Vitamin D levels and deficiency with different occupations: a systematic review

Although there has been recent improvement in vitamin D status in the North American population, the prevalence of insufficiency remains high. About 70%–97% of Canadians are estimated to demonstrate vitamin D insufficiency [1] and approximately 40% of the US population are estimated to be vitamin D deficient [2]. While vitamin D has traditionally been shown to be involved in calcium homeostasis [1] and bone health [2], recent evidence suggests several roles not limited to the musculoskeletal system. An inadequate level of vitamin D has been linked to a number of diseases including metabolic disorders, autoimmune conditions, psychiatric, respiratory and cardiovascular disorders, and cancers as well as osteoporosis and osteomalacia [2‐4]. The widespread systemic effects of vitamin D have been attributed to the ubiquitous expression of vitamin D receptors in various organ systems [2, 5].

Vitamin D is synthesized in vivo when solar ultraviolet B (UVB) radiation interacts with the precursor molecule, 7-dehydrocholesterol, in the skin [2, 6, 7]. Another important source of vitamin D is dietary intake and supplementation, although endogenous production is estimated to account for 90% of total vitamin D in healthy individuals, and any activity that reduces sunlight exposure will tend to reduce vitamin D levels [2, 8]. Whether from endogenous production or dietary sources, vitamin D is subsequently transported in the blood (bound to vitamin D-binding protein) to the liver where it is hydroxylated to 25-hydroxyvitamin D (25-(OH)D) [7]. 25-(OH)D is further converted to the metabolically active form, 1α, 25-dihydroxyvitamin D (1α, 25-(OH)₂D), primarily in the kidneys [9]. In the present study, the term ‘vitamin D’ was used in the context of status, i.e., deficiency, insufficiency or sufficiency, while 25-(OH)D referred to serum levels of the metabolite.

Because of its half-life in blood of ~2–5 weeks, the circulating concentration of 25-(OH)D has been used as a measure of vitamin D status in individuals [6]. Serum 25-(OH)D concentrations lower than 50 nmol/L (20 ng/mL) appear to be detrimental to bone health [10]; however, optimal serum 25-(OH)D concentrations have not been established with respect to other outcomes, and there is lack of agreement on how deficiency should be defined, or how best to conduct population-based screening for vitamin D deficiency [11]. Despite a lack of consensus on optimal levels, it is becoming abundantly evident that vitamin D deficiency and its associated untoward health outcomes are a worldwide phenomenon [12, 13].

As vitamin D synthesis is highly dependent on sunlight, factors and conditions associated with decreased time spent outdoors can be expected to adversely impact vitamin D status. Shiftwork represents work that occurs outside the traditional 9 a.m. to 5 p.m. workday and may include evening or night shift work, with or without rotating shifts. An estimated 28% of working Canadians [14], 17% of Americans [15], and 22% of European workers [16] work outside the traditional 9 a.m. to 5 p.m., Monday through Friday schedule. Shiftwork has been epidemiologically associated with a number of health conditions, including sleep disturbances, cardiovascular disorders, gastrointestinal and digestive problems, and increased cancer risk, among others [17‐19]. Little is known about the role of vitamin D deficiency with respect to the adverse health outcomes related to shiftwork, although vitamin D deficiency has been previously postulated as a mechanism of shiftwork-related cancers [20]; shiftworkers can plausibly be expected to have lower serum vitamin D levels due to reduced exposure to sunlight or altered dietary intake of vitamin D-rich foods. Additionally, other occupational groups (e.g. indoor workers) may be at risk of vitamin D deficiency through this same mechanism of reduced sunlight exposure, and it has been previously noted that there has been comparatively little research into the health of indoor nonindustrial workers [21].

