The prevalence of vitamin D deficiency among patients with type 2 diabetes seen at a referral hospital in Kenya

Introduction The prevalence of diabetes mellitus is rising at an alarming rate, calling for more insights into its pathogenetic mechanisms, and other factors involved in its progression. The prevalence of vitamin D deficiency is higher in diabetic compared to non-diabetic patients, and is associated with poor glycaemic control. This has not been documented among diabetic patients in Kenya. Aims: to determine the prevalence of hypovitaminosis D among type 2 diabetic patients at Kenyatta National Hospital in Nairobi, Kenya. Methods We recruited type 2 diabetic patients on follow-up at Kenyatta National Hospital. Measurements of height, weight and waist/hip ratios were taken. We drew 6mls of peripheral blood to determine vitamin D, zinc and HbA1c levels. Results A total of 151 participants were recruited, with 69.5% females and mean age of 58.2 years. Hypertension was found in 72.8% of the participants, and obesity in 37.7%. The mean HbA1c levels were 8.46%, and 62.9% had poor glycaemic control. The mean vitamin D level was 31.40ng/ml. Vitamin D deficiency and insufficiency was found in 38.4% and 21.9% of the participants respectively. We found a significant inverse correlation between vitamin D and glycaemic control (r = -0.09, p = 0.044) and vitamin D and BMI (r = - 0.145, p = 0.045). Conclusion In this study population on long-term follow-up for diabetes, there was high prevalence of vitamin D deficiency. This forms a basis for further management of patients with poor glycaemic control. Further studies are needed to document the causal association between poor glycaemic control and vitamin D deficiency.


Introduction
Type 2 diabetes mellitus, characterized by pancreatic β-cell dysfunction and peripheral insulin resistance is now highly prevalent in developing countries. It is associated with significant morbidity and mortality due to microvascular and macrovascular complications.
Diabetes affects over 300 million people worldwide, and its prevalence is expected to double by 2030, with low and middleincome countries carrying up to 80% of the burden [1]. Therapies for type 2 diabetes have dramatically improved over the last two decades, but the rising burden calls for more insights into prevention and management of the disease. Lifestyle changes such as weight loss, dietary modifications, exercise and cessation of smoking have been shown to improve glycaemic control and reduce complications of type 2 diabetes [1]. Vitamin D receptors are present in the β cells of the pancreas and vitamin D has been linked to insulin secretion regulation [2]. Vitamin D levels may have an inverse relationship with HbA1C and low levels of vitamin D show a correlation to increased incidence of type 2 diabetes [3]. Vitamin D deficiency is as highly prevalent as diabetes in the general population. Emerging evidence suggests that vitamin D is involved in the aetiology and pathogenesis of diabetes mellitus. Several studies have implicated vitamin D deficiency in the development and progression of diabetes, while high plasma vitamin D is related to lower risk of developing diabetes in high risk patients. Vitamin D deficiency is involved in central pathogenetic mechanisms of diabetes mellitus; it affects insulin sensitivity and β-cell function [4]. In this study, we aimed to assess vitamin D levels among type 2 DM patients on follow-up in an outpatient clinic in a large referral hospital, and document the association between hypovitaminosis D and glycaemic control.

