Study area and period

Institution based cross sectional study was conducted from March 19 to May 19, 2018, in Ilu Abba Bora Zone of Oromia Region,south western Ethiopia. Metu Town is the administrative center of the zone which is 600 km south west of Addis Ababa. The study was conducted in two governmental hospitals (Metu Karl Referral Hospital and Darimu Hospital) both gives follow up care for diabetic and NCD patients.

In the current study, all  adults  diagnosed  with  type  2  diabetes mellitus who were on regular follow up at MKRH and Darimu hospital NCD follow up units were considered as a source population and selected adults diagnosed with type 2 diabetes mellitus who were on regular follow up at MKRH and Darimu hospital NCD follow up units were a study populations. Critically ill and patients with severe mental illness were excluded from the study.

Sample size determination and sampling technique

Sample size was separately calculated for the outcome (Gycemic control) and for each associated factors using previously published parameters and the largest sample size was taken using the assumption of 50% poor glycemic level which was reported in a study conducted another part of Ethiopia (Minyahil A et al., 2014) 95% level of confidence (α), 80% power (ß) and 10% non-response rate, the final sample size was 422. The final sample size was allocated proportionally to two hospitals based on their number of diabetic patients on follow-up accordingly 756 and 338 patients were in a regular follow-up in MKRH and darimu hospital respectively accordingly 291 and 132 patients were proportionally taken from Metukarl hospital and Darimu Hospital respectively. Finally study participants were selected using systematic random sampling technique. 

Data collection method and tools

Semi-structured interviewer administered questionnaire was used to collect socio-demographic. Patients health profile, including; duration of disease, type of treatment, co morbidity, family history, and Fasting blood sugar was reviewed from patient follow up card. Diabetic health belief  was assessed using 26 questions,  5 questions each  assessing the perceived susceptibility to diabetes complications and perceived severity constructs,  perceived benefit and barrier to  diet and physical regimen was assessed using 4 and  8 questions respectively likewise self-efficacy towards following dietary and physical activity recommendation was assessed by 4 questions. The other section of the tool assess emotional and active or instrumental family and non-family members support which was modified from “The Diabetes Social Support Questionnaire-Family Version: developed in 2002 (Greca et al., 2002). Dietary adherence status was assessed using 12 item scale (Worku et al., 2015) which is  modified from eight-item Morisky medication adherence scale (MMAS-8) which was developed in 2008 (Morisky, 2011; Woods et al., 2011) )(World Health Organization(WHO), 2009).

Data processing and analysis

Data were cleaned, edited, coded and entered to Epi data version 3.1 and was exported to SPSS windows version 20 for analysis. Exploratory data analysis and Descriptive statistics including proportion, percentage, ratios, frequency distribution, mean and standard deviation was used to describe the data. Wealth index was constructed using the principal component analysis(PCA) form 27 items after checking all assumptions. Bivariate logistic regression analysis were done to see the association between individual explanatory and outcome variables; variables with P-value <0.25 were a candidate for multi variable logistic regression. Odds ratio with 95% C.I was used to measure the strength of association between dependent and independent variables. P value <0.05 was used to declare level of statistical significance.
The scores of each diabetic health belief construct were summed up, to come up with a single scoring for each construct, for every respondent. Participants were labeled to have high  or  low  level  of individual constructs based on correct response of mean value as a cutoff, social support scores of every respondent  were also summed up. Participants were labeled to have poor or good social support based on mean value.  Dietary adherence status score was computed by taking the mean values of 12 questions assessing patients’ dietary recommendation adherence.

Ethical consideration

Ethical clearance and permission were obtained from the Research Review Board of Jimma University and offered to illuababor zone Health office and respectivehospital administrative officials, information about the objective of the study, confidentiality, autonomy and justice were explained for the participants during data collection. Informed verbal consent was obtained from each study participant.

Operational and standard definitions

Good glycemic control

Fasting blood glucose of 70–130 mg/dl (29).

Poor glycemic control

Fasting blood glucose of<70 mg/dl and >130mg/dl (29).

Acute glycemic control

In the current study, FBS measured during the last follow up visit was considered as recent glycemic control.

Adult

A person who has reached 18 years old.

Diet

This refers to the recommended diet for people with DM who should eat small meals spread throughout the day (at least 5/day), eat fruits and vegetables daily, and eat foods high in fibers and whole grain but low in fats, sugars and carbohydrates that have high glycemic index (30).

Physical activity

This refers to bodily movement produced by the contraction of skeletal muscle that requires energy expenditure in excess of resting energy expenditure (Westerterp, 2013).

Non adherent to dietary recommendation

These are those who scored the mean and above the overall dietary recommendation score.

Adherent to dietary recommendation

These are those who scored below the overall mean dietary recommendation score.

Clinical characteristics of study participants

The mean illness duration since diagnosis was 6.9 (SD±5.3) years. One hundred and eighty two (46.4%) patients reported that they had family history of DM and nearly half of the respondents (46.9%) had additional co-morbidity. More than half of patients (52.3%) used oral hypoglycemic agent. One hundred and sixty nine (43.1%) of patients never missed their diabetic follow up appointment in the past three month. Additionally more than one fourth of participants, 105 (26.8%) reported that they did not attend diabetic education while 47 (12%) attended regularly. In the current study, one fourth (25.5%) of diabetic patients were member of Ethiopian diabetic association, the mean BMI was 27 (SD±13) and 44.1% of respondents had normal BMI while 19.6% of respondents were obese. Furthermore 62 (17.3%) of the respondents had abdominal obesity (Table 2 and Figure 1).

