Alcohol Consumption and the nature of Alcohol related problems in Botswana: A Premiliminary report by J. Finlay and R.K. Jones Introduction Public accounts over the last few years concerning the consumption of alcohol in Botswana have tended to suggest a widespread increase. The Daily News, Botswana's only newspaper, has increasingly published items dealing with excessive consumption. For example, we are informed (June 25, 1982, No. 121) that the North East District Council's model bye-laws should prohibit the intake of alcohol by those under 18 years of age and alcohol should not be sold to those already inebriated.! Similarly, the same paper (August 10 1982, No. 157 and June 2, 1982, No. 104) recently urged the Bakwena to drink lessZand the Central District youth to stop 'corrupting' themselves with alcohol and 'dagga') Other indications of a /trowing Concern over a perceived problem of increasing alcohol consumption include a special 'Round Table' radio programme4 and various pronouncements from special interest groups. The difficulty of assessing industrial attrition, loss of productivity and deficit to the economic section is greater than it would be in a more developed country. Similarly alcohol-related social problems such as child neglect and personal abuse are difficult to assess. Sporadic declaration by public figures directed against drunkenness and public affray occur both on Radio Botswana and in Th~DailY NewS. Not infrequently politicians and church leaders lament the decline of traditional constraints exercised by the community on individual behaviour and perceive alcoholism as a 'social pathology'. With increasing urban drift (177% increase between 1971 and 1981)5 and rising unemployment (83,400 out of the total population in 1980 were employed in the formal sector)6 alcoholic intake per capita is e"xpected to increase. Thus, from the above references, it would appear that an alcohol problem exists and is growing in Botswana or at least there is enough general public concern to warrant a more detailed investigation into the matter. It is the primary purpose of this paper to describe the overall rate of alcohol consumption and the reported consequences it has for the road safety and health sectors of the country. Methodology The information reported was obtained from publications of the Ministry of Health, the Ministry of Works and Communications, the Department of Customs and Excise, and from interviews with the commercial producers of alcohol and principal importers of alcohol to Botswana. In addition, extensive information has been obtained from Haggblade's study of local and commercial beer brewing in Botswana.7 Chemical analysis of alcoholic content was obtained from various SOurces. Alcohol Consumption Commercial beer and Chibuku figures were obtained from the manufactu~er.s while the imported wines and spirits were taken from external. trade . sta~lstlcS. of the Department of Customs and Excise. The data on Khadl (a Wine-like drink) and "bojalwa" (the home-brewed sorghum beer) were estima~~d from. information cgllected by Haggblade in rural and urban sample surveys of traditIOnal village brewers. From these surveys, Haggblade found that 75% of the country's home brewers p~rchased commercial malt to brew their beer and that an average brewer used 10 kilograms of malt to make 100 litres of beer. The total quantity of home-brewed sorghum beer was estimated from the amount of malt sold in 1981. It is further estimated that there are 400 liquor outlets in Botswana. Health Problems Data on alcohol-related health problems were abstracted from the most recent statistical publications of the Ministry of Health and from discussions with technical personnel working for the Ministry of Health. Traffic Accidents Data on alcohol-related traffic accidents were obtained from transport statistics of the Ministry of Works and Communications and from discussions with the Botswana National Police Force. Results Alcohol Consumption in Botswana As seen in Table I more than 137 million litres of alcoholic beverages are estimated to have been consumed in