UVB improved vitamin D levels in 3 diseases, controls improved more – PHD Nov 2013

Narrow-band Ultraviolet B Exposures Improve Vitamin D Balance Trials Involving Dermatological and Haemodialysis Patients and Healthy Subjects

MERI ALA-HOUHALA
ACADEMIC DISSERTATION To be presented, with the permission of the Board of the School of Medicine of the University of Tampere, for public discussion in the Small Auditorium of Building M, Pirkanmaa Hospital District, Teiskontie 35, Tampere, on November 22nd, 2013, at 12 o'clock.
UNIVERSITY OF TAMPERE

ACADEMIC DISSERTATION University of Tampere, School of Medicine
Tampere University Hospital, Department of Dermatology and Venereology and Department of Internal Medicine National Graduate School of Clinical Investigation Finland
Supervised byProfessor emeritus Timo Reunala University of Tampere Finland
Professor Erna Snellman University of Tampere Finland
Reviewed by Professor emeritus Aarne Oikarinen, University of Oulu Finland
Docent Pauli Karhapaa University of Eastern Finland Finland
Copyright ©2013 Tampere University Press and the author
Cover design by Mikko Reinikka, Layout Sirpa Randell
Acta Universitatis Tamperensis 1870 Acta Electronica Universitatis Tamperensis 1351
ISBN 978-951-44-9260-0 (print) ISBN 978-951-44-9261-7 (pdf)
ISSN-L 1455-1616 ISSN 1456-954X; ISSN 1455-1616 http://tampub.uta.fi
Suomen Yliopistopaino Oy - Juvenes Print Tampere 2013

VitaminDWiki suspect that disease consumption of vitamin D caused the blood response
When filling up two buckets with water, it is hard to fill up the bucket with leaks nearly as much

Note: having technical difficulties uploading large PDF files on new web host

1 LIST OF ORIGINAL PUBLICATIONS (in the PDF and some are in VitaminDWiki)

This thesis is based on the following papers, which will be referred to in the text by the Roman numerals I-IV:

I Vahavihu K*, Ala-Houhala M*, Peric M, Karisola P, Kautiainen H, Hasan T, Snellman E, Alenius H, Schauber J, Reunala T.
Narrowband ultraviolet B treatment improves vitamin D balance and alters antimicrobial peptide expression in skin lesions of psoriasis and atopic dermatitis. Br J Dermatol 2010; 163:321-8.
II Ala-Houhala MJ, Vahavihu K, Hasan T, Kautiainen H, Ylianttila L, Viljakainen HT, Snellman E, Reunala T.
Comparison of narrowband ultraviolet B exposure and oral vitamin D substitution on serum 25-hydroxyvitamin D concentration. Br J Dermatol 2012; 167:160-4.
III Ala-Houhala MJ, Vahavihu K, Hasan T, Kautiainen H, Snellman E, Karisola P, Dombrowski Y, Schauber J, Saha H, Reunala T.
Narrow-band ultraviolet B exposure increases serum vitamin D levels in haemodialysis patients. Nephrol Dial Transplant 2012; 27:2435-40.
IV Ala-Houhala MJ, Vahavihu K, Snellman E, Hasan T, Kautiainen H, Karisola P, Dombrowski Y, Schauber J, Saha H, Reunala T.
A narrow-band ultraviolet B course improves vitamin D balance and alters cutaneous CYP27A1 and CYP27B1 mRNA expression levels in haemodialysis patients supplemented with oral vitamin D. Nephron Clin Pract 2013; 124:17-22.
in VitaminDWiki as UVB increases vitamin D even if poor kidney by changing CYP27B1 genes in skin – Dec 2013

* = Equal inputs from both authors.

