https://orcid.org/0000-0002-4906-9610
Figures
Abstract
Background
Alcohol consumption continues to be a public health problem in Ethiopia. Previous investigations have been conducted on alcohol consumption in Ethiopia; however, these investigations were limited to specific localities, which could not represent the existing alcohol consumption in different parts of Ethiopia. Besides, the spatial variation of alcohol consumption was not well investigated in the previous studies, which could hinder the implementation of effective intervention towards alcohol consumption. Hence, this study aimed to determine the spatial distribution and determinant factors of alcohol consumption in Ethiopia.
Methods
Secondary data from the 2016 Ethiopian demographic health survey was used in this study. A total of 44,023 weight samples were included using a stratified two-stage cluster sampling technique. The spatial variation of alcohol consumption was analyzed using ArcGIS version 10.7.1. The statistical significance of alcohol consumption clusters were identified using Kuldorff’s SaTScan version 10.2. A multi-level analysis was also conducted to identify factors associated with alcohol consumption using STATA version 14.
Results
In this study, 33.15% (95%CI: 32.5–33.8) of the study participants consumed alcohol with statistically significant spatial variation across regions of the country. Traditional religion (AOR = 13.7; 95%CI: 2.68–70.3), Regional variations (Amhara region—AOR = 3.56; 95%CI: 1.85–6.8, living in a low proportion of community literacy (AOR = 1.84; 95%CI: 1.1–3.18), cigarette smoking habit (AOR = 15.82; 95%CI: 4.31–58.1), and chewing Khat (AOR = 2.98; 95%CI: 1.22–7.27) were positively linked with alcohol consumption. Hot spot areas of alcohol consumption were found in Tigray, Amhara, and some parts of Oromia regions. The statistical significance of the primary clusters was also observed in Tigray and Amhara regions.
Conclusions
We found that one-third of Ethiopia’s population is consuming alcohol. Having a cigarette smoking habit, chewing khat, high proportion of community literacy, and traditional religion were associated factors for alcohol consumption. Therefore, the federal government of Ethiopia, and ministry of health, and other concerned bodies should work in collaboration to decrease the proportion of people consuming alcohol.
Citation: Daba C, Debela SA, Gasheya KA, Endawkie A, Gebrehiwot M (2025) Spatial variation and determinant factors of alcohol consumption in Ethiopia: Spatial and multilevel analysis of Ethiopian demographic and health survey. PLoS ONE 20(1): e0309943. https://doi.org/10.1371/journal.pone.0309943
Editor: Ayele Mamo Abebe, Debre Berhan University, ETHIOPIA
Received: January 15, 2024; Accepted: August 21, 2024; Published: January 8, 2025
Copyright: © 2025 Daba et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Data Availability: All relevant data are within the manuscript.
Funding: The author(s) received no specific funding for this work.
Competing interests: The authors have declared that no competing interests exist.
Introduction
Alcohol consumption is a significant cause of premature mortality and morbidity globally; causing 1.78 million deaths in 2020 [1]. The Global Burden of Disease (GBD) report indicates that about 1.34 billion individuals consumed harmful amounts of alcohol [1]. The consumption of alcohol at any level is linked to health impairments caused by various diseases, such as liver cirrhosis, breast cancer, tuberculosis, and injuries [2–4]. The World Health Organization (WHO) has identified more than 300 diseases and adverse conditions that can result from irresponsible alcohol consumption patterns [5]. Beyond health impairments, alcohol consumption is a significant cause of adverse pregnancy outcomes, such as low birth weight, preterm birth, spontaneous abortion and others, which need urgent interventions [6, 7].
The consumption of alcohol is more prevalent in Sub-Saharan Africa [8]. According to the WHO report, the magnitude of harmful alcohol consumption was found to be 78.9% in Sub-Saharan Africa [5]. Evidence from different countries also revealed that the magnitude of alcohol consumption is as high as 57.9%-76.0% in Nigeria [9, 10], 53.7% in Sri Lanka [11], and 51.4% in Uganda [12]. In Ethiopia, the prevalence of alcohol consumption varied from 16.7% to 72.6% [13, 14]. Evidence from a systematic review and meta-analysis revealed that the magnitude of alcohol consumption could be as high as 44.16% in Ethiopia [15] and 48.23% in Gondar [16]. Evidence from these studies showed that being male [9, 17], being cigarette smoker [18, 19], having lower socioeconomic status (SES) [20–22], and being a follower of a Christian religion (51.6%) [16] were determinant factors for alcohol consumption.
