The association of food insecurity with diabetes control and self-care in diabetes type 2 patients: A cross-sectional study

Background Food insecurity can increase risks of health and nutritional problems, leading to diculties in self-care and poor glycemic control in diabetic patients. This study assessed food insecurity and its association with diabetes control and self-care in type 2 diabetes patients. Methods In this cross sectional study, 148 adults with type 2 diabetes participated. Food insecurity and self-care were determined using the 18-item USDA household food security status questionnaire and self-care (SDSCA) questionnaire. Data were analyzed using logistic and linear regression tests (SPSS 20 software). Results Thirty-seven percent of the participants were food-insecure and signicantly more likely than food-secure participants to have poor glycemic control (OR = 3.02; CI: 1.45–2.65). No signicant association was found between food-insecurity and overall self-care score. Results of the present study showed that food-insecurity has a direct association with poor glycemic control in type 2 diabetes patients. On the other hand, food-secure group had higher scores of following healthy diet and physical activity. Since economic status was signicantly better in the food-secure group, it can be postulated that nancial problems will eventually lead to non-compliance with the recommendations and therefore poor glycemic control. Policy strategies to increase access to diabetes-appropriate foods may reduce socioeconomic inequalities in glycemic control.


Introduction
Food insecurity is de ned as "limited or unreliable access to adequate and healthy food in terms of nutritional value or limited ability to access acceptable foods in appropriate social ways" (1)(2)(3).
People with food insecurity are at risk for diets that are less varied and of lower quality, insu cient intake of micronutrients, iron de ciency anemia, and low consumption of fruits and vegetables (16)(17)(18)(19)(20). Food insecurity also interferes with self-care advice for diabetes (21). Researchers have always reported a relationship between food insecurity and low self-esteem with diabetes control (21)(22)(23)(24). Also, diabetic patients with food insecurity have higher HbA1c levels than those with food security (24)(25)(26)(27)(28). On the other hand, diabetics with food insecurity are also at risk for hypoglycemia (26).
In a study, it was shown that food insecurity was associated with poor glycemic control and had no signi cant relationship with self-care behaviors (29). In Iran, the relationship between food insecurity and diabetes has also been studied. In a study in Tehran, there was no signi cant relationship between food insecurity and blood sugar and lipid pro le (30). In another study in newly diagnosed diabetic patients, as well as in a case-control study, a signi cant and direct relationship between food insecurity and diabetes control was observed (31,32).
Given the limited studies in Iran and considering the relatively widespread prevalence of food insecurity (26,33) and the growing prevalence of type 2 diabetes in Iranian society (34), the present study aimed to investigate the status of food insecurity and its association with diabetes control and self-care in patients with type 2 diabetes.

Study population
This was a cross-sectional study on 148 type 2 diabetic patients between 30 and 65 years old, who had a medical record at Eghlid county Health Center and whose HbA1c level were recorded during the last two weeks. Eghlid is a county in Fars Province in Southwest Iran. Patients were excluded if they any speci c diseases (cardiovascular disease and stroke, cancer, acute kidney disease) were present, as well as pregnant and lactating women.
In the period of May to September of 2018, eligible patients were enrolled and the study procedure was explained for them on the phone. They were invited to three community health centers in Eghlid County to complete the questionnaires. Written consent was obtained and demographic questionnaire, food security questionnaire and self-care questionnaire were completed by patients with the supervision of nutritionists. The most recent HbA1and fasting blood sugar (FBS) abstracted from the medical record and HbA1c ≥ 8.5% was classi ed as undesirable glycemic control. Weight and height were measured and BMI was calculated.