Understanding the levels of vitamin D and prevalence of vitamin D deficiency in various occupational categories can inform public health attempts to reduce vitamin D deficiency and ensure improved population health outcomes. The link between vitamin D levels and occupation has previously been explored in the published literature, and low levels of vitamin D have been demonstrated in some occupational groups with expected low exposure to sunlight. For example, a recent study in bakers concluded that vitamin D insufficiency was very common, especially in night workers [22]. A large study from Korea found that the risk of vitamin D deficiency was significantly increased for shift work and office work [23]. An Indonesian study likewise showed that vitamin D deficiency may occur in women with indoor occupations [24]. In contrast, other evidence confirms that outdoor workers have comparatively high serum vitamin D levels [25‐27]. However, the association between occupational factors (e.g. shiftwork, indoor work, work activities) and vitamin D levels is far from clear in the literature, and vitamin D deficiency in working populations does not seem to be entirely explained by sunlight exposure. A Japanese study on shiftworkers comparing serum 25-(OH)D levels in fixed daytime workers to rotating workers with or without night shift demonstrated no significant differences [28]. Additionally, vitamin D deficiency can be prevalent among certain groups of workers, for example health care workers, even in regions with high sunshine exposure [29].

To our knowledge, there has been no previous systematic attempt to examine the effect of occupation on vitamin D status, or determine the prevalence of vitamin D deficiency in different occupational groups, despite the importance of work in most workers’ lives. Additionally, we are unaware of any current guidelines on screening for vitamin D deficiency or vitamin D supplementation, which include specific guidance for workers, or risk stratification elements based on occupational factors. The present article aims to provide evidence to address these gaps.

The objective of the present study is to investigate serum vitamin D levels, and prevalence of insufficiency and deficiency in different occupational categories to identify groups of workers at particular risk of vitamin D deficiency or insufficiency.

The electronic database search generated 1991 records after deduplication, of which 87 primary studies (all in English) were considered potentially relevant, based on title and abstract screening, and available as full text journal articles. An additional 17 articles were identified by searching reference lists of previously identified articles or reviews. Of this total of 104 articles, 33 were ultimately excluded after review of the full-text article, leaving 71 articles to be included for further quantitative analysis as shown in ‘Additional file 2: Study selection’. The total number of subjects for all included studies was at least 53,345 (one study did not report on the number of subjects) and the sample sizes ranged from 4 to 10,646 subjects per study as shown in Table 1. The overall proportion of participants who were women was 65%. Publication dates ranged from 1971 to 2016. Based on the latitudes of study locations, the included studies spanned a range of latitudes from 3° N (Indonesia) to 64° N (Reykjavik, Iceland) in the Northern hemisphere, and 23° S (Sao Paulo, Brazil) to 30° S (Porto Alegre, Brazil) in the Southern hemisphere (Table 1 and Additional file 3: Figure S4A).

Thirty-five of 71 articles described studies performed in health care workers (physicians, nurses, hospital employees, health sciences or medical students, and other health professionals). Three studies were performed in coal miners, 5 in lead/smelter workers and 6 in shiftworkers. Eleven of the 71 papers described groups of outdoor workers; 19 studies were performed in indoor or office-based workers. Some of the primary studies included subjects of more than one occupational setting and such studies were categorized under more than one occupational group depending on the occupations described by the authors, as shown in Table 1. Each occupational category that we examined had 3 or more primary studies, therefore permitting further quantitative analysis. Studies in which we could not establish a well-defined occupational setting were excluded (Table 1).

We extracted data from the included studies based on season of the year in which the study was conducted, assay type (measure of validity) and intra- and inter-assay CV (indication of reliability) in order to assess study quality; ‘unknown’ indicates that a study did not describe the parameter in question (Table 2). As shown in Table 2, 43% of included studies were of high quality, 37% were of medium quality, and 20% were of low quality. Regarding assay types, 40% of reports employed a radioimmunoassay (RIA) technique to assess serum vitamin D levels, 14% used a competitive protein binding assay, a chemiluminescence assay technique was employed in 13% of studies, 11% assayed vitamin D levels via the high performance liquid chromatography (HPLC) assay, while 4%, 2% and 1%, respectively, employed the enzyme-linked immunosorbent assay (ELISA), liquid chromatography and radioceptor techniques.

