Hypothesis—Ultraviolet-B Irradiance and Vitamin D Reduce the Risk of Viral Infections and thus Their Sequelae, Including Autoimmune Diseases and some Cancers†
This paper is part of a special issue dedicated to Professor Hasan Mukhtar on the occasion of his 60th birthday.
*Corresponding author email: email@example.com (William B. Grant)
Many viral infections reach clinical significance in winter, when it is cold, relative humidity is lowest and vitamin D production from solar ultraviolet-B irradiation is at its nadir. Several autoimmune diseases, such as multiple sclerosis, type 1 diabetes mellitus and asthma, are linked to viral infections. Vitamin D, through induction of cathelicidin, which effectively combats both bacterial and viral infections, may reduce the risk of several autoimmune diseases and cancers by reducing the development of viral infections. Some types of cancer are also linked to viral infections. The cancers with seemingly important risk from viral infections important in winter, based on correlations with increasing latitude in the United States, an index of wintertime solar ultraviolet-B dose and vitamin D, are bladder, prostate, testicular and thyroid cancer, Hodgkin’s and non-Hodgkin’s lymphoma, and, perhaps, gastric cancer. The evidence examined includes the role of viruses in the etiology of these diseases, the geographic and seasonal variation of these diseases, and the time of life when vitamin D is effective in reducing the risk of disease. In general, the evidence supports the hypothesis. However, further work is required to evaluate this hypothesis.
Provide feedback or get help You are viewing our new enhanced HTML article.
If you can't find a tool you're looking for, please click the link at the top of the page to "Go to old article view". Alternatively, view our Knowledge Base articles for additional help. Your feedback is important to us, so please let us know if you have comments or ideas for improvement.
The understanding of the health benefits of vitamin D is growing rapidly. In addition to the well-known calcemic effects such as reduced risk of rickets (1) and bone fractures (2), there are many noncalcemic benefits. The well-known ones include prevention of cancer and autoimmune diseases (3–10). Vitamin D supplementation reduces total mortality rate (11).
A topic of current interest is the role of vitamin D in reducing the risk of diseases caused by bacterial and viral infections. The role of vitamin D in combating tuberculosis has been known for years, with Niels Finsen receiving the Nobel prize in 1903 for his therapeutic results with ultraviolet irradiation for lupus vulgaris, a skin form of tuberculosis (12). The mechanism involved is induction of cathelicidin, an antimicrobial polypeptide (13). There is one human cathelicidin, which is also known as hCAP18, LL-37 (a peptide of 37 amino acid residues), and FALL-39, hereafter called LL-37. (Cathelicidins are a family of mammalian proteins containing a C-terminal cationic antimicrobial domain that becomes active after being freed from the N-terminal cathelin portion of the holoprotein .) LL-37, in addition to its bactericidal effect, plays many roles in inflammatory and tissue remodeling processes. It stimulates angiogenesis, induces proliferation of lung epithelial cells, accelerates wound closure of the airway epithelium and provokes cytokine release (e.g. interleukin-8) and cell migration. LL-37 can also neutralize lipopolysaccharide, a heteropolymer associated with organic dust, produced by gram-negative bacteria (15). Also, calcitriol downregulates monocyte Toll-like receptor (TLR) expression. TLRs are inducers of inflammation and act as important triggers for such events as sepsis or autoimmune disease exacerbation (16).
Epidemiologic findings on the role of vitamin D in reducing the risk of viral infections are rather recent. Cannell et al. (17) hypothesized that the winter peak of epidemic influenza activity is due in part to the lowest serum calcidiol levels at that time. The mechanisms proposed to explain the role of vitamin D in reducing the risk of influenza in summer included the production of LL-37 (18). In response to this report, Aloia and Li-Ng published poststudy analysis of the incidence of common colds and influenza among a group of 208 postmenopausal black women with a mean age of 60 ± 6 years living in New York State (19). The women were enrolled in a 3-year randomized clinical trial. Half were given 800 IU day−1 of vitamin D3 the first 2 years and 2000 IU day−1 the third year, and half were given a placebo. Every 3 months, the women were interviewed and asked whether they had experienced a cold or influenza in the previous 3 months. Twenty-six women taking the placebo reported having at least one of these illnesses, compared with seven taking 800 IU day−1 and only one taking 2000 IU day−1. Shortly thereafter, a study was published reporting the correlation of meteorological parameters with respiratory syncytial virus on the basis of weekly case numbers. Solar UVB radiation explained 13% of the variance in Miami, Florida, 5% in Buffalo, New York and 0.6% in Winnipeg, Manitoba, with temperature and relative humidity explaining much of the rest of the variance (20). Together, these three reports provide epidemiologic support for the hypothesis that vitamin D reduces the risk of respiratory viral infections.
