A GEOGRAPHICAL ANALYSIS OF DIABETES MELLITUS IN OYO STATE, NIGERIA

Abstract

The substantial increase in the incidence rates of non-communicable diseases such as Diabetes Mellitus (DM) is closely linked to the social and economic changes that accompany development. Most research on DM focused on epidemiological analysis of individual-level predictors within populations, while few studies on the spatial pattern and its association with the level of development exist. This study, therefore, analysed the geographical pattern of DM in relation to the level of development in Oyo state, Nigeria.

The Human ecology of disease and epidemiological transition theories provided the framework. The design is both hospital-based and cross sectional. Stratified random sampling technique was used to collect primary data from the adult population in households of the 33 Local Government Areas (LGAs) of Oyo state. A structured questionnaire covering self-reported DM, family history of DM, socio-economic, behavioural, environmental and lifestyle risk factors was administered to 1,619 respondents using the Neuman’s approach. DM cases from 2000 to 2014 were obtained from the eight health facilities with DM registries in Oyo state: University College Hospital, Ibadan, Ring Road State Hospital, Ibadan; University Health Services (Jaja Clinic), Ibadan; Jericho Nursing Home, Ibadan; General Hospital, Moniya; Baptist Hospital, Oyo; Baptist Medical Centre, Saki and Bowen Teaching Hospital, Ogbomoso. The hospital records provided information on the location of these cases. Development indicators (Internally generated revenue, percent of population that is urban, medical facilities, medical personnel, hospital beds, primary and secondary schools, primary and secondary school teachers, commercial banks, manufacturing industries, hotels, households with access to electricity, and households with access to pipe borne water) at LGA levels were extracted from published sources. Descriptive statistics, Moran’s I, Local Getis Ord (hotspot analysis), Stepwise regression, Principal component analysis and Pearson correlation were used for data analyses at p<0.05.

There was a total of 2,723 DM (male: 1,334 and female: 1,323) cases from hospital records, from 2000 to 2014 while 13.7% of the survey respondents were diabetic. The DM rates varied from Ibadan North (29.3/10,000), to Surulere (0.1/10,000) LGAs while self-reported DM varied from Iseyin (42.2%) to Ogbomoso North (4.0%) LGAs. There was a significant positive spatial autocorrelation in DM (I=0.30; z=4.6). The DM hotspots were found in Akinyele (z=3.6), Ibadan North (z=3.6), Ibadan Northwest (z=3.9) and Lagelu (z=3.2) LGAs. A hotspot for self-reported DM was found in Iwajowa (z=3.1) LGA. Proximity to bus stop (b=-0.18; t=-2.2) and percent urbanisation (b=0.08; t=2.2) significantly influenced the spatial distribution of DM rates (R2=0.34; F=5.8). Family history of DM (b=0.63; t=5.9) was significant in the explanation of the spatial pattern of self-reported DM (R2=0.53; F=35.3). The index of development showed that Ibadan North LGA was the most developed (1.78) while Ibarapa North LGA was the least developed (-0.41) in the state. There was a significant positive correlation between DM and the level of development (r=0.68) among the LGAs.

Diabetes mellitus incidence varies significantly across the local government areas and was closely related to the level of development. Therefore, area-specific prevention and intervention programmes may be required to reduce diabetes mellitus in Oyo state.

Keywords: Diabetes mellitus, Human ecology, Epidemiological transition theory, Spatial pattern, Development indicators