1.1 Participants information
A total of 733 participants were continuously registered as inpatients at the Second Affiliated Hospital of Fujian Medical University hospital during the summer months of 2019 and 2022 (July, August, and September) and were randomized to this study. All 531 of these enrolled met the diagnostic criteria for type 2 diabetes. Meanwhile, 202 volunteers were randomly enrolled as healthy controls. The exclusion criteria were maintenance or high-dose vitamin D supplementation, using sunbeds, non-diabetic neuropathies, diabetic ketosis, age < 25 years or > 75 years, long-term alcohol abuse, pregnancy, kidney or liver insufficiency, perioperative, severe infection, neoplasm, and blood disease. Participants in the control group were required to meet HbA1c 5.6%, with fasting glucose < 5.6 mmol/L and no history of diabetes. All participants were informed and signed consent forms. The study has been approved by the Ethics Committee of the Second Affiliated Hospital of Fujian Medical University.
1.2 Indicator measurement
Each participant was tested by the same physician with the same experience following standard procedures. All inspections and tests were carried out in a comfortable and quiet laboratory. All participants in the study completed a questionnaire on the frequency of outdoor activities. All participants underwent fundus examination under the same ophthalmologist, and the ophthalmologist issued a diagnosis.
Standing height and weight were measured under fasting, without shoes, on the same one-piece scale. The BMI value was obtained by calculating the participant's weight (kg) divided by the square of height (m2). The blood pressure of the participant's right arm was measured with a mercury sphygmomanometer after 30 minutes of sitting and resting. Participant blood samples were collected during the summer months, July, August, and September, which can avoid seasonal differences in light. Venous blood was collected from each participant in the morning after a 10–12 hour fast. An automated assay (Maglumi 4000; China) was used to measure vitamin D according to the instruction of the manufacturer. We used high-performance liquid chromatography to detect HbA1c (D10; Bio–Rad, Berkeley, CA). We used automatic biochemical analyzers (Cobas 8000; Roche, Germany) uniformly to measure fasting plasma glucose, liver function [aspartate transaminase (AST), alanine transaminase (ALT)], kidney function (serum creatinine), and blood lipids [low-density lipoprotein (LDL) cholesterol, high-density lipoprotein (HDL) cholesterol and total cholesterol (TC)]. Urinary albumin concentration and urinary creatinine were evaluated using immunohypertensive assays. This study was calculated from the ratio of urinary albumin (mg) to urinary creatinine (g) urinary albumin-creatinine ratio (UACR). All glomerular filtration rates were obtained by the formula "Ccr= {[140-age (years) × weight (kg)]/[0.818 × serum creatinine (Scr, µmol/L)]} for men and Ccr × 0.85 for women".
1.3 Followed up for one years
All diabetes without retinopathy enrolled were followed up by readmission after one year. The same ophthalmologist reassessed all follow-up participants for fundus examination. The follow-up participants were divided into two subgroups based on the presence or absence of new-onset DR ( the new-DR group and the non-DR group).
1.4 Statistical analysis
Statistical analysis was performed on most of the data using SPSS version 19.0. Baseline characteristics were expressed as the mean (SD). One-way analysis of variance (ANOVA) was used for two-way comparisons between the three groups. The count data of baseline characteristics were compared using the chi-square test. Data were compared between the two subgroups using a t-test. Vitamin D was added to the logistic regression model to control for possible confounders. A ROC curve was used to make a cut-off value for vitamin D in people with diabetes. Pearson analysis was used to compare the correlation between vitamin D and HbA1c. P < 0.05 was considered statistical significance.