Comparison of clinical characteristics and dietary intakes according to phenotypes of type 2 diabetes mellitus in South Korea: a cross-sectional study

Ethics statement

The study was conducted in accordance with the ethical principles outlined in the Declaration of Helsinki and obtained clearance from the Institutional Review Boards of Changwon Fatima Hospital (approval No. 17-04). All participants provided informed consent.

1. Study design

This study was a cross-sectional study and compared two groups based on their diabetes phenotypes. It was described in accordance with the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) statement, available at https://www.strobe-statement.org/.

2. Setting and participants

T2DM patients between the age of 19 and 60 years were recruited in this study. Participants were visited to Changwon Fatima Hospital (Changwon, Korea) for diabetes management from April 2017 to July 2018. Recruitment was carried out through publicity at the endocrine center within the hospital. The number of participants was determined using on an effect size of 0.9, a significance level of 0.05, and a power of 0.9. A total of 36 participants who understood the contents and purpose of the study and voluntarily wished to participate were enrolled. T2DM was diagnosed as glycated hemoglobin level ≥ 6.5% or fasting blood glucose level ≥ 126 mg/dL. The participants were assigned to lean diabetes mellitus group (LDM; BMI < 25 kg/m², n = 18) and obese diabetes mellitus group (ODM; BMI ≥ 25 kg/m², n = 18). Given that Asians have a higher risk of CVD than Caucasians, obesity has been defined differently in Asia compared to other regions. Specifically, the World Health Organization (WHO) Asia-Pacific Guidelines recommend a low BMI cutoff point of 25 kg/m² for obesity [11]. Thus, lower BMI cutoff points have been used in studies conducted in Asia, including Korea [12]. Patients with thyroid, cerebrovascular, gallbladder, or gastrointestinal diseases, gout, depression, and mental conditions, such as porphyria, alcohol addiction, schizophrenia, and drug addiction, were excluded from the study. Additionally, those who were under prescription of weight control drugs, who were pregnant or lactating, and who participated in commercial diabetes programs within 30 days from the start date of this study were excluded.

3. General characteristics, health-related lifestyle, and medical history

The general characteristics, health-related lifestyle, and intake of medication of all participants were analyzed through questionnaires and medical interviews. The general characteristics of age, marital status (married/not married), occupation (office worker, production worker, service, business, housewife, etc.) were assessed. The health-related lifestyle was categorized as follows: smoking status (never-smoked, ex-smoker, current smoker), alcohol consumption (yes, no), physical activity (yes, no), intake of dietary supplements. Data on diabetic complications (hypertension, hyperlipidemia, etc.) were collected and multiple responses were used.

4. Anthropometric and body composition measurements

The participants were asked to undress any accessories and shoes and stand in an upright position with comfortable clothes to conduct anthropometric and body composition measurements using bioimpedance analysis program (InBody 720; InBody Co., Ltd.). BMI was calculated by dividing the body weight (kg) by the square of height (m). According to the WHO Asia Pacific and the Korean Society for the Study of Obesity (KSSO), obesity was defined as BMI of > 25 kg/m². Skeletal muscle mass (SMM), body fat mass (BFM), percent of body fat mass (PBF), intracellular water, extracellular water, and total body water were measured and used in this study. As suggested by the KSSO, waist circumference ≥ 90 cm and ≥ 85 cm were considered as abdominal obesity for male and female, respectively. While standing comfortably in a vertical position with arms stretched, waist circumference was measured between the lowest lower rib and iliac crest, and hip circumference was measured at the widest part of the pelvis. Systolic blood pressure and diastolic blood pressure were measured using a standard electric pressure gauge (FT 500; Jawon Medical) in a relaxed state. All measurements were made by a single trained researcher.

5. Nutrition intake assessment

To investigate the daily intake of nutrients, such as energy, carbohydrate, protein, and fat, 24-hour recall method was used. Prior to assessment, a trained clinical nutritionist educated the study participants on the recording method and explained accurate observational measurements and food material recordings to the study participants. After assessment, the exact amount of food intake was investigated through one-on-one inter-views with the participants for the names of food consumed in a total of 3 days, including 2 days on weekdays and 1 day on weekends. The collected data was analyzed using the Computer Aided Nutritional Analysis Program 5.0 (CAN-Pro 5.0; The Korean Nutrition Society) to calculate average daily intake of nutrients. Nutrient intake was compared to the standards for each sex and age group as suggested in the 2020 Dietary Reference Intakes for Koreans (KDRIs) by the Ministry of Health and Welfare and Korean Nutrition Society. Nutrient intake was then converted into percentage to calculate KDRIs % [13].

6. Dietary habit evaluation

The dietary habits of the participants were investigated using a questionnaire that consisted of a total of 13 items on the number of meals per day, eating breakfast, regularity of mealtime, mealtime duration, frequency of snack, and frequency of eating out. A questionnaire developed for adults in Korea [14] was modified and used for this study.

7. Food intake frequency evaluation

A draft version of food intake frequency questionnaire was prepared based on the food intake frequency of the National Health and Nutrition Examination Survey. The questionnaire was then revised according to the purpose of this study [15]. The items of food intake frequency questionnaire were subdivided into 62 categories, including carbohydrates (7 types), legumes and potatoes (5 types), meat and seafoods (14 types), vegetables and seaweeds (13 types), fruits (11 types), dairy products (3 types), beverages (6 types), and instant foods (3 types). For each food item, the amount of intake per serving was presented. The frequency of intake was divided into “rarely,” “6–11 times a year,” “once a month,” “2–3 times/month,” “once a week,” “2–3 times/week,” “4–6 times/week,” “once a day,” “twice a day,” and “three times a day.” In this study, frequency of food intake was converted into servings/per day. The frequency of nutrient intake was analyzed using CAN-Pro version 5.0 for experts to calculate nutrient intake.

8. Evaluation of obesity-related quality of life

Much research mainly evaluated health-related quality of life using Europe version quality of life (EuroQol) questionnaire. EuroQol comprises of two parts, the health states descriptive system and visual analogue scale. The Korean version of those tools has been modified culturally and translated [16]. The Korean version quality of life (KOQOL) was established to evaluate obesity related quality of life after sufficient testing for reliability and validity. The Korean version of obesity-related quality of Life, developed by Park et al. [17], was used. The reliability analysis results of KOQOL showed that Cronbach's alpha coefficient was 0.838. The questionnaire consisted of a total of six sub-domains and 15 items with four items on psychosocial health, three items on physical health, three items on task, two items on daily living, two sex-related items, and one food-related item. Each item was scored on a 4-point Likert scale (one point: “not at all,” two points: “sometimes,” three points: “often,” and four points: “always”). A higher score indicated a lower quality of life.

9. Biochemical examination

On the second visit, approximately 10 mL of blood was collected from the brachial vein after 12 hours of fasting. Blood was centrifuged at 3,000 rpm for 15 minutes to separate the serum, which was then used to test various biochemical parameters. Fasting glucose, triglycerides, total cholesterol, high-density lipoprotein-cholesterol, low-density lipoprotein-cholesterol, aspartate aminotransferase, and alanine aminotransferase levels were measured using an enzymatic quantification kit (AM201, AM157S, AM202, A203ST, AM103-K, AM102, Asan set Assay kit; Asan Pharmaceutical Co.) [18]. Atherogenic index was calculated using total cholesterol and HDL-C according to Lauer et al.’s formula [19]. Serum insulin concentration was measured using a commercial kit (80-INSHU-E01.1, Insulin EIA; ALPCO Co.), and homeostasis model assessment of insulin resistance (HOMA-IR), which is an indicator of insulin resistance was assessed using a formula suggested by Bradford [20].

10. Statistical analysis

The study results were analyzed using the Statistical Package for Social Science Statistics (SPSS version 26; IBM Corp.). Data are presented using mean ± standard deviation, and the chi-square test was conducted to compare the characteristics and dietary habits of the participants. When the expected counts were less than 5, analysis was performed using Fisher’s exact test. The Mann-Whitney U-test conducted to compare differences in anthropometric measurement and body composition, nutrition intake, food intake frequency, and biochemical parameters between the two groups. The obesity-related quality of life variables was tested for Normality Test using Shapiro-Wilk analysis and independent t-tests were performed. A P-value of < 0.05 was considered statistically significant.

1. General characteristics of the participants

The general characteristics of the participants are shown in Table 1. The participants were divided into two groups according to BMI. A total of 18 participants were included in each of the ODM and LDM groups. There was no significant difference between the two groups in general characteristics, but the prevalence of diabetic complications was significantly higher in the ODM group. The prevalence of hypertension was significantly higher in the ODM group (P = 0.011), and the prevalence of dyslipidemia also tended to be high.

Table 1.

General characteristics of the participants

2. Anthropometric measurements and body composition

Anthropometric measurements and body composition per group are shown in Table 2, for male, the weight of the ODM group was 89.1 kg and that of the LDM group was 66.3 kg, showing that the weight of the ODM group was higher (P = 0.001). Accordingly, the BMI of male in the ODM group was 29.8 kg/m², and that of the LDM group was 23.0 kg/m², which was higher in the ODM group (P < 0.001). SMM, BFM, PBF, protein, and mineral were significantly higher in the ODM group than in the LDM group in male, but only BFM and PBF were statistically significantly higher in the ODM group in female (P < 0.050).

Table 2.

Anthropometric measurements and body composition of the participants

3. Percentage of nutrition intake of Recommended Nutrient Intake (RNI)

Table 3 shows the percentage of nutrition intake compared to the Korean Dietary Reference Intakes per group. There were significant differences in the intake of vitamin C and iodine between the ODM and LDM groups (P < 0.05). Energy intake of each group was lower compared to that of Estimated Energy Requirements (EER). The rate of carbohydrate intake was 181.3% and 185.3% in the ODM and LDM groups, respectively, which were higher than the RNI. The rate of protein intake was 120.7% in the ODM and 107.7% in the LDM, both groups intake more carbohydrate and proteins than RNI. There was no significant difference in the percentage of macronutrient intake between the two groups.

Table 3.

Percentage of nutrition intake of Korean RNI of the participants

4. Dietary habits

The dietary habits of the participants are shown in Table 4. For the number of meals per day, most of the participants had three meals per day (47.2%, n = 17), followed by those who had two meals per day (33.3%, n = 12), and irregular number of meals (19.4%, n = 7). Approximately 44.4% of the total participants answered that they always had breakfast. Comparing the eating habits of the two groups, significant differences were found in meal-times duration and types of eating out.

Table 4.

Dietary habits of the participants

The LDM group showed a difference from the ODM group in that the mealtime was '20 minutes' the most (83.3%) (P = 0.006). These findings indicate that obese diabetic patients had a shorter mealtime duration than the lean diabetic patients. And 55.6% of the participants mainly consumed “Korean dishes” for eating out. In the LDM group, 77.8% of the participants mainly consumed “Korean dishes,” which was higher than that in the ODM group. In the ODM group, 27.8% of the participants mainly consumed “fast food,” which was significantly higher than in the LDM group (0.0%).

5. Food intake frequency

The frequency of food intake is shown in Table 5. Food was divided into 11 groups, and one serving per day was treated as one for average intake. The intake of rice cakes and chicken was higher in the ODM group than in the LDM group (P = 0.050). The intake of pork, fruits such as tangerines, grapes, strawberries, oranges and yogurt were higher in the LDM group (P < 0.050). In both the groups, the intake frequency of mixed grains was more than once a day, suggesting a tendency to choose food with high levels of dietary fiber among carbohydrate foods. For meats and eggs intake frequency, the intake frequency of eggs as protein source was the highest at 0.4 and 0.5 in the ODM and LDM groups, respectively. And it was confirmed that fruit intake was low in both the groups. For milk and dairy products, the intake frequency of milk was 0.2 in the ODM group and 0.3 in LDM group, which was less than once a day.

Table 5.

Daily food intake frequency of the participants

6. Korean version of obesity-related quality of life

Obesity-related quality of life of the participants is shown in Table 6. The total score was 28.2 in the ODM group, which was significantly higher than in the LDM group (23.7), suggesting that the quality of life was lower in the ODM group (P = 0.043). Analysis of the sub-domains showed that there were statistically significant differences in work-related health (P = 0.014) and diet distress (P = 0.007) domains between the two groups.

Table 6.

Obesity-related quality of life of the participants

7. Biochemical parameters

The biochemical parameters of the participants are shown in Table 7. There were no significant differences in biochemical parameters between the two groups.

Table 7.

Biochemical parameters of the participants

Limitations

This study had some limitations. Data were self-reported based on 24-hour recall for nutrition intake assessment. The amount of food intake recorded by the participants may be less than the actual intake. As a small cross-sectional study, it is difficult to generalize the results due to the small number of participants, which complicates the establishment of a clear causal relationship.

Previous studies have showed that classifying obese patients according to the location or amount of adipose tissue may be appropriate. However, BMI is correlated with body fat percentage, so it may be suitable for this study [40]. Despite these limitations, this study demonstrated an association between the prevalence of CVD, bad dietary habits, and obesity when T2DM patients were classified according to phenotype. In addition, it is meaningful in that it compared dietary habits, nutrient intake, and obesity-related quality of life in Korean diabetic patients. The results of this study will help to implement more effective medical nutrition treatment reflecting dietary habits to prevent CVD in T2DM patients with obesity.

Conclusion

In conclusion, obese diabetes patients had an increased risk of CVDs, bad dietary habits, and low obesity-related quality of life. Therefore, based on these findings, personalized nutrition treatment considering the risk and differences in dietary habits according to the phenotype of diabetes would be necessary for treatment of diabetic patients. Personalized nutrition treatment, including strategies for selection of food and nutrient balance and dietary education, would be helpful for effective treatment of diabetes. Future studies on the effects of personalized treatment in patients according to the obesity status in T2DM patients and related mechanisms would be necessary.