Questionnaires
A demographic questionnaire was used to collect general demographic information, including age, sex, ethnicity, education, and employment status. Household economic status was determined using a questionnaire including 9-item home appliances and was classi ed owning less than 3-item as poor economic level, 4 to 6-item as moderate economic level, and more than 7-item as wealthy economic level.
The 18-item USDA household food security questionnaire was used to assess food security over the past 12 months (35). This questionnaire has been validated during previous studies in Iran (36). Rating of 18item USDA household food security status questionnaire was a below: positive rate to answers "often true", "sometimes true", "almost every month", "some months", and "yes" and zero score to responses "not true", "does not know or refused", "only 1 or 2 months", and "no". Finally, scores 0-2 were categorized as food secure, scores 3-7 as food insecure without hunger, scores 8-12 as food insecure with moderate hunger, and scores 13 and higher as food insecure with severe hunger.
Collecting self-care behaviors was assessed using the Summary of Diabetes Self Care Activities (SDSCA) with 14 questions that assesses patients' self-care quality in the last seven days (37). Questions include diet: 5 questions, exercise: 2 questions, blood sugar control: 2 questions, foot care: 5 questions, and regular medications: 1 question. To score the questionnaire, for each question of self-care behaviors, one point was considered per day if it was done and zero points were considered if it was not done. The highest score in this tool is 98, which indicates the highest quality of self-management. The validity and reliability of this questionnaire have been studied by Anbari et al (38).
Weight was measured using a Seca scale with an accuracy of 100 g with minimal coverage and no shoes, and height was measured by a tape without shoes with an accuracy of 0.1 cm by a nutritionist. BMI was calculated by dividing weight (in kilograms) by height squared (in square meters).

Statistical analyses
All analyzes were performed using SPSS version 20 and the signi cance level was considered to be 0.05.
The relationship of food quality and self-care with food security was tested by linear regression analysis. Logistic regression was used to investigate the association between glycemic control and food security with a 95% con dence interval.

Results
Of the 155 eligible patients, 148 subjects completed the interviews. Findings of the food insecurity prevalence in diabetic participants are given in Table 1. According to this table, 63% of households had food security, 23.9% had food insecurity without hunger, 10.2% had insecurity with moderate hunger, and 2.9% had insecurity with severe hunger. The majority of patients with good control had food security (72.8%) and the majority of patients with poor diabetes control had some degrees of food insecurity (53.8%).  Table 2 shows the characteristics of participants according to food security status. According to the table, there was a signi cant relationship between food security and marital status, so that the percentage of married people was higher in the food secure group than in insecure group. There was also a signi cant association between food security and economic status, education, HbA1c, and body mass index, with a higher percentage of poor, sub-diploma, patients with uncontrolled diabetes (HbA1c≥ 8.5%) and overweight and obese patients in the food insecure group compared with the secure group. Weight, body mass index, fasting blood sugar and HbA1c were also signi cantly higher in the food insecure group than in the secure group.  Table 3 shows the mean values of self-care subscales in food secure and food insecure groups.
Following a healthy diet and doing regular physical activity was signi cantly higher in the food secure group than in the food insecure group. However, the overall self-care score was not signi cantly different between the two groups. Table 4 shows the relationship between poor glycemic control (HbA1c≥ 8.5%) and food insecurity using logistics regression tests. With increasing degrees of food insecurity, the probability of poor glycemic control increased. In the crude model, poor glycemic control had a signi cant direct association with moderate and severe insecurity (OR: 1.15; CI: 0.74-1.35 and OR: 2.30; CI: 0.62-2.63; respectively). However, after adjustment for confounders, only the association between severe insecurity and poor glycemic control remained signi cant (OR: 3.02; CI: 1.45-2.65). Table 5 shows the relationship between self-care and food insecurity with HbA1c using linear regression.
After adjustment for confounders, it was observed that self-care has a signi cant inverse relationship with HbA1c levels (P= 0.016) and food insecurity had a direct relationship with HbA1c levels (P= 0.009). That is HbA1c decrease by 0.68 units given a unit increment in self-care score, and increased by 0.41 units given a unit increment in self-care score.