There is also a growing body of literature examining the role of vitamin D in inducing LL-37 in response to bacterial or viral infections. The process of producing cathelicidin involves the microbe or viral activation of the enzyme 25(OH)D3-1-α-hydroxylase in macrophages, other immune cells or epithelial cells, which increases local calcitriol levels from circulating calcidiol, which then activates the cathelicidin antimicrobial peptide (CAMP) gene (21). Early work showed that 1,25-dihydroxyvitamin D3 (calcitriol) induced CAMP gene expression in isolated human keratinocytes, monocytes and neutrophils, and human cell lines (22). This finding also held for myeloid cells (23) and many cancer cell lines (24). Serum calcitriol levels are reduced in smokers (25), which may explain why smokers tend to have lower bone mass density and are more susceptible to respiratory tract infections (26).
There are also viral infections that peak in summer, such as human papilloma virus (27,28) and herpes simplex (27). The reason for the summertime peak is likely immunosuppression by UV irradiance (29). However, such viral infections are not discussed further in this article.
Viral infections have many sequelae. These include autoimmune diseases such as multiple sclerosis (MS), type 1 diabetes mellitus and asthma, and some forms of cancer. Vitamin D reduces the risk for these diseases. However, the conventional viewpoint does not include the possibility that the action of vitamin D is to reduce the risk of viral infection. This work proposes the hypothesis that vitamin D, through antimicrobial and antiviral actions, prevents the onset of these diseases.
Epidemiologic evidence for a role of vitamin D
The introduction to this report discussed recent findings for influenza, the common cold and respiratory syncytial virus. Many other viral infections have peak activities in winter—not only respiratory tract infections but also gastroenteritis and diarrhea (30–32), meningitis (33) and scarlet fever (34). The fact that internal cancers also have wintertime peaks indicates that systemic immune system defenses are involved. Table 1 summarizes the data for some viral infections with clinical symptoms.
Table 1. Evidence for bacterial and viral infections and UVB/vitamin D for viral infections with clinical symptoms.
|Bronchitis||(20)||(20)||(20)|| || |
|Herpes simplex virus|| || ||(27)|| ||(36)|
|HIV|| || || || ||(37)|
|Mononucleosis||(38)|| ||(39)|| || |
Pandemic influenza appears to have seasonality similar to that for epidemic influenza (43). If so, insuring that the populations have high calcidiol levels in winter might be much more effective in limiting the spread of pandemic influenza than would mass vaccinations. One reason is that developing such vaccines takes about a year, and preparing enough vaccine for the entire population is difficult (44). Also, vaccine may have only a 50% effectiveness rate (45). Often the elderly are targeted, but children, who come in contact with many other children at school, bring the influenza virus home (46).
Many autoimmune diseases develop because of an immune system response that may have started to attack a viral infection but instead attacked healthy body tissues. Davidson and Diamond (47) define such diseases as “a clinical syndrome caused by the activation of T cells and B cells, or both, in the absence of an ongoing infection or other discernible cause.” They may develop some time after an infection, which may not have reached the stage of clinical symptoms. They can also develop after vaccinations (48). I will review several of the autoimmune diseases that have received the most study.
The increase in the prevalence of MS with increasing latitude has been known since shortly after World War I (49) and was confirmed in later war periods (50). Incident UVB and vitamin D status were proposed to explain this geographic variation in 1974 (51) and more comprehensively in 1997 (52). Latitude also correlated with MS prevalence in Australia (53).
