Cardiometabolic syndrome and associated factors among Ethiopian public servants, Addis Ababa, Ethiopia

Non-communicable diseases (NCDs) are increasingly becoming the global cause of premature death encompassing cardiovascular diseases (CVDs), cancer, respiratory diseases and diabetes mellitus. However, cardiometabolic risk factors in the general population, especially among the high-risk groups have rarely been assessed in Ethiopia. The study aimed to assess the prevalence of metabolic syndrome, its components and associated factors among staff in the Ethiopian Public Health Institute (EPHI). An institutional-based cross-section study was conducted from March to June 2018 among EPHI staff members. A total of 450 study participants were involved in the study, and the World Health Organization NCD STEPS survey instrument version 3.1 was used for the assessment. The biochemical parameters were analyzed by using COBAS 6000 analyzer. Statistical package for the social science (SPSS) version 20 was used for data analysis. Both bivariate and multivariate logistic regression analyses were used to identify associated risk factors. p value < 0.05 was considered for statistical significance. The overall prevalence of metabolic syndrome was 27.6% and 16.7% according to IDF and NCEP criteria respectively, with males having greater prevalence than females (35.8% vs 19.4%). Central obesity, low high-density lipoprotein (HDL) and hypertension had a prevalence of 80.2%, 41.3%, and 23.6%, respectively. In multivariate analysis increasing age and having a higher body mass index (25–29.9) were significantly associated with metabolic syndromes. The magnitude of metabolic syndrome was relatively high among public employees. Preventive intervention measures should be designed on the modification of lifestyle, nutrition and physical activities, and early screening for early identification of cardiometabolic risks factors should be practised to reduce the risk of developing cardiovascular diseases.

ethics and research committee (DRERC). All study participants provided written informed consent. The identity of participants was not revealed, and an identification number was allocated. All methods used were also performed by the relevant guidelines and regulations.

Results
General characteristics of the study participants. A total of 450 (232 males, 218 females) study participants were included from all staff members of EPHI. In this study with 46% and 36% of males and females, the study participants were between 29 and 39 years, respectively. Half of the study participants were married and completed college/University completed and around 24% of the study participants during the study period had less than 1500 birr income per month ( Table 2).

Distribution of behavioural, clinical and biochemical characteristics.
About two-thirds of the study participants declared a moderate level of physical activity, but about 4% were smokers, 67% consumed alcohol, 4% chew Khat, and less than 1% consumed fruits and vegetables according to WHO criteria. In this study, 2% had hyperglycemia (≥ 126 mg/dl), 24% had high blood pressure, 19% had high serum triglycerides level (≥ 150 mg/dl), and 25% had a high LDL cholesterol level (≥ 130 mg/dl). Significant differences were  www.nature.com/scientificreports/ observed between male and female and smoking status, alcohol intake status, khat chewing status, BMI, blood pressure, CO-based on IDF, and CO based on NCEP-ATPII (p < 0.05) (  (Table 4).
Associated factors with raised blood pressure, raised blood glucose, dyslipidemia and central obesity. The results of the logistic regression analyses for raised blood pressure, raised blood glucose, dyslipidemia, and central obesity based on IDF are presented in

Risk factors of metabolic syndrome.
In multivariate analysis, sex, age, BMI, raised blood glucose; raised blood pressure and dyslipidemia were shown to be significant risk factors for metabolic syndrome based on IDF criteria (

