Patients
In this case-control study, 60 eligible patients with chronic LBP persistent in more than three months enrolled in the case group from the reference outpatient clinics of Tabriz University of Medical Sciences for further evaluation during a period of 6 months from March to August 2018. Initially, for each case with LBP, one control was selected from the same outpatient clinics without LBP who were matched based on age and body mass index (BMI). Inclusion criteria were as follows: age older than 20; BMI greater than 18.5. On the other hand, the patients were excluded based on these criteria: LBP duration less than 3 months; lumbosacral deformities; intervertebral disc herniation; osteoporosis; rheumatoid arthritis; uncompensated renal, and liver diseases. The eligible individuals were categorized in 3 subgroups based on age (18-40, 40-60, more than 60 years old) and weight (normal, overweight, obese).
A simple random sampling method was applied. Based on observations in a similar study (10), 6.2 and 6.9 were considered as the standard deviation of the case and control groups respectively and 3.6 for the effect size. With a significant level of 0.05, the power of 0.8 and using a two-way test, the sample size was estimated as 53 individuals in each group. Considering 15% chance of loss, the sample size was estimated to be 60 in each group and 120 in total.
Physical activity
Physical activity of the participants was assessed by the short-form IPAQ (International Physical Activity Questionnaire) (15). Three categories of physical activity were suggested: low, moderate, and high (16).
Vitamin D measurement
Fasting peripheral venous blood (3 ml) was collected from participants in both groups. The level of 25-hydroxyvitamin D3 was measured by ELISA. All the steps were performed exactingly according to the instructions of the kits (Euroimmun, Germany).
Anthropometric measurements
The weight was measured by a Seca 813 digital scale. BMI was calculated via dividing the weight (kg) by the square of height (m2).
LBP assessment
Modified Oswestry Questionnaire was used for functional evaluation of LBP in individuals of the case group. This questionnaire included 10 questions regarding pain severity, LBP during self-care activities, lifting objects, walking, sitting, standing, sleeping, social life, and leisure time. Each question had 6 items, based on Likert scoring system, where 0 reflected "No Problem" while 5 represented "Severe Problem" (17, 18). LBP severity was also recorded by Visual Analog Scale (VAS) in which zero referred to "No Pain" and 10 denoted "Very Severe Pain" (19).
Statistical measurement
Measurement data were introduced as a mean ± standard deviation. The two study groups were compared using Student's t-test or the Chi-square test. Logistic regression analyses were used to determine whether 25 hydroxyvitamin D level was significant factors associated with the presence of LBP. Finally, Partial correlation analyses were employed for analyzing correlation. Age, sex and BMI served as control variables. Values of P < .05 were considered as statistically significant. Statistical analysis was conducted using SPSS 17.0 software (SPSS Inc., Chicago, IL).