The global prevalence of vitamin D deficiency has reached an alarming proportion. This trend has elicited a significant amount of research interest to elucidate the potential causes of vitamin D deficiency and insufficiency in order to advance interventional strategies to ameliorate the associated risks [2, 46]. Several studies have demonstrated that populations worldwide, including those located in sunny regions of the world, are at risk of vitamin D deficiency [47, 48]. Some vulnerable demographic groups of the population, including pregnant women [49, 50], the elderly [51], hospitalized patients and other institutionalized groups [12, 52, 53], and certain occupations [45, 54‐56] who are predisposed to receive low or no exposure to sunshine, may be at a greater risk of developing vitamin D inadequacy.

In general, we found that vitamin D levels in most occupational groups fell well below the levels considered optimal for health [13, 57]. Additionally, we observed a higher prevalence of vitamin D deficiency in all occupational groups examined than the reported population burden of vitamin D deficiency in multiple populations, suggesting that workers may be particularly vulnerable to vitamin D deficiency [23, 41]. For instance, 25-(OH)D among US subpopulations demonstrated that adult women had mean serum 25-(OH)D levels of 62 nmol/L and 75 nmol/L, respectively, in winter and summer [58]. Adult men had mean serum 25-(OH)D of 70 nmol/L and 82 nmol/L in winter and summer, respectively [58]. In contrast, Platz et al. [59] reported that among healthcare professionals in the US, the mean serum 25-(OH)D was 46 nmol/L. The prevalence of vitamin D deficiency was 30% [60] in the adult population in the US compared to 38% among medical residents [61, 62]. Other determinants of vitamin D deficiency include culture [63], geography [64, 65], genetics [66], disease states [67], diet [6] and age [68], and such other determinants will have to be considered alongside occupation as part of a comprehensive assessment of vitamin D status.

Among the occupations considered in the present study, indoor workers and shiftworkers demonstrated very low levels of serum 25-(OH)D and high rates of vitamin D deficiency and insufficiency. The relative risks of vitamin D deficiency and insufficiency in these two worker groups were also very high compared to the other worker groups. Indoor workers spend a high proportion of their working hours indoors without sunshine exposure. Additionally, indoor workers working conventional working hours would be expected to get their sunlight exposure during mornings and evenings, when sunlight intensity is relatively low. Since vitamin D is produced through sunshine and adequate UV exposure, sunlight deprivation in indoor workers may put them at greater risk of developing vitamin D deficiency and its accompanying health risks. Consistent with this assumption, our study showed that workers in an indoor setting displayed a lower level of vitamin D relative to their outdoor counterparts. Further, medical residents and healthcare students, who spend a considerable amount of time indoors, had vitamin D levels almost identical to the level of vitamin D observed in specified indoor workers (Fig. 1).

Our study also demonstrated that shiftworkers are at the highest risk of vitamin D deficiency or insufficiency when compared with other occupational groups. Shiftworkers make up about 20% of the workforce in developed countries [69]. Shift work may include rotational daytime shifts or overnight shifts. Kimlin and Tenkate [70] reported that workers with permanent night shifts receive less exposure to daytime solar UVB. This may result in a lower vitamin D level in shiftworkers relative to the general population. Additionally, shiftworkers with low sunlight exposure may depend on exogenous vitamin D for proportionately more of their total vitamin D requirements. These hypotheses are consistent with the findings of the present review, which revealed that shiftworkers had a low vitamin D level with a comparatively large proportion of workers with deficiency or insufficiency. The low level of serum 25-(OH)D seen in shiftworkers has been suggested to be associated with their predisposition to various diseases including cancers [71], musculoskeletal disorders, and cardiovascular disease [2, 72, 73].

