The study was a randomised, single-centre, double-blind, placebo-controlled trial including healthcare professionals employed in psychiatric and somatic hospitals. The participants were suffering from moderate seasonality with SAD symptoms. The participants were randomized to either 70 μg vitamin D or placebo for a period of 12 weeks during the winter period. The dose of 70 μg vitamin D was chosen in accordance with the previous study by Jorde et al. in which the lowest dose of two active drug arms (70 μg or 140 μg) significant improved BDI, without significant differences in side effects as compared to the placebo group.
Healthcare professionals were recruited from somatic and psychiatric hospitals in the Region of Southern Denmark. The study was carried out by The Research Unit, Department of Mental Health Services, Esbjerg, during the period from October 1 2011 to March 31 2012.
The study population
Healthcare professionals working in one of the above-mentioned departments were invited to participate in the study via direct e-mail, flyers and posters. If the invited healthcare professionals wished to participate, information were sent to them along with the Danish version of the SAD questionnaire, Seasonal Pattern Assessment Questionnaire (SPAQ-SAD). SPAQ-SAD is validated in both psychiatric patients and healthy individuals, and question 2 (The Seasonality Score) in the Danish version is the most important and best studied item of SPAQ-SAD. In the Seasonality Score, the individuals report the effect the changing of the seasons has on them in six areas (sleep length, social activity, mood, weight, appetite and energy level) on a scale of 0 to 4, with 0 being "no change" and 4 being "extremely marked change". The test-retest correlation for Seasonality Score ranges from 0.65 to 0.87, and the Seasonality Score has a good internal consistency, Cronbach’s alpha = 0.82. We decided to invite participants for a baseline visit if they rated 8 or more in question 2 in the SPAQ-SAD questionnaire. We chose a cut-off at 8 in order to include enough participants and gain power; thus, we included participants with less severe symptoms equivalent to moderate seasonality with SAD symptoms. Thus we did not use the recommended definition by Kasper et al.. Diagnostic interviews were not planned due to the public health perspective focusing on symptoms of SAD rather than the full SAD diagnosis. Participants were eligible if they were 18–65 years old and had signed a written informed consent form. Exclusion criteria were any form of schizophrenia, bipolar affective disorder, sarcoidosis, tuberculosis and pregnancy, or an intake of more than 10 μg vitamin D per day or allergy to the content of the pills. Women of childbearing potential were required to use effective contraception, and a negative human chorionic gonadotropin (HCG) pregnancy test for each female participant was also required at baseline. Furthermore, participants were excluded if they had a baseline serum 25(OH)D < 10 nmol/L or > 160 nmol/L, serum calcium > 1.40 nmol/L, serum phosphate < 1.50 nmol/L (females) or < 1.60 nmol/L (males aged 18–49 years) and <1.35 nmol/l (males > 49 years), estimated glomerular filtration rate (eGFR) < 60 mL/min/1.73 m2 or serum parathyroid hormone (PTH) > 9.2 pmol/L.
Eligible participants received information on the study by the research nurse. The study drug was handed out along with instructions regarding the blood samples after informed consent was obtained. A consultant psychiatrist determined whether all inclusion criteria and no exclusion criteria had been met and enrolled the participants.
Randomisation and blinding
The tablets were produced by Vimenco, Denmark, and were identical in size, smell and taste. The participants were randomised using blocks of four into the intervention group or the control group. The allocation sequence was computer generated by the Hospital Pharmacy Funen, Denmark during the labelling procedure and was concealed for staff and researchers involved in the trial.
The primary outcome was the sum of the Danish version of the validated self-reported questionnaire Structured Interview Guide for the Hamilton Depression Rating Scale, Seasonal Affective Disorders (SIGH-SAD).
Secondary outcome and explorative outcome measures
Secondary outcomes were the sum of the Danish version of the validated self-reported World Health Organization-Five Well-Being Index (WHO-5). Weight, waist circumference, blood pressure, absenteeism from work and 25(OH)D were exploratory outcome measures.
The primary outcome was rated at each visit using a 24-item SIGH-SAD, a scripted version of the Hamilton Depression Rating Scale modified to better reflect the atypical symptomatology of SAD. This version of the SIGH-SAD consists of the HDRS 17-item scale plus the first 7 atypical items (i.e. excluding reverse diurnal variation). The Danish version was translated from English and back-translated under supervision by Claus Martiny PhD, Psychiatric Research Unit, Frederiksborg General Hospital.
Primary, secondary and explorative outcomes were assessed at baseline and at 12 weeks after baseline. Additionally, skin type and socio-demographic parameters were recorded at baseline. Known side effects of vitamin D supplementation, e.g. fatigue, muscle spasm, pain, nausea and constipation as well as severe and adverse events were systematically recorded yes or no at baseline and endpoint. The use of nutritional supplements and medication including vitamin D was recorded at baseline and at the 12-week follow-up.
Treatment adherence was assessed by the counting the amount of tablets that the participants returned at 12 weeks.
Blood samples were collected at baseline and 12 weeks later. The blood was analysed for C-reactive protein (CRP), ionised calcium, phosphate, creatinine, eGFR, PTH and 25(OH)D (to be reported only at the end of the trial). PTH analyses were performed on Immulite 2000 from Siemens using a 2-sided chemiluminescent enzyme-labelled immunometric method. 25(OH)D were analysed on ISYS equipment from IDS-Nordic, which measures the total amount of vitamin D2 and D3.
Differences between the intervention and control groups with regard to demographic and clinical characteristics at baseline were tested using Fisher’s exact test for categorical variables and Wilcoxon rank sum test for numerical variables. For each group, the mean SIGH-SAD level was calculated at baseline and after 12 weeks. The difference between the two groups with regard to increase of SIGH-SAD sums over time was compared using the Wilcoxon rank sum test. In addition, differences between the two groups in each of the secondary and exploratory outcomes were tested using Wilcoxon rank sum test. Correlations between 25(OH)D and SIGH-SAD and WHO-5 were measured using Spearman correlation coefficients. The analyses were conducted according to intention to treat principles.
All analyses were conducted using STATA/IC 11 (StataCorp LP, College Station, Texas, USA).
We expected a mean reduction of 3 points (standard deviation (SD) 6) on the SIGH-SAD according to the smallest clinical important difference in the primary outcome variable, consistent with a clinical effect of 0.5. Assuming an alpha of 0.05 and a power of 0.90, a sample size of 85 in each group was calculated. However, only 43 were randomised because of exclusion criteria and few responders. Only 34 completed the study after drop out, i.e. 20% of the number needed to give meaningful results according the power calculation. The achieved sample size reduced the opportunity to measure a difference of 6 on the scale used.
The study was approved by the Danish Data Protection Agency, the Ethics Committee for the Region of Southern Denmark (ID 301115) and the National Board of Health (Eudract: 2011-002585-20). The complete protocol can be acquired by contacting the corresponding author.