Although numerous investigations have been conducted on alcohol consumption [15, 16, 18], they are limited to specific locations. This could not represent the current alcohol consumption in Ethiopia. Besides, the spatial distribution of alcohol consumption was not well investigated in the previous studies, which could hinder the implementation of effective intervention strategies for alcohol related mortality and morbidity. Therefore this study aimed to answer the following research questions: (i) what is the magnitude of alcohol consumption in Ethiopia? (ii) What are the individual and community level factors which contribute to alcohol consumption in Ethiopia? (iii) Where are the hotspot areas of alcohol consumption in Ethiopia?. The results of this study will provide essential evidence that can inform substance abuse control program planners, policymakers, and healthcare providers. This evidence will be valuable in designing and implementing evidence-based interventions aimed at reducing the burden of alcohol consumption-related mortality and morbidity in Ethiopia and other similar settings.
Methods and materials
Study setting and data source
This study used secondary data from the 2016 Ethiopian Demographic Health Survey (EDHS). The EDHS data pool was conducted by the collaboration of EPHI, Federal Ministry of Health (FMoH) and Central Statistical Agency. This survey was conducted in all regions of Ethiopia namely Tigray, Amhara, Oromia, SNNPR (Southern Nations, Nationalities and People Region), Afar, Somali, Gambela, Benishangul Gumuz, Harari, Sidama, Southwest Ethiopia, and two administrative cities (Addis Ababa and Dire Dawa city) [23].
Sampling procedures and populations
A stratified two-stage cluster sampling method was applied to the 2016 EDHS. The process of stratification involved categorizing the eleven regional states and two city administrations of Ethiopia into urban and rural zones. A total of 84,915 Enumeration Areas (EA) were identified at the initial stage. Of these, 645 EA (202 in urban and 443 in rural areas) were included. At second stage, twenty-eight households per cluster were systematically selected. The enumeration areas that have no latitude and longitude were dropped. Moreover, detailed information about the sampling procedures is available on the DHS website in the 2016 EDHS report [23]. We incorporated the weighted sample of 44,022 study participants.
Dependent variable
The dependent variable of the current study was alcohol consumption using two EDHS questions. Have you ever taken a drink that contains alcohol? and “during the last 30 days, how many days did you have a drink that contains alcohol?”. Therefore, respondent who consumed alcohol were categorized as “Yes”, coded as “1” otherwise “No”, coded as “0”.
Independent variables
Individual independent factors, such as the sex of the respondent, age of the household head, religion, marital status, family size, chewing Khat, and cigarette smoking were included in this study. Similarly, community-level factors, such as community-level poverty, community-level media exposure, community-level literacy, residence, and region of the study participants were also considered.
Operational definition
Community-level literacy.
The proportion of household heads that are educated in primary school and above categories were summed and divided by the total in each cluster. Those household heads scored at or above the mean value of educational level were classified as high community-level literacy, while those scored below the mean score value were classified as low community-level literacy. The mean value was used as the cut-off point for classification because data was normally distributed [24].
Community-level media exposure.
Was determined using three survey questions: “how often do you have read newspaper or magazine?; how often do you listen to the radio?; and how often do you have watching television? The responses were “not at all”, “at least once a week” and “more than once a week” for each question. The proportions of household heads that are exposed to television, radio, and newspaper were summed and divided by the total media exposure value. Those household heads scored at or above the mean score of the media exposure were classified as high community-level media exposure, while those scored below the mean score value were classified as low community-level media exposure [25].
Community-level poverty.
Was generated by aggregating the individual characteristics in the clusters (poor, middle and rich). The proportion of household heads classified within the poor wealth index were summed and divided by the total in each cluster. Those households heads scored at or above the mean score value of the wealth index were classified a high poverty level, while those below the mean score value were categorized as low poverty level. The mean value was used as the cut-off point for classification because data was normally distributed.