The interpretation of our analysis is complicated by the fact that shiftworkers may work at various times of the day in a fixed or rotating pattern. Additionally, shiftworkers may have a greater proportion of their nonworking hours during daylight periods, and spend more time outdoors during nonworking days than indoor office workers. Another factor is that shiftworkers may be outdoor or indoor workers. Furthermore, shiftworkers may differ in dietary intake, use of vitamin D supplements, or other lifestyle factors from non-shiftworkers, and considerable variability may exist with regard to the amount of time spent outdoors. Nevertheless, a compelling finding from the present review is that shiftwork appears to be a strong predictor for vitamin D deficiency.

Another determining factor shown to impact vitamin D status in shiftworkers is the season during which vitamin D levels are measured. Ito et al. [28] demonstrated that, during the summer, the amount of ambient solar UVB can compensate for time confined working indoors in shiftworkers. However, in the winter, vitamin D levels were lower in shiftworkers who work fixed night shifts compared to the shiftworkers who work strictly daytime shifts. This finding was confirmed by Romano et al. [45], who showed that nighttime shiftworkers had lower vitamin D levels compared with daytime workers during spring. Taken together, these observations suggest that, although shiftworkers may be at greatest risk of vitamin D deficiency, spending sufficient time outside and obtaining sufficient UV exposure has the potential of alleviating this risk.

The lifestyle and nature of work of many healthcare professionals may suggest less opportunity to be exposed to daytime solar UVB. Thus, it is reasonable to assume that healthcare employees would be at risk of vitamin D deficiency. Additionally, working hours may be particularly long during the early training period for most healthcare professionals, where sun deprivation due to long working hours can be exacerbated by additional time spent indoors studying. Consistently, average vitamin D levels in healthcare students and medical residents were shown in the present review to be in the deficient range, and average serum vitamin D levels were significantly lower than those of practicing physicians, nurses, or other healthcare professionals. Indeed, the average serum vitamin D levels in medical residents and healthcare students were close to those of specified indoor workers. For healthcare professionals, our study revealed a high prevalence of vitamin D deficiency among healthcare students and medical residents. This is alarming, as students and residents are generally young adults, and vitamin D deficiency during early adulthood may decrease peak bone density and lead to an increased risk of osteopenia or osteoporosis in later life, as well as other long-term health impacts associated with suboptimal vitamin D status [74]. Additionally, vitamin D sufficiency in young healthcare professionals may be a surrogate marker for other healthy behaviours (e.g. outdoor exercise and good nutrition), and a high burden of vitamin D deficiency in trainees in the health disciplines should prompt enhanced educational measures on the importance of adequate vitamin D, as well as an examination of the underlying training-related factors which may contribute to vitamin D deficiency.

Practising physicians, nurses and other healthcare workers had average serum vitamin D levels significantly higher than students and residents, although the average vitamin D levels in all three groups were still in the insufficient range. The prevalence of vitamin D deficiency or insufficiency was also very high among all healthcare professionals with the exception of the other healthcare employee group. The latter group comprised employees with diverse work environments and lifestyles, which suggests that some may have more exposure to solar UVB not considered typical of many healthcare professionals. This could account for the relatively low prevalence of vitamin D deficiency or insufficiency in the other healthcare employees group.

The large proportion of studies on healthcare workers may in part reflect the use of healthcare workers as a convenience sample of young, presumed healthy individuals. Additionally, several large population-based studies in health professionals (e.g. Nurses’ Health Study, Physicians’ Health Study) have been performed, and publications arising out of these cohorts are included in our analysis. However, it is concerning that among healthcare workers, such a high prevalence of vitamin D deficiency exists. This may reflect a number of occupational factors including long working hours mainly indoors [61, 62], shiftwork and a tendency for healthcare workers to neglect their own health [29, 75, 76]. The extent to which healthcare workers, in general, are aware of their own vitamin D status is unknown, and it is conceivable that targeted interventions aimed at identifying and treating vitamin D deficiency in health care workers may be beneficial. Of note, some healthcare professions (e.g. nursing) have a high proportion of female workers of childbearing age, and there may be ancillary health benefits beyond those to healthcare workers themselves.

Surprisingly, our data also demonstrate that among healthcare professionals, only healthcare students had an increased risk to develop vitamin D deficiency. However, all healthcare professionals had an elevated risk of vitamin D insufficiency, in keeping with the 25-(OH)D levels in these worker categories.