Botswana during 1981. When converted to Htres of pure alcohol, it is seen that over 5 million litres were consumed in Botswana during the year. This represents 5.7 litres per capita or 10.3 litres of pure alcohol for the estimated 54.8% of the population aged 15 and over (513,257). Age group estimation was taken from the 1971 Census Projectionsl4 - (constant fertility, declining mortality and no permanent emigration), as applied to the 1981 Census.!5 Effects of Alcohol on Health Although reference to the destruction of individual and community health and welfare 'MiSalluded to in the Introduction to this paper actual quantitative evidence of this is much more difficult to find in reports available to us. The following information was therefore taken frO(ll such reports and interviews. A. Mortality I~ is ~stimated ~hat almost all deaths among persons over 25 in 1979 from cirrhosIs of the li~er were due to alcoholism.!6 This represents 3.7% (35 out of 956) of all deaths In persons aged 25 or over during the lear,, 7 However this is (~~~joablY higher than data reported for 1977 (1.6%),1 1975 (2.4%),19a~d 1974 2 Morbidity Since reasons for out-patients visits to health facilities are neither specific enough nor available by age, it was impossible for the authors to use this sizeable body of information. Thus, attention was turned to in-patient morbidity, or hospital discharge diagnoses from the general hospitals of Botswana, both government and private. In Table II, data on alcohol dependence, chronic liver disease, and certain other possibly alcohol-related diagnoses for those 15 years of age and over are shown.21 Thus, these 476 diagnoses are seen to account for only 1.1% of all hospital discharges. Among men of this age group, the percentage discharges is slightly higher (3.15%). Whereas data from general hospitals appear to show little evidence of alcohol-related morbidity, discharge data from the mental hospital at Lobatse is perhaps more instructive. The two reported relevant categories of alcoholic psychosis and alcoholism account for 12.6% (118 of 937) of all discharges and 18.0% (108 of 601) of all male discharges from the country's mental hospital.22 In Table III, relevant reasons for out-patient visits to mental health clinics are inspected. Here it is seen that 5.9% of all visits, and 9.0% of all male visits, are for reasons related to problems with alcohol.23 In data fr.om the annual report of the Lobatse Mental Hospital for 198024 alcohol-related reasons for visits to rural mental health clinics accounted fOI 8.3% of all visits. Despite this relatively low rate, when all schizophrenias were lumped together, alcohol-related visits were the third most frequently reported reason for attending these rural mental health clinics. Traffic Accidents and Alcohol Use According to the most recent reported figures for causes of motor vehicle accidents (1980)25 the category 'Intoxication' accounted for 154 of the total 2253 motor vehicle accidents that year - or 6.8% of all traffic accidents. This was down slightly from the similar figure of 197626 which was 9.1% (164 of 1800). However, the 154 alcohol-related traffic accidents of 1980 accounted for 28 of the 116 deaths due to traffic accidents, or 24% of all deaths. This represents an increase over the 10 deaths (out of 106) attributed to accidents involving alcohol during 1979. Discussion From the data presented, we thus see that nearly 6 litres of pure alcohol were available for consumption by every person in Botswana during 1981. It is worth noting that 'consunption' is actually calculated from alcoholic beverages produced and/or imported into Botswana during 1981 just as are comparable figures of other countries. However, in contrast to the bottled alcoholic drinks of developed countries, the village-made wines (khadi) and beers (bojalwa) that contribute heavily to consumption in Botswana have limited shelf lives. In the case of the factory- made sorghum beer (Chibuku), this is estimated at a maximum of five days after which the beverage 'goes off' and is no longer fit for consumption. Nevertheless, factory managers of Chibuku consider such wastage to be negligible.27 From the many anecdotes told of villagers drinking until all the beer is finished, the authors surmise that there is also little wastage of home-made brew. In Table IV, Botswana consumption is comrred with the annual average consumption of several developed countries in 1968-70.2 Frem this it is seen that the wine-drinking countries of France and Italy, and the beer-drinking country of Germany are the greatest consumers and that Botswana consumption equals that of Sweden and is more than that of Southern Ireland and 3 Norway. However, it is well documented that different drinking patterns of countries may skew the results of a valid comparison of consumption rates. Haworth and Serpe1l2~ noted in relation to Zambia, for example, that more than 50% of the populations they studied were 'abstainers' and that only 15% were 'regular heavy drinkers'. While the neighbouring Zambian study made no attempt to estimate the quantity of local or traditional alcoholic beverages, this has been possible in Botswana. Thus 65 million Iitres of home-brewed sorghum beer was estimated to have been produced during 1981. It has been estimated that over 5,000 households are dependent on home-brewing30 and that 31% of consumer expenditure on alcoholic drinks is for purchases of sorghum beer)1 This represents a 'bojalwa' annual consumption of 131 Iitres for every person in Botswana who is 15 and over, or more than one large cup of traditional beer per day for every male and female adult living in Botswana. In addition, the quantity of tinned and bottled beer consumed annually in Botswana represents 98 tins per year for every adult. Finally, it should be recognised that the alcoholic content of home-made beverages, and particularly the wines, is not always consistent and thus the alcoholic content estimates employed in this paper are only approximate. Depending on what is added to the drink, it is estimated that 'khadi' may vary anywhere from 2 to 20% alcohol content per volume. Obviously then, the overall consumption figures reported herein can only be approximations. Turning away from the quantitative aspects of alcohol consumption in Botswana, what then are the effects of alcohol drinking on the country's population? It must be emphasised that only negative consequences of alcohol drinking were investigated in preparing this report. However, in one village survey of alcohol consumption patterns, nearly two-thirds of those who drank reported that they did so because they considered alcohol to be 'a food,)2 Indeed the sorghum- based beer is reputed to have a considerable nutritive value. In his treatise on Sorghum, Doggett33 estimates that whole grain sorghum contains 9.5% protein and further states that "it would seem that ordinary sorghum beer is a drink of low alcoholic content but with enhanced levels of Vitamin B and is a useful dietary supplement when taken in moderate amounts". One figure given for protein content of unclarified sorghum beer is 2.6%)4 The principal effect of aicohol on health in Botswana which can be compared to those found elsewhere is the mortality from cirrhosis of the liver. Where other diseases (such as the tropical diseases of bilharzia and malilri~) . may be ruled out, cirrhosis is ~enerally attributed to excessive mta~e of. alcohol in the over 25 age group. From this population, the 35 de':lths. from cirrhosIs. repres«:nt a specific death rate of 12.0 per 100,000 population. ThiS figure was obtamed us10g the 1979 population estimate from the 1971 census. The 1981 census indicates that all estimates from the 1971 census may have been low. T~us the actual speci~ic death rate for cirrhosis in the 25+ age group may have been slightly lower - say, 10 the 11,3 - 11,5/100,000 range. However, in addition it should be ac~nowledged that reported cirrhosis death rates in other recent years were much lo~er 5 an~ that an overall average for the past eight years was 7.1. In Table V, thiS rate, IS .comp.ared wi~h that of seve~al developed countries in 1972)6 Thus, Botswana s CirrhosIs rate IS seen to