This was the first study to investigate Vitamin D levels in patients living in northern latitudes who were seeking care in an outpatient clinic for their LBP. These patients had Vitamin D levels comparable to those in the general population in Denmark. Therefore, our hypothesis was not supported. Firstly, low levels of Vitamin D were not related to the clinical symptoms. Secondly, Modic changes were less common in people with low levels of Vitamin D. The current study could not support the existence of a specific subgroup that would benefit from treatment aimed at normalising Vitamin D levels.
The results of this study show that Vitamin D deficiency is not common in LBP patients, who are similar to those included in this study. The trend of an increase in Vitamin D levels from March until May is consistent with people being exposed to more sunlight during the northern Europe spring [16, 18]. Vitamin D levels of LBP patients in the current study reflect those observed in the general Danish population. Mean Vitamin D levels of blood donors, who are supposed to be representative of at least the healthier segment of the general population in Denmark, have been reported to be 85 nmol/L (range: 26–163 nmol/L) in summer and 45 nmol/L (range: 13–128) in winter . Previous studies by Glerup et al. have reported mean Vitamin D levels in healthy Danish women of 47.1 nmol/L (+/− 4.6) in a control group  whereas Brot et al. reported 53.5 nmol/L in a group of 280 middle-aged women regularly exposed to sun and not taking vitamin supplements .
A sufficient level of Vitamin D in serum is continuously discussed in the literature and it is still subject to debate [9, 10, 18]. Previous studies that reported low levels of Vitamin D in populations with LBP used different reference criteria for normal values >100 nmol/L (>40 ng/ml)  and > 22.5 nmol/L , which makes comparison between studies difficult. Haroon et al. investigated 231 consecutive patients in general clinics using a reference criterion for a normal value of >53 nmol/L . Of the 231 patients, 8 had backache, 6 (75%) of whom had Vitamin D deficiency. Our results differ from the findings of Haroon et al., but comparison is difficult since only eight patients had backache.
Multiple factors affect the ability to synthesise Vitamin D, such as latitude, skin pigmentation, clothing, exposure to sunlight, obesity, malnutrition, and protein deficiency [13, 16, 24]. These factors complicate the direct comparison of Egyptian and Saudi Arabian people with people from countries in northern latitudes. Individual thresholds exist for the need for certain minerals and vitamins in humans. It is possible that thresholds differ across ethnic groups for the development of musculoskeletal symptoms due to a deficiency in certain minerals and vitamins, including Vitamin D.
The available methods for analysing 25-OH-D show significant differences in levels depending on which method is used [32, 33]. Lips et al. compared three different methods of measuring Vitamin D levels . A difference of 80% between Competitive Protein Binding (CPB) assay was seen when compared with High-Performance Liquid Chromatography (HPLC), while Radioimmunoassay (RIA) gave intermediate values . The LC-MS/MS method that we used in this study is reported to be the most reliable method . Despite this, variability in results between reporting laboratories can be up to 15% .
Our results did not support our hypothesis, and are in contrast to the findings in several previous studies on both chronic musculoskeletal pain and LBP, which suggests there might be several musculoskeletal symptoms related to low levels of Vitamin D [17, 19–23]. However, the lack of association between LBP and low levels of Vitamin D has been shown previously: Heidari et al. found significantly lower Vitamin D levels in patients with non-specific skeletal pain, but, similar to our findings, not in the group of back pain patients . In another study, Lotfi et al. found no associations between Vitamin D levels and back pain intensity or duration of LBP .
The lack of association we found between widespread pain and low Vitamin D levels is in contrast to the findings of McBeth et al. who reported that people with chronic widespread musculoskeletal pain have a 50% increased risk of low levels of Vitamin D .
We found no association between Vitamin D levels and myopathy-related symptoms. This could be due to either a small group of patients in the moderate/severe-group (n = 17 (11%) and only one with severe Vitamin D deficiency), or the fact that the symptoms were self-reported and not objectively measured. Glerup et al. found that most myopathy-related symptoms are present at Vitamin D levels below 25 nmol/L and that the first symptom most often reported is fatigue and not pain or paresthesia , which could account for the lack of association in our study.
Our finding of a negative correlation between Vitamin D deficiency and high BMI has previously been reported and can be explained by a storage of Vitamin D in adipose tissue [9, 16, 34]. Further investigations aimed at examining whether there is a relation between LBP, high BMI and levels of Vitamin D could be interesting in the search for relevant subgroups within LBP patients.
Unexpectedly, Modic changes were more common in people with normal levels of Vitamin D compared with those with low levels. According to our theoretical framework, people with Modic changes were expected to have low levels of Vitamin D due to an increased susceptibility to inflammation  and/or to micro-fractures in the vertebrae due to increased levels of parathyroid hormone . Only serum 25-OH-D were measured in this study. Therefore, it is not possible to draw conclusions about increased bone turnover in the investigated group of LBP patients, although levels of < 50 nmol/L have been shown to cause an increased level of parathyroid hormone . Our initial theoretical framework failed to explain the possible relationship between Vitamin D and Modic changes. Perhaps, the increased risk of micro-fractures related to cortical bone due to hyperparathyroidism  does not apply to the vertebra end plates, which mainly consist of hyalin cartilage, and therefore the significant finding was perhaps simply a random and sample-specific artifact.
The mechanisms in the development of Modic changes are still not sufficiently known. Current significant findings suggest that there could be a link between Vitamin D and Modic changes perhaps related to inflammation. Further studies are needed on the subject.
The strengths of this study are that participants were recruited from consecutive patients, standardised institutional methods were used for collecting self-reported data, and blood samples were taken on the same day as filling in the self-report questionnaires. However, while the included patients might be representative of non-specific LBP patients at The Spine Center of Southern Denmark, it is not known how comparable they are with patients in other settings.
Our study does not support an identifiable sub-group of people with LBP and Vitamin D deficiency who could benefit from treatment with Vitamin D. This is in line with a Cochrane review that concludes that there is not sufficient evidence to support treatment of chronic pain with Vitamin D . The study had a cross-sectional design. A randomised control trial is needed to further investigate whether people with non-specific LBP would benefit from Vitamin D treatment. Furthermore, the design does not allow us to make conclusions regarding adequate thresholds of 25 (OH)D in non-specific LPB patients. This is still a relevant research question.
LBP is a common disorder throughout the world with two-thirds of all adults experiencing back pain at some time in their life . The group of non-specific LBP patients is a heterogeneous group with multiple factors contributing to the symptoms experienced by the individual patient. Many different biological and psychosocial factors can contribute to LBP symptoms [36, 37], and therefore, it is difficult to isolate a single reason for non-specific or chronic LBP in a given individual. Perhaps the underlying mechanisms of non-specific LBP differ in different regions of the world and perhaps these mechanisms differ across ethnic groups.