Data management and analysis
The extracted data was cleaned, coded, and analyzed using STATA version 14. Sampling weights were also carried out before any analysis. Spatial autocorrelation, interpolation, and SatScan analyses were performed to determine the spatial distribution of alcohol consumption in different regions of Ethiopia. Besides, multilevel logistic regression analysis was employed to identify factors associated with alcohol consumption. Detailed procedures are indicated as follows. In the bivariable analysis, the variables with p-values less than 0.2 were candidate for a multivariable model [26].
Spatial analysis
ArcGIS version 10.7.1 was used to determine the spatial analysis of alcohol consumption among the 2016 EDHS clusters. The proportions of alcohol consumption cases in each survey cluster were calculated and appended to the latitude and longitude coordinates of the selected enumeration area. Global Moran’s I was used to determine whether alcohol consumption was dispersed, clustered, or randomly distributed in Ethiopia. Moran’s I value close to -1, + 1, and 0 indicate a dispersed, clustered, and random distribution of alcohol consumption, respectively. A p-value of Moran’s I less than 0.05 showed the presence of spatial autocorrelation. Getis-OrdGi* statistics was used to identify the hotspot area of alcohol consumptions and how spatial autocorrelation differs across study locations. The output with a high GI* implies a “hotspot”, whereas a low GI* denotes a “cold spot”. The spatial interpolation technique was also performed to predict alcohol consumption for un-sampled areas from the sampled cluster.
Spatial SaTscan analysis
Spatial Satscan statistics were performed using Kuldorff’s SaTscan version 10.1 software. Spatial scan statistics were employed using a Bernoulli-based model to identify statistically significant clusters of alcohol consumption among study participants [27]. Household heads who consumed alcohol were considered as cases and those household heads who did not consume alcohol were considered as controls. The result was reported using both table and figure and areas with a high log-likelihood ratio and p-value less than 0.05 were considered to high risk of alcohol consumption. The primary and secondary clusters were identified, assigned p-values, and ranked based on their likelihood ratio test based on the 999 Monte Carlo replications.
Multi-level analysis
Multilevel analysis was conducted with four models. The first model (I or null model) contains only outcome variable, which help to determine the variability of alcohol consumption in the community. Model II contains only individual-level factors whereas Model III contains only community-level factors. The final model (model IV) contains both individual and community-level factors. In multivariable analysis, the variables p-value less than 0.05 were considered as factors associated with alcohol consumption. The log-likelihood was used for model comparison and fitness and the results showed that Model IV was considered the best–fitted model. Besides, Intra-Class Correlation (ICC), median odds ratio (MOR), and proportional change in variance (PCV) were calculated to measure the variation of alcohol consumption among the household heads or clusters. The ICC is determined using the formula—, Where, VA represents the area-level variance [28–30]. The
, where VA donates the area level variance [30]. The PCV measures the proportion of total observed individual variation that can be attributed to differences between clusters. The PCV is calculated as—
, where, Vnull represents the variance of the initial model, while VA represents the variance of the model with more terms [30].
Ethical approval
An authorization letter to download EDHS dataset was obtained from DHS portal https://dhsprogram.com/data/dataset_admin/index.cfm) after requesting. The requested data was treated confidentially and was used only for the study purpose. No attempt was done to contact any individual respondent or household included in the survey. The dataset and all methods and data were performed based on Helsinki principles research guidelines.
Results
Socio-demographic characteristics
In this analysis, weight samples of 44,022 study participants were included. More than three-fourths (81.8%) of the study participants were male. The majority (88.3%) of the study participants were rural residents. 74.9% of the study participants had no formal education. Similarly, majority of the study participants were aged between 25–64 years (94.2%) followed by ≥ 65 years (4.4%) (Table 1).
Spatial autocorrelation of alcohol consumption
The prevalence of alcohol consumption spatially varied across regions of Ethiopia, as shown in the cluster pattern. We found a Global Moran’s I value of 0.4 with statistically significant variation (p < 0.001). The Z-score value was found to be 13.3 which indicates that there was less than a 1% probability that this clustered pattern has been the result of a random chance (Fig 1).
Prevalence and spatial distribution of alcohol consumption in Ethiopia
The prevalence of alcohol consumption in Ethiopia was found to be 33.15% (95%CI: 32.5–33.8). As shown in Fig 2, the spatial distribution of alcohol consumption varied across the region. The spatial distribution of alcohol consumption has been observed in the Sidama, SNNP, Amhara, Tigray, Gambela, Benishangul gumuz, and some parts of the Oromia region (Fig 2).
