The present study is a quasi-experimental pretest-posttest with a control group. The statistical population of the study was overweight women in Kangavar, Iran, who had referred to the sports clubs and health centers in 2019. According to previous studies and considering α = 0.05 and β = 0.20, the sample size for each group was determined as 15, considering the possible loss of 20%. Using the Random Allocation Software, 30 subjects who met the inclusion criteria were randomly divided into two equal groups of intervention and control. The allocation of subjects was performed by a blinded person not involved in any stages of the study. Also, subjects and data analyzers were blind to allocation.
Inclusion criteria for entering the research were: willingness to participate in research, female gender, age range 50 - 18 years, BMI ≥ 28, a minimum of basic literacy, and the lack of psychological disorders requiring medication based on the fifth edition of the diagnostic and statistical manual of mental disorders (DSM-5). Exclusion criteria included non-response to more than 20% of the questions, absence from more than two training sessions, and simultaneous participation in other treatment programs.
This study was based on a combination of cognitive-behavioral therapy guidelines, acceptance and commitment, and dialectical behavior therapy. The necessary permission was obtained from the ethics committee of the Islamic Azad University of Sanandaj. To conduct the study, the study aims were explained to all participants, and a written consent was obtained. We also assured the confidentiality of the participants’ information and requested the participants to read the questions carefully and select the answer that seems more appropriate.
The participants in the intervention group received ten sessions (90-minute) of eating disorder-based mental health training program once a week by the researcher who had previously participated in the relevant training courses, but the control group received no intervention. After the sessions, post-test was performed in both groups. The content of this training program is briefly presented in
Table 1. To measure the positive and negative components of mental health, we used the 28-item Goldberg General Health Questionnaire (GHQ) and the nine-item Lutz Positive Mental Health Questionnaire (PMHQ) (
18,
19). The 28-question form of the Mental Health Questionnaire was designed by implementing factor analysis on its long form (
20). The questions in this questionnaire examine a person's mental state in the last month and include symptoms such as abnormal thoughts and feelings and aspects of observable behavior that emphasize the here and now situation. The 28-item GHQ consists of four sub-tests, each of which has seven questions. The questions of each subtest are listed in order as follows: questions 1 to 7 are related to the physical symptoms subtest; questions 8 to 14 are related to the anxiety and insomnia subtest; questions 15 to 21 are related to the social dysfunction subtest; and questions 22 to 28 are related to the depression subtest. All items of the GHQ have four options and there are two types of scoring methods for these options. In the first method, the test options are scored as (1, 1, 0, and 0), and as a result, the individual score will vary from zero to 28. The second method is the Likert scoring method, according to which the test options are scored as (1, 2, 3, and 4), and as a result, a person's total score will vary from zero to 84. In both methods of scoring, a lower score indicates better mental health. Goldberg and Williams (
18) reported that the reliability of this questionnaire by the split–half method was 0.95. After administering the questionnaire to 72 students in Hong Kong, Chan reported an internal consistency coefficient of the questionnaire using the Cronbach's alpha method of 0.93 (
21). Robinson and Price asked 103 patients who had previously had a heart attack to complete a GHQ twice at 8-month intervals (
20). Analysis of the results showed a reliability coefficient of 0.90. Taghavi evaluated the reliability of the GHQ based on three methods of test–retest, split-half, and Cronbach's alpha, which obtained reliability coefficients of 0.93, 0.70, and 0.90, respectively (
22). In this study, the reliability of the questionnaires by Cronbach's alpha was calculated as 0.7. Also, concurrent validity and factor analysis methods were used to evaluate the validity of the mental health questionnaire.
3.1. Positive Mental Health Questionnaire
This nine-item questionnaire was first developed by Lukat et al. (
19). The questions, according to the general definition of positive mental health, are behavior-oriented and personal-oriented. Person-oriented questions emphasize the stability of a person's overall personality patterns in many situations, while behavior-oriented questions emphasize the pattern of person's behavior in different situations. It also measures internal factors such as emotional factors and psychological factors compared to external factors such as social support and cooperation. The final version, presented by Lukat et al. (
19) is reduced to nine questions, rated on a Likert scale from 1 (incorrect) to 4 (correct). The positive mental health scale has a positive correlation with life satisfaction (r = 0.75) because high scores indicate higher satisfaction. The reliability of test-retest in samples 1, 2, and 3 was 0.81, 0.77, and 0.74, respectively, and Cronbach's alpha coefficient between different groups was between 0.82 and 0.93. Discriminant validity was also evaluated and confirmed (
19). The validity of this questionnaire was confirmed in Iran by Naghavi et al. (
23).
To analyze the data, SPSS software version 21 and to remove the pre-test effect, MBox, Lambday Wilkes, univariate and multivariate covariance were used. A P-value < 0.05 indicated a statistically significant difference.