INTRODUCTION

Diabetes Mellitus (DM) is one of the four major non- communicable diseases (NCDs) causing a high morbidity and mortality, globally. It is a metabolic disorder of multiple eti- ologies characterized by chronic hyperglycemia induced from defects of insulin secretion and action or both (1).

Long standing elevated blood glucose leads to micro and macro vascular complications (2) and becomes a serious health problem unless early screened (3). Complication from undiagnosed diabetes could lead to significant decline in quality of life (4) and have a higher risk for premature death

(5) unless prevented.

Globally, the magnitude of diabetes has been increasing among adults; According to International diabetes Federation Atlas report, as of 2017, there are451 million people living with diabetes, with projections as tall as 693 million by 2045

(6). Domestically, the prevalence of diabetes is higher in urban than rural areas (7).

In 2014, about 179.2 million people lived with undiagnosed DM worldwide with Africa having the highest percentage compared to other regions; about 62.3% of the people with the diseases do not know the effects, and about 13.4 million were undiagnosed (8; 9; 10).

In Ethiopia, the magnitude of diabetes mellitus is increasing. According to the WHO report, the number of cases docu- mented in 2000 (800,000), is rising and that it would hit an estimated 1.8 million by 2030(11, 12). Evidence from studies conducted in Ethiopia: in Gondar and Bahir Dar city were 2.3% and 10.2% individuals lived with undiagnosed DM, respectively (13, 14). Another study conducted in 2014 in Ethiopia showed, about 1,603,100 people (75.1% of popula- tion) were undiagnosed for diabetes mellitus (1, 15).

However, different factors, not quite understood by the com- munity, contributed to risk of diabetes development. Alt- hough undiagnosed diabetes is prevalent, it was not ad- dressed well in west Ethiopia. So far, nothing has been done at community level. Therefore, this study aims to investigate the magnitude and associated factors of undiagnosed diabetes among middle aged adult urban residents in west Ethiopia.

METHODS

Study design and setting:

Acommunity-based cross-sectional study was conducted purposively in the hub of western Ethiopian Town, Nekemte, which is located 328km from Addis Ababa. It is divided into

6 sub cities administratively with an altitude ranging from 1960 to 2170 Meters above sea level. Its average annual rain- fall and temperature ranges are 1854.9mm and 14oc to 26oc, respectively. The total population of the city projection in 2017 was estimated to be 117,819, of which 51 % (60,088) of them were adults.

Study period: Study was conducted on 266 people in their mid -adulthood from 1March 2019-1 August 2019.

Sample size: The minimum sample size was calculated using single proportion formula, by taking the prevalence of ab- dominal obesity the most common component of metabolic syndrome with 19.6% among healthy Ethiopian adults (16). Hence with a margin of error of 5%, confidence level of 95%, and 10% gnawing away, we had minimum sample of 266 par- ticipants.

Sampling techniques: Within decision the appropriate sam- pling method was identified for specific area and study partici- pants. Accordingly, randomly one commune/kebele was select- ed by lottery method from six kebeles and one kebele not adja- cent to the other was selected purposively. Totally two kebeles were selected. Each study participants were selected through simple random sampling techniques.

Eligibility: Adults aged 40-65years who were eligible to par- ticipate in the study were asked to undergo diagnosis and re- spond questionnaires to be included in the study. While who were already on medication for NCDs, pregnant, lactating, serious mental conditions, bariatric surgery and physically dis- ables were excluded.

Measurements: Data was collected using structured self- administered questionnaire, and anthropometric measurements take of each participant. Fasting blood sugar (FBS) level was determined using samples taken early in the morning, with readings ≥126 mg/dLbeing classified as diabetes and 100- 125mg/dl, pre-diabetes. In addition, other biomarkers like cho- lesterol level and blood pressure other biomarkers like choles- terol level and blood pressure were also collected to assess common associated risk factors of diabetes.

64

Analysis: The data was analyzed using SPSS version 24 (IBM corporation, NY, USA). Frequency, percentage, and de- scriptive summaries were used to explain the amount of study participants in the analysis. Descriptive statistics was used to summarize and describe various sample characteristics as well as the association between high blood glucose and other risk factors. The binary regression computed the crude OR and vari- ables with p values less than 0.2 were entered into multivaria- ble logistic regression model to control potential confounding effects in the model. The strength of associations between inde- pendent and outcome variables was assessed using AOR with a 95% CI and p values ≤0.05 were considered statistically signifi- cant predictors of undiagnosed DM.

Ethical review and confidentiality:

Permission was sought from the Institutional Review Board (IRB), Institute of Health, Jimma University (Approval No.IHRPGD/596/2019) to conduct this study. The households willing to participate in the study signed consent form. Confi- dentiality of the respondents was ensured, and each household had its own identification number. Subjects were free to partici- pate in the study without any coercion.