2 ABBREVIATIONS in PDF

3 ABSTRACT

Narrow-band ultraviolet B (NB-UVB) phototherapy is used to treat dermatological diseases such as psoriasis and atopic dermatitis. Some previous studies have suggested that it also increases serum 25-hydroxyvitamin D (25(OH)D) concentrations. On the other hand, most patients with chronic kidney disease (CKD) requiring dialysis are known to have insufficient vitamin D. We therefore conducted trials to assess how short NB-UVB courses could affect serum 25(OH)D concentrations in dermatological and haemodialysis patients in winter, when little UVB from the sun is available for vitamin D synthesis. In addition, we compared the effects of an NB-UVB course and oral vitamin D supplementation on serum 25(OH)D concentrations in healthy subjects.
In the first trial (I),

89% of the patients with psoriasis,
94% of those with atopic dermatitis and
53% of the healthy subjects

were found to have baseline vitamin D insufficiency (serum 25(OH)D <50 nmol/L).
A course of 15 whole body NB-UVB exposures significantly increased serum 25(OH)D (p <0.001), by 59.9 nmol/L in the psoriasis patients, 68.2 nmol/L in the atopic dermatitis patients and 90.7 nmol/L in the healthy subjects. PASI (psoriasis area and severity index) and SCORAD (severity scoring of atopic dermatitis) improved significantly (p <0.001), but no correlation with the increase in serum 25(OH)D was found. Expression of antimicrobial peptides (AMPs), cathelicidin and human (3-defensin 2 (HBD2) was high in the psoriasis skin lesions. After 6 NB-UVB treatments cathelicidin had increased further, while HBD2 expression had decreased. NB-UVB caused a marked but non-significant decrease in the cytokines interleukin (IL)-1(3 and IL-17 in the psoriasis lesions. It was concluded that, in addition to a significant improvement of psoriasis and atopic dermatitis, NB-UVB treatment effectively corrects vitamin D insufficiency. It also increases cathelicidin and lowers HBD2 levels in healing psoriasis and atopic dermatitis skin lesions. This effect may be mediated by the improved vitamin D balance and the local cytokine network.

In the second trial (II) healthy adult hospital employees and medical students having serum 25(OH)D below 75 nmol/L were randomly given either a course of 12 whole body NB-UVB exposures or 20 jxg of oral cholecalciferol daily for 4 weeks. The baseline serum 25(OH)D concentrations were similar in both groups: 52.9 nmol/L in the 33 NB-UVB-treated subjects and 53.5 nmol/L in the 30 treated with oral cholecalciferol. The mean increase in serum 25(OH)D was 41.0 nmol/L in the NB-UVB group and 20.2 nmol/L in the cholecalciferol group, the difference being significant at 2 weeks (p = 0.033) and at 4 weeks (p <0.001). Two months after the treatments the 25(OH)D concentrations had decreased in both groups but were still clearly higher than the baseline values. It was concluded that 12 NB-UVB exposures given over 4 weeks increase the serum 25(OH)D concentration significantly more than do daily doses of 20 jxg oral cholecalciferol. A short, low-dose NB-UVB course is therefore an effective way of improving the vitamin D balance in winter, and the response is still evident 2 months after the course.

In the third trial (III) fifteen haemodialysis patients and twelve healthy subjects received nine upper body NB-UVB exposures. Mean serum 25(OH)D levels before NB-UVB were 32.5 nmol/L in the dialysis patients and 60.2 nmol/L in the healthy subjects (p <0.001). After eight NB-UVB exposures serum 25(OH)D had increased by 13.8 nmol/L (43%; p <0.001) in the dialysis patients and 1,25-dihydroxyvitamin D (1,25(OH)2D) by 3.3 pmol/L (27%; p = 0.002). Serum 25(OH)D in the dialysis patients was still 10% higher two months after NB-UVB exposures than initially. The mRNA expression level of CYP27B1, an enzyme needed for the final hydroxylation of vitamin D to its active metabolite, was examined in skin biopsy specimens and was found to have increased after NB-UVB exposures relative to the level in non-treated healthy subjects (p = 0.04). It was concluded that a short course of NB-UVB exposures significantly increases serum 25(OH)D and 1,25(OH)2D in dialysis patients, but the effect is short-lasting suggesting that patients need cyclic NB-UVB exposures to maintain their improved vitamin D concentrations.
In the fourth trial (IV) fourteen haemodialysis patients and fifteen healthy subjects receiving oral cholecalciferol supplements of 20 jxg daily were given nine whole body NB-UVB exposures. Given baseline serum 25(OH)D concentrations of 57.6 nmol/L in the dialysis patients and 74.3 nmol/L in the healthy subjects the NB-UVB course increased serum 25(OH)D significantly (p <0.001), by 14.0 nmol/L in the former and 17.0 nmol/L in the latter. The dialysis patients showed significantly increased baseline CYP27B1 levels and decreased CYP27A1 levels in the skin relative to the healthy subjects. It was concluded that a short NB-UVB course is an efficient way of improving the vitamin D balance in dialysis patients who are receiving oral vitamin D supplementation. The increased cutaneous CYP27B1 levels in the dialysis patients suggest that the loss of the renal activity of this enzyme is at least partially compensated for in the skin.
To conclude, NB-UVB exposures were shown to be an efficient way of increasing the serum 25(OH)D concentration in dermatological and dialysis patients and in healthy subjects in winter. A short NB-UVB course was shown to increase serum 25(OH)D in healthy subjects significantly more than did daily supplementation with 20 ug oral cholecalciferol. NB-UVB courses offer a new possibility for improving the vitamin D balance in dialysis patients who have an insufficiency of this vitamin.