Discussion
MetS is a constellation of different risk factors associated with a 5-fold increase in the incidence of Type 2 diabetes and a 2-3-fold increase in the incidence of CVDs 16 . Based on our findings the prevalence of overweight or obese, of were found higher than other studies conducted in Ethiopia national survey (1.2% obese and 5.2% overweight) 17 , northern Ethiopia, Mekele (4.1% obese and 26% overweight) 18 and Northwest Ethiopia, Jimma obese (5.1%) and overweight (10.4%) 19 . Hypertension, the third most prevalent component (23.6%) for metabolic syndrome was higher than those reported in earlier studies 15.8% conducted in Ethiopia in 2015 national survey 17 , 9.3% at Gilgel Gibe field research center 20 and 20% in male and 14% in female among working adults in Addis Ababa 21 . Possible explanations for the difference are stress conditions, lifestyle and genetic differences, and alcohol consumption status might have contributing factors. Our study participants had a high prevalence in alcohol consumption status as compared to other studies. The association of alcohol consumption with an increased incidence of hypertension was explained by different studies 22,23 . This showed that appropriate interventions were needed to reduce the burden of alcohol use, which could help to lower blood pressure levels 24 . The prevalence of hypertension was also higher when compared to a study done in Angola, 17.9% 25  www.nature.com/scientificreports/ www.nature.com/scientificreports/ www.nature.com/scientificreports/  32 . This may be due to biochemical tests used to define the prevalence of diabetes. In our study, we had used only fasting blood glucose but the study done by Gebremariam et al. 18 used a combination of FBG and HgA1c, which results in observed prevalence differences.
Dyslipidemia, especially low HDL levels with 41.3% was the second most prevalent finding in our study participants. The prevalence of low HDL in our study is in line with other studies 31,33,34 . On the contrary, a higher prevalence of low HDL was observed in the Ethiopian national survey (68%) and finding among public employees in northern Ethiopia (71.3%) 18,32 . Environmental factors, physical activity status, nutrient intake and sample size and age of study participants may be used as part of an explanation for this difference.
Regarding the prevalence of hypertriglyceridemia, that is (19.3%) in our study is nearly similar to a result reported in the Ethiopian national survey. But higher prevalence is found in different studies 18,34,35 . Dietary intake, level of physical activity, lifestyle difference, and level of awareness may be part of a possible explanation for this variation.
Abdominal obesity drives the development of cardiometabolic risks through altered secretion of adipocyte-derived active substances called adipokines, including free fatty acids, adiponectin, interleukin-6, tumour necrosis factor-alpha, and plasminogen activator inhibitor-1, and through exacerbation of insulin resistance and associated cardiometabolic risk factors 36 . In the present study, we have found that elevation of waist circumference based on IDF criteria was the most prevalent (80.2%) that was the superior component to yield a larger magnitude for metabolic syndrome. This result is higher than the community-based study done among Andean highlanders (75.9%) 37 and the study done in South African Asian Indians who found a prevalence of (73.1%) even though harmonized criteria were used 38 . This may be due to differences in sample size, level of physical activity difference, and dietary intake. Concerning sex-difference, it is noted that males had a higher frequency of central obesity (87.9%). The reason for this difference may be the majority (65%) of female participants were younger as compared to males (35%) and central obesity increases with increasing age 39 .
Findings from this study showed that the prevalence of metabolic syndrome among staff members of EPHI was 16.7% using NCEP ATP III criteria while the IDF criteria yielded a higher prevalence of 27.6%. This higher prevalence of metabolic syndrome based on IDF criteria was due to a higher prevalence of central obesity which is one of the pre-request criteria for defining metabolic syndrome. Based on IDF criteria our result was fairly comparable to studies conducted in different regions and countries [40][41][42][43][44] . The prevalence of metabolic syndrome in our study was less than from other studies 35,[45][46][47] . Differences in the age of study subjects, sample size, socioeconomic status, residence & lifestyle, dietary intake, and physical activity may contribute to the different prevalence of metabolic syndrome in these different studies.
