ABSTRACT The present study was undertaken to determine the prevalence of known cases of diabetes and hypertension among adults in Sana’a city. Thus 1080 persons aged 20–85 years were selected for interview using a multistage random sampling technique. The crude prevalence of known diabetes was 6.57% (95% CI: 5.2–8.2) and of known hypertension 13.5% (95% CI: 11.5–15.6). The age-standardized prevalence for the age range 30–64 years was 9.75% (95% CI: 7.55–11.95) for diabetes and 17.1% (95% CI: 15.0–19.2) for hypertension. These results provide an estimate of the prevalence of known diabetes, hypertension and related vascular events in a well-defined urban community.

RESUME La présente étude a été réalisée pour déterminer la prévalence des cas connus de diabète et d’hypertension chez des adultes dans la ville de Sanaa. Ainsi, 1080 personnes âgées de 20 à 85 ans ont été sélectionnées pour un entretien en utilisant la technique de l’échantillonnage aléatoire à plusieurs degrés. La prévalence brute du diabète connu s’élevait à 6,57 % (IC 95 % : 5,2-8,2) et de l’hypertension connue à 13,5 % (IC 95 % : 11,5-15,6). La prévalence standardisée pour l’âge dans la tranche d’âge 30-64 ans était de 9,75 % (IC 95 % : 7,55-11,95) pour le diabète et de 17,1 % (IC 95 % : 15,0-19,2) pour l’hypertension. Ces résultats fournissent une estimation de la prévalence du diabète, de l’hypertension connus et des évènements vasculaires qui y sont liés dans une communauté urbaine bien définie.

Introduction

Methods

The sample size was calculated using Epi-Info, version 6.02 taking into consideration the following criteria: total target population (i.e. adult population aged ³20 years living in Sana’a city): 464 000 people; desired precision: 2.5%; expected frequency: 3.5% for known diabetes and hypertension, estimated since no previous study has been conducted in the country; sample size with 95% confidence interval (CI): 1294 people.

The sampling procedure was based on a multistage random technique. In the first stage, Sana’a city was divided into four geographic zones; one-fourth of the sample was drawn from each zone. The second stage involved the random selection of one “election circle” from each geographic zone. In the third stage each selected election circle was divided into streets. The fourth stage included the random selection of four streets within each election circle: it was planned to interview 80 participants from each street (40 males and 40 females). The fifth stage was the systematic random selection of houses from each street and one participant (³20 years old) from each household. The survey team included 16 volunteers in each geographic zone (a total of 64 volunteers), who were final-year students in the Department of Sociology, University of Sana’a. These future social researchers were chosen because they were trained in fieldwork and were able to sensitize people to the aims of the study and to encourage their participation. They visited the randomly selected households in groups of four volunteers and inquired about diabetes, hypertension, heart attack and stroke, using standard terms. The volunteers were taught how to administer a predesigned questionnaire. Details taken included the participant’s name, age, sex, place of birth, current residence, standard of education reached, presence of diabetes mellitus (and age at which diabetes was diagnosed), presence of hypertension, ischaemic heart disease or stroke, and any family history of diabetes, hypertension, ischaemic heart disease or stroke in the parents, siblings, spouse or offspring. The questionnaire format was verified by validity and reliability testing. Reliability was ensured by pretesting a subsample of 100 subjects. The survey team was provided with a list of specific drugs used in the management of diabetes mellitus, hypertension, ischaemic heart diseases and stroke.

Table 2 Basic characteristics of the study population (n = 1080)

The prevalence rates and 95% CI of known diabetes mellitus and hypertension in the study population are given in Table 3. The crude prevalence rate of known diabetes mellitus was 6.57% (95% CI: 5.2–8.2), with a slightly higher rate in males than in females (7.8% versus 5.5%). Using the world population as the standard [6,7], the age-standardized prevalence rate of known diabetes mellitus in a truncated age range of 30–64 years for the Yemeni standard population for the year 2000 was 9.75% (95% CI: 7.55–11.95) with a slightly lower rate in males (9.4%, 95% CI: 7.9–10.9) compared to females (11.8%, 95% CI: 9.9–13.7).

Table 3 Crude and age-standardized prevalence rates of known (previously diagnosed) diabetes mellitus and hypertension in the adult population in Sana’a, Republic of Yemen

The coexistence of known cases of diabetes mellitus, hypertension, ischaemic heart disease or stroke in the study population is shown in Figure 1. It indicates that hypertension was more frequent than diabetes mellitus (14% versus 7%) and that both had co-morbidity with each other (3%) and with ischaemic heart disease (2%) and stroke (1%).

Table 4 Prevalence of known cases of diabetes, hypertension, ischaemic heart disease and stroke among the relatives of index subjects (n = 1080)

This population-based survey in the capital city has elicited important information about the prevalence of known cases of diabetes mellitus, hypertension and vascular events of the heart and brain among the study population and their relatives. Table 4 indicates that a history of known diabetes was present in 507 subjects and their relatives with a peak prevalence in fathers (33.4%), then mothers (18%), and to a lesser extent in the index subjects (14%) and their brothers (12.4%) and spouses (12%). A history of known hypertension was reported in 711 subjects and relatives with nearly equal frequency in mothers (24.3%) and fathers (23.8%), then index subjects (20%) followed by their spouses (13.4%). A history of known ischaemic heart disease was reported by 246 participants with a peak prevalence in fathers (36.2%), followed by mothers and brothers (18.3% each), then in the index subjects themselves (11%). A history of known stroke was reported by 81 subjects and relatives with a peak prevalence in fathers (34.6%), followed by mothers (17.3%), brothers (16%) and spouses (12.3%). The prevalence among index subjects was lower (9.9%).

Figure 2 Age at diagnosis of diabetes (early onset <40 years versus late onset ³40 years) in index subjects and their parents (n = 71)
EODM = early onset diabetes mellitus; LODM = late onset diabetes mellitus

Of the diabetic participants, 35 had diabetic parents and 36 did not (Figure 2). Of the 35 diabetic parents, 25 (71.4%) had late onset diabetes (onset ³40 years), while 29 (83%) of their diabetic offspring had early onset diabetes (onset <40 years). The diabetic offspring of non-diabetic parents had an approximately equal opportunity of having either early onset diabetes mellitus (n = 17, 47%) or late onset diabetes mellitus (n = 19, 53%).