4 INTRODUCTION

Vitamin D insufficiency is common worldwide (Holick 2007, Holick & Chen 2008). In the Nordic countries and Britain this condition affects people especially during the winter when vitamin D synthesis induced by the sun is negligible (Burgaz et al. 2007, Hyppoen & Power 2007, Grant et al. 2011). The desirable concentration of serum 25-hydroxyvitamin D (25(OH)D), which is the best indicator of vitamin D status, is still under debate (Holick & Chen 2008, Pludowski et al. 2013), but a concentration below 75 nmol/L is considered to be insufficient for bone fracture prevention (Dawson-Hughes et al. 2005, Bischoff-Ferrari et al. 2006). In addition to osteoporosis, low serum 25(OH)D concentrations have recently been associated with a risk of colorectal cancer and cardiovascular disease (Gandini et al. 2011, Pittas et al. 2010). Vitamin D is also known to affect skin inflammation and innate or adaptive cutaneous immune responses (Schauber & Gallo 2008, Pludowski 2013).
Narrow-band ultraviolet B (NB-UVB) phototherapy is used to treat dermatological diseases such as psoriasis and atopic dermatitis (Bandow & Koo 2004, Patel et al. 2009) and both broad-band ultraviolet B (BB-UVB) and NB-UVB treatments seem to increase serum 25(OH)D concentrations (Guilhou et al. 1990, Vahavihu et al. 2010). On the other hand, most chronic kidney disease (CKD) patients on dialysis are known to suffer from vitamin D insufficiency (LaClair et al. 2005, Bhan et al. 2010, Nigwekar et al. 2012) and their response to supplementation, as evaluated in terms of increased 25(OH)D, seems to be rather slow (Quinibi et al. 2010, Kandula et al. 2011). The effect of NB-UVB on vitamin D balance in dialysis patients has not been studied previously.
The main topics of the present work were to examine the effect of NB-UVB phototherapy on the vitamin D balance of dermatological patients in winter, and to assess in healthy subjects whether NB-UVB exposures improve vitamin D balance better than does oral vitamin D supplementation. Further topics were to study the effect of NB-UVB exposures on CKD patients treated with haemodialysis and to investigate the effects of NB-UVB on the enzymes hydroxylating vitamin D, antimicrobial peptides and cytokines in the skin.