RESULTS

Socio-demographic and lifestyle characteristics: Out of two hundred sixty-six undiagnosed participants, majority (62.8%) of them were females and more than half (54.89%) were living below poverty threshold (<1.25dollar/day). The average age of adults in our study was 52.2 years, with participants aged 41-48 years accounting for 54.5%.Regarding lifestyle, majority (75.2%) of them live a sedentary life, and about 40.6% have fragmented sleep types, 24.8% had history of alcohol intake, 1.1% are current khat chewers and 2.3% smoke cigarette. (Table 1).

Prevalence of undiagnosed Diabetes:

The prevalence of pre-diabetes (impaired fasting blood glu- cose) and diabetes of the participants was 4.89 % and 2.25 %, respectively (Table 2). Participants with elevated fasting blood sugar (FBS>126mg/dl) were linked to Wollega Univer- sity Specialized Hospital chronic care unit for further diagno- sis and follow up.

Table2. Description of fasting blood sugar by sex, west Ethiopia, 2019 (n=266)

Factors associated with undiagnosed diabetes mellitus: From 266 participants, 7.14 % were newly diagnosed, of which the actual diabetes accounts for 2.25%. The magnitude of diabetes is shown to significantly be associated with sleep disorders, sedentary lifestyle, increased: waist circumference, waist to height ratio, BMI, blood pressure, TG and HDL on binary analysis. The multivariate logistic regression analysis showed that only sleep related problems, increased BMI and high blood pressure were independently associated with diabetes (Table 3).

The mean fasting blood glucose level was 99.7(29.60 mg/dl) with (95%CL: 96.12, 103.27; p<0.0001). The prevalence of diabetes significantly increased with high BMI (6%) when compared to participants with BMI <25 Kg/m2 (1.1%) by a factor of AOR: (4.87 (1.01, 23.45), P=0.048). More than half (69.3 %) of the study participants have central obesity (high waist circumference).

66

The prevalence of diabetes was shown to be higher in these participants (7.14 %) as compared to 0.8 % of the participants with normal/low waist circumference (AOR=1.61 (1.14, 18.53), P =0.702). However, it was shown not to be significantly asso- ciated (Table 3).

In addition, our study revealed those participants with sleep apnea had 3.5 (OR=41.37 CI= (1.02, 11.81, p=0.046) times higher chance of having diabetes than those with normal range of sleeping hours. Participants with sleep apnea AOR: (0.19 (0.05, 0.80), P=0.023) and elevated blood pressure >130/85mmHg AOR: (5.22 (1.67, 16.33), P=0.005) were sig- nificantly associated with undiagnosed DM (Table 3).

DISCUSSION

The current magnitude of diabetes mellitus is 7.14%. This re- sult is slightly higher than the estimated Ethiopian prevalence of DM by IDFA (5.2%) (17). and studies conducted on some urban residents of Ethiopia like Gonder city (5.1%) (17), Des- sie Town (6.8%), (18),Mizan-Aman Town (6.5%) (19), and in Hosana, south Ethiopia (5.7%) (20).

Contrary to the above comparison, the magnitude of undiag- nosed diabetes is low when compared with a study conducted on 2013 on HIV/AIDS patients taking HAART in Ethiopia (8% )(21), whereas, in Jimma town 15% had Impaired Glucose Tolerances (12). Likewise, the prevalence of undiagnosed DM was lower than studies done in North India, Punjab (8.3%)

(22), Pakistan (26.3%)(23) , Bangladeshi (9.7%) (24) and pre- vious studies in African Countries( 25,26, 27,28,) This differ- ence might be due to variations in socio-demographic and life- style behavior factors. Different scholars agree that a sleep dis- order is highly associated with diabetes. For instance, diabetic patients often have a high prevalence of obstructive sleep apnea (OSA) (29). Clinical studies have shown an increase in serum glucose in patients with OSA, independent of obesity (30, 31). In this study, we observed an independent association (P=0.023) between high fasting blood glucose and sleep apnea.

CONCLUSIONS

The magnitude of undiagnosed diabetes mellitus among adult urban residents was found to be high. On multivariate analysis it was shown that having a high body mass index, sleep disor- der and elevated blood pressure were significantly associated

with diabetes mellitus. Therefore, age targeted community- based education on early detection and prevention of diabetes, as well as its complications are significant to save adult life.

67

ACKNOWLEDGEMENT

We thank Jimma University, participants, data collectors, Che- leleki health center, Wollega University specialized hospital, and Nekemte municipal for their cooperation for the study.

Competing Interest:

All the authors declare that they have neither financial nor non- financial competing interests.