Prevalence of alcohol consumption within past 30 days
Out of lifetime alcohol use (14,595; 33.15%), 13,521 (30.7%) of the study participants were drink the alcohol within past 30 days. Out of 30.7% of drink alcohol within past 30 days, majority of study participants were drink the alcohol 1–10 days (58.3%) followed by 11–20 days (Table 2).
Hot spot analysis of alcohol consumption
Significant high proportion (hotspot) areas of alcohol consumption were indicated in red color and they were observed in Amhara, Tigray, Oromia, Benishangul Gumuz region, and Addis Ababa. The low proportion (cold spot) areas of alcohol consumption were observed in Somali, Sidama, SNNP, Gambela, Dire Dawa, and Harari regions of Ethiopia as indicated in blue color (Fig 3).
Spatial interpolation of alcohol consumption
According to ordinary Gaussian Kriging interpolation results, the predicted high proportion area for alcohol consumption was extremely high ranging from 63.5% - 79.8% in Amhara and Tigray regions, which was presented in red color. On the other hand, the lower predicted alcohol consumption has been observed in Somali, Afar, Harari, Dire Dawa, Sidama, and some parts of Oromia and SNNPR (Fig 4).
Spatial scan statistical analysis
A total of 257 clusters were statistically significant; of which 252 were most likely (primary) and 5 secondary clusters. The primary cluster spatial window (red color) was in Tigray and Amhara regions that were centered at 11.793026N, 37.584665E with 342.42km radius, a Relative Risk (RR) of 5.81 and Log-Likelihood Ratio of 5999.04, at p-value <0.001. It indicates that the household head living in the window red color had a 5.81 times higher risk of drinking alcohol than the household head living outside the window (Fig 5). Similarly, the secondary statistically significant clusters were found in Oromia and southwest Ethiopia regions that were centered at 9.584129N, 41.866112E with 0.46km radius, a relative risk of 2.38, and log-likelihood ratio of 27.22 at p-value < 0.001; which revealed that the household head living in the second window had 2.38 times higher risk of consuming alcohol than household head living outside of the spatial window (Table 3).
Multilevel logistic regression analysis
Random effect model analysis.
The log odds of alcohol consumption were statistically significant among the community. According to the intra-class correlation (ICC) in the null model, 86% of the variation in alcohol consumption was attributed to community-level factors. The null model also had the highest median odds ratio (MOR) value of 11.74, indicating that if we randomly select study participants from two different clusters, the study participants in the cluster with a high risk of alcohol consumption had 11.74 times higher odds of alcohol consumption than study participants in the cluster with a lower risk of alcohol consumption. Similarly, the percentage of PCV increases from 73.3% in Model I to 84.7% in Model III, which indicates that the third model was best in explaining the variations of alcohol consumption (Table 4).
Factors associated with alcohol consumption.
Based on the final model of multilevel multivariable logistic regression, religion, region, community level literacy, and behavioral factors, such as cigarette smoking and chewing Khat, were factors associated with alcohol consumption. The odds of alcohol consumption were thirteen times higher among traditional religion followers as compared to protestant religion followers (AOR = 13.7; 95%CI: 2.68–70.3). Household heads in the Amhara region were three times more likely to drink alcohol as compared to these in other regions (AOR = 3.56; 95%CI: 1.85–6.8). The odds of alcohol consumption were almost two times higher among these with low literacy as compared to participants with high literacy (AOR = 1.84; 95%CI: 1.1–3.18). Similarly, the odds of alcohol consumption were fifteen times higher among cigarette smokers as compared to their counterparts (OR = 15.82; 95%CI: 4.31–58.1). Those study participants who chew Khat were also three times more likely to consume alcohol as compared to those who did not chew Khat (AOR = 2.98; 95%CI: 1.22–7.27) (Table 3). On the other hands, the odds of alcohol consumption were 89% times lower among muslim religion followers as compared to their counterparts (AOR = 0.11; 95%CI: 0.04–0.36) (Table 5).