Lead exposure, either short- or long-term, has the potential to influence the metabolism of vitamin D [77, 78]. The impact of lead on serum 1α, 25-(OH)₂D levels is attributed to the inhibitory effect of lead on cytochrome P450 in the proximal tubules of the kidney, which mediates the hydroxylation of 25-(OH)D to the dihydroxy metabolite. In adults occupationally exposed to lead [78], Mason et al. demonstrated increased serum 1α, 25-(OH)₂D levels in lead-exposed workers compared with a referent group who were not occupationally exposed to lead. In contrast, Greenberg et al. [79] did not demonstrate any effect of lead exposure on serum 1α, 25-(OH)₂D levels.

Of all the occupational groups considered in our study, lead/smelter workers had the highest level of circulating 25-(OH)D. The mean 1α, 25-(OH)₂D levels found in lead/smelter workers in our study was 139.73 ± 57.51 pmol/L, consistent with the average serum 1α, 25-(OH)₂D levels found in the general population [80]. This may imply that the lead/smelter workers did not demonstrate compromised renal hydroxylase activity. The number of studies on lead/smelter workers and the number of study participants were, however, very few relative to the other occupational groups examined except coalminers. Moreover, the studies were older and the nutrition of the subjects at the time may have been different compared to subjects in more recent studies. Furthermore, the studies on lead/smelter workers presented in this report also are of low quality; they did not demonstrate any relationship between serum 25-(OH)D levels and blood lead levels. Thus, modern and high quality studies that account for all confounders of the relationship between lead exposure and serum 25-(OHD) and 1α, 25-(OH)₂D levels are warranted to establish if there is an association between lead exposure and vitamin D status.

Likewise, the literature is lacking in high quality studies that describe the relationship between coalminers and serum vitamin D levels. Our review found only three reports [81‐83] comprising 314 subjects. Coalminers can be underground miners or surface miners. Underground miners experience reduced exposure to sunlight and, following on from the previous discussion, may be at increased risk of low vitamin D status compared to surface miners. Surprisingly, our findings revealed that vitamin D status in underground miners was not significantly different from surface miners. The lack of appreciable difference in these two group of miners may mean that exposure of the underground group to sunlight in-between shifts could be sufficient to maintain serum vitamin D levels. In keeping with this view, Shuster et al. [81, 82] showed that, in the summer and winter seasons, serum vitamin D levels were not significantly different between underground and surface workers. However, in the summer months, serum vitamin D levels were higher than the corresponding levels seen during the winter months. These findings were in agreement with those demonstrated by Sarikaya et al. [83] in underground and surface miners.

The angle at which the sun rays impact the skin, which is a function of latitude, determines the amount of vitamin D production [64]. The more oblique the angle, the lesser the amount of vitamin D synthesized [84]. At latitudes beyond 35°, vitamin D production declines [65]. However, geography did not seem to be an obvious determinant of the difference in vitamin D level, as evidenced in the apparent lack of impact of latitude on vitamin D levels (Additional file 3: Figure S4A-C), and this is consistent with other published literature. A recent study in the US demonstrated that, for a large proportion of the year (March – October), serum vitamin D status was independent of geographical latitude [64]. This study further described that latitude becomes limiting only during the winter months (November – February). In our review, indoor vs. outdoor work and the amount of time exposed to sunlight seemed to be the dominant determinant of vitamin D levels.

