Poor sleep quality and its predictors among type 2 diabetes mellitus patients attending Jimma University Medical Center, Jimma, Ethiopia

Objectives The aim of this study is to determine the prevalence of poor sleep quality and its associated factors among people with type 2 diabetes mellitus at Jimma University Medical Center, Jimma, Ethiopia 2018. Comparative cross-sectional study was employed on 198 participants (99 cases and 99 controls). Data were collected using Pittsburgh Sleep Quality Index (PSQI) and analyzed using SPSS version 20. Variables with p value ≤ 0.05 in multivariable logistic regression were treated as significant predictors of poor sleep quality. Results The prevalence of poor sleep quality was 55.6% among people with type 2 diabetes mellitus and 32.3% among controls. Longer duration of diabetes since diagnosis (> 10 years) [AOR = 4.88 CI (1.27, 18.66) p = 0.021], co-morbid hypertension [AOR = 3.2, CI (1.16, 8.84) p = 0.025], poor glycemic control [AOR = 3.16 CI (1.2, 8.27), p = 0.02] and current khat chewing [AOR = 3.06, CI (1.04, 8.98), p = 0.042] were factors significantly associated with poor sleep quality. The prevalence of poor sleep quality was significantly higher among people with diabetes than those who didn’t have diabetes (controls). Poor sleep quality may bring about mental impairment and reduce working capacity of individuals with diabetes mellitus. Therefore, diabetes mellitus patients need to have heath education about risk factors for poor sleep quality.


Introduction
Sleep is a basic biologic function which is essential for life [1]. It is defined as unconsciousness from which the person can be aroused by sensory or other stimuli [2]. It is the time when the body secretes many important hormones that affect growth, sleep, regulate energy, metabolic and endocrine functions [3,4].
Good quality sleep is one of the most important things that we need to keep our body healthy which really requires adequate duration, appropriate timing and regularity, and the absence of sleep disorders [5]. Jemere et al. BMC Res Notes (2019) 12:488 which imposes a higher caregiver burden and impairing quality of life for both patients and their bed-partners. Insufficient sleep is associated with many negative health outcomes and increased risk of mortality. It impairs cognitive performance, and substantially reduces workplace productivity through absenteeism [15].
According to the findings of different studies, female gender, low income, longer disease duration, poor glycemic control, and presence of hypertension increases the risk of poor sleep quality among people with type 2 diabetes [10,[16][17][18].
Even though DM is associated with many health problems, there is scarcity of data that determine the magnitude and determinants of poor sleep quality among people with type 2 diabetes. Therefore, we conduct this study so as to fill the above gap by determining the prevalence of sleep quality and its associated factors among people with type 2 diabetes.

Study design, area and period
Institution based comparative cross-sectional study was conducted at Jimma University Medical Center (JUMC) from April 5 to May 5, 2018. Jimma University Medical Center is one of the oldest university hospital in Ethiopia, which serves as a referral unit and training center for both medical and health science students. The Hospital serves for a total of 2594 people with type 2 diabetes during the time of data collection.

Source and study population
All adult Type 2 DM patients who were enrolled to JUMC and healthy individuals who came to the hospital for routine purpose were the source populations. Type 2 DM patients who were attending JUMC and age and sex matched relatively healthy individuals present at the time of data collection were the study populations.

Eligibility criteria
All adult cases who had complete medical records and controls who didn't have family history of DM were included. Those people with type 2 diabetes who had diabetes ketoacidosis, psychiatric disorder, other chronic illnesses, working in night shifts in the last month, and those pregnant and lactating mothers were excluded.

Sample size and sampling procedure
The sample size of the study was calculated using two population proportion formula, by considering the proportion of poor sleep quality among cases (P1 = 53.4%) and controls (P2 = 29.5%) in another similar study [8], 95% confidence level and 90% power. After adding 10% non-response rate, the sample size for each group was 99, yielding the total sample size of 198. Consecutive sampling technique were used to select cases and convenient sampling technique for controls.

Data collection procedure
Interviewer administered questionnaire was used to collect the data. The PSQI which provides information about sleep quality over a month's time was used for sleep quality. It is a 19-item self-report measure which has seven subscales. These subscales are added in order to determine a global sleep quality score (GSQ). The GSQ score ranges from 0 to 21, in which higher scores (PSQI > 05) indicate poor sleep quality [19]. Weight and height were measured using digital weight scale and stadiometer respectively. Blood pressure (BP) was also measured using automated digital BP monitor. Data were collected by two trained BSc nurses.

Operational definitions
Good glycemic control: a 3 month average fasting blood glucose measurement between 70 and 130 mg/dL. Poor glycemic control: average blood glucose measurements on three consecutive visits is > 130 or < 70 mg/dL.
Individuals who use any of the substances such alcohol, khat, and cigarette smoking as at least once in their life time were classified as ever user. Participants who consumed psycho-stimulant substances such as alcohol, khat, and cigarette smoking at least once within the last 30 days was classified as current user.

Data analysis procedure
Data were checked for completeness, entered into Epidata version 3.1 and exported to SPSS version 20.0 for analysis. Bi-variable and multivariable logistic regression analysis were done to determine the association between different explanatory variables and poor sleep quality. All variables with a p-value < 0.25 in the bi-variable analysis were entered into multivariable logistic regression model. We used p < 0.25 as a cutoff point to select candidate variables of the final model so as to improve the chances of retaining meaningful confounders. Odds ratio and its 95% confidence interval were estimated for potential predictors of poor sleep quality, which were included in the final model. A p-value ≤ 0.05 was used to declare statistically significant.

