer survivors. Health Psychol . 2012; 31( 5): 548– 551. Google Scholar CrossRef Search ADS PubMed  22. Lu Q, Wong C, Gallagher M, Tou R, Young L, Loh A. Expressive writing among Chinese American breast cancer survivors: A randomized controlled trial. Health Psychol . 2016. 23. Gross JJ, John OP. Individual differences in two emotion regulation processes: Implications for affect, relationships, and well-being. j Pers Soc Psychol . 2003; 85( 2): 348– 362. Google Scholar CrossRef Search ADS PubMed  24. Kwong SL, Chen MSJr, Snipes KP, Bal DG, Wright WE. Asian subgroups and cancer incidence and mortality rates in California. Cancer . 2005; 104( 12 Suppl): 2975– 2981. Google Scholar CrossRef Search ADS PubMed  25. Cohen J. Statistical Power Analysis for the Behavioral Sciences . 2nd ed. Hillsdale, NJ: Erlbaum; 1988. 26. Cella DF, Tulsky DS. Quality of life in cancer: Definition, purpose, and method of measurement. Cancer Invest . 1993; 11( 3): 327– 336. Google Scholar CrossRef Search ADS PubMed  27. Wan C, Zhang D, Yang Zet al.   Validation of the simplified Chinese version of the FACT-B for measuring quality of life for patients with breast cancer. Breast Cancer Res Treat . 2007; 106( 3): 413– 418. Google Scholar CrossRef Search ADS PubMed  28. Muthén LK, Muthén BO. Mplus User’s Guide . 7th ed. Los Angeles, CA: Muthén & Muthén; 1998–2015. 29. Hu LT, Bentler PM. Cutoff criteria for fit indexes in covariance structure analysis: Conventional criteria versus new alternatives. Struct Equ Model . 1999; 6( 1): 1– 55. Google Scholar CrossRef Search ADS   30. Frattaroli J. Experimental disclosure and its moderators: A meta-analysis. Psychol Bull . 2006; 132( 6): 823– 865. Google Scholar CrossRef Search ADS PubMed  31. Cella D, Eton DT, Fairclough DLet al.   What is a clinically meaningful change on the Functional Assessment of Cancer Therapy-Lung (FACT-L) Questionnaire? Results from Eastern Cooperative Oncology Group (ECOG) Study 5592. j Clin Epidemiol . 2002; 55( 3): 285– 295. Google Scholar CrossRef Search ADS PubMed  32. Merz EL, Fox RS, Malcarne VL. Expressive writing interventions in cancer patients: A systematic review. Health Psychol Rev . 2014; 8( 3): 339– 361. Google Scholar CrossRef Search ADS PubMed  33. Brucker PS, Yost K, Cashy J, Webster K, Cella D. General population and cancer patient norms for the Functional Assessment of Cancer Therapy-General (FACT-G). Eval Health Prof . 2005; 28( 2): 192– 211. Google Scholar CrossRef Search ADS PubMed  34. Rogers LQ, Courneya KS, Anton PMet al.   Effects of the BEAT Cancer physical activity behavior change intervention on physical activity, aerobic fitness, and quality of life in breast cancer survivors: A multicenter randomized controlled trial. Breast Cancer Res Treat . 2015; 149( 1): 109– 119. Google Scholar CrossRef Search ADS PubMed  35. Lu Q, Yeung NC, You J, Dai J. Using expressive writing to explore thoughts and beliefs about cancer and treatment among Chinese American immigrant breast cancer survivors. Psychooncology . 2016; 25( 11): 1371– 1374. Google Scholar CrossRef Search ADS PubMed  36. Wei M, Su JC, Carrera S, Lin SP, Yi F. Suppression and interpersonal harmony: A cross-cultural comparison between Chinese and European Americans. j Couns Psychol . 2013; 60( 4): 625– 633. Google Scholar CrossRef Search ADS PubMed  37. Hall GC, Ibaraki AY, Huang ER, Marti CN, Stice E. A Meta-analysis of cultural adaptations of psychological interventions. Behav Ther . 2016; 47( 6): 993– 1014. Google Scholar CrossRef Search ADS PubMed  38. Stanton AL, Kirk SB, Cameron CL, Danoff-Burg S. Coping through emotional approach: Scale construction and validation. j Pers Soc Psychol . 2000; 78( 6): 1150– 1169. Google Scholar CrossRef Search ADS PubMed  © The Society of Behavioral Medicine 2018

Journal

Annals of Behavioral MedicineOxford University Press

Published: Jan 25, 2018

There are no references for this article.

You’re reading a free preview. Subscribe to read the entire article.


DeepDyve is your
personal research library

It’s your single place to instantly
discover and read the research
that matters to you.

Enjoy affordable access to
over 18 million articles from more than
15,000 peer-reviewed journals.

All for just $49/month

Explore the DeepDyve Library

Search

Query the DeepDyve database, plus search all of PubMed and Google Scholar seamlessly

Organize

Save any article or search result from DeepDyve, PubMed, and Google Scholar... all in one place.

Access

Get unlimited, online access to over 18 million full-text articles from more than 15,000 scientific journals.

Your journals are on DeepDyve

Read from thousands of the leading scholarly journals from SpringerNature, Elsevier, Wiley-Blackwell, Oxford University Press and more.

All the latest content is available, no embargo periods.

See the journals in your area

DeepDyve

Freelancer

DeepDyve

Pro

Price

FREE

$49/month
$360/year

Save searches from
Google Scholar,
PubMed

Create lists to
organize your research

Export lists, citations

Read DeepDyve articles

Abstract access only

Unlimited access to over
18 million full-text articles

Print

20 pages / month

PDF Discount

20% off