«The Mediating Effect of Acculturation on the Effectiveness of Culturally Adapted Cognitive Behavioral Therapy with Mexican Americans Suffering From ...»
plan that was used in this study.
Research question #1: “Is there a relationship between acculturation level and depression scores among Mexican Americans?” was answered using independent t test and simple linear regression analyses. Research question #2: “Is CACBT effective in reducing depression scores among Mexican Americans?” was answered using independent and paired t-test and simple linear regression. Research question #3: “Does acculturation level mediate the effectiveness of CACBT treatment?” was answered using OLS multiple regression.
adapted cognitive behavioral therapy varies according to acculturation level in Mexican American adults diagnosed with depression. To investigate this problem, the study was divided into three parts. First, the relationship between acculturation level and pretest depression scores among Mexican Americans being treated with CACBT was studied.
Second, the study examined whether CACBT was effective in reducing posttest depression scores. Finally, analyses examined the mediating role of acculturation on the effectiveness of CACBT treatment. This chapter begins with descriptive and correlation analysis of the data.
DESCRIPTIVE ANALYSESDescriptive analyses included running a set of analyses to examine differences between treatment and comparison groups on selected demographic, cultural and clinical characteristics. In addition, a set of analyses was run to examine differences between participants and nonparticipants on selected demographic, cultural and clinical characteristics. Finally, descriptive statistics considered the frequency distribution of demographic, cultural and clinical characteristics for the treatment and comparison groups.
In order to rule out sampling biases between the treatment and comparison groups, a set of analyses examined differences for selected demographic (age, gender, and education), cultural (acculturation and generation status), and clinical characteristics (psychiatric medications, previous therapy, medical conditions, and comorbid psychiatric diagnosis). No significant differences were found between the groups on demographic, cultural or clinical characteristics. Further analyses examined the differences between participants and nonparticipants (due to attrition, non eligibility, or incomplete data).
These analyses examined differences for selected demographic (age, gender, and education), cultural (acculturation and generation status), and clinical characteristics (psychiatric medications, previous therapy, medical conditions, and comorbid psychiatric diagnosis). No significant differences were found between participants and nonparticipants.
Demographic Characteristics Table 5-1 details the demographic characteristics of the sample, including gender, education, age and marital status. The sample consisted predominantly of female participants (85.2%). The treatment group was comprised of 87.5% female participants and the comparison group was comprised of 81.8% female participants. With regard to level of education, 18.5% reported having less than a high school diploma, 24.7% reported having their high school diploma or GED, 23.5% reported having some college, and 9.9% reported having a college degree. The education distribution for the sample
Age 18 – 29 30.9% (25) 27.4% (13) 36.4% (12) 30 – 39 21.0% (17) 22.9% (11) 18.2% (6) 40 – 49 19.8% (16) 18.8% (9) 21.2% (7) 50+ 28.4% (23) 31.2% (15) 24.2% (8) Marital Status
reported in the 2000 U.S. Census Bureau (U.S. Census Bureau, 2000). The mean age of the total participants for this study was 41. The mean age for participants in the treatment group was 41. The mean age for participants in the comparison group was 37. With regard to age percentages, 30.9% of the participants reported being 18 to 29, while 28.4% reported being older than 50. The largest age group for the treatment group was 50+ (31.2%) and the largest age group for the comparison group was 18 to 29 (36.4%). In addition, 23.5% of the participants reported being married or cohabitating, 32.1% reported being separated or divorced, and 44.4% reported being single including widows or widowers. The largest marital status group for the treatment group was single (47.9%) and the largest marital status group for the comparison group was separated (42.4%).
Cultural Characteristics Table 5-2 provides descriptive statistics for measures related to culture included in the study: acculturation level, generation status, and language preference. As shown in Table 5-2, approximately 38% of the participants scored as being very Mexican oriented (acculturation Level I) on the ARSMA-II, while approximately 28% scored as being Mexican oriented to approximately balanced bicultural (acculturation Level II). The smallest percentage of the participants (9%) scored in the slightly Anglo oriented (acculturation Level IV) or very assimilated/Anglicized (acculturation level V) of the ARSMA-II. On average, participants of the treatment group were lower acculturated than participants of the comparison group (50% versus 21% acculturation level I). These
participants reported being born in Mexico (First Generation) and approximately 33% reported being born in the United States with either parent being born in Mexico or another country (Second Generation). Approximately 17% of the participants reported being third, fourth or fifth generation U.S.-born. A higher percentage of participants of the treatment group were of First Generation status than of the comparison group (54% versus 39%). When asked to select language preference for therapy, approximately 42% of the participants preferred therapy in English, approximately 56% preferred therapy in Spanish, while approximately 3% preferred therapy in both English and Spanish. A higher percentage of treatment group participants (65%) preferred therapy in Spanish than comparison group participants (42%). The high number of participants preferring therapy in Spanish may be due to large numbers of predominantly Spanish-speaking participants (Sue et al., 1991) or a large number of participants who preferred therapy in their native language because they prefer to express emotional content in their native language (Altarriba & Bauer, 1998; Altarriba & Santiago-Rivera, 1994).