Viral infections are an important risk factor for MS. The Epstein–Barr virus is associated with many who develop MS (54,55). The virus is also a risk factor for infectious mononucleosis, and many who develop MS contracted infectious mononucleosis in their youth (56). Once MS develops, UVB and vitamin D can reduce the severity of the disease. The mechanism seems to be vitamin D-induced apoptosis of CD4 T lymphocytes (57).
The season associated with development of MS is also a clue to the role of vitamin D. A population-based case–control study in Tasmania found that higher UV exposure in winter appeared more important than higher exposure in summer (58). A study of monozygotic twins born in the United States, for which one of the twins developed MS, assessed nine measures of solar UV irradiance in childhood. Sun tanning was an important risk reduction factor (odds ratio [OR] = 0.40; 95% confidence interval [CI], 0.19–0.83; P = 0.01), as was spring UVB exposure (OR = 0.25; 95% CI, 0.07–0.89; P = 0.03) (59).
The data for American veterans at the time of entry into World War II and the Korean Conflict (50) are also useful for establishing the seasonality of risk for MS. July solar UVB levels in the United States are asymmetrical (60) for two important reasons—the surface elevation is generally higher from the Rocky Mountains to the west, and the stratospheric ozone layer is thinner because of the prevailing westerly winds crossing the Rocky Mountains and pushing the tropopause higher while doing so. However, latitude is presumably an index for wintertime UVB and vitamin D because solar zenith angle at that time is much more important than is surface elevation or stratospheric ozone. In a second-order regression with respect to latitude, using the data for the continental states other than California (which was divided into two states), the adjusted R2 is 0.76 (P < 0.001), whereas that for July UVB doses is 0.29 (P < 0.001) (see also Grant and Holick ).
Type 1 diabetes mellitus
Type 1 diabetes is another autoimmune disease linked to viral infections (62,63) and low vitamin D intake (64) and low serum calcidiol levels (65). A significantly increased risk was observed for illnesses in the neonatal period (OR = 1.61; 95% CI, 1.06–2.44), most of which were infections and respiratory difficulties (62). Exclusive breast feeding as the initial feeding method was significantly protective (OR = 0.65; 95% CI, 0.45–0.94) (62). Breast feeding helps build the immune system of the infant through a variety of mechanisms (66). Viral childhood diseases (OR = 4.29; 95% CI, 1.57–11.74) and bottle feeding (OR = 1.83; 95% CI, 1.08–3.09) were directly correlated with type 1 diabetes; an inverse correlation was found for vitamin D administration during lactation (0–14 years) (OR = 0.31; 95% CI, 0.11–0.86) and for history of scarlet fever in both sexes and age groups (OR = 0.19; 95% CI, 0.08–0.46) (63). Prevalence increased with increasing latitude in Australia (67). The highest incidence rates in Spain occurred in fall and winter (68). Thus, type 1 diabetes shares the same epidemiologic features associated with MS.
Asthma is another autoimmune disease associated with viral infections before diagnosis (69–74). Maternal intake of vitamin D during pregnancy reduces the risk of developing asthma at age 3 years (75,76). However, vitamin D supplementation in infancy correlated with increased risk of asthma in adulthood (OR = 1.35; 95% CI, 0.99–1.8) (77). Thus, there appears to be a protective effect in utero but an adverse effect in infancy. The health-related quality of life of asthmatic children in Spain was highest in summer and lowest in fall (78), although not at a statistically significant level. Other factors that affect risk and health-related quality of life for those with asthma include active or passive exposure to cigarette smoke (79) and airborne pollutants including environmental tobacco smoke, outdoor air contaminants and increased pollen burdens (80). The strongest evidence that vitamin D reduces the risk of viral infections, thereby reducing the risk of asthma, appears to be that viral infections are associated with about half of all cases of adult asthma (71,74,81).
Thus, there is good epidemiologic evidence that several types of autoimmune diseases have origins in viral infections and that solar UVB and vitamin D affect the risk of incidence and severity of the symptoms. Because vitamin D reduces the risk of viral infections, it follows that vitamin D reduces the risk of these diseases by reducing the risk of viral infections with clinical symptoms. The same hypothesis applies to several other autoimmune diseases. Table 2 presents the evidence for the autoimmune diseases.
Table 2. Evidence for viral infections and UVB/vitamin D for autoimmune diseases.
|Type 1 diabetes||(62,63)||(67)||(78)||(75,76)|| |
|Atopic eczema||(85)||(86)||(87)|| || |
|Lupus||(47,93–95)|| || || || |
|Inflammatory bowel disease||(96,97)||(98–100)||(101,102)||(103–105)|| |
There is strong ecological and observational evidence that solar UVB irradiance and vitamin D play important roles in reducing the risk of many types of cancer (3–10,106,107). Most of the ecological or observational studies are based on either summertime (5) or annual average (106) solar UVB doses. Vitamin D seems to be more effective in combating cancer in the latter stages. One study found that annual average solar UVB was much more strongly inversely correlated with cancer mortality rates than incidence rates (106). There is also one recent randomized prospective double-blind placebo-controlled study that found a 77% reduction in all-cancer mortality rates between the ends of the first and fourth years (6). Another recent study, based on single serum calcidiol measurements up to 12 years prior to death by cancer found significant inverse correlations for breast and colorectal cancer but not all-cancer mortality rates (108). The mechanisms whereby vitamin D reduces the risk of cancer are generally well known (109,110).
There are, however, several recent studies that are not in agreement with the UVB/vitamin D/cancer theory. Studies based on calcidiol from stored sera several years prior to detection of cancer often do not find an inverse correlation between calcidiol levels and cancer incidence. A study in Nordic countries found both lower and higher than average calcidiol levels correlated with increased risk of prostate cancer (111). Stored calcidiol levels among Finnish smokers were directly correlated with risk of pancreatic cancer (112). A study in Linxian, China found serum calcidiol directly correlated with risk of esophageal cancer (113). The general lack of inverse correlation between stored sera and prostate cancer in many studies led, in part, to the present manuscript. The finding for pancreatic cancer is in opposition to many other studies on pancreatic cancer based on ecological studies with respect to indices for solar UVB (5,106,114) and oral intake (115). I have proposed that the finding regarding esophageal cancer in China was due to confounding from immunosuppression by solar UV, thereby increasing the risk of esophageal cancer from human papillomavirus (116; W.B. Grant, unpublished). Oral sources of vitamin D are generally more important than solar UVB irradiance in Nordic countries. Milk is fortified with vitamin D in Sweden, and dairy products are correlated with prostate cancer risk in Sweden (117) and pancreatic cancer in the San Francisco Bay Area (118). While these explanations are plausible, confirmation will await further evaluation.
There are several problems associated with using stored sera to investigate the role of vitamin D in reducing the risk of cancer. One, just discussed, is confounding with other factors. Another is that measurements are generally made a number of years prior to diagnosis of or death from cancer. As vitamin D seems to be much more effective in fighting cancer near the time of detection (6,106,119,120), sera stored many years prior to cancer outcome may have little relevance to the outcome. In addition, the measurements are generally one-time determinations and do not take into account changes in vitamin D intake and production over a period of many years. The ecological approach, on the other hand, is based on solar UVB over many years, with the possible confounding factor of migration (106). Another approach that is based on integrated lifetime solar UVB irradiance is the use of incidence of nonmelanoma skin cancer (NMSC). This approach was tested in a meta-analysis of cases of second solid tumor after development of NMSC (121). In doing so, it was important to adjust the relative risk by the effects of smoking in each study population by using lung cancer incidence rates. NMSC mortality rates over a 15-year period in continental Spanish provinces were used in another ecological study (122). In that study, it was reported that NMSC mortality rates were significantly inversely correlated with 17 types of cancer. It was recently pointed out that as the number of NMSC deaths was small, that was not a good index to use. In addition, I realized that I had not performed a multiple linear regression analysis to determine the contributions from smoking or to see whether latitude might be a better index. Thus, I reanalyzed the data using multiple linear regressions. A total of 15 types of cancer had either latitude or NMSC as a significant factor. Bladder and uterine corpus are now removed from the list of UVB/vitamin D-sensitive cancers for Spain from this study. Table 3 summarizes the findings in terms of numbers. More recently, a linkage study in sunny countries found that NMSC was correlated with reduced risk of solid tumors other than lip or skin cancer (123,124), and a study in The Netherlands found that diagnosis of skin cancer was correlated with reduced risk of prostate cancer (125). Thus, this stronger measure of lifetime UVB irradiance is very often found inversely correlated with many types of cancer.