5 REVIEW OF THE LITERATURE

5.1 Vitamin D

5.1.1 Historical background

The first clinical descriptions of the devastating bone-deforming disease rickets was produced in the 17th century by Daniel Whistler (1645) and Francis Glisson (1650), as reviewed by Rajakumar et al. (2007). The lack of sunlight and its association with rickets in children was first recognized by Jerdzej Sniadecki in 1822, but one hundred years passed before Kurt Huldschinsky (1919) observed that exposure to ultraviolet B (UVB) radiation from a mercury arc lamp or sunlight was a means of preventing and treating rickets. In the early 1930s the US government set up an agency to provide recommendations for parents concerning the beneficial effect of sensible exposure to sunlight for the prevention of rickets (Holick & Chen 2008).
It was common practice in the 19th century to give children cod liver oil to prevent and cure rickets. The first scientific report to recognise rickets as a nutritional deficiency disease was published by Edward Mellanby in 1918 (Rajakumar et al. 2007) and 7-dehydrocholesterol was identified as being the precursor of vitamin D in the 1920s through co-operation between Alfred Hess, Adolf Windaus and Otto Rosenheim (Wolf 2004). Hess and Harry Steenbock discovered that UV-irradiated food, particularly whole milk containing butterfat, had an antirachitic potency, and thereafter the chemical structure of vitamin D (ergocalciferol) was subsequently identified in 1931, as cited by Nigewekar et al. (2012).
The fortification of milk with vitamin D was effective in eradicating rickets in the United States and Europe in the 1930s, but then an unfortunate outbreak of hypercalcaemia in infants occurred in Great Britain in the 1950s that was blamed on the overfortification of milk with vitamin D. This resulted in a ban on the fortification of dairy products with vitamin D in Europe (Holick & Chen 2008). In Finland, vitamin D supplements have been systematically given to children since the 1930s, initially in the form of cod liver oil. The Ministry of Social Affairs and Health recommended the fortification of liquid milk products, margarines and cheese spreads with vitamin D in 2003 (Lehtonen-Veromaa et al. 2008).

5.1.2 Photosynthesis of vitamin D

Humans receive vitamin D from exposure to sunlight, from their diet and from dietary supplements. The major source, however, is the skin when exposed to sunlight (UVB radiation), which under normal circumstances contributes more than 90% of the serum concentration of vitamin D (Reiracht 2006). Solar UVB radiation (wavelength 280 to 315 nm) penetrates the skin and converts 7-dehydrocholesterol to previtamin D3 in basal and suprabasal layers (Fig. 1; Holick 2007) by means of a photochemical reaction, which activates its maximum spectral effectiveness at about 297 nm (Lehmann & Meurer 2010).
This previtamin D3 then undergoes heat isomerization to form 25-hydroxyvitamin D (25(OH)D). This takes several hours (Webb 2006, Holick 2007) and the circulating concentration of 25(OH)D reaches its maximum level 12-24 hours after UVB exposure (Lehmann & Meurer 2010). Alternatively, the previtamin D3 can be photoisomerised further into one of two inert isomers, or back to 7-dehydrocholesterol (Webb 2006). Since any excess previtamin D3 or 25(OH)D is destroyed by sunlight, further exposure to sunlight does not cause 25(OH)D intoxication (Holick 2007).
There are several factors that limit cutaneous 25(OH)D synthesis and it is in any case difficult to say how much solar radiation is needed to achieve and maintain an adequate vitamin D status as it depends on the solar zenith angle, latitude, time of year and day, weather, person's age, skin type, clothing and activity (Webb 2006, Holick & Chen 2008, Lehmann & Meurer 2010). There is a seasonal variation in circulating 25(OH) D (Hypponen & Power 2007, Holick & Chen 2008) and the angle at which the sun strikes the earth has a dramatic effect on the numbers of UVB photons that reach the surface. Consequently, little if any 25(OH)D synthesis occurs when the zenith angle is increased during the wintertime and in the early morning and late afternoon (Holick & Chen 2008). A radiative transfer model has been used to calculate that poleward of 51° latitude there is at least some period of the year when no vitamin D synthesis can occur, but no limit has been placed on exposure times (Webb 2006).
Sensible sun exposure can provide an adequate amount of 25(OH)D and this can be stored in body fat and released during the winter, when 25(OH)D cannot be produced. Exposure of the arms and legs for 5 to 30 minutes between the hours of 10 a.m. and 3 p.m. twice a week is often adequate (Holick 2007). Exposure to one minimal erythema dose while wearing only a bathing suit is equivalent to the ingestion of approximately 500 jxg of ergocalciferol (Holick 2007) and people with abundant exposure to sunlight can easily exhibit a serum 25(OH)D >150 nmol/L (Vieth 2007). Ageing is associated with decreased concentrations of 7-dehydrocholesterol, the precursor of 25(OH)D in the skin (Holick & Chen 2008).
Figure 1. Synthesis of vitamin D starts in the skin and is completed in the kidneys. CYP27A1 in the liver and CYP27B1 in the kidneys are the major hydroxylating enzymes in the cascade. Adapted from Holick 2007.

