**3. Results**

## *3.1. Demographic Characteristics of Patients—Section A*

The data were collected from 350 patients with a median age of 45 years (range: 40–51 years). The results presented in Table 1 show that there were more male patients (*n* = 202; 57.7%) than female patients (*n* = 148; 42.3%). More than 90% of the patients were married. A majority of patients had received a secondary education (*n* = 200; 57.14%), while some were illiterate (*n* = 69; 19.7%), and others were graduates and postgraduates (*n* = 81; 23.14%). A majority of patients in the study were overweight (*n* = 167; 47.7%), some were obese (*n* = 115; 32.9%), some had a normal weight (*n* = 56; 16%), and others were underweight (*n* = 12; 3.4%). A significant association was observed between the overall DA of patients and their educational status (*p* = 0.034). However, the overall DA was not significantly associated with gender (*p* = 0.142), marital status (*p* = 0.413), or BMI (*p* = 0.666). The frequency, percentage, and ranked mean score for each item are presented in Table 2.


**Table 1.** Sociodemographic Characteristics.

BMI, Body mass index.


**Table 2.** General dietary attitude of patients with type 2 diabetes based on ranking analysis.

DM, Diabetes mellitus; SD, Standard deviation.

### *3.2. Patients' General Dietary Attitude Towards Food—Section B*

No outlier problem was detected in the overall DA score variable as z-score values (−2.30–2.22) were less than the absolute value of 4. The mean DA score of 16 items was 3.94 + 0.87. Based on the mean score, the DA was categorized into having a positive attitude and having a negative attitude. There was a majority of patients with a negative DA (*n* = 198; 56.6%) compared with those with a positive DA (*n* = 152; 43.4%). The result of the one-sample chi-squared test showed that the overall DA of patients with type 2 diabetes was inappropriate (χ<sup>2</sup> = 6.04 (1), *p* = 0.014). The positive and negative attitude when compared within the subgroups (identified by EFA) showed that the patients also had an inappropriate DA towards food selection (*p* = 0.003), healthy choices (*p* = 0.005), food restraint (*p* < 0.001), health impact (*p* < 0.001), and food categorization (*p* = 0.033). These results are presented in Table 3.


**Table 3.** Comparison of Positive and Negative Dietary Attitude in Subgroups identified by exploratory factor analysis (EFA).

DM, Diabetes mellitus; PDA, Positive Dietary Attitude; NDA, Negative Dietary Attitude; \* statistically significant at the 5% level of significance.

### *3.3. Patients' General Dietary Attitude Towards Food—Section B*

Backward elimination with the conditional approach retained six items in the final model. The values of model chi-squared and Hosmer–Lemeshow tests were 81.80 (*p* < 0.001) and 20.02 (*p* < 0.001), respectively, which showed that the fitted model was appropriate at the 95% confidence interval (CI). Overall, the model correctly classified 71.4% of patients. The odds ratio for red meat was 2.43 (*p* < 0.001). Converting the odds ratio into a probability showed that the consumption of red meat was 70.84% greater in patients who said "yes" they are eating a healthy diet. Dairy products had an odds ratio of 1.408 (*p* = 0.015), which showed that the consumption of dairy products was 58.38% greater in patients who said "yes" they are eating a healthy diet. The odds ratio for rice was 3.472 (*p* < 0.001). The probability results showed that consumption of rice was 77.63% greater in patients who said "yes" they are eating a healthy diet. Junk food had an odds ratio of 2.347 (*p* < 0.001), showing that the consumption of junk food was 70.12% greater in patients who said "yes" they are eating a healthy diet. The odds ratio for soups and sauces was 1.383 (*p* = 0.040). The probability results showed that the consumption of soups and sauces was 58.03% greater in patients who said "yes" they are eating a healthy diet. Fruits had an odds ratio of 1.416 (*p* = 0.024). Converting the odds ratio into a probability showed that the consumption of fruits was 58.60% greater in patients who said "yes" they are eating a healthy diet. However, for foods such as white meat, bakery products, cereals, sweets and snacks, drinks, vegetables, boiled or grilled meals, olive oil, and canned food, there was no statistical significance (*p* > 0.05). These results are presented in Table 4.


**Table 4.** Binary Logistic Regression Analysis using the Backward Conditional Approach for the Dietary Attitude of Patients with Type 2 Diabetes towards Specific Food Items.

\* Significant at the 5% level of significance; † non-significant variables.