Serum 25-(OH)D levels vary widely according to the season of the year in which the studies or collection of samples are conducted [25, 85], thus establishing season of the year as a confounder of serum vitamin D level [84]. Several lines of evidence suggest that there is a high variability in seasonal vitamin D levels across the globe [86, 87]. In a study on a normal Japanese population (adults without any abnormal biochemical data shown on routine medical check-up; in particular, those not suffering from parathyroid or calcium-related diseases, based on biochemical measurements and clinical assessment), Ono et al. [88] showed that mean serum 25-(OH)D levels were lowest in winter and spring, and peaked in the summer and the beginning of autumn. These findings were consistent with data from healthy postmenopausal women in New Zealand presented by Bolland et al. [89]. Accordingly, vitamin D deficiency was more prevalent in the spring and winter relative to the level of deficiency seen in summer and autumn [64]. These findings confirm the dependence of vitamin D status on season. Devgun et al. [85] also demonstrated in both indoor and outdoor workers that serum 25-(OH)D levels varied significantly according to season, being lowest in the spring and winter, and highest in the late summer and the beginning of autumn, in agreement with previous data [88, 89]. More importantly, they showed that vitamin D levels in outdoor workers were higher relative to indoor workers for all seasons but more pronounced in early winter.

In this systematic review, we could not perform a meta-analysis on all occupational groups to determine the effect of seasonality on vitamin D status because most of the included studies did not report on the season in which serum vitamin D level was assessed. We suggest that seasonality should be routinely reported in future studies on vitamin D levels to try to address this gap. Nonetheless, from the included studies which reported on seasons in indoor and outdoor workers, our data showed that vitamin D levels in outdoor workers were higher relative to indoor workers in all four seasons (Fig. 3). An unexpected finding in the present study was that the level of 25-(OH)D in the winter was comparable to the summer and autumn values in outdoor workers. Endogenous vitamin D synthesis is a function of UV radiation in the wavelength range of 280–320 nm, which in turn depends on season and latitude [64, 65, 85, 90]. At high latitudes (>35° N) UV radiation becomes almost negligible in the winter months, which consequently affects vitamin D production [85]. In contrast, at lower latitudes, UV radiation is not limiting during the winter months, which suggests that vitamin D synthesis can proceed all year long [64]. An alternative explanation may be that it is possible to accumulate sufficient vitamin D stores to get through the winter, but that the stores are depleted by springtime. Therefore, the lowest levels are seen in spring.

In the present study, the location of the studies that contributed the highest amount to the pooled mean 25-(OH)D level in outdoor workers in the winter were at latitudes <35° N [91]. Thus, outdoor workers in these regions may not experience huge seasonal variation in vitamin D synthesis compared to indoor workers. Together with the fact that there were few studies that contributed to the analysis of serum 25-(OH)D level dependence on season, this could account for the relatively high level of 25-(OH)D in outdoor workers during the winter season.

Most adults in the general population globally have vitamin D inadequacy [13, 46]. A significant proportion of the adult population in Europe, the US and Canada have vitamin D deficiency [2, 92]. Despite the relationship between sunlight exposure and vitamin D levels, vitamin D deficiency is reportedly also prevalent among populations living in sunny climates including the Middle East, Africa, Australia, India and South America [2, 93, 94]. Population-level prevalences of vitamin D deficiency have been reported as 59% in the Canadian population [95], 52% in the Danish population [96], and 40% in the US population [97, 98]. Our systematic review suggests that occupation is a major determinant that may contribute to suboptimal vitamin D levels and that workers in some occupations have lower average levels of vitamin D and a higher prevalence of deficiency compared to the general public. Indoor workers, shiftworkers, medical residents, healthcare students, practising physicians and coalminers have a particularly high prevalence of vitamin D deficiency. However, most occupational groups considered in this review, with the exception of lead/smelter workers, had a moderate to high burden of vitamin D deficiency or insufficiency.

Individuals who work predominantly indoors and shiftworkers are at risk of developing vitamin deficiency or insufficiency. Despite a lack of consensus on optimal levels of vitamin D for health, vitamin D insufficiency and deficiency are common in the occupational groups investigated, and some workers should be considered an at-risk group for vitamin D deficiency. Further high quality studies are needed to explore the relationship between occupation and vitamin D status. The assumption that trainees in the health care disciplines represent a convenience sample of “healthy” adults may not always be true. Guidelines on screening for vitamin D deficiency and supplementation strategies in vulnerable groups should include consideration of occupation.