5.1.3 Dietary sources of vitamin D

Vitamin D comprises two closely related substances of nutritional importance: vitamin D3 (cholecalciferol) and vitamin D2 (ergocalciferol). Vitamin D3 is formed from its precursor 7-dehydrocholesterol, which is found in ample amounts in the skin of humans and animals. Vitamin D2 is formed by UV radiation from its precursor ergosterol, which is present in plants, yeast and fungi (Lehmann & Meurer 2010). Plants are a poor source of vitamin D2, however.
There are a few foods that naturally contain vitamin D including fish (e.g. salmon, mackerel and herring), fish oils (including cod liver oil), mushrooms and egg yolk (Holick 2007, Hypponen & Power 2007, Lehmann & Meurer 2010). Farmed fish have a lower vitamin D content than wild fish, so that 100 g of farmed salmon contains 2.5-6.3 fig of vitamin D while 100 g of wild salmon has 15-25 fig (Holick 2007).
The range of vitamin D intake required to ensure maintenance of wintertime vitamin D status in the vast majority of 20-40-year-old adults, considering a variety of sun exposure preferences, is between 7.2 and 41.1 fig daily (Cashman et al. 2008). The dietary intake of vitamin D in Finland, evaluated using a validated food frequency questionnaire, is approximately 10 fg daily in elderly people (Viljakainen et al. 2006) and 2-6 fig daily in adolecents (Tylavsky et al. 2006).

5.1.4 Synthesis of vitamin D and calcium homeostasis

Vitamin D (D2 or D3) from the skin and diet is metabolized in the liver to 25-hydroxyvitamin D (25(OH)D, calcidiol), which is used to determine a patient's vitamin D status (Holick 2007, Pludowski et al. 2013). 25(OH)D is then metabolized in the kidneys to its active form, 1a,25-dihydroxyvitamin D (1,25(OH)2D, calcitriol). These hydroxylations are enabled by the enzymes vitamin D-25-hydroxylase (25-hydroxylase, CYP27A1) and 25-hydroxyvitamin D-1a-hydroxylase (1a-hydroxylase, CYP27B1), respectively (Fig. 1; Holick 2007, Lehmann & Meurer 2010). The serum 25(OH)D concentration regulates the levels of these enzymes by a negative feedback system, which implies that the baseline concentration of 25(OH)D is an important factor for how a person responds to an oral dose of vitamin D (Holick & Chen 2008).
Vitamin D synthesized in the skin or received from the diet can be stored in fat cells and later released from them. Obesity is associated with vitamin D deficiency, which is believed to be due to sequestration of vitamin D by the large body fat pool (Hypponen & Power 2007, Holick & Chen 2008). In the circulation vitamin D is bound to carrier proteins, in particular, vitamin D-binding protein (VDBP). 25(OH)D quickly enters the circulation, where it has a half-life of about 15 days (Holick 2007, Lehmann & Meurer 2010). A negative relationship exists between serum 25(OH)D and serum parathyroid hormone (PTH). The threshold for serum 25(OH)D, at which serum PTH starts to rise is reported in most surveys to be about 75 nmol/L (Lips 2006).
25(OH)D bound to VDBP is transported to the kidneys and hydroxylated to hormonally active 1,25(OH)2D. This 1a-hydroxylation of 25(OH)D to 1,25(OH)2D is tightly regulated by PTH and also to some extent by calcium, phosphate, calcitonin, fibroblast growth factor 23 and 1,25(OH)2D itself. 1,25(OH)2D has biological effects in the kidneys, but it also is transported by VDBP to other vitamin D receptor (VDR)-positive target tissues, mainly the bones, intestines, and parathyroid gland. The main effect of 1,25(OH)2D is to increase the absorption of calcium from the small intestine (Lips 2006, Lehmann & Meurer 2010), although it also mobilizes osteoclastic activity (Holick & Chen 2008). It has a serum half-life of 10-24 hours (Lips 2006, Lehmann & Meurer 2010). 1,25(OH)2D induces the expression of the enzyme 25-hydroxyvitamin D-24-hydroxylase (CYP24A1), which catabolizes both 25(OH)D and 1,25(OH)2D to biologically inactive, water-soluble calcitroic acid (Fig. