Validity and reliability study of diabetes fatalism scale in turkish patients with type 2 diabetes

Background. Turkey is one of the them that has the highest prevalence in Europe, with about one in every seven adults diagnosed diabetes mellitus. By 2035, Turkey will have the highest number of people with type 2 diabetes in Europe, at almost 12 million. Mortality rates have increased with the increasing prevalence of type 2 diabetes, especially in the younger population, such that half of the deaths come from those under sixty. The beliefs and mental state of patients with chronic illnesses like diabetes can affect disease outcomes and the patients’ self-management. Self-care and diabetes medications are important components in improving the di sease outcome, though many studies have shown that these activities can be negatively related to fatalism about the disease state. The aim of this study was to investigate the reliability and validity of the Turkish version of the Diabetes Fatalism Scale (DFS), which was developed by Egede. Materials and methods. This was a methodologi cal study. The scales were administered to a total of 139 patients with type 2 diabetes. The content and construct validity of the scale were assessed. The construct validity was evaluated using confirmatory factor analysis (CFA), and the reliability was assessed in terms of internal consistency. Results. In terms of the population tested, 54.7 % of the participants were men, 73.4 % were married, 54 % had another disease, 18 % were high school graduates, the average age was 50.20 ± 16.82 years, the average duration of diabetes was 19.31 ± 14.25, and mean glycated hemoglobin (HbA1c) level was 7.06 ± 0.65 %. It was found that the Kaiser-Meyer-Olkin (KMO) measure of sampling-size adequacy was 0.770, indicating an adequate size, and the chi-square value was 1078.402. When the fifth item was excluded from the study and the analysis was repeated, the KMO coefficient was 0.802 and the chi-square value was 1020.244, p = 0.000. The Cronbach’s alpha value reached 0.806, indicating a good internal consistency. The Cronbach’s alpha values of the other subscales also seemed to be at a very good level. Conclusions. Our study showed that the DFS is a valid and reliable scale for the Turkish society. DFS-T is a suitable scale for health professionals to use to assess the fatalism of diabetic patients in Turkey.


Introduction
Type 2 diabetes mellitus (DM) is one of the global chronic disease has been increasing every last decades, now 1/7 adults is living with DM and 3/4 of cases are from low or middle class country [1]. Turkey is one of the them that has the highest prevalence in Europe, with about one in every seven adults diagnosed DM. By 2035, Turkey will have the highest number of people with type 2 DM in Europe, at almost 12 million [2].
Mortality rates have increased with the increasing prevalence of type 2 DM, especially in the younger population, such that half of the deaths come from those under sixty [2]. The beliefs and mental state of patients with chronic illnesses like DM can affect disease outcomes and the patients' self-management [3]. Self-care and diabetes medications are important components in improving the disease outcome [4], though many studies have shown that these activities can be negatively related to fatalism about the disease state [5].

Оригінальні дослідження / Original Researches
Fatalism is about a decrease in healthy behaviors known to improve diabetes-related outcomes, like eating healthy foods, exercising, and not smoking [6,7]. Fatalism is the belief that the person has no means of control over their own symptoms, and they become totally dependent on a higher power, mainly God [8]. Many people with different religions in the world share a lot of common cultural and religious beliefs and values, such as fatalism [9]. There is little attention given to how a person with a chronic illness, in general, and diabetes, specifically, responds to fatalism and its subsequent implications on disease self-management. Patients succumbing to fatalistic views about the disease believe that it is something coming from God and that they do not have power and control over it [9,10].
This belief affects the ability of diabetes patients to cope with their disease and adopt self-care practices, but the absolute effects on the self-management of diabetes are unclear [10,11]. Diabetes fatalism is characterized by hopelessness, perceptions of despair, and powerlessness [12]. Many studies about diabetes fatalism found that diabetes fatalism is related to unexpected health outcomes, glycemic levels which cannot be control, and a lower quality of life [12][13][14]. Fatalism is associated with individuals' coping response and spiritual beliefs and also type 2 DM experience [13]. Previous studies findings show that type 2 DM fatalism is also related to poor self-care and glycemic control [13,14]. It reflects patients opinions whether and to what extent they can control the outcomes of their deeds [14,15]. DM fatalism has been studied in ethnic groups such as African-Americans [12,13], South Asians in UK [16], Latinos/ Hispanic [3] and Iranians [17] of which exhibited increased fatalistic beliefs and practices, thereby jeopardizing their diabetes self-care and glycemic control.
According to the findings of an international comparative field study carried out recently, unlike people from Canada, US and Norway, half of the people participating in the study in Turkey acknowledged that "they have very little to change the course of their own life" [18]. There are fatalism scales developed in the international literature in many different fields [19]. It is observed that the studies on fatalism in Turkey were mostly carried out in the field of religion [20].
The fatalism studies in Turkey are quite limited although it is a Muslim country with a strong belief in destiny. Studies that determine the fatalism status of health-related patients will shed light on this area. Because fatalism is effective in many areas from health to social life in Muslim countries. L.E. Egede and C. Ellis developed the Diabetes Fatalism Scale (DFS) to determine the level of a diabetic patients' fatalism [12]. The DFS scores measure the quality of life, selfcare, healthy lifestyle choices, blood sugar testing, and the glycated hemoglobin (HbA1c) levels of DM patients [5,12].
The purpose of this study was to validate the Turkish version of DFS.

Sample
This methodological-type research stud consist of 139 Type 2 diabetes patients in Turkey. The patients were chosen randomly and were selected in accordance with the following criteria: aged older than 18 years; have had a DM diagnosis for at least one year; were not pregnant; were fully oriented and conscious; diabetic patients who did not have problems with vision, hearing, or using their hands; and who could read, write, speak, and understand Turkish. The faceto-face interview method was used for collecting data by researhers and it lasted 20 minutes The study has three steps: 1) the adaptation to Turkish language; 2) testing the content validity; 3) performing psychometric analyses.

Linguistic Validity and Assessment of the Data
In step one, the original scale was translated from English to Turkish by who can speak both languages well. First it translated from Turkish to English then the scale in Turkish back translated in English. Second, items in original original scale compared with items in back translated scale. After exa mining the compatibility between the original English scale and the translated one, the Turkish scale finalized.

Content Validity
In step two, to test content validity the Turkish scale sent to an expert panel (included of one endocrinologist, two diabetes nurses, one psychiatrist, three nurse academicians) who have studied on type 2 DM. This expert panel evaluated the items of scale in terms of distinctiveness, understandability and appropriateness for the purpose. Davis technique was used to test content validity based on the views of experts [12].
In step three, the Turkish scale was piloted on 10 patients with type 2 DM who included study criteria to obtain initial assessment of the Turkish scale and all items in Turkish scale was understood by patients.

Measurement Instruments
Two form were used to collect the data, one was socio-demographic form that consist of questions such as gender, age, educational level and DM duration. And other was the scale DFS to measure the level of diabetes fatalism of patients with type 2 DM. DFS has 12 items that are scored on 6-point likert with ranging from 1 = strongly disagree to 6 = strongly agree and three sub-scales that are "perceived self-efficacy (powerlessness)", "religious and spiritual coping (hopelessness)" and "emotional distress (despair)". Higher scale total scores represent more fatalistic attitudes of patients with type 2 DM towards diabetes. The DFS has a good internal consistency (Cronbach's alpha = 0.804) [12].

Data Analysis
Descriptive statistics, exploratory factor analysis (EFA), confirmatory factor analysis (CFA), the multiple fit indexes of chi-square goodness, being the goodness of fit index (GFI), the adjusted goodness of fit index (AGFI), the comparative fit index (CFI), the standardized root mean square residuals (SRMR), and the root mean square error of approximation (RMSEA) were used to analysis the data by using SPSS 24.0 and AMOS software packages.

Ethical Consideration
Permission was obtained via e-mail from L.E. Egede and C. Ellis, who developed the scale to adapt the diabetes fata lism scale to Turkish. The purpose of the study was ex- plained to patients with type 2 DM who were voluntary to participate in the study and met the study criteria and written consent has been obtained.

Results
In this study, 54.7 % of the patients were men, 73.4 % were married, 54 % had another disease, 18 % were graduated from high school, the average age of patients was 50.20 ± 16.82 years, the average duration of DM was 19.31 ± 14.25 years, and the mean HbA1c was 7.06 ± 0.65 % ( Table 1).

Exploratory Factor Analysis
The Kaiser-Meyer-Olkin (KMO) coefficient and Bartlett's test were measured to determine whether the data set was fit for factor analysis. It was found that the KMO measure of sampling adequacy value was 0.770, and the chisquare value was 1078.402. Since the KMO coefficient was higher than 0.6 (KMO > 0.6) and the Bartlett's test of significance level was lower than 0.05 (p < 0.05), the data set was found to be fit for factor analysis. According to the total explained variance table, it can be said that the 12 items were weighted into groups of four factors because the number of components is four. The first of these four factors alone accounted for 40.6 % of the total variance, the second factor accounted for 17.7 %, the third accounted for 12.4 %, and the fourth accounted for only 8.9 %. These four factors together made up 79.6 % of the total variance. The rotated factor matrix that examined how the items were weighed in each factor showed that it was necessary to exclude the fifth item from the scale because it alone constituted the fourth factor ( Figure 1).
When the fifth item was excluded from the study and the analysis was repeated, the KMO coefficient became 0.802