Discussion
Alcohol consumption continues to be a significant cause of premature mortality and morbidity in Ethiopia. This study aimed to determine the spatial distribution and factors associated with alcohol consumption using the EDHS 2016 dataset. The prevalence of alcohol consumption was found to be 33.15%. The current finding was lower than the previous study findings in different regions of the world, such as Ethiopia (72.6%) [13, 14], New Zealand (95%) [31], India (49.7%) [32], Nigeria (57.9%) [9], Sri Lanka (53.7%) [11], Uganda (51.4%) [12], and Nigeria (76.0%) [10], and Ghana (39.5%) [33]. In the contrary, the current finding was greater than these reported in Kenya (10.8%) [34], Ambo, Ethiopia (27%) [35], and systematic review and meta-analysis finding in Ethiopia (23.86%) [15]. But, it is in agreement with the global status report on alcohol and health in Sub-Saharan Africa (32.7%) [5, 36]. This difference might be due to the difference of the study population and sample size. Majorities of the previous focused on the students only, which could vary the prevalence of alcohol consumption. In this study, the spatial distribution of alcohol consumption was high among Amhara, Tigray, and some parts of the Oromia region. This might be because majority of the community in these regions are more agrarian where local alcoholic beverages are highly produced, which could increase alcohol consumption in the area [37].
The odds of alcohol consumption were three times higher among Orthodox religion followers as compared to protestants. Similarly, the odds of alcohol consumption were thirteen times higher among traditional religion followers as compared to protestants. On the other hand, Muslim religion followers have even 89% lower odds of alcohol consumption compared to protestants. The finding is in agreement with studies conducted in Nigeria [9] and Brazil [17]. This consistency might be due to the fact that alcohol consumption is culturally and socially accepted by Orthodox religious followers in Ethiopia.
The odds of alcohol consumption were almost two times higher among respondents living in low proportion of community literacy as compared to high proportion of community literacy. The association might be due to low level of awareness about the health effects of alcohol consumption are high among household heads living in low community’s literacy. In other words, household heads living with a high proportion of community literacy might assist household heads in obtaining better knowledge about the health effects of alcohol consumption that could decrease alcohol consumption.
Once more, study participants that cigarette smoking habits had fifteen times higher odds of alcohol consumption than those who had no cigarette smoking habit. This finding is supported by other reports [8, 17, 18]. Smoking cigarettes can significantly influence alcohol consumption in the communities. The co-occurrence of these behaviors is common due to shared addictive qualities. Nicotine in cigarettes can enhance the pleasurable effects of alcohol, leading to increased alcohol consumption.
Similarly, study participants that chew Khat had three times higher odds of alcohol consumption as compared to their counterparts. Chewing khat can have a notable impact on the communities’ alcohol consumption. Social settings where chewing khat occurs can normalize alcohol consumption. The communities may also use alcohol to counteract the effects of khat. These findings highlight the complexity of substance use behaviors among communities, necessitating a comprehensive approach to address multiple risk factors simultaneously.
Strengths and limitations of the study
This study was conducted in all regions of Ethiopia with a large sample size. Multilevel logistic regression analysis and spatial analysis were used to determine the spatial variation of alcohol consumption and its associated factors, which could identify the specific areas for intervention. However, as EDHS data was collected using a cross-sectional study design, it could limit the causality of predictors on alcohol consumption. The quantity of alcohol consumption was not determined in this study since the data did not contain this information. Additionally, the prevalence of alcohol consumption was also not determined in the current study.
Conclusions
The prevalence of lifetime alcohol consumption was found to be high in Ethiopia. Religion, region, and community literacy level were factors associated with alcohol consumption. Behavioral factors, such as chewing khat and cigarette smoking, were also identified as significant predictors of alcohol consumption. Spatial analysis showed that there was a significant clustering of alcohol consumption in Amhara, Tigray, and some parts of the Oromia region. Therefore, the ministry of health, religious leaders, and other concerned bodies should provide compressive public health interventions that target hot spot areas. Substance abuse control law enforcement should be strengthened in Ethiopia.
Acknowledgments
We would like to express our deepest heartfelt thanks to the “Measure DHS program” for providing the data for this study.
