The aim of our study is to investigate how household socioeconomic status and neighborhood support system are associated with adherence to dietary recommendation among persons with T2DM in Ghana. Hospital based cross-sectional survey was conducted among 530 persons living with DM in Brong Ahafo Region (BAR), Ghana. Single population proportion formula
was used to determine the sample size for this study. The letter ‘n’ in the formula denotes the study sample size, ‘Z’ denotes normal standard distribution of 1.96 for 95% confidence interval, ‘P’ is the true population proportion of adherence to dietary recommendation among DM persons in the study area (Brong Ahafo region) and ’e’ is standard error (5%). Previous study in Brong Ahafo region Ghana, reported prevalence of adherence to dietary recommendation as 68.5% [11]. Substituting these values in the equation above, the sample size n was calculated as
However, for the event of non-response and registration error, a contingency sample of 60% was considered in the sampling, therefore the final sample was increased to 0.6*332=531.2≈532.
Individuals 18 years and above who were diagnosed with T2DM by physicians, using the American diabetes association (ADA) diagnostic and classification guideline 2011[12], and counseled to follow recommended dietary guidelines for at least 3 months and over were recruited into the study. Participants’ 70 years and above who could not answer interview questions, intellectually deficient, and severely ill were excluded from this study. Pregnant and lactating mothers were also excluded. Simple random sampling was used to select 6 hospitals, and the eligible participants consecutively recruited using systematic random sampling.
Ethical Approval
The study protocol was approved by Ghana Health Service Ethics Review Committee (GHS-ERC008/08/18) and Tehran University of Medical Sciences Ethics Review board (IR.TUMS.VCR.REC.1397.409). Each participant was requested to sign an informed consent form before participating. This research project was performed in accordance with the Declaration of Helsinki.
Assessing Patients Demographic Characteristics Anthropometry Measurements and Clinical Parameters
Age, diabetes-duration, medications intakes and other demographic characteristics were assessed using structured questionnaires. Weight and height were measured and recorded to the nearest 0.5kg and 0.5m using adult weighing scale and stadiometer respectively. These measurements were taken while participants were in light clothes without shoes, and were in standing position. Body mass index (BMI, kg/m2) was calculated by dividing weight in kilograms with height in meters square. Systolic and diastolic blood pressures were measured using manual sphygmomanometer and stethoscope, and the reading recorded to the nearest 0.5mmHg after participants were allowed to relax for 5 or more minutes.
Assessing Socioeconomic Status
We assessed participants’ socioeconomic status using composite wealth index. This proxy indicator was used because participants were unwilling to tell us their disposable household income they often earn through sales and salaries per month. Using this method, we asked participants to name items and properties they possess and use in their homes including fixed assets like land and building, and movable assets like vehicles. We then used principal component analysis (PCA) to extract participants’ socioeconomic status from this wealth index. The extracted socioeconomic status was categorized into three quintiles: - poorest, middle and richest quintiles to represent participants’ socioeconomic status. After the extraction, the percentage of total variance explained by the three factors was 35.6%
Neighborhood Support System
We assess neighborhood support system by using structured questionnaire. Participants were asked to self-report on a continue scale, how frequent they received support in the form of materials gifts, cash, in kind or volunteerism from friends, relatives, love ones, or from religious organizations like churches ,mosques, or from cooperate institutions in their societies. Participants who reported ‘’very frequently’’ were classified to have high social support, those who reported ‘’frequently’’ were said to have moderate social support, and those who reported ‘’less frequently’’ were said to have low social support system.
Assessing Dietary Intakes
We assess dietary intakes using three separate 24-hour dietary recall questionnaires on three different days (Monday, Wednesday, and Saturday). This was done to accurately predict participants’ dietary intakes in the three separate days and sum them up to form their average dietary intake. In the three different 24-hour dietary recalls, participants were asked to report details of all foods and beverages they took in each day preceding the interview. They were asked to report detail of the foods they ate, the preparation method, type of oil added, the portion size served and the actual amount they ate. The information obtained were summed up and analyzed with Ntri.IV software to obtain participants’ average foods and nutrients intakes.
Assessing Participants’ Alcohol intake
WHO 10-items alcohol use disorder identification scale (AUDITs-10)[13] was also used to assess Participants’ alcohol intake level. Participants’ were asked to respond to the 10 points in AUDITs scale, ranging from ’How often do you have drink containing Alcohol? To ‘’Has a relative, friend, doctor, or other health care worker ever been concerned about your drinking and suggested that you cut it down?’The responses obtained from them were also summed up to represent participants’ alcohol intake status. These questionnaires were also pretested among 20 participants (chronbach alpha of 0.55).
Assessing Participants’ Smoking status
Fagerström 6-iterms nicotine dependency test scale was used to assess participants smocking status[14]. Participants were asked to respond to 6-iterms in the Fagerström nicotine dependency test scale, ranging from ‘How soon after you wake up from bed do you smoke your first cigarette?’ to ‘Do you smoke even if you are so ill that you are in bed most of the day?’ The responses obtained were also summed up to form participants’ total smocking status. These questionnaires were also pretested among 20 people (chronbach alpha of 0.55).
Assessing Physical Activity levels
Participants’ physical activity (PA) level was measured using International Physical Activity short form Questionnaires (IPAQ). Response from participants’ for IPAQ were also categorized into ‘low’ (<600 metabolic equivalent (MET)/h per week), ‘moderate’ (between 600 and 3000 MET/h per week) and ‘high’ physical activity level (>3000 MET/h per week) according to the IPAQ scores[15] to represent PA levels.
Assessing Adherence to Recommended Dietary Guidelines
Perceived Dietary Adherence Questionnaires (PDAQ) for Persons with T2DM was used to assess adherence to dietary recommendation [16]. These questionnaires consist of nine items and seven point likert’s scale designed to generate information about adherence to recommended dietary guidelines among patients. These seven point likert’s scale questionnaires have a range between 0 and 7. Zero point mean non-adherence to the PDAQ, and 7 point means highest adherence. Participants’ responses from the nine items were summed up to form participants’ total adherence to dietary recommendation score. These questionnaires were pretested among 20 participants (chronbach alpha of 0.95).
Statistical analysis
IBM SPSS version 22.0 (SPSS, Chicago, IL, USA) was used in all data analysis. Data normal distributions were checked with Kolmogorov-Smirnov test. Descriptive statistics were used to describe participants’ demographic characteristics, while Pearson correlation used to test the correlation of these variables with socioeconomic status and adherence to dietary recommendation. Finally multiple linear regression models were used to assess the association of household socioeconomic status and neighborhood support system for adherence to dietary recommendation. Multiple linear regression model looks at the association between predictor variables on one dependent variable in an equation Y=a+b1X1+b2X2+b3X3+b4X4+b5X5+bkXk+e, where ‘a’ is the regression constant, b is the regression coefficient, X1…….Xk are the independent variables and ‘e’ is the variance of the population mean distributions. The assumption for using this statistic in our study is that our dependent variable (Adherence to dietary recommendation) is normally distributed and has equal variance around the mean. The independent variables also have linear relationships with no multicollinearity. Furthermore, our sample size is fairly large and could be said to have fair representation of the larger population. In our analysis, we set all variables significant at 0.05 alpha levels.