Data quality control
To assure data quality, pre-tested interviewer administered questionnaire was used to collect data. A 2 day training was given for data collectors regarding the purpose of the study and measurement techniques. The questioner was translated to Afaan Oromo and Amharic language and then retranslated back to English to maintain its consistency. To minimize bias, respondents were assured to keep their response confidential.

Results
A total of 198 participants (99 cases and 99 controls) were included in the study making a response rate of 100%. The male to female ratio was 1.11. The mean age of Type 2 DM patients was 50.14 years and that of controls was 49.9 years (p = 0.872). More than one-third (35.9%) of respondents were in the age range of 40 to 50 years. Twenty-five percent of people with T2DM and 26.3% controls had no formal education. Majority, 83.8% of T2DM patients and 76.8% of controls were married. Regarding monthly income, the average monthly income of the controls was higher than the cases (Table 1). and 11 (11.1%) of controls had over-weight. Eight (8.1%) of cases and 2 (2%) of controls were found to be obese. One-third (33.3%) of the cases and 3 (3%) of controls had hypertension at the time of data collection.

Prevalence of poor sleep quality
The overall prevalence of poor sleep quality was 55.6% for cases 95% CI (45.5, 64.5) and 32.3% for that of controls 95% CI (23.2, 41.4). The mean GSQ score was 6.58 for cases and 3.61 for controls (p = 0.000). The average bed time of the participants was 10:01 pm for T2DM patients and 10:05 pm for non-diabetes individuals. The mean total time to fall asleep was 33 min for cases and 21 min for controls. The mean actual sleep time 6.2 h for T2DM group and 6.9 h for controls.
The seven components of sleep quality were also assessed. Accordingly, 16 (16.2%) of cases and 4 (4%) of controls rate their subjective sleep quality as very bad.
Majority, (78.8%) of cases and 65.6% of controls had ≤ 7 h of sleep per night. The habitual sleep efficiency was < 65% in 26.3% of cases and in 6.1% of controls ( Table 2).

Predictors of poor sleep quality
The independent predictors of poor sleep quality among T2DM patients were: duration of DM > 10 years [ Table 3).

Discussion
This study revealed a high prevalence of poor sleep quality among T2DM patients (55.6%) than non-diabetes (controls) (32.3%). This is in agreement with studies conducted in USA (55%) [20], Korea (49%) [21], Kenya (53.4%) [8], and Trinidad and Tobago, (63.9%) [22]. However, this finding is higher than the finding of another similar study in Iran (38%) [9]. The result of this study is lower than studies conducted in Turkey (77.4%) [16], Saudi Arabia (72%) [17], USA (84%) [14] and Sudan (97.1%) [13]. The possible reason for this discrepancy might be difference in PSQI cutoff point, socioeconomic demands and recruitment criteria. This study also revealed that, prevalence of poor sleep quality differed significantly according to disease duration and glycemic control. Type 2 DM patients who had longer disease duration (> 10 years) and/or poor glycemic control (FBS > 130 mg/dL) had higher risk of poor sleep quality than those with shorter duration of disease and/ or good glycemic control. This finding is in line with the study conducted in China [10] and Saudi Arabia [17]. This might be related to the occurrence of diabetes complications as duration of diseases gets longer. Moreover, diabetes patients with poor glycemic control will develop nocturia, which leads to frequent awakenings, to result in poor sleep quality [23].
Those T2DM patients with co-morbid hypertension had increased risk of poor sleep quality than who didn't have hypertension. This is in line with the study conducted in Brazil [23] and Turkey [16]. This might be due to the increased risk of emotional disorders such as anxiety and depression among DM cases with comorbid hypertension compared to DM cases only, which are the risk factors of poor sleep quality [24]. Current khat chewers' participants had increased risk of poor sleep quality than non-chewers. This finding agrees with the study conducted in Jimma town [25]. This could be the result of sympathomimetic effects of khat [26]. Khat (Catha edulis) is a stimulant with effects similar to amphetamine, because the main active ingredient of khat (cathinone), is an amphetamine like substance [26]. Effects of cathinone are mediated through decreased dopamine uptake by nerve terminals, increased dopamine release and inhibition of monoaminoxidase [25,26]. Those all processes might lead to poor sleep quality, due to persistent stimulation of post synaptic neurons following high level of dopamine in synaptic cleft.
Increased secretion of catecholamine's might also further increase blood glucose level by activation of glycogenolysis in skeletal muscles and liver, which might further exacerbate the glycemic control to increase poor sleep quality. Moreover, pesticides like DDT sprayed on khat, might inhibit pancreatic secretary activity by reducing calcium permeability of β cells, which would also elevate blood glucose level and result in an increased risk of poor sleep quality [25][26][27][28][29].

Limitations of the study
First, the sample size was limited. Second, glycemic control was measured with FBS level rather than the glycated hemoglobin due to budget constraint. Despite these limitations, our study clearly showed the magnitude of poor sleep quality and its predictors in our country, which is not well investigated so far.