diabetes in newly diagnosed zanzibari patients: 1986 1 98 - NCBI

DIABETES IN NEWLY DIAGNOSED ZANZIBARI PATIENTS: 1986 1 98 Mohammed H. Makame, MD, and Eugene S. Tull, DrPH Zanzibar Town, Zanzibar

Few data exist on the impact of diabetes mellitus, particularly the insulin-dependent subtype, in many parts of Africa. The importance of diabetes as a public health problem in the East African Islands of Zanzibar was assessed through prospective registration of all newly diagnosed diabetic individuals who attended the diabetic clinic at Mnazimmoja Hospital from January 1986 to December 1989. A total of 323 diabetic patients, 192 men and 131 women, were diagnosed. Two hundred fiftythree (78.3%) individuals had noninsulindependent diabetes mellitus (NIDDM), 61 (18.9%) had insulin-dependent diabetes mellitus (IDDM), and 9 (2.8%) had diabetes of uncertain type. Two hundred twenty-six (70%) of the patients were town residents, 62 (19.2%) had office jobs, and 84 (26%) were laborers. The majority of the patients presented with the classic symptoms of diabetes. A positive family history of diabetes was found in 35 (13.8%) and 4 (6.6%) of the NIDDM and IDDM cases respectively. Hypertension was diagnosed in 29 (11.5%) of all NIDDM individuals while obesity was present in only 41 (16.2%) of all NIDDM patients. These data suggest that diabetes is a problem of major public health importance in the Islands of Zanzibar. (J Nati Med Assoc. 1993;85:621-625.) Key words * insulin-dependent diabetes mellitus noninsulin dependent diabetes mellitus * uncertain diabetes * Zanzibaris From Mnazimmoja Hospital, Zanzibar Town, Zanzibar, and the Department of Epidemiology, Graduate School of Public Health, University of Pittsburgh, Pittsburgh, Pennsylvania. Requests for reprints should be addressed to Dr Mohammed H Makame, Diabetes Research Ctr, Rangos Research Ctr, 3460 Fifth Ave, 5th Fl, Pittsburgh, PA 15213. JOURNAL OF THE NATIONAL MEDICAL ASSOCIATION, VOL. 85, NO. 8

Diabetes mellitus is increasingly becoming an important and common disease in the developing world and especially in Africa. It is currently estimated that there are about 40 million diabetic individuals in the developing countries.' In these countries, noninsulindependent diabetes mellitus (NIDDM) is the most prevalent form of the disease, accounting for approximately 90% of diabetes cases.2 Rapid changes in the sociodemographic profile of developing populations, including rural to urban migration, and increased life expectancy, have led to marked increases in middle-age population groups. These factors, which are associated with an increased occurrence of NIDDM,3 will cause the estimate of diabetes prevalence in these countries to be revised to about 65 million in a little more than a

decade.'1 On the continent of Africa, a number of epidemiological studies on the occurrence of diabetes mellitus have been conducted in countries located from north to the south between Egypt and South Africa and from west to east between Cote d'Ivoire and the Indian Ocean Islands of Zanzibar off the East African coast.4-'2 In a recent review of diabetes in Africa, estimates of the disease in sub-Saharan countries were reported to range from 0% to 5.7%.7 Studies in some sub-Saharan populations suggest that diabetes may account for as much as 5% of all medical admissions.'3'15 While most of the research on the occurrence of diabetes in Africa has focused on NIDDM, little data exist on the frequency and impact of IDDM, particularly for black African populations. A few studies in Nigeria'6 and the Sudan'7 have reported prevalence estimates for IDDM. However, data on the incidence of the disease in black Africa is lacking despite the recent proliferation of IDDM registries around the world'8 as a result of the efforts of the World Health Organization (WHO) Multinational Project for Childhood Diabetes.19 In the East African Islands of Zanzibar, diabetes mellitus is fast becoming a disease of major public 621

DIABETES IN ZANZIBAR

health importance. Although ranked 14th on the list of most common diseases diagnosed in Zanzibar hospitals during the years 1982 to 1987, diabetes mellitus accounted for annual percentages of 1.9, 2.9, 3.7, 2.1, 4.0, and 3.2, respectively, of all hospital deaths due to specific causes during the same years.20 To date, there has only been two published reports about diabetes mellitus in Zanzibar,12'21 and these have provided very little data on its epidemiology, and no estimates of the rates of occurrence of the disease. In this article, we present epidemiological data on newly diagnosed Zanzibari diabetic patients who were diagnosed during the years 1986 to 1989, and estimate, for the first time, the incidence of IDDM in a black East African population.