Table 3. Summary of the regression results (122) giving the number of statistically significant findings and using the Bonferroni criterion (P < 0.05/n) where n is the number of independent variables.
|Males plus females||21||5||13||15|
Although most cancers are caused by factors related to carcinogens, smoking, alcohol consumption and dietary factors (126,127), viral infections also increase risk for several cancers. The well-known ones include several for which viral infections are the primary risk—cervical cancer (128), Hodgkin’s lymphoma (129) and nasopharyngeal cancer (130). However, viral infections are a risk factor for several other cancers. These cancers include, but are not limited to, anal cancer (131), bladder cancer (132–135), esophageal cancer (116), hepatocellular carcinoma (136), oral cancer (137) and prostate cancer (138–141).
Prostate cancer was one of the first cancers hypothesized to have risk reduced through solar UVB irradiance and vitamin D intake (142). However, the geographic variation of prostate cancer mortality rates in the United States is different from that of many of the other cancers for which UVB and vitamin D are risk-reduction factors. Prostate cancer mortality rates are more closely tied to a symmetrical latitudinal pattern, whereas other cancers such as breast and colon cancer have an asymmetric variation, highest in the northeast and lowest in the southwest (143). Prostate cancer mortality rates in the United States have a geographic variation that is similar to that for MS, which Schwartz (144) recognized early. The adjusted R2 for a second-order regression with respect to latitude, using data for 1970–1994 (143) for the continental states, is 0.44 (P < 0.001), whereas that for July UVB doses is 0.08 (P < 0.001). Genital bacterial and viral infections are more common in youth and early adulthood (145), and higher levels of solar UVB exposure in youth are significantly correlated with reduced risk of prostate cancer (107). Viral infections can give rise to inflammation (136), an important risk factor for prostate cancer (146,147). Prostate cells have reduced ability to convert calcidiol to calcitriol because the activity of the 1α-hydroxylase enzyme decreases as the cells turn cancerous (148,149), thereby reducing the effectiveness of vitamin D later in life.
Regression results for other cancers are presented in Table 4. Cancers with a significant correlation with either latitude or January temperature are bladder, gastric, prostate, testicular, thyroid cancer and Hodgkin’s lymphoma. The evidence supporting a role of viral infections for these cancers is summarized in Table 5. Many of the cancers not listed or related to viral infections exacerbated by solar UV and immunosuppression, such as breast (165,166), colon and ovarian (167) cancer, do not have viral infections as important risk factors. One exception is the finding of viruses associated with colon cancer after kidney transplants (168).
Table 4. Multiple linear regression results for mortality rates for 1950–1969.
|Bladder||M||0.56,*|| || ||0.21, 0.01||0.59,*|| ||0.22, 0.01||0.84, 64,*|
|0.67,*|| || ||0.35,*||0.72,*|| || ||0.83, 78,*|
|F|| || ||−0.38,*|| ||0.52,*|| ||0.26, 0.02||0.59, 24,*|
|Gastric||M|| ||−0.53,*||−0.49,*|| || ||0.52,*|| ||0.67, 33,*|
|0.41, 0.01|| ||−0.49,*|| || ||0.50,*|| ||0.52, 19,*|
|F|| ||−0.48,*||−0.50, 0.001|| || ||0.54,*|| ||0.62, 27,*|
|0.42, 0.005|| ||−0.51, 0.004|| || ||0.54,*|| ||0.56, 22,*|
|Hodgkin’s lymphoma||M||0.68,*|| || || ||0.36, 0.005|| || ||0.39, 16,*|
|0.66,*|| || || ||0.41, 0.004|| || ||0.32, 12,*|
|F|| || ||−0.86,*|| || ||0.31, 0.02|| ||0.50, 24,*|
|NHL||M||0.50,*|| || ||0.42, 0.001|| || || ||0.38, 15,*|
| ||−0.43, 0.001|| ||0.47,*|| || || ||0.34, 13,*|
|F|| ||−0.56,*|| ||0.49,*|| || || ||0.47, 22,*|
|0.56,*|| || ||0.44,*|| || || ||0.45, 20,*|
|Prostate cancer||M||0.64,*|| || || || || || ||0.40, 33,*|
| ||−0.64,*|| || || || || ||0.39, 32,*|
|Testicular||M|| ||−0.59,*|| || || || || ||0.33, 21,*|
|0.58,*|| || || || || || ||0.32, 20,*|
|Thyroid||M|| ||−0.29, 0.02|| ||0.50,*|| || || ||0.27, 9.4,*|
|−0.06, 0.66|| || ||0.43, 0.002|| || || ||0.16, 5.7, 0.006|
|F||0.63,*|| || ||0.31, 0.006|| || || ||0.44, 21,*|
| ||−0.58,*|| ||0.40, 0.001|| || || ||0.42, 19,*|
Table 5. Evidence for viral infections in winter and UVB/vitamin D for cancer.
|Gastric||(153)||(5,106,122)|| || || |
|Hodgkin’s lymphoma||(129,154) ||(5,122)||(39,155,156)|| || |
|NHL||(157)|| || || || |
|Testicular||(161,162)|| || || || |
|Thyroid||(163)||(106,122)|| ||(164)|| |
The primary mechanism whereby vitamin D reduces the risk of bacterial and viral infections reaching the point of clinical symptoms is through induction of the antimicrobial peptide LL-37.
Using keratinocytes, Wang et al. (22) first demonstrated that calcitriol induces LL-37. Other findings supporting this conclusion were reported for production of hCAP18, on the skin with calcitriol (169), and for myeloid cells (23). Later, other reports generalized the finding to several cancer cell lines (24), as well as to cystic fibrosis bronchial epithelial cells (170).
Liu et al. (171) recently reported an experiment showing how vitamin D plays a role in the immune response to Mycobacterium tuberculosis. They demonstrated that when TLRs 1 and 2 were activated by exposure to M. tuberculosis, they responded by upregulating vitamin D receptors and Cyp27B1, the enzyme that converts calcidiol to calcitriol, in monocytes. After calcitriol levels increased, LL-37 was upregulated as well. LL-37 is effective against M. tuberculosis, which explains why black Americans, who generally have lower calcidiol levels than white Americans (172) have a greater risk of tuberculosis. Calcitriol also modulates T-cell responses (173).
Whereas these reports demonstrated a beneficial effect of cathelicidin against bacteria, there is good evidence that it is also effective against viruses (174). LL-37 expressed by conjunctival and corneal epithelia was effective against both bacteria and viruses (175). LL-37 presumably punches a hole in the envelope. LL-37 has been beneficial against the herpes simplex virus (36), as well as influenza (35). Also, CAMP inhibits replication of human immunodeficiency virus type 1 (HIV-1), thereby probably contributing to local protection against HIV-1 infection (37).
Summary and conclusion
Vitamin D, through induction of human cathelicidin, LL-37, reduces the risk of many types of autoimmune disease and several types of cancer. Vitamin D helps the innate immune system fight bacterial and viral infections to prevent them from reaching the stage of clinical symptoms. This hypothesis helps explain the seasonality of many viral diseases, as well as the seasonal variation in incidence and exacerbation of several autoimmune diseases and cancers. Although more studies should verify these findings, the hypothesis appears strong enough now to add support for increasing the recommendations for vitamin D production from UVB irradiance (176), food fortification (177) and supplementation (178). Given the overall positive benefits associated with calcidiol levels of 60–80 ng mL−1 or higher (61), there appears to be little health risk associated with doing so now.
Disclosure— I receive funding from the UV Foundation (McLean, VA) and the Vitamin D Society (Canada) and the European Sunlight Association.