Seasonal Variation of Type 1 Diabetes Diagnosis in Children, Adolescents and Young Adults in Kenyatta National Hospital -retrospective Cohort Study

Abstract

Background-The most commonly diagnosed endocrine disorder in adolescents, children and young adults is T1DM. The prevalence of T1DM in Kenya is rising with an estimated 1,694 children and adolescents having T1DM (IDF Diabetes Atlas 2019). Seasonal variation is an environmental factor that triggers type 1 diabetes onset in genetically predisposed children. Studies done show an increased type 1 diabetes diagnosis in cold seasons for particular age groups compared to warm seasons. Objectives-The study aimed to determine the seasonal variation of T1DM diagnosis in children, adolescents and young adults in diabetes clinic in KNH between 2009 to 2021 and its correlation with the various age categories and geographical patterns. Methodology-This was a retrospective cohort study carried out at diabetic clinic at Kenyatta National Hospital among children, adolescents and young adults who at the point of diagnosis with T1DM were aged between 0 - <25 years between 2009 to 2021.Demographic data (age, sex), age at diagnosis with T1DM, month and season of diagnosis and patient’s geographical origin were recorded in a data collection form. The outcomes of interest were the seasonal variation in the diagnosis of T1DM based on sex, age and age categorization, geographical origin, month and season of diagnosis of T1DM.Data was analyzed based on age, sex and month, season of diagnosis and geographical origin. Age categorization of 5 year intervals at 0-4 years, 5-9 years, 10-14 years, 15-19 years and 20-24 years were used and seasonal and geographical origin were compared. Categorical variables were summarized into proportions and continuous variables into medians where applicable. Statistical analytical methods using available software R Studio was used to inference any association and strength or significance of the associations thereof in seasonality of diagnosis of T1DM. Statistical tests were interpreted at 5% level of significance (p value less or equal to 0.05) and findings presented in form of tables, graphs and charts. Results-A total of 1,250 patient charts were identified by health records using ICD 10 code for T1DM diagnosis. Out of these 379 were consecutively sampled and included in the study and analyzed. The median age was 9 years with an interquartile range of 3-12 years. There were 201(53%) males and 178(47%) females. The data was grouped into six age categories ages: 10-14 years were the majority at 35.1% (n=133) followed by those aged 0-4 years at 29% (n=110). Participants aged 20-24 years were the least at 4.5% (n = 17).Four Kenyan seasons were used as per Kenya Meteorological Department namely MAM or “long rains” season, OND or “short rains” season , JJA &S or “cool dry” season and JF or hot and dry xiii season. Among the four seasons, JJA&S had the highest diagnosed cases at 37.5% followed by MAM at 27.4%. JF season had the least diagnoses at 16.4%.In the 12 months of the year, March had the highest percentage of cases at 12.1% followed by August and September at 11.3% each. The month of November had the least diagnoses at 3.2%.To determine seasonality of the data decomposition of additive time series to was done using two approaches. The first was a time series plot of the monthly decomposed data for each year from 2009 to 2021 and the second involved time series plot of the four Kenyan seasons decomposed data for each year from 2009 to 2021.In both, repeating patterns of highs (peaks) and lows (troughs) was observed in the seasonal panel related to the months and seasons of the year, which suggested seasonality in the data. A significant statistical association between gender and season of diagnosis was established (p<0.01) with more females in the MAM season. There was also an association between the various age categorization at diagnosis and seasonality (p=0.04).It was also established that there was a significant association between the place of residence (urban vs rural) and season of diagnosis (p=0.004). Conclusion-In this study we observed seasonality of T1DM in children, adolescents and young adults attending KNH diabetic clinic. This is the first study exploring seasonality of T1DM in Kenya. The study alludes to the role of environmental factors in triggering T1DM diagnosis. The study also provides a basis for evaluation of the time varying environmental variables such as rainfall precipitation, temperature, seasonal viral and bacterial infections and specific geographical points of locations in regards to onset of T1DM diagnosis.