MATERIALS AND METHODS Zanzibar Islands consist of the larger Unguja Island, its sister Island of Pemba, and about 21 other smaller islands in the Indian Ocean, off the East African coast. The Islands enjoy an equatorial type of climate with heavy rains (masika) during the months of March to May, and short rains (vuli) in November. The 1988 census has estimated the population of Zanzibar to be approximately 640 578. Data on all new onset diabetic patients who were diagnosed or referred to the Mnazimmoja hospital in Zanzibar Town, the only referral hospital for diabetes in Zanzibar, was collected for the period of January 1986 to December 1989. Information on age, sex, residence, occupation, education, date of diagnosis, type of diabetes, onset, duration and type of symptoms, and family history of diabetes were recorded on special questionnaires completed by the examining physician. Past or current histories of smoking and alcohol consumption also were determined. Blood pressure measurements were taken with mercury sphygmomanometers. Fasting or random blood glucose levels were tested by using Ames glucometers or the Corning Colorimeter-253 (Coming Medical and Scientific, Medfield, Massachusetts). When necessary, 2-hour specimens also were analyzed with the Corning Colorimeter-253. Routine urinalysis was done microscopically and with the Ames-NMultistix (Miles Inc, Tarrytown, New York). Body weight in kilograms (in light clothes) and height in meters were measured and used to calculate the body mass index (BMI) (weight/height [kg/m2]). Individuals then were grouped into those who were underweight, normal weight, overweight, and obese (BMI 30, respectively). Patients had 622

ophthalmological examinations (including funduscopy) performed at the eye clinic at the same hospital. The diagnosis and classification of hypertension and diabetes mellitus was based on WHO criteria.22'23 Diabetic patients were classified according to the following subtypes: NIDDM, IDDM, protein-deficient pancreatic diabetes (PDPD), and fibrocalculus pancreatic diabetes (FCPD). Patients who could not be classified into these categories were said to be of uncertain type. Diabetic individuals of uncertain diagnosis included those nonketonic-prone young patients who presented with very high levels of blood glucose unresponsive to normal insulin dosages and without clear-cut features of malnutrition-related PDPD and FCPD, those who defaulted, and those whose whereabouts were unknown. Data analysis involved calculation of relevant proportions and frequencies of variables of interest. When appropriate, incidence rates were computed and 95% confidence intervals (CI95) based on the position distribution were calculated.24 Denominator populations for the incidence rates were based on the 1988 census obtained from the Tanzanian Census Commission. Data on urinary ketone body levels were not analyzed as only a few patients had their urine samples tested due to recurrent shortages of testing strips. Non-Zanzibari individuals and visiting Zanzibaris not residing in the country who were diagnosed with diabetes during the 4-year study period were not included in the analysis.

RESULTS A total of 323 new diabetic patients (192 men and 131 women; male/female ratio = 1.5: 1) aged between 4 to 73 years were diagnosed during the period of study: 116 in 1986, 98 in 1987, 40 in 1988, and 62 in 1989. Of these, 253 (78.3%) were noninsulin dependent, 61 (18.9%) were insulin dependent, and 9 (2.8%) had diabetes of an uncertain type. Overall, the prevalence of diabetes increased with age, with a peak in the age group 30 to 39 followed by a gradual decline with increasing age (Figure 1). The mean age (± SD) of NIDDM patients was 44 years (± 10.0), and the mean age of IDDM patients was 27.6 years (± 13.4). All IDDM patients were diagnosed under 25 years of age with approximately half of the cases (50.8%) occurring prior to age 20. Only one (1.6%) case of IDDM was diagnosed before age 10-a 4-year-old daughter of a recently diagnosed male diabetic patient. For the 4-year study period, the average annual incidence of IDDM among Zanzibari JOURNAL OF THE NATIONAL MEDICAL ASSOCIATION, VOL. 85, NO. 8


95 60 ~~~~~~~~49

100 50

64

.R 80

5

~40

~~~~~44

o 60

.0

0~~~~3

32 28

~~~~~~~~~~33

040 6)

28

28

27

z 30

i6 z

z

10

0

0-9 10-19 20-29 30-39 40-49 50-59 60+ Age Groups(Years)

Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec Month

Figure 1. Age distribution of newly diagnosed

Figure 2. Monthly diagnoses of new diabetic

patients 1986-1989.