References
- 1. Bryazka D, Reitsma MB, Griswold MG, Abate KH, Abbafati C, Abbasi-Kangevari M, et al. Population-level risks of alcohol consumption by amount, geography, age, sex, and year: a systematic analysis for the Global Burden of Disease Study 2020. The Lancet. 2022;400(10347):185–235. pmid:35843246
- 2. Liu Y, Nguyen N, Colditz GA. Links between alcohol consumption and breast cancer: a look at the evidence. Women’s health. 2015;11(1):65–77. pmid:25581056
- 3. Taylor B, Irving H, Kanteres F, Room R, Borges G, Cherpitel C, et al. The more you drink, the harder you fall: a systematic review and meta-analysis of how acute alcohol consumption and injury or collision risk increase together. Drug and alcohol dependence. 2010;110(1–2):108–16. pmid:20236774
- 4. Roerecke M, Vafaei A, Hasan OS, Chrystoja BR, Cruz M, Lee R, et al. Alcohol consumption and risk of liver cirrhosis: a systematic review and meta-analysis. The American journal of gastroenterology. 2019;114(10):1574. pmid:31464740
- 5.
WHO. World Health Organization. Global status report on alcohol and health 2018: World Health Organization; 2019.
- 6. Henriksen TB, Hjollund NH, Jensen TK, Bonde JP, Andersson A-M, Kolstad H, et al. Alcohol consumption at the time of conception and spontaneous abortion. American journal of epidemiology. 2004;160(7):661–7. pmid:15383410
- 7. Ornoy A, Ergaz Z. Alcohol abuse in pregnant women: effects on the fetus and newborn, mode of action and maternal treatment. International journal of environmental research and public health. 2010;7(2):364–79. pmid:20616979
- 8. Adeloye AO-IOOEOAD. Substance use among adolescents in sub-Saharan Africa: A systematic review and meta-analysis. South African Journal of Child Health. 2018;12.
- 9. Lasebikan VO, Ola BA. Prevalence and correlates of alcohol use among a sample of Nigerian semirural community dwellers in Nigeria. Journal of addiction. 2016;2016. pmid:27195170
- 10. Ijomanta IN, Lasebikan VO. Lifetime and 12 months prevalence of alcohol use and alcohol use disorders among soldiers residing in a military community in Ibadan. Substance Use & Misuse. 2016;51(6):722–32.
- 11. Abeysena C, Hapugoda C. Prevalence of alcohol, tobacco and illicit narcotic substances usage and associated factors among patients attending mental health clinics conducted by a group of hospitals in Sri Lanka. 2018.
- 12. Kuteesa MO, Weiss HA, Cook S, Seeley J, Ssentongo JN, Kizindo R, et al. Epidemiology of alcohol misuse and illicit drug use among young people aged 15–24 years in fishing communities in Uganda. International journal of environmental research and public health. 2020;17(7):2401. pmid:32244722
- 13. Gebremariam TB, Mruts KB, Neway TK. Substance use and associated factors among Debre Berhan University students, Central Ethiopia. Substance abuse treatment, prevention, and policy. 2018;13(1):13. pmid:29622021
- 14. Abate SM, Chekol YA, Minaye SY. Prevalence and risk factors of psychoactive substance abuse among students in Ethiopia: A systematic review and meta-analysis. Annals of medicine and surgery. 2021;70:102790. pmid:34540211
- 15. Ayano G, Yohannis K, Abraha M, Duko B. The epidemiology of alcohol consumption in Ethiopia: a systematic review and meta-analysis. Substance abuse treatment, prevention, and policy. 2019;14(1):1–16.
- 16. Yismaw S, Kebede H. Prevalence and associated factors of alcohol consumption among college students in Gondar Town, Northwest Ethiopia. Sci J Public Heal. 2015;3(4):453–9.
- 17. Reisdorfer E, Büchele F, Pires ROM, Boing AF. Prevalence and associated factors with alcohol use disorders among adults: a population-based study in southern Brazil. Revista Brasileira de Epidemiologia. 2012;15:582–94. pmid:23090305
- 18. Getachew T, Defar A, Teklie H, Gonfa G, Bekele A, Bekele A, et al. Magnitude and predictors of excessive alcohol use in Ethiopia: Findings from the 2015 national non-communicable diseases STEPS survey. Ethiopian Journal of Health Development. 2017;31(1):312–9.
- 19. Murphy A, Roberts B, Stickley A, McKee M. Social factors associated with alcohol consumption in the former Soviet Union: a systematic review. Alcohol and Alcoholism. 2012;47(6):711–8. pmid:22813540
- 20. Allen L, Townsend N, Williams J, Mikkelsen B, Roberts N, Wickramasinghe K. Socioeconomic status and alcohol use in low-and lower-middle income countries: a systematic review. Alcohol. 2018;70:23–31. pmid:29723827
- 21. Klein H, Sterk CE, Elifson KW. The prevalence of and factors associated with alcohol-related problems in a community sample of African American women. Journal of addiction. 2016;2016.
- 22. Poelen EA, Scholte RH, Engels RC, Boomsma DI, Willemsen G. Prevalence and trends of alcohol use and misuse among adolescents and young adults in the Netherlands from 1993 to 2000. Drug and alcohol dependence. 2005;79(3):413–21. pmid:16039803
- 23.
Central Statistical Agency (CSA) [Ethiopia] and ICF. Ethiopia Demographic and Health Survey 2016. Addis Ababa, Ethiopia, and Rockville, Maryland, USA: CSA and ICF. 2016
- 24. Liyew AM, Teshale AB. Individual and community level factors associated with anemia among lactating mothers in Ethiopia using data from Ethiopian demographic and health survey, 2016; a multilevel analysis. BMC Public Health. 2020;20:1–11.
- 25. Sisay D, Ewune HA, Muche T, Molla W. Spatial distribution and associated factors of institutional delivery among reproductive-age women in Ethiopia: the case of Ethiopia demographic and health survey. Obstetrics and Gynecology International. 2022;2022. pmid:35795329
- 26. Tryon WW. Evaluating statistical difference, equivalence, and indeterminacy using inferential confidence intervals: an integrated alternative method of conducting null hypothesis statistical tests. Psychological methods. 2001;6(4):371. pmid:11778678
- 27. Kulldorff M. A spatial scan statistic. Communications in Statistics-Theory and methods. 26(6): p. 1481–1496.1997.
- 28. Austin PC, Merlo J. Intermediate and advanced topics in multilevel logistic regression analysis. Statistics in medicine. 2017;36(20):3257–77. pmid:28543517
- 29. Weinmayr G, Dreyhaupt J, Jaensch A, Forastiere F, Strachan DP. Multilevel regression modelling to investigate variation in disease prevalence across locations. International Journal of Epidemiology. 2017;46(1):336–47. pmid:27864412
- 30. Merlo J, Chaix B, Ohlsson H, Beckman A, Johnell K, Hjerpe P, et al. A brief conceptual tutorial of multilevel analysis in social epidemiology: using measures of clustering in multilevel logistic regression to investigate contextual phenomena. Journal of Epidemiology & Community Health. 2006;60(4):290–7. pmid:16537344
- 31.
Mason K. Alcohol use in New Zealand: key results of the 2007/08 New Zealand alcohol and drug use survey: Ministry of Health; 2009.
- 32. Bute J, Kori S, Arora V, Sahasrabuddhe A. Prevalence and pattern of alcohol consumption among males attending primary health care setting using AUDIT in rural Indore, Madhya Pradesh, India. Int J Community Med Public Heal. 2018;5(10):4461.
- 33. Aboagye RG, Kugbey N, Ahinkorah BO, Seidu AA, Cadri A, Akonor PY. Alcohol consumption among tertiary students in the Hohoe municipality, Ghana: analysis of prevalence, effects, and associated factors from a cross-sectional study. BMC psychiatry. 2021;21(1):431. pmid:34479516
- 34. Jenkins R, Othieno C, Ongeri L, Kiima D, Sifuna P, Kingora J, et al. Alcohol consumption and hazardous drinking in western Kenya—a household survey in a health and demographic surveillance site. BMC psychiatry. 2015;15:1–10.
- 35. Birhanu A, Mekuria M. Prevalence of alcohol use disorders and associated factors among Ambo Town Community, Central Ethiopia: a community based cross-sectional study. J Addict Res Ther. 2019;10(4):1000389.
- 36. Peacock A, Leung J, Larney S, Colledge S, Hickman M, Rehm J, et al. Global statistics on alcohol, tobacco and illicit drug use: 2017 status report. Addiction. 2018;113(10):1905–26. pmid:29749059
- 37. Tafere G. A review on traditional fermented beverages of Ethiopian. J Nat Sci Res. 2015;5:94–102.