Figure 1. Structural equation model with standardized path coefficients of Diabetes Fatalism Scale dimensions
and the chi-square value became 1020.244, with p = 0.000. In the new total explained variance table lacking the fifth item, 11 items were weighed into groups of three factors. The first of these four factors alone accounted for 43.4 % of the total variance, the second accounted for 19.3 %, and the third accounted for 13.6 %. These three factors together accounted for 76.2 % of the total variance. In the rotated component matrix given below, items 6, 7, 8, and 9 are weighted in the first factor, items 10, 11, and 12 are in the second factor, and items 1, 2, 3, and 4 are weighted in the third factor (Table 2 and Table 3).
To test the construct validity of the factors extracted using confirmatory factor analysis (CFA). CFA using AMOS v22 software was undertaken using maximum likelihood estimation. In the confirmatory factor analysis (CFA) mul tiple Mìžnarodnij endokrinologìčnij žurnal, ISSN 2224-0721 (print), ISSN 2307-1427 (online) Оригінальні дослідження / Original Researches creteria including goodness of fit index (GFI), adjusted goodness of fit (AGFI), root mean squared error of approximation (RMSA), Comparative Fit Index (CFI) were used to assess the fit of model to the model to data. Guidelines for testing model fit followed guidance, the goodness of fit index (GFI) ≥ 0.90, CFI ≥ 0.90, SRMR < 0.05 and the root mean square error of approximation (RMSEA) < 0.05 [14].
The results evidenced aduequated model fit: x2/SD = 3.12, GFI = 0.92, AGFI = 091, CFI = 0.90, RMSEA = 0.11 and SRMR = 0.080. When the inner consistency coefficients of the scale were examined for the emotional distress, which corresponded to the first subscale before the exclusion of the fifth item from original scale, the Cronbach's alpha value was 0.613. Because the acceptable consistency index was   (Table 4).

Discussion
DFS scale African American population, general United States of America, British and British South Asians and Lubnan has been validated in the studies. Turkey is geographically diverse as Africa and the Middle East is in contact with Europe. In this study, which has a Muslim population of Turkey and we have demonstrated the results in a different is geography. The validity and reliability of a tool used for measuring the fatalism of diabetic patients was tes ted in this study. The KMO value was determined through exploratory factor analysis according to the literature [15].
A KMO value of less than 0.50 indicates that the sample size is not sufficient. The KMO value in our study was 0.770, and this conclusion suggests that the sample size was large enough for the study. In the factor analysis of the scale items, it was determined that the fifth item was subdivided separately from the original scale. In this study, we therefore decided to exclude the fifth item from the analysis because it did not contribute to the subscales. After excluding the fifth item, the Cronbach's alpha value of the emotional distress subscale reached 0.806. Otherwise, the other scale items remained similar to the original scale. It was determined that the ratio of the chi-square statistics obtained from the confirmatory factor analysis to the degree of freedom (χ 2 /df) was 3.12; the RMSEA was 0.119; the GFI value was 0.925; and the CFI value was 0.905.
In particular, CFI and GFI values of a model that are 0.90 or higher mean that the model is well adapted [16,17]. If the RMSEA value is less than or equal to 0.08, it means that the fit is good. Cronbach's alpha indicates the internal consistency of the scale. The Cronbach's alpha value of the emotional distress subscale, which corresponds to the first subscale, was considered an unacceptable level without removing the fifth item from the scale. After this fifth item was excluded, the value reached 0.806.
The expression corresponding to item #5 of "diabetes causes a lot of suffering for me" showed that the Turkish society does not perceive diabetes as an emotional distress. This finding suggests that the effects of the religious beliefs of the community that we are studying on emotional distress were different from the original audience of this questionnaire. The diabetic personality, which is caused by DM, can be influenced by the religious beliefs of people [18]. In the validity and reliability study of the scale in the Arab society in Lebanon, it is observed that all of the items of the scale retain their current status or it is determined that removal of material from the scale [19]. Nabols and friends found that society in Lebanon is fatalistic. These findings said that Turkish society is less fatalistic than Lebanese society. L.E. Egede concludes that fatalism is multidimensional and influences people's religious beliefs, values and spirulence as well as their experience of illness, their sense of diabetes, and their personal coping responses. These results may have been obtained as a feature of the language.
Diabetic fatalism may be different in this study because we have measured the reliability and validity in a society that has different values than the values of the society where the scale was developed. When resignation is regarded as a religion doctrine and is used as a coping mechanism, it may be easier to accept the disease. For this reason, the disease may not be perceived as an emotional stress. L.E. Egede indicated in his studies that distress levels increased with the severity of the illness and with the burnout caused by self-care [20]. Religion and spirituality played an important role in coping with illness, and in African societies, illness comes from God. This finding showed a similarity between Asian society and African society. These findings further support the validity of the DFS-T. Some Muslims believe in destiny and predetermined life events that occur beyond a person's control [20]. They believe that their disease comes from God, and they are patient about their illnesses, meaning that they do not consider that their disease causes a lot of suffering for them. The "religious and spirituality coping" and "perceived self-efficacy" subscales of the Turkish version are the same as the original scale. This shows a similarity between Asian society and Western society.

Conclusions
As a result of the this study, DFS is a validate and reliability scale for Turkish type 2 patients to assess the fatalism in diabetes mellitus.