be higher than those of the U.K. and Southern' Ireland. though nowhere near the cirrhosis death rate of the U.S.A. and the European countries which consume more alcohol. While every health wor:ker has a. tale ~t. alcohol, abuse in her community, very few of these ~, appear, 10 the patient statistics received from clinics and hospitals. Probably thiS IS because the actual reason for the clinic visit may be due to a gastric 4 disturbance or an injury. In Zambia, it was reported that 55% of the assault cases involved alcohol)7 Thus, neither out-patient nor general in-patient statistics provide much information on morbidity. As seen in Table II, however, morbidity from alcohol-related conditions in 1979 was three times higher for men than it was for women. For specific information on illness from alcoholism, then, it is necessary to examine the mental health statistics. Thus, as reported, 18% of all males in the country's mental hospital during 1979 were there for reasons related to alcohol. It should be noted, however, that this percentage is based upon 'discharge' data and that since Botswana was in the process of shifting from a traditional hospital approach to their present community approach to mental health care in 1979, there were an unusually large number of discharges that year. A recent examination of the mental hospital records over a ten year period by one of the authors indicated that alcohol was implicated in 9.1% of the hospitalised patients during this time)8 Since data were not yet published for all out-patient mental health clinics in 1980, it is not clear whether the increase from 5.9% to 8.3% between 1979 and 1980 for alcohol-related psychiatric out-patient visits is real or not. Just over 9% of all traffic accidents in Botswana during 1980 were attributed to 'intoxication'. While not particularly an alarming figure, it must be remembered that these were accidents for 'reported' alcohol involvement. As is true for most countries of the world, it is believed that alcohol involvement in motor vehicle accidents is grossly under-reported in Botswana. In Zambia, only 2% of the road traffic accidents are reported as alcohol-related. 39 Realising that alcohol involvement in road accidents is probably under-reported, it must then be admitted that finding 24% of the road accident fatalities to involve alcohol is 'alarming' - unless, of course, there is a bias toward reporting alcohol involvement in only those accidents involving fatalities. Conclusions and Recommendations 1. By world-wide standards, the estimated per capita alcohol consumption in &>t~ is. not high. 2. In Botswana, little has been reported about the pattern of alcohol consumption among the population. 3. Reported general health statistics do not implicate alcohol as a serious 'health' proolem in Botswana though it is responsible for a sizeable portion of mental health illness. 4. Alcohol abuse in Botswana constitutes a serious hazard to road safety. 5. Alcohol abuse represents a problem for some segment of every society, no matter how small. To further define the dimensions of the problem and to identify those factors important to the design of appropriate educational programmes for its prevention and control, further studies of population practices, beliefs, and attitudes regarding alcohol use in Botswana are required. Specifically, the following are recommended: a. A country-wide sample survey concerning beliefs, attitudes, and practices of the population in Botswana. b. A special school survey of youth to determine the age of adoption and other precursors to alcohol use in Botswana. 5 c. A special survey of selected occupational groups to investigate the impact of alcohol use on productivity at the worksite and its related implication for the economic development of the country. d. A literature survey of applied educational programmes to combat alcohol abuse in other countries of Africa and the world. e. The implementation of an appropriate educational intervention and its use in a controlled study of its effect on the prevention and control of alcohol abuse in selected rural and urban areas of Botswana. 6 TABLE I Alcohol Productiona and IrnportatioJ> for Botswana in 1981 (Alcohol Consumption) f percent9 litre. li tre. of Type of Produced Alcohol Pure Alcohol Product or I.ported Content c 941,435 8eer 17,117.000 5.5% lIine: Unforti fied 620,000 11% 68,200 Fortified 531,000 16% 84,960 Spirits 603,000 43% 259,290 Khadid 26,000,000 6% 1,560,000 Sor9hu. 8:er: Factory d 27,679,000 3% 833,070 Ho.. -ude 65,000,000 2.4% 1,560,000 TOTAL 137,640,000 - 5,306,955 Notes a. Alcoholic beverages produced in Botswana are bottled beer, factory and home- made sorghum beer, and khadi (a type of home-made wine). b. All commercial spirits and wines are imported into Botswana. Figures for these products in this Table were obtained from the Statistics section of the Botswana Government of Customs and Excise.9 To confirm the overall accuracy of these figures, quantities of alcohol products imported by the two principal commercial liquor companies of Botswana were also obtained.lU c. Practically all tinned and bottled beer consumed in Botswana is produced by Kgalagadi Breweries. "Beer" figures were obtained from them)l In addition 937,302 lit res of bottled beer were imported in 1981 and 74,506 Iitres of cider.J2 d. Figures for the home-made wine (khadi) and the home-made sorghum beer (bojalwa) were taken from the Ministry of Commerce and Industry study by Haggblade.13 e. Figures for the factory-made sorghum beer (Chibuku) were obtained from the Botswana Breweries Ltd., brewers of "chibuku" beer. f. All beverage figures were rounded to the nearest thousand. g. The percentage of alcohol content was taken from manufacturers, by chemical analysis, and by expert estimates. 7 TABLE II Patient Discharges from General Hospitals in 1979 by Possible Alcohol-Related and All Other Diagnoses for the 15 Year and Over Age Group a,.b. All Patients Male Patients Only ~iagnosis No. (l) Ho. (l) Possible Alcohol-Related Illness : 1. Nutri tiona! ) 98 ) 59 Deficiencies 2. Alcohol Dependence 133 ) 109 ) 3. Physical Malnutrition due to Mental 1 ) ) Illness )(1.1 ) )( 3 .2) 4. Ulcer of the Sto.ach 62 ) 42 ) and Ouodenu. 50 Chronic liver Disease 148 ) 109 ) c 60 nOther" Poisoning 34 ) 24 ) All Other 43,408 ( 9809) 10,581 ( 96.8) Illness TOTAL 43,884 (100.0) 10,925 (JOO.O) Notes: a. Data for this table were abstracted from Medical Statistics, 1978-79.21 b. The small number of diagnoses for non-specific age groups were assumed to be in this 15+ age category. c. "Other" poisons (other than medicinal, parafin, mushroom, and animal bites) are included as "possibly" related to alcohol. 8 TABLE III Out-Patient Visits to Mental Health Clinics for Alcohol-Related Reasons During 1979* Reasons for Vist All Patients I IIde Patients Only • 0. (l:) b• (l:) Alcohol-Related: Deliriu. tre.ens 44 ) 33 ) I. 2. Korukov's 2 ) 0 ) psychosis 3. Alcoholic 62 ) (5.9) 31 ) (9.0) 4. hallucination Chronic • . 251 ) 209 ) dcoholi •• s. Acute 37 ) 10 ) drunkenness All Other 5,301 ( 94.1) 2,859 ( 91.0) TOTAL 6,697 (100.0) 3,142 (100.0) Notes: *Data for this table were abstracted from Table 6.4 of Medical Statistics, 1978-79.23 9 TABLE IV A Comparison of 1981 Alcohol Consumption in Botswana with Several Developed Countries in 196&-1970* Country li tres of Pure Alcohol per Capita France 16.4 Italy 14.0- lIest Geruny 11.3 Checkoslovakia 8.0 Canada 6.5 United Kingdol 6.2 U.S.A. 5.8 80tswana 5.7 Sweden 5.7 Southern Ireland 4.5 Finland 4.1 Norway 3.4 Note: * Data for the countries in this table were taken from Schmidt's paper on "Cirrhosis allld Alcohol Consum2%tion"Figure 2 which appeared in Alcoholism: New Knowledge and New Responses. Botswana's figure was calculated from data in Table I of the current paper. 10 TABLE V Comparison of Average* Cirrhosis Rate in Botswana with Rates of Selected Developed Countries foe 1971-72** Cirrhosis "ortalit, per Countr, 100,000 Population 25 Years of Age Ind Over Frlnce 57.2 Ih1, 52.1 Nest Gerun, 39.6 U.S.A. 28.6 Checkos10vakia 28.1 Canada 19.6 S..eden 15.6 "or ..a, 7.6 Finland 7.5 Botswana 7.1 Southern Ireland 7.0 United Kingdol 5.7 Notes: * Obtained by averaging cirrhosis rates for 1974, 1975, 1977 and 1979 foe their res- pective 25+ populations. ** Data for the other countries in this table were taken from Schmidt's paper on "Cirrhosis and Alcohol Consumption,,)6 II References and Notes I. The Botswana Daily News, No.121, 25 June 1982. 2. The Botswana Daily News, No.157, 10 August 1982. 3. The Botswana Daily News, No.104, 2 June 1982. 4. Radio Botswana. "The Round Table". A discussion about availability, its use and abuse, and government policy regarding it in Botswana. November 1981• .5. The Central Statistics Office. Tablel, "Enumerated Population - 1981 and 1971 Censuses by Census District" in Statistical Bulletin, Vol.7, No.1. Ministry of Finance and Development Planning, Botswana. March 1982. 6. The Central Statistics Office, Ibid., Table 5, "Estimated Total Formal Sector Employment by Economic Activity". 7. Haggblade, S. Sorghum Beer: The Impact of Factory Brews on a Cottage Industry. Mimeographed report submitted to the Botswana Ministry of Commerce and Industry, 15 July 1982. 8. Ibid. 9. Botswana Department of Customs and Excise. External Trade Statistics, 1981 unpublished figures made available to the authors by the Statistics Section. 10. Data on alcohol imports were made available to the authors by Segwana Liquor Distributors and Botswana Liquor Manufacturers, both of Gaborone, Botswana. 1982. 11. Kgalagadi Breweries. Data on bottled and tinned beer production for 19&1. Gaborone, Botswana. July 1982. 12. Botswana Department of Customs and Excise. (J 981) op.cit. 13. Haggblade, op.cit. 14. Report on the Population Census, 1971, Central Statistics Office, Botswana. August 1972. 1.5. Central Statistics Office, 1982, op.cit. 16. Information informally obtained from the Chief Government Pathologist. 17. Medical Statistics Unit, Ministry of Health, Botswana, Medical Statistics, 1978-79, Gaborone, 1981. Table 5.11. 18. Medical Statistics Unit, Ministry of Health, Botswana. Medical Statistics, 1977, Gaborone, 1978. 19. Medical Statistics Unit, Ministry of Health, Botswana. Medical Statistics, 197.5, Gaborone, 1976. Table 3.5. 12 20. Medical Statistics Unit, Ministry of Health, Botswana. Medical Statistics, 1974, Gaborone, 1975. Table 29. 21. Medical Statistics Unit, Medical Statistics, 1978-79, Table 5.10. 22. Ibid., Table 6.3 23. Ibid., Table 6.4 24. The Lobatse Mental Hospital, Annual Report, 1980, in Mimeographed Report of the Botswana Ministry of Health, 1981. 25. Ministry of Works and Communications, Transport Statistics, 1980, Government Printer, 1981. 26. Ministry of Works and Communications, Transport Statistics, 1980, Government Printer, 1980. 27. Informal discussion with the managers of Botswana Breweries Ltd., manufacturers of Chibuku, Gaborone, August, 1982. 28. Schmidt, W. "Cirrhosis and Alcohol Consumption: An Epidemiological Perspective", in Edwards, G. and Grant, M. (eds.) Alcoholism: New Knowled~e and New Responses, London: Croom Helm, 1977. Figure Two. 29. Haworth, A. and SerpelJ, R. "A Summary of the Final Report on Phase I", Community Response to Alcohol-Related Problems in Zambia, Community Health Research Unit, Institute for African Studies, University of Zambia, 1981. 30. Egner, B. and Klausen, A.L. Poverty in Botswana, National Institute of Research, University of Botswana, 1980. 31. Haggblade, op.cit. 32. A student survey made by a Health Education student in September, 1981. 33. Doggett, H. Sorghum, London: Longman, 1970, pp.231-237 3/j. Food and Agriculture Organisation/U.S. Department of Health, Education and Welfare (J968). Food Composition Table for Use in Africa, Item 1566. p.219. Rome/Bethesda. 35. From publications of Medical Statistics for 1974, 1975 and 1977 mortality from cirrhosis/IOO 000 25+ populations were as follows: 7.2 in 1974, 4.7 in 1975, and 4.1 in 1977. Some variation in reporting of institutional deaths between years is believed to also have an effect on calculated cirrhosis rates. 36. Schmidt, op.cit. Fig. 2. 37. Haworth and Serpell, (1981), op.cit. p.15. 38. Jones, R.K. A Survey of Lobatse Mental Hospital Records, Mimeographed, 1982. 39. Haworth and Serpell, op.cit., p.3 13