patients 1986-1989.

children ages 0 to 19 was 2.1/100 000 (CI95 = 1.4, 2.9). About 70% (226) of all the patients were town residents; 19.2% (62) had office jobs, and 26% (84) were laborers. The majority of women 64.2% (84) were housewives. Two hundred twenty-eight (70.6%) patients had Koranic and basic primary school education while 28 (8.7%) had Koranic school education only. A positive family history of diabetes was found in 13.8% (35) and 6.6% (4) of the NIDDM and IDDM patients, respectively. Past or current history of alcohol consumption was found in only 10.7% (27) of the patients while that of cigarette smoking was detected in 22.5% (52). All patients with history of alcohol consumption or smoking were males. The classic symptoms of diabetes (polyuria, thirst, hunger, weightlessness, and tiredness) were present in 90.7% (293) of all patients. Three (4.9%) of the IDDM patients died of diabetic ketoacidosis a few days after diagnosis. The time of onset of symptoms and subsequent diagnosis ranged from 1 to 15 weeks, with a median of 8 weeks. Figure 2 shows a seasonal distribution pattern in the diagnosis of both IDDM and NIDDM patients. Generally, incident cases increased during the cool season of August to December. No clear pattern was observed with the time of onset of symptoms. Random blood glucose levels ranged from 8.7 mmol/L to 24.4 mmol/L (average 13.7 mmol). Hypertension was observed in 29 of 253 (11.5%) NIDDM patients, 7 males and 22 females. The mean (± SD) systolic and diastolic blood pressures were 124 mmHg (±21) and 88 mmHg (± 14) in NIDDM patients and 110 mmHg (±16) and 68 mmHg (±13) in IDDM patients, respectively. None of the IDDM cases had

hypertension. About 43.5% (10) of NIDDM patients were of normal weight. Only 16.2% (41) of all NIDDM patients were obese, most of whom (70.7%; n = 29) were women. Most of the IDDM patients were underweight. Retinopathy was detected in 24.1% (61) of NIDDM patients and 6.6% (4) of IDDM patients. The table summarizes some of the clinical features of the newly diagnosed diabetics.

JOURNAL OF THE NATIONAL MEDICAL ASSOCIATION, VOL. 85, NO. 8

DISCUSSION The results of this study clearly show that diabetes is a disease of public health importance in the Islands of Zanzibar. Our findings are similar to those of Swai et a125 in a study of new diabetic patients diagnosed at the Muhimbili Medical Centre, in Dar-Es-Salaam, mainland Tanzania. Diabetes was found to be more common among Zanzibari males than females, and NIDDM seemed to present itself at an earlier age in East African countries than in Western countries. Similarly, obesity did not seem to be a common finding in newly diagnosed Zanzibari diabetic patients, in contrast with diabetic individuals in Western populations.26 The education levels and employment status of our diabetic patients was similar to those of the general Zanzibari population. This is contrary to what was found by Swai et a125 and Bella27 who reported a higher proportion of patients who had received minimal or no formal education compared with the general population in

Tanzania and Nigeria respectively. Only one IDDM patient under the age of 10 years was diagnosed during the whole period of the study. This may be the result of underascertainment due to early deaths of patients before diagnosis as a result of ignorance of the disease among medical workers and in 623


TABLE. CLINICAL FEATURES OF NEWLY DIAGNOSED DIABETIC PATIENTS (1986-1989) IDDM NIDDM Clinical Feature Mean age at diagnosis 27.6±13.4 44±10 (years ±SD) Classic symptoms 97.8% 89.9% Family history of diabetes 6.6% 13.8% Retinopathy 6.6% 24.1% Hypertension 0.0% 11.5% Obesity (BMI > 30) 0.0% 16.2% Abbreviations: SD = standard deviation, I DDM = insulin-dependent diabetes, NIDDM = noninsulindependent diabetes mellitus, and BMI = body mass index.

