A Trial to Study the Effect of Long Term Vitamin D Supplementation on Insulin Sensitivity

Overview

Type 2 diabetes is a major public health problem in India with an estimated a prevalence of approximately 4% and 12% in rural and urban areas respectively. Accumulating evidence suggests that serum cholecalciferol levels may be inversely related to the prevalence of diabetes, insulin resistance and metabolic syndrome. The trials available on the effect of Vitamin D supplementation on insulin/glucose metabolism have been conducted using small sample sizes in different subgroups document variable results (significant decrease in HbA1c concentration with insulin concentration in hemodialysis patients; insulin levels lower with oral Vitamin D in gestational diabetes; no effect of Vitamin D on serum insulin levels in post menopausal women). A double blind randomized controlled trial conducted at our institute using 3,60,000 IU of cholecalciferol over 6 weeks documented improvement in OGIS index of insulin sensitivity. We therefore, plan to study the long term effect of vitamin D supplementation on peripheral insulin sensitivity.

Volunteers will be recruited from amongst the preventive health check subjects, neighboring offices, hospital employees and corporate associates of SBISR after obtaining a detailed informed consent. The study will include 300 subjects with the following selection criteria. INCLUSION CRITERIA 1. 35 to 75 years of age 2. Waist circumference (WC) ≥78 cm in men and ≥ 72 cm in women Recruited subjects will be randomized into two groups on the basis of random numbers generated by a computer after pre-stratification by gender (to allow equal representation of both sexes) and baseline Vitamin D levels. The baseline information will be recorded on standardized questionnaire include age, education, marital status, annual household income, family history, smoking and alcohol use. A complete physical examination will be done including height, weight, waist circumference and blood pressure. Weight will be recorded on a manual weighing scale (sensitivity 500 g), height using SECA stadiometer (sensitivity 0.1 cm), waist circumference (WC) at the midpoint between the lower rib and iliac crest using a measuring tape (sensitivity 0.1 cm), and blood pressure using an OMRON electronic instrument (sensitivity 1 mmHg; accuracy 3 mmHg) validated in an earlier trial. Height, WC and weight will recorded with light clothing and without shoes. Three serial BP recordings from the right arm will be taken after 10 minutes of rest at 10 minute intervals in the sitting posture as per WHO recommendations. The mean of the three recordings will used for analysis. Daily sun exposure will be calculated by taking a detailed history of sun exposure separately during summers and winters and the type of clothing worn. Sun exposure will be calculated as hours of exposure per day x percentage of body surface area (BSA) exposed (calculated according to Wallace rule of nine). A detailed dietary assessment by the 24 hour recall method will be done at baseline to determine the dietary intake of calcium, phosphate and Vitamin D. Consenting subjects will be advised to fast from 8 pm of the previous night. 10 ml of blood will be drawn at baseline for the following tests between 9 am to 10 am on the next day. * 25 (OH) vitamin D levels * Serum calcium and phosphorus * Liver function tests including total proteins, SGOT, SGPT and alkaline phosphatase * Serum creatinine levels The 3-hour Oral GTT will be done after a loading glucose dose of 75 grams with blood sampling at 0, 120 and 180 minutes. The sampling arm will be warmed for 20 minutes prior to sampling to obtain arterialized venous samples. The blood samples will be immediately transported to the laboratory where plasma will be separated for all samples immediately. The samples will be labeled with a random numbers (to avoid revealing the time sequence of the samples to the laboratory) and stored at -70 C until assay. The subjects in the treatment group will be given cholecalciferol 120,000 IU monthly orally for one year (\~ 4000 IU per day). This dose has been documented to be safe in an earlier review.The drug will be advised and provided to the subjects at enrollment. The control group will receive a placebo which will be identical in taste, color and texture. The subjects will be advised to continue their normal routine and lifestyle and to report immediately for any vomiting, headache, blurring of vision, abdominal pain, muscle cramps hematuria or hospital admission. For the purpose of the trial hypercalcemia would be defined as \> 2.65 mmol/L, Urinary calcium: creatinine ratio and ultrasonography will be done if clinically indicated. Any volunteer developing any adverse effect would be treated as recommended and free of charge.. Compliance will be monitored by means of a home diary. A mid-term follow-up will be done at 6 months after recruitment. Instructions will be reinforced at the 6 month follow-up especially for subjects with poor compliance or other deviations. Final assessment will be done at 1 year (with margin of +10 days) after recruitment. Follow up assessment for any subject with a febrile illness will be postponed till 5 days after recovery from fever. A detailed dietary assessment by the 72 hour recall method will be done at baseline and at the two follow up visits to determine the dietary intake of calcium, ,phosphate and Vitamin D. All investigations done at baseline will be repeated at the two follow-up visits. The investigators and the laboratory will be blinded to the random allocation and the code will be broken only after the reports are available for all subjects. Outcome measures: The oral glucose sensitivity index as calculated by the Mari's formula as a measure of the post prandial insulin sensitivity will be evaluated as the primary outcome variable and the changes in the index will be compared between the supplemented and unsupplemented groups. The HOMA and QUICKI indices will be similarly analyzed as secondary outcomes indicating the fasting insulin sensitivity. The incidence of adverse effects will also be compared between the two groups as a secondary outcome. Biochemical Analysis The sample will be centrifuged at 3000 rpm and the serum/plasma would be stored at -70 C. Serum 25(OH) D will be measured by solvent extraction followed by radioimmunoassay (Diasorin, USA). Plasma Insulin will be measured by double antibody IRMA (IMMUNOTECH, France). PTH levels will also be estimated using the double antibody IRMA (IMMUNOTECH, France). 5% of the samples will be re-run for quality control. Sample size considerations It is estimated that a sample size of 98 subjects in each group will be required to detect a change in OGIS (Oral Glucose insulin sensitivity; primary outcome variable) of 10% with an alpha error of 0.05 and a power of 90% based on data available from the earlier trial conducted at the institute on the subject. To account for attrition over one year 150 subjects in each group will be recruited. Data Analysis Data entry and analysis will be done using Epi-Info 2002 and SPSS v13.0 software. The values for plasma glucose and insulin at the specified times will be used to calculate the following indices OGIS calculated using the Mari's formula as described elsewhere will be the primary outcome variable. HOMA and QUICKI indices will be calculated from the fasting insulin and glucose samples. The student's't' test for paired values will be used to estimate the pre and post intervention differences in each group. The crude mean difference in two groups will be compared using the unpaired't' test. Non- compliant subjects and trial deviates will be analysed in the group allocated for the 'Intention to treat' analysis and will be excluded for the "per protocol analysis" for studying the differences between groups. The outcomes will be adjusted for known confounders including age, waist circumference, BMI, baseline insulin sensitivity, changes in Vitamin D and PTH levels. The subjects will be stratified by their baseline 25 (OH) D levels into deficient and non-deficient for analysis. Regression analysis will be used to study the relationship of baseline characteristics with the outcome variables.