The high prevalence of metabolic syndrome has been linked to urbanization, westernization, nutritional and epidemiological transition 48 . Our result was also showed lower prevalence than the recent study conducted in Northern Ethiopia involving public employees in Mekele, which found a prevalence of metabolic syndrome was 40% using IDF criteria 18 . The explanation for this discordant may be due to the environmental and sampling methods in which we had used random sampling. However, the finding in this study was higher than other community-based studies using both NCEP ATP III and IDF criteria 17,19,21 . Our result had found comparably higher prevalence than studies conducted among adults in the rural area of West China (10.8%) and health professionals in Brazil (4.5%) 49,50 . This could be due to differences in socioeconomic backgrounds, lifestyle variations and ethnic differences. www.nature.com/scientificreports/ www.nature.com/scientificreports/ The result also showed that the prevalence of metabolic syndrome, based on IDF criteria was 35.8% in males and 18.8% in females. This was in line with the study reported in Colombia who observe that the prevalence of metabolic syndrome in males was three times higher than in females 33 . The possible explanation for the higher prevalence of metabolic syndrome in males is due to the majority of female participants are younger as compared to males 39 . We have also found older age was significantly associated with metabolic syndrome. The other possible explanation for higher metabolic syndrome prevalence in males can be because of central obesity which was the primarily prevalent component in males (72%) than females (28%). However, in contradiction with our result, other studies have reported a higher prevalence of metabolic syndrome among females. The prevalence of metabolic syndrome was also found significantly higher in older age, which is in line with other studies 19,51 . The reason is that ageing is characterized by a progressive deterioration in physiological functions and metabolic processes that generate reactive oxygen species as a by-product of biological oxidation. The oxidative damage of reactive oxygen species induces cellular dysfunction, which plays an important role in many pathological conditions like chronic low-level inflammation-induced metabolic syndrome 52   ]. These were also in line with other studies 19,33,51 . Overweight characterized by unbalanced energy intake and expenditure could result in continued elevation of blood glucose level 53,54 . Thus, it further results in hyper-secretion of insulin and leading to insulin resistance over time. Once insulin resistance occurs in different target organs for metabolic process dysregulation could be initiated such as lipid profile abnormalities, endothelial dysfunction, and inflammatory reactions 55,56 .
This result revealed that smoking habits, alcohol consumption, physical activity status and serving of fruit and vegetables per day were not individual predictors for metabolic syndrome. These findings were consistent with other researcher's reports 57,58 . Another finding was in contrary found that smoking, alcohol use, fruit, and vegetable consumption were statistically significant factors for metabolic syndrome 48 . The discordant between smoking and alcohol use with these research findings might be due to the amount and type of alcohol and smoking products taken by the study populations might be different. However, sex, age, BMI, raised blood pressure, raised blood glucose, dyslipidemia and raised hsCRP had statistical significance with metabolic syndrome in bivariate analysis. After adjusting confounders in logistic regression only age, BMI, elevated blood glucose, high blood pressure and dyslipidemia were independent predictors for metabolic syndrome. This was also in line with different studies 50,59 . Three fourth of the participants had at least one component for metabolic syndrome. The prevalence of central obesity expressed as increased waist circumference was the first ranked abnormality, followed by low HDL and raised blood pressure which is acquiesced with many researchers 35,60 . It is assumed that the luxurious lifestyle lies behind abdominal obesity and dyslipidemia for the most prevalent components of metabolic syndrome 61 . High prevalence of abdominal/central obesity, low HDL and raised blood pressure emphasizes the susceptibility of the study population to CVD and Type 2 DM, especially in older age. Controlling body mass and fat with better physical activity and an appropriate diet are important to reduce the risk of CVDs 62 . The prevalence also has been linked to urbanization, westernization, nutritional and epidemiological transition and this calls for urgent action by the policymakers and health managers to further emphasize the need for routine screening for all the components of Metabolic syndrome. www.nature.com/scientificreports/

Conclusion
From this study, it is possible to conclude the following: the prevalence of metabolic syndrome and its components were significantly high among the study population as compared to other studies like the country national survey. Central obesity, followed by dyslipidemia and hypertension were the most frequent components of metabolic syndrome. The prevalence of hypertension was found substantial as compared to the national survey report. Being male, over 39 years old, overweight, raised blood pressure elevated fasting blood glucose and dyslipidemia were significantly associated with metabolic syndrome. Twenty-four percent of the study participants were free from any risk factors for metabolic syndrome. About 16.7% of the study participants had ≥ 3 risk factors based on NCEP ATP III defining criteria.
Limitation of the study. The study employed a cross-sectional study design which could not conclude causality and effects. Moreover, this finding may not be generalized to a broader Ethiopian population since our study participants were an employee of a specific organization.

Data availability
The whole data supporting this study are included within the manuscript.