the population. It also may be due to much belief in witchcraft and traditional healers as evident in the long period lapse between symptom onset and subsequent diagnosis of diabetes. The estimated age 0 to 19 incidence of 2.1/100 000 for the Zanzibari population is higher than might be expected for a low-incidence nonwhite African population, but it is still lower than similar rates reported for black populations in the United States28 and Caribbean.29 It has been suggested that IDDM susceptibility genes more common in the white population have become admixed into United States and Caribbean black populations, and thus may account for the higher rates of IDDM seen in western hemisphere blacks compared to African black populations.30 It is also possible that the moderate rate of IDDM seen in the Zanzibari population also might result from admixture with Arab (white) populations who have traded and settled in the Zanzibar islands for centuries. Future studies to determine the degree of underascertainment of IDDM cases in Zanzibar and to estimate the incidence of IDDM in other black African populations will be necessary to better assess differences in IDDM incidence in sub-Saharan black African populations. Hypertension, as reported in some other studies in Africa,31'32 is not uncommon in Zanzibari NIDDM patients. The low proportion of diabetic individuals with current or past history of alcohol consumption is probably due to the fact that the majority of Zanzibaris (about 98.8%) are Muslims and religious practice prohibits them from consuming alcohol. Seasonality in the presentation of both NIDDM and IDDM patients is similar to that found in Dar-EsSalaam25 but with a steep rise in the month of December. This seasonal trend of presentation of the 624

disease also has been reported by researchers in several other countries.27'33-35 No clear pattern was observed with the time of onset of symptoms, probably due to inaccuracies in recalling past events. The size of the problem of diabetes in Zanzibar is evident. It is now necessary for the Ministry of Health in Zanzibar to put more effort into the training of physicians, paramedicals, and nurses on diabetes and its management. Standardized management guidelines should be formulated and used in all hospitals. Although home blood glucose monitoring is not feasible for most Zanzibari diabetic patients, the government should ensure that insulin, oral hypoglycemic agents and other essential diabetic supplies are available at least at the referral hospital and major district hospitals. This will enable the provision of basic diabetic care. The population also should be educated on diabetes, and their support to the Diabetes Association of Zanzibar should be solicited. The development of a noncommunicable disease control program and inclusion of diabetes as one of the target diseases, should be one of the priorities of the Ministry of Health, Zanzibar. This measure will be an implementation of what has been recommended by WHO.23 Such a measure would be both logical and cost-effective.36'37 Measures to encourage national and international collaboration are important in stimulating further research in these Islands, which are virgin to diabetes research workers. Acknowledgments The authors thank M Rajab M/A, M Haji, and A Haji of Mnazimmoja Hospital, Zanzibar Town, Zanzibar, for their help in data collection. They also thank Professor Alberti KGMM of the School of Clinical Medical Sciences, New Castle upon Tyne, United Kingdom, and Ron J. Vergona, MSc of the Diabetes Research Center, Pittsburgh, Pennsylvania, for reviewing the manuscript and providing valuable comments and recommendations.

Literature Cited 1. Bajaj JS. Guest editor's introduction: diabetes mellitus in developing countries. IDF Bulletin. 1986;31:6-7. 2. Zimmet P. Type 2 (non-insulin-dependent) diabetes-an epidemiological overview. Diabetologia. 1982;22:399-41 1. 3. Federation News: African region; first Pan-African congress on diabetes. IDF Bulletin. 1986;31:119-120. 4. Sandwith FM. Diabetes in Egypt. BMJ. 1907;11:1059. 5. Zmirou D. Epidemiologie du diabete en Cote d'lvoire (Thesb). Grenoble: Universite de Grenoble; 1979. 6. Goldberg MD, Marine N, Ribeiro F, Campbell GD, Vinik AL, Jackson WPU. Prevalence of glycosuria and diabetes among Indians and Bantu. S Afr Med J. 1969;43:733-738. 7. McLarty DG, Pollit C, Swai ABM. Diabetes in Africa. JOURNAL OF THE NATIONAL MEDICAL ASSOCIATION, VOL. 85, NO. 8


Diabet Med. 1990;7:670-684. 8. Dodu SRA. The incidence of diabetes in Accra (Ghana). A study of 4000 patients. West Afr J Med. 1958;7:129-134. 9. Tulloch JA. The incidence of glycosuria and diabetes mellitus among hospital outpatients in Uganda. East Afr Med J 1964;41 :572-580. 10. Steel JM, Mngola EN. Diabetes in Kenya. Trop Doct. 1974;4:184-187. 11. Lester FT. Childhood diabetes mellitus in Ethiopians. Diabetic Med. 1986;3:278-280. 12. Makame MH. Diabetes in Zanzibar. Practical Diabetes Digest. 1991;2:160. 13. Corrigan CB, Ahren B. Ten years experience of a diabetes clinic in northern Tanzania. East Afr Med J. 1987;64:772-780. 14. Kinnear TWG. The pattern of diabetes mellitus in a Nigerian teaching hospital. East Afr Med J. 1963;40:288-294. 15. Castle WM, Wick ACB. A follow-up of 93 newly diagnosed African diabetics for 6 years. Diabetologia. 1980;18:121-123. 16. Osuntokun BO, Akinkugbe FM, Francis TI, Reddy S, Osuntokun 0, Taylor GOL. Diabetes mellitus in Nigerians: a study of 832 patients. West Afr J Med. 1971;20:295-312. 17. Elamin A, Omer MIA, Hofvander Y, Tuvemo T. Prevalence of IDDM in school children in Khartoum, Sudan. Diabetes Care. 1989;12:430-432. 18. Rewers M, LaPorte RE, King HOM, Tuomilehto J. Trends in the prevalence and incidence of diabetes: insulindependent diabetes mellitus in childhood. World Health Stat 0. 1988;41 :179-190. 19. WHO DIAMOND Project Group. The multinational project for childhood diabetes. Diabetes Care. 1990; 1 3:1062-1068. 20. Statistical Tables for Health Planners and Administrators. Basic Statistical Data for Decision Makers in the Health Sector. Zanzibar: Ministry of Health; March 1987. Department of Planning, Administration and Finance health bulletin no 47. 21. Makame M, Rajab M, Haji A. Clinical features of new Zanzibari diabetics: a preliminary report for the year 1987. Diabetes Res Clin Pract. 1988;5(suppl):S210. 22. Report of a WHO Expert Committee. Arterial hypertension. WHO Technical Report Series. 1978;628:1 -58. 23. World Health Organization Study Group. Diabetes mellitus. WHO Technical Report Series. 1985;728:10-24. 24. Snedecor GW, Cochran WG. Statistical Methods. Ames,

JOURNAL OF THE NATIONAL MEDICAL ASSOCIATION, VOL. 85, NO. 8

Iowa: Iowa State Univ Press; 1980. 25. Swai ABM, Lutale J, McLarty DG. Diabetes in tropical Africa: a prospective study, 1981-1987. I. Characteristics of newly presenting patients in Dar-Es-Salaam, Tanzania, 19811987. BMJ. 1990;330:1103-1106. 26. Multi-centre study. UK prospective diabetes study. IV. Characteristics of newly presenting type 2 diabetic patients male preponderance and obesity at different ages. Diabet Med. 1986;5:154-159. 27. Bella AF. A prospective study of insulin-dependent diabetic Nigerian Africans. J Natl Med Assoc. 1992;84:126-128. 28. LaPorte RE, Tajima N, Dorman JS, Cruickshanks KJ, Eberhardt MS, Rabin BS, et al. Differences between blacks and whites in the epidemiology of insulin-dependent diabetes mellitus in Allegheny County, Pennsylvania. Am J Epidemiol. 1986; 1 23:592-603. 29. Tull ES, Roseman JM, Christian CLE. Epidemiology of childhood IDDM in the US Virgin Islands from 1979 to 1988: evidence of an epidemic in the early 1980s and variation by degree of racial admixture. Diabetes Care. 1990;14:558-564. 30. MacDonald MJ. Lower frequency of diabetes among hospitalized Negro than white children: theoretical implications. Acta Genet Med Gemellol. 1975;24:1 19-126. 31. Mhando PA, Yudkin JS. The pattern of diabetic complications in African patients in Dar-Es-Salaam. Trop Geogr Med. 1 980;32:31 7-323. 32. Omar MAK, Asmal AC. Patterns of diabetes mellitus in young Africans and Indians in Natal. Trop Geogr Med. 1984;36:133-138. 33. Fleegler FM, Rogers KD, Drash A, Rosenbloom AL, Travis LB, Court JM. Age, sex, and season of onset of juvenile diabetes in different geographical areas. Pediatrics. 1 979;63:374-379. 34. Durruty P, Ruiz F, Garcia de los Rios M. Age at diagnosis and seasonal variation in the onset of insulin-dependent diabetes in Chile (Southern Hemisphere). Diabetologia. 1 979;63:374-379. 35. Gamble DR, Taylor KW. Seasonal incidence of diabetes mellitus. BMJ. 1969;111:631-633. 36. WHO Expert Committee. Community prevention and control of cardiovascular disease. Geneva. WHO Technical Report Series. 1986;732:7-61. 37. Zimmet PZ. Primary prevention of diabetes mellitus. Diabetes Care. 1988;1 1:258-262.

625