Severe vitamin D deficiency is associated with frequent exacerbations and hospitalization in COPD patients

The prevalence of COPD is increasing and this has a heavy influence on healthcare costs [1], particularly because of the frequent disease exacerbations (AECOPD) and hospital admissions. The disease control is far from being reached and dietary factors are acknowledged as one of the several factors contributing to COPD [2].

Vitamin D deficiency was highly prevalent in a general US population in the third National Health and Examination Survey [3]. According to Janssens et al. [4], vitamin D deficiency occurs in over 60% of patients with severe COPD, and is quantitatively related to disease severity. Epidemiological studies indicate that decreased vitamin D is associated with increased frequency of respiratory infections not only in COPD patients, but also in healthy people [4],[5]. This may be due to the involvement of vitamin D in both innate and adaptive immunity regulation [6],[7]. A general population study by Skaaby et al. [8], demonstrated a significant inverse association between vitamin D status and death caused by diseases of the respiratory and digestive system and by endocrine, nutritional and metabolic diseases. Nevertheless, the role of vitamin D in AECOPD is still debated. In a secondary analysis of a study performed in exacerbation-prone COPD patients [9], no association between baseline vitamin D levels and subsequent risk of acute exacerbations was found; negative results have also been reported in a primary care setting [10]. A single center randomized trial on 182 COPD patients [11] demonstrated that vitamin D supplements were able to reduce COPD exacerbations only in the 30 subjects with severe deficiency. These findings leave open the question of the role of vitamin D deficiency and the benefit of its correction in COPD.

We performed a retrospective observational cohort study in COPD patients not taking vitamin D supplements, to evaluate if low vitamin D levels is associated with severe airway obstruction, annual FEV1 decline, disease exacerbations and hospital admission over the course of one year.

Patients were selected among 229 consecutive COPD patients, with any range of severity according to the GOLD classification [2], who presented for a scheduled visit at our Respiratory Clinic during the period October 2011-March 2012. Inclusion criteria were age over 40 years, a post-bronchodilator ratio of forced expiratory volume in 1 second (FEV₁) to vital capacity (VC) <0.7, and at least one year follow-up in our clinic. Exclusion criteria were AECOPD in the last month (n = 19), current treatment with vitamins and dietary supplements (n = 46) and lack of availability for data on lung function/exacerbations/hospitalization the year previous to the inclusion in the study (n = 67), resulting therefore in 97 patients to be included in the study.

The study was approved by the Institutional Review Board (CEI N. 414) and written informed consent was obtained from each patient.

At the enrolment visit, patients underwent clinical examination, recording of symptoms, smoking habits, medication use, lung function tests, venous blood sampling for nutritional assessments. Subjects were classified as current, former- and never-smokers, according to self-reported smoking history. Body mass index (BMI) was calculated as weight divided by height squared (kg/m²).

Comorbidities were recorded on the basis of prior diagnosis and current treatment for: systemic arterial hypertension, diabetes, dyslipidemia, anxiety and/or depression, chronic kidney disease, cerebrovascular disease, osteoporosis, obstructive sleep apnoea (OSA), any type of active malignant tumor. The diagnosis of pulmonary hypertension, heart disease and chronic heart failure had to be supported by symptoms and clinical and echocardiographic examination [12].

The medical records of the patients were collected and reviewed retrospectively, to gain information relative to the previous year:

Lung function tests were measured using the Baires System (Biomedin, Padua, Italy). The values of VC, FEV1, FEV1/VC ratio, and maximal midexpiratory flow-rate (MEF₅₀) were computed. VC, FEV1 and MEF₅₀ were expressed as% of the predicted value [16]. Annual FEV₁ decline was calculated as percent of the pre-bronchodilator starting value.

Nutritional assessment consisted of: serum levels of ferritin, folic acid and vitamin B12, measured using the chemiluminescent micro-particle immunoassay (Architect System, Abbott diagnostic division, Longford, Ireland), vitamin D (25-hydroxyvitamin D) levels, measured by the RIA method (25OH Vitamin D total-Ria-CT Kit, DIAsource ImmunoAssay S.A., Louvain, Belgium). Vitamin D levels were regarded as normal (≥30 ng/ml), mild-moderate deficiency ( ≥10, but <30 ng/ml), severe deficiency (<10 ng/ml) [17].

The results of the study demonstrate that most COPD patients had vitamin D deficiency. Severe deficiency was associated with frequent exacerbations, documented in our clinic, and hospitalization during the previous year. These findings were consistent even after adjustments for comorbidities and other nutrient deficiencies. Vitamin D levels were not related to either lung function at the study enrolment or lung function decline in the preceding year.