#### **4. Discussion**

Our study showed that patients with type 2 diabetes had an overall inappropriate DA. Subgroup analysis also showed an inappropriate DA of patients towards food selection, health impact of food, healthy choices, food restraint, and food categorization. In addition, the patients had a poor DA towards the consumption of red meat, rice, soup and sauces, dairy products, and junk food. The results of our study also showed that for the majority of patients, food selection and health impact of food were not important, and this is consistent with the findings of a study conducted in Egypt [6]. This may be because of deeply rooted cultural beliefs and values, which may pose a difficulty for patients' adherence to food selection and consumption of foods having a health impact. The role of cultural attitudes and behaviors towards food in the management of diabetes cannot be neglected [15]. This is consistent with our study results, as the attitude of patients with diabetes towards food is influenced by a strong cultural attitude. Most of them stated that the selection of food, its health impact, healthy choices, food restriction, and food categorization are not important to them. The Saudi cultural barrier factor towards food selection and its consumption and health impact has also been supported by a local study [16]. In our study, a majority of the patients stated that they do not like to eat diet food, nor do they like to stay away from foods that contain sugar. Moreover, only one-fifth of the patients indicated that they feel guilty after eating oily foods. These findings are supported by research conducted by Buttar et al. [17].

A study conducted by Ntaate [18] among patients with type 2 diabetes from Uganda reported a positive DA (82%) towards diet. In contrast, in our study, the patients not only had an overall inappropriate DA, but also an inappropriate DA towards the consumption of red meat, rice, soup and sauces, dairy products, and junk food. Most of the patients in our study were unaware of the caloric content in the food they were consuming. This can be attributed to their literacy level; in our study, 57.2% of the patients had received a primary and secondary education, while approximately 20% were illiterate. This fact is supported by studies that also stated that literacy is an important

influential factor, because patients with low literacy have difficulty reading food labels and estimating potion sizes [19–21].

Therefore, to achieve the DA goals, a patient empowerment approach should be used. Since an empowerment approach is a social phenomenon, when a patient is empowered with necessary knowledge about lifestyle modification, outcomes of disease if not controlled, etc., he/she shows a more responsible attitude with better self-efficacy towards diabetes care [22,23]. The empowerment approach in dealing with type 2 diabetes is highly recommendable because it brings about changes in the behavior of the patient that is deeply rooted in their daily routine. Healthcare providers should be well-informed about patient attitudes and beliefs towards diabetes to design tailored educational and salutary programs for a specific community [24].

Imparting nutritional education is a perilous component of diabetes care, especially for the self-management of the disease. Thus, for better diabetes care, patients should be referred to dietitians who should assess their attitude towards food in general, and towards various foods such as meat, rice, junk food, etc., and suggest tailored dietary self-management strategies. To facilitate behavioral dietary changes, this assessment should be individualized and patient-centered, and it must be based on a patient's cultural beliefs, norms, psychosocial status, and literacy, as these factors have been identified as a barrier to reaching nutritional therapy goals [25]. Along with these efforts, the authorities in the Kingdom of Saudi Arabia should provide diabetes self-management educational programs on a regular basis, with special emphasis on diet and its related components. Such educational programs have been found to have an encouraging impact on patient behaviors. However, to achieve a long-term positive effect on behavior modification, sustained reinforcement is needed, which can be achieved using a patient empowerment approach [26].

There are some limitations to this study. The research design was cross-sectional, which itself has methodological limitations, so it cannot be used to analyze behavior over a period of time. The study was conducted in the central region of the KSA, and although the eating habits do not vary much within the eastern, southern and northern regions of the KSA, there is still a need for a national DA assessment program. Another limitation is that we were unable to compare the self-prepared DAQ with the gold standard; doing so might have helped us to study the DA of the patients with diabetes in more detail to devise strategies for better patient care. Nonetheless, the study provided important points: The results can be generalized as we used a systematic random sampling technique for the selection of patients, the DA questionnaire was reviewed by experts in the field, it successfully passed the psychometric analysis, and we can say that it is a valid and reliable questionnaire for assessing and evaluating the DA of patients with type 2 diabetes. This is the first study conducted in the KSA related to assessing and evaluating the DA of patients with type 2 diabetes. Therefore, the results can serve as a baseline for similar studies conducted in the KSA and references can be extended to the neighboring Gulf Cooperation Council (GCC) countries.