Table 5-3 details the descriptive statistics for the clinical variables, including:
PHQ-9 score, psychotropic medications, and previous therapy. On average, study participants tended to be moderately depressed (Mean=15) for both the treatment group and comparison group. Approximately 20% of the participants had mild depression, while 26% had severe depression. A higher percentage of comparison group participants reported mild depression (24% versus 17%). A higher percentage of treatment group participants reported severe depression than comparison group participants (29% versus 21%). With regard to taking psychotropic medication at the time of their first therapy session, approximately 27% of the participants reported taking psychotropic medications, while approximately 68% reported they were not taking any psychotropic medications.
Treatment group participants and comparison group participants reported use of psychotropic medications at approximately the same rates. Approximately 40% of the participants reported having a history of previous therapy, while approximately 56% reported never participating in therapy. A higher percentage of comparison group participants reported history of previous therapy (49%) than treatment group participants (33%).
Most of the participants had below a college education. In addition, most of the participants were single. Overall this sample was similar to the demographic characteristics of the general Mexican American population. Related to cultural characteristics, study participants tended to be predominantly low acculturated, with 67% scoring as being Level I and Level II acculturation on the ARSMA-II. A higher percentage of treatment group participants were of Level I and Level II acculturation (75%) than comparison group participants (55%). Finally, more study participants stated they preferred to receive therapy in Spanish (56%), with the largest percentage being in the treatment group (65%).
The clinical profile of study participants was that a higher percentage was moderately depressed (31%). There was no significant difference in mean depression scores between treatment and comparison group. In the overall sample, 27% reported taking psychotropic medications at the onset of treatment. Treatment group participants and comparison group participants reported taking psychotropic medications at about the same rates. In addition, more than half of the participants reported never having previous therapy.
CORRELATION ANALYSISA set of correlation analyses were run to test the strength of the association between independent variables and the dependent variable. Pearson correlation analysis was used for continuous measure variables. Spearman Rho correlation analysis was used
Table 5-4 reveals patterns that emerged from these correlation analyses. Related to acculturation, Table 5-4 shows that study participants in the comparison group tended to be more acculturated than study participants from the treatment group (rs=.258, p.05).
In addition, older study participants were less acculturated (r=-.434, p.01). As expected, those with a higher generation status were more acculturated (rs=.721, p.01). Related to generation status, comparison group participants were more likely to be of higher generation status (rs=.268, p.05). Younger participants were also more likely to be of higher generation status (rs=-.222, p.05). Moreover, language preference seems to be an important factor. For example, older participants were more likely to prefer therapy in Spanish (rs=.281, p.05). Lower acculturated participants were more likely to prefer therapy in Spanish (rs=-.666, p.01) and participants with higher education were more likely to prefer therapy in English (rs=-.231, p.05). In addition, participants who preferred therapy in Spanish were less likely to report having a history of previous therapy (rs=.347, p.05).
Several significant correlations were associated with the clinical variables.
Related to age, younger participants were less likely to use psychotropic medications (rs=-.342, p.05) and had more education (rs=-.283, p.05). Several correlations were found related to reporting a history of previous therapy. Those participants who reported using psychotropic medications were more likely to report a history of previous therapy
Research Question 1: Effect of Acculturation on Depression The first research question examines whether there is a relationship between acculturation level and depression scores among Mexican Americans. An independent ttest and simple linear regression analyses found no significant relationship between pretest depression scores and level of acculturation.
Table 5-5 considers the direct effect between acculturation and pretest depression scores. Model 1 shows there was no significant relationship between acculturation and depression scores (R2=0.00, F=0.01). In fact, this model does not explain any of the variance in pretest depression scores.
Model 2 includes raw acculturation scores, demographic predictors age (40-49) and marital status (married). These predictors were selected for inclusion due to the significance found in the correlation matrix. For the predictors age and marital status, only the category that was found to have a significant relationship was loaded into the model. The addition of these selected demographic predictors to the main independent variable for acculturation in Model 2 does not significantly improve the amount of variance explained in posttest depression scores (R2=0.00, F=1.05). That is, acculturation is not significantly related to depression, controlling for selected demographic predictors.
Model 3 tested the direct effect of selected cultural predictors, controlling for acculturation, and age. Model 3 included the main independent variable of acculturation, the demographic predictor age (40-49) and the predictors generation status (First Gen)
preference, only the category that was significantly correlated was loaded into the model due to the limited number of allowable predictors (Tabachnick & Fidell, 1983). The inclusion of cultural predictors in Model 3 does not significantly improve the amount of variance explained in posttest depression scores (R2=0.04, F=0.73).