The present study was a cluster randomized controlled trial carried out for a period of 3 years from January 2021 to December 2023, conducted in rural areas of Kolar (Latitude: 13°07′48.00″ N and Longitude: 78°07′48.00″ E). Kolar is one of the districts in Karnataka, India. Kolar has more than 36 rural schools, out of which, 20 were randomly selected. The sample size was estimated based on the mean difference in depression score determined in a study by Libuda et al., in which an average scale variance score of 13.3 (SD) was obtained. Expecting a reduction of 20% in Beck Depression Inventory scores in the intervention group compared to in the placebo group, with 80% power and an alpha error of 5%, the estimated sample size per group was 96. Expecting a dropout rate of 5% during the study and considering that the design had a cluster design effect of 1.5, the final sample size achieved in each group was 150 [
15]. A total of 20 rural schools in Kolar were selected. Each school was considered as a cluster. The study was carried out during the summer months of February and March for the included schools and the geographical location did not differ amongst the schools. Block randomization was performed using the online software randomization.com. There were two groups after randomization: one group received a high dose of vitamin D at 2000 IU per day, and the other received a lower dose of 250 IU per day, along with calcium. All adolescents aged 14–19 years were included in the study after consent and assent were obtained. The adolescent boys and girls were screened, and those with the following were excluded from the study: any pre-existing mental health illnesses such as previously diagnosed severe depression or a history of suicidal tendencies or attempts; previously confirmed renal abnormalities; previously confirmed cardiac abnormalities; and previously confirmed neurological disorders, such as epilepsy. To assess sociodemographic status, a pretested, semi-structured questionnaire was used. To assess depression, the Beck Depression Inventory (BDI-II) was used, which contains 21 items rated on a Likert scale. Participants with any form of depression were included in the study. When using the Beck Depression Inventory scale, the scores for each of the 21 questions are added up. The highest possible score for the whole test is sixty-three, and the lowest possible score is zero. The determination of various categories of depression is based on the summed-up scores: 0–10 means that the mood fluctuations are considered normal, 11–16 indicates minimal depression, 17–20 indicates mild depression, 21–30 indicates moderate depression, 31–40 indicates severe depression, and more than 40 indicates extreme depression. The present study used the cutoff values as per BDI-II, where 0–13 was considered minimal depression, 14–19 mild depression, 20–28 moderate depression, and 29–63 as severe depression [
16]. All school children were interviewed by the Associate Professor of the Department of Community Medicine, who had prior experience in using this BDI scale. Venous blood was taken by an experienced lab technician and analyzed by Central Diagnostic Laboratory Services, Biochemistry Department, SDUMC, SDUAHER, Kolar. Baseline vitamin D3 levels were assessed via blood sampling using VITROS Immunodiagnostic products with a 25-OH Total reagent pack. The diagnostic cut-off levels of serum vitamin D (ng/mL) are classified as follows: deficiency < 20 ng/mL, insufficiency 21–29 ng/mL, sufficiency > 30 ng/mL, and toxicity > 150 ng/mL [
17]. The present study used the Indian Academy of Pediatrics classification for the diagnostic cut-off of vitamin D levels, in which less than 12 ng/mL is considered deficient, 12–20 ng/mL is considered insufficient, 20–100 ng/mL is considered sufficient, and more than 100 ng/mL is considered toxic [
18]. Block randomization was conducted by an expert from the Department of Community Medicine using the free online software randomization.com. Vitamin D3 supplementation was achieved by administering 60,000 IU of vitamin D3 once a month for two months, totaling 2000 IU per day, and by administering 500 mg of calcium with 250 IU of vitamin D3 per day for 9 weeks. A total of 2250 IU of vitamin D was taken by the vitamin D intervention arm. The other arm received 500 mg of calcium in addition to 250 IU of vitamin D3 per day for 9 weeks. Both vitamin D3 and calcium were administered at a safe dose, and the fear of toxicity was minimal. Both tablets were taken in front of the investigating team to ensure adherence. After 9 weeks of intervention, vitamin D3 levels were again assessed using the methods previously described. From the start of the trial, the participants were contacted every week by the pediatric team. A take-part book was given to all participants to examine drug adherence, a medical examination was carried out at the start of the study, follow-up was carried out weekly, and any side effects were documented. When depression was found in a student, even after the trial, they were referred to a psychiatrist. Flow chart of the study participants from recruitment to completion of the trial has been mentioned as in
Figure 1. The study commenced after obtaining approval from the Central Ethics Committee (SDUAHER/Res.Proj.173/2020-21). Informed written consent/assent was obtained from the school children after informing them about the benefits and risks of the study. Autonomy was maintained by the study participants, as participation in the trial was voluntary. Confidentiality was also maintained, as the participants’ names and personal details were not recorded. The study was registered in CTRI (REF No/2021/03/042355). All collected data were entered into Microsoft Excel and analyzed using SPSS v 22 (IBM Corp, Armonk, NY, USA). Sociodemographic data are expressed using descriptive statistics, such as frequencies and percentages. Inferential statistics, such as Chi square, were applied to determine associations between variables. Pre-intervention and post-intervention differences were assessed using
t-tests, summarized as the mean and standard deviation (SD), with a statistically significant difference defined as a
p value of less than 0.05. Both an intention-to-treat (ITT) analysis and a per-protocol analysis were conducted and are reported separately.
p < 0.05 was accepted as statistically significant.