Discussion
In this sample of Iranian adults with type 2 diabetes, 37% of patients were food-insecure. The majority of patients with well-controlled diabetes were food-secure and the majority of patients with poor glycemic control had some degrees of food insecurity. On the other hand, there was a signi cant association between food insecurity and glycemic control. As food insecurity increased, the likelihood of poor glycemic control increased. This association with sever insecurity remained signi cant after adjustment for confounding factors. In this study, although food security was not signi cantly associated with overall self-care scores, "following a healthy diet" and "doing regular physical activity" behaviors were signi cantly higher in the food secure group than in the food insecure group. There was also a signi cant inverse association between self-care and HbA1c, as well as a positive association between food insecurity and HbA1c.
Findings on the prevalence of food insecurity in diabetic patients referred to Eghlid health center showed that 37% of households had food insecurity. In a cross-sectional study in Shiraz on 135 newly diagnosed diabetic patients, the prevalence of food insecurity was 66.7% (32). In a study in Canada, the relationship between food insecurity and diabetes was examined and the results showed that food insecurity was more prevalent among diabetics than non-diabetics (9.3% vs. 6.8%) (39). A study by Mohammad Aria et al. showed that the prevalence of food insecurity was 85.3% in diabetics and 67.1% in non-diabetics (31).
It seems that the reason for the difference in the prevalence of food insecurity in different studies is due to differences in the population in terms of geographical, socio-economic and cultural conditions as well as food habits.
With increasing food insecurity, the probability of poor glycemic control increased. Other studies in this eld have also reported a signi cant association between food insecurity and decreased glycemic control (40)(41)(42)(43). Adults with food insecurity provide most of their diet through high-calorie and low-nutrient foods, which diabetic patients are recommended to avoid, such as re ned carbohydrates, added sugar and fat. These foods are cheaper than balanced calorie foods such as fruits, vegetables, and dairy products (44, 45). People with food insecurity are at risk of diets with less variety, lower quality, lack of micronutrients, and low intake of fruits and vegetables, which can lead to poor glycemic control in diabetics.
In self-care scales, following a healthy diet and doing regular exercise was signi cantly higher among food secure group than food insecure group, but the overall self-care score was not signi cantly different between the groups. In some studies, there was no association between food insecurity and diabetes selfcare (29,46). While in some other studies, there was a signi cant relationship between food insecurity and lack of self-care behaviors (22,27,47).
In the present study, self-care was inversely related to HbA1c and food insecurity was directly associated with HbA1c. Similar results were found in some other studies (48). Improving self-care behaviors is the rst step in helping patients better control their disease. It also emphasizes the design and strengthening of interventions related to self-care behaviors and helps caregivers to better control the patient and reduces the associated complications.
One of the weaknesses of this study is the cross-sectional design of the present study, so we will not be able to interpret the results of this study in the form of a causal relationship. The use of self-administered questionnaires is another weakness of this study, and therefore the actual behaviors of patients may not be reported. In this study, we could control for important confounders. Also, using the USDA food security questionnaire, which is one of the best options for measuring household food security, is another strength of this study.

Conclusion
Results of the present study showed that food-insecurity has a direct association with poor glycemic control in type 2 diabetes patients. On the other hand, food-secure group had higher scores of following healthy diet and physical activity. Since economic status was signi cantly better in the food-secure group, it can be postulated that nancial problems will eventually lead to non-compliance with the recommendations and therefore poor glycemic control. Policy strategies to increase access to diabetesappropriate foods may reduce socioeconomic inequalities in glycemic control.

Competing interests
The authors declare that they have no competing interests.

Funding
This research received no speci c Grant from any funding agency in the public, commercial, or not-forpro t sectors.
Contributions MM and AA-J designed and conceived the research. MM recruited participants and collected data. AD analyzed the data and interpreted the results. MM, AA-J and AD drafted the manuscript. All authors read and approved the nal manuscript.