Vitamin D deficiency represented the major nutritional disorder in our study population, being mild-moderate in about 60% and severe in as high as 36% of the patients. A high prevalence of vitamin D deficiency in COPD patients has been previously reported [4],[9],[11],[19], but never as high as in our patients. This may depend on several reasons. First, vitamin D was deliberately assessed in the winter season, when levels are supposed to be lowest [20], especially to the latitude of our city [21],[22]. Anyway, it is presumable that our subjects were deficient throughout the year, as in Italy, at variance with other countries [23], there is no food fortification with vitamin D. It has been estimated that a summer vitamin D level around 40 ng/mL is needed to achieve a 20 ng/mL concentration the following winter [24] and similar variations appear to exist in COPD patients, with approximately 35% lower values in winter compared to summer [25]. The lowest winter vitamin D values may contribute to the marked impact of winter season on both frequency and outcomes of COPD exacerbations [26],[27].

These findings are in agreement with prior observations that low vitamin D levels are associated with increased frequency of respiratory infections in both COPD patients and healthy adults [4],[5],[28]. However, the role of vitamin D deficiency in AECOPD is still debated. A recent study on exacerbation-prone COPD patients found no association between baseline vitamin D levels and subsequent risk of AECOPD [9]. However, according to Heulens et al. [29], this negative finding might depend on the fact that some of the patients with worse clinical conditions were taking vitamin D supplements. Of course, once supplemented, vitamin D levels no longer reflect the underlying COPD severity. Actually, excluding those taking supplements from the analysis, Heulens et al. [29] could demonstrate that patients with vitamin D levels below 10 ng/ml had the shortest time to first exacerbation and experienced the highest number of AECOPD. Moreover, pooling together patients with vitamin D levels ranging from 10 to 30 ng/ml, they found a dose–response relationship for exacerbations number very similar to that found in our study. Another recent study by Puhan et al. [10] reported no relation of severe vitamin D deficiency with exacerbations and no effect of vitamin D supplements on AECOPD. Such negative result may depend on several reasons. First, vitamin D deficiency was less severe than that observed in our study patients (15.5 versus 12 ng/dl). Second, although patients taking vitamin D supplements had been excluded at enrolment, it was not specified if patients were supplemented during the biannual follow-up. Furthermore, different vitamin D assays were used in these studies and this could, at least in theory, yield different results and therefore vitamin D levels in absolute values might be more difficult to compare between studies. A recent study on 12,041 individuals from the Danish general population by Skaaby et al. [30], found that vitamin D status was significantly inversely associated with COPD but had no influence on COPD incidence. This finding suggests that vitamin D deficiency is a consequence rather than a cause of COPD. The effect of vitamin D supplementation in COPD is still debated. High dose vitamin D supplementation has been found to to decrease AECOPD number, but only in patients with severe deficiency [11] and to improve inspiratory muscle strength and maximal oxygen uptake [31].

At variance with other findings [19],[32], in our patients vitamin D levels were not related to FEV₁ and to the GOLD-COPD class, which is based on FEV₁ value, neither were they predictive of FEV₁ decline. This could simply be due to the fact that the size of such effect is so small that a larger population would have been needed to demonstrate it. Actually, both the studies of Persson et al. [19] and Berg et al. [32] had populations about five times larger than ours, and patients were well distributed among the GOLD classes, while most of our patients were in GOLD class II. Interestingly, in univariate analysis the levels of vitamin D were significantly related to both FEV₁/VC and MEF₅₀, but these correlations disappeared after adjustments for potential confounders.

The major weakness of the present study was its retrospective design, as the vitamin D levels were measured at the end of the observation period, where events (exacerbations) and lung function decline were recorded. Therefore a potential reverse causation effect, with exacerbations lowering levels of vitamin D, cannot be excluded even if no such relation has been described in the literature [25]. Furthermore, a study in independently living community-dwelling subjects of similar age, with different degrees of vitamin D deficiency, reported steady levels over a one year period [33] and therefore it might be speculated that the levels of vitamin D measured in the present study could be similar to the previous year values. However, in our opinion, a retrospective study, as compared to a prospective one, avoids the criticism of not treating patients with severe deficiency.

The results of the present study demonstrate that vitamin D deficiency is very common in COPD patients, and indicate that severe deficiency is associated with higher probability having frequent exacerbations and hospitalization and this association is independent of patient characteristics and comorbidities. The prevention of exacerbations is a major treatment goal of COPD and the benefit of vitamin D supplementation, particularly during the winter season, is an intervention that warrants further assessment. However randomized controlled trials are needed to clarify whether the observed association in the present study is causal and is really attributable to vitamin D deficiency.