Methods
This was a descriptive cross-sectional study, conducted among type 2 diabetic patients attending the diabetes outpatient clinic at Kenyatta National Hospital. The diabetes outpatient clinic runs once a week, and caters for 150 to 200 patients per week. It is manned by nurses, podiatrists, medical residents and consultants. Kenyatta National Hospital is a level 6 (National Referral Hospital) in Nairobi, Kenya, that handles referrals and specialized cases from all over the country. It has an inpatient bed capacity of 1500, and serves as a university teaching hospital for the University of Nairobi. We included all type 2 diabetic patients above 18 years of age who consented to the study.
Patients were selected consecutively to join the study, until the sample size was attained. We excluded patients with renal disease, as evidenced by estimated glomerular filtration rate of < 30ml/min/1.73m 2 calculated from serum creatinine, patients with liver disease (if Alanine Aminotransferase was > 5 times the upper reference limit) and patients taking multivitamin supplements, steroids, anticonvulsants (especially barbiturates and phenytoin). We also excluded patients who had taken multivitamins or vitamin D supplements within 6 months of the study. Kenya experiences two seasons (rainy and dry season), with the sun present in both seasons.
This has little effect on Vitamin D in the bodies, hence no significant seasonal variations.
A pre-designed questionnaire was used to collect data on patient's age, gender, duration since diagnosis of diabetes, other comorbidities, drug history, any diabetic complications and current treatment for diabetes mellitus. We performed anthropometric measurements of all eligible study participants in light clothing and without shoes. These included height, weight, hip and waist circumferences. Height and weight were measured to the nearest 10 th cm and kg respectively. Body Mass Index (BMI) was calculated from the two measurements, in kg/m 2 . Hip circumference was taken as the greatest circumference at the level of greater trochanters (widest portion of hips) on both sides. The waist circumference was measured as the smallest horizontal girth between costal margins and iliac crests at minimal inspiration. The waist/hip ratio was calculated.
Vitamin D was analyzed using High-Performance Liquid Chromatography (HPLC) method. We used the agilent 1100 HPLC system with a quartenary pump. Four milli-liters of blood were centrifuged and serum separated for use in the analysis. The concentration of vitamin D was measured in ng/dl, with ≥ 30ng/dl as sufficiency, 20.1-29.9ng/dl as insufficiency and < 20ng/dl as deficiency. It was extracted from serum using acetronitryl with 0.4% acetic acid. This was injected into a chromatographic column, Polaris C18-A 3µ 150 x 2.0mm. HPLC uses isocratic mode with eluent (MeCN: 0.4% Acetic acid) as the mobile phase. The flow rate was 0.3mls/min and oven temperature 30 degrees Celsius, and detector was UV-Vis at wavelength of 280nm. The amount of vitamin D was determined by matching the retention of pure standard and a calibration code.
Data was coded, entered into SPSS version 23.0 (IBM), and cleaned.
Continuous variables like age, BMI, Waist/hip ratio, vitamin D, were expressed in means ±SD, or median (Interquartile range). Categorical variables like vitamin D insufficiency and deficiency were analysed as proportions, n (%). Comparison and correlation of vitamin D levels and glycaemic control were done using paired students' t-test, with level of significance set at p < 0.05. Chi-squared tests were used to analyze relationships between categorical variables such as gender and vitamin D states (deficiency or insufficiency). We performed a bivariate analysis to see the relationship between vitamin D levels and glycaemic control. Ethical approval was sought from the University of Nairobi/Kenyatta National Hospital Ethics and Review Committee, and patients were consented accordingly (ethical approval reference number KNH-ERC/A/207).  Table 3 illustrates the complete laboratory characteristics of the patients, while Table   4 shows the correlation between selected patient characteristics and vitamin D status. Bivariate analysis with Pearson correlation showed a significant inverse correlation between vitamin D levels and HbA1c

Relationship between vitamin D and glycaemic control: in the present study, a a significant negative correlation existed between vitamin D levels and glycated haemoglobin. A significant inverse correlation between glycaemic control and vitamin D levels has been
reported by several authors [10,16]. It appears that the inverse relationship is not limited to diabetes only, as Kostoglou-Athanassiou et al [10] reported a significant negative correlation between vitamin D and glycated haemoglobin in controls who were non-diabetic. On the other hand, some studies have found no significant correlation between vitamin D and glycaemic control [6].
A cross-sectional Finnish study found an inverse association between vitamin D levels and fasting insulin, fasting glucose and 2 hour glucose tolerance results [17]. This provides further evidence that low serum vitamin D may be associated with impaired glucose metabolism. An inverse association has been described between vitamin D levels and insulin resistance, especially at vitamin D levels between 16 and 36ng/ml [18]. Longitudinal studies have also reported a correlation between vitamin D levels and progression to overt diabetes among pre-diabetic and normal individuals [19]. A follow-up cohort of non-diabetic patients for 10 years revealed an inverse relationship between serum vitamin D levels and future glycaemic and insulin resistance [19].
There's no robust evidence on the role of supplementation of vitamin D on glycaemic control. One study conducted among African Americans showed that vitamin D supplementation significantly increased glycaemic control [20]. In one study, supplementation of vitamin D conferred a lower risk of diabetes to the participants when followed over time [3,21]. The National Health and Nutrition  [11]. This inverse association between vitamin D and BMI is possibly connected to glycaemic control. There is robust evidence that poor glycaemic control is correlated with low vitamin D levels. Further, obesity is associated with poor glycaemic control, which may partly explain the low levels of vitamin D. Whether any mechanistic association exists between vitamin D and obesity remains to be elucidated.
Study limitation: our study cannot however show the causal association between glycaemic control and vitamin D levels. We therefore recommend further case-control and interventional studies.

Conclusion
We found a high prevalence of vitamin D deficiency and insufficiency in our study population. The glycaemic control and body mass index correlated negatively with vitamin D levels. In long-term diabetics with poor glycaemic control, looking for other correlates of glycaemic control like vitamin D is prudent.
What is known about this topic  The relationship between vitamin D deficiency and poor glycemic control is well documented in previous studies.