In the present study, level of glycemic control was assessed using FBS; the mean FBS was 183.4 mg/dl and accordingly 64.5 % of respondents had inadequate glycemic control (Figure 2).

Diabetic patient’s dietary and physical activity recommendation adherence

The  present  study  showed  51.8%  of  the  respondents were non-adherent to dietary recommendation while 38% of participants were non adherent for physical activity recommendations and overall non adherence accounted for 67.9% (Figure 3).

In the present study patients diabetic health beliefs was assessed, nearly three fourth of participants had high perceived susceptibility, while diabetic patients with high perceived severity, perceived barrier and self-efficacy accounts 54.6, 43.9 and 60.7% respectively (Table 3). 

In multivariable logistic regression analysis, patient’s occupational  status,  family history of DM, co-morbidities, physical activity recommendation adherence and written diabetic instruction were an independent predictors of diabetic diabetic Acute glycemic control (Table 4). Patients who were government employers had less odds than unemployed patients [AOR=0.39 (0.22-0.68)], similarly patients who had diabetic family history had 0.34 less odds than patients who did not know their family history [AOR=0.34 (0.18-0.62)]; patients who had another co-morbidities were 3.7 times likely to have poor glycemic control [AOR= 3.72(1.39-9.94)] furthermore patients who were adherent to physical activity recommendations were 54% less likely to have poor glycemic control.

Good glycemic  control  among  T2DM  patients  prevents short-term complications, decrease risk of long-term complications and decrease health care resource use and costs. In the current study, nearly two third (64.5%) of respondents had poor (inadequate) glycemic control.

This study also reported higher dietary and physical activity non adherence, and this could be one of the possible justification for the high proportion of poor glycemic control, on the other hand the result is comparable with a study finding from Gonder (Solomon et al., 2015), however  the finding is lower than findings from Tikurambesa, Malysiaand India (Ababa et al., 2018; Almutairi et al., 2013; Gopinath et al., 2013) and higher than a finding from northern Ethiopia, Mekelle, which reported 48.7% had poor glycemic control (Tadele et al., 2016), the discrepancy might arise fromin socio-cultural, economic and dietary consumption difference among study populations, in addition the variation might be due different blood glucose measurement modality used. In this study, fasting blood sugar was used to monitor patient’s blood glucose level while most of referenced studies used glycated hemoglobin, furthermore the variation might arise from difference in level of urbanization between conducted comparatively in less urbanized area. It is well known that urbanization is commonly associated with motorized way of life and unhealthy eating pattern like consumption of ready-to-eat food products containing high levels of salt, sugar and fat (Tadele et al., 2016), which are the common factors adversely affect patients glycemic control.

In the current study patients occupational status was found to be an independent predictor of patients Glycemic control, the result is in line with finding from Jimma University Teaching Hospital, which also indicated a positive correlation between patient’s occupational status and glycemic control. The observed association might be due to low educational achievements among unemployed patient’s; it is known that education offers opportunity to learn more about health generally and glycemic control particularly. Likewise, the result is also supported by low economic level of unemployed patients which directly interfere with patients capacity to afford medicine and strictly follow healthy dietary and physical activity recommendations.

The present study point out that patients who know their family diabetic history were found to less likely to have poor glycemic control; the result is incoherent with the study done in Saudi Arabia and Malaysia; both reported significantly higher proportion of poorly glycemic control among patients who know their family history of diabetes (Almutairi  et   al.,   2013;   Riyadh    and  Alzaheb, 2018); the disagreement might arise from the possible lifestyle and family attachment difference between study setups. In the current study area, family members commonly lived in a common shelter and there was also a strong social attachment which facilitates exchange of health information between family members, in view of the fact that diabetic self-care is a continuous activity, it’s likely for other family members to gain diabetic self-care knowledge’s. 

This study also revealed that presence of other co morbidities was associated with poor glycemic control; the association can be explained by the possible effect of co-morbidities in precipitating hyperglycemia among diabetic patients, likewise the result might also be due to likely diversion of patient’s attention from the routine diabetic self-care to management of acute co–morbidities symptoms, which likely impose adverse economic,  psychological   and  physical  pressure which likely had undesirable effect on patients adherence towards health recommendations and later leads to inadequate glycemic control.

In this study, adherence to physical activity recommendation was found to be an independent predictor of diabetic patients Acute glycemic control. The result is similar with the study conducted in Tepi (southern Ethiopia) (Biru, 2017); Netherlands (Jan-Willem VD and Luc JCV , 2015) and Yemen (Rita de Cássia, 2011). A healthy lifestyle with regular physical activity are very important tools in reaching and maintaining an adequate glycemic control in patients with diabetes. The benefits of exercise for glycemic control are largely explained by an increase in whole-body insulin sensitivity furthermore exercise also lowered glycemic variability throughout the day, indicating a decline in the frequency and/or amplitude of glucose fluctuations.

Although  this  study  provides basic information addressing knowledge gap, institutional based nature of the study might not infer for other diabetic patients. The cross-sectional nature of thestudy design does not show temporal relationship or causality, additionally in this study, contrary to scientific plausibility, diabetic health belief and patients dietary adherence were not found tobe predictors of patients glycemic level, this might  arise from sample size, study design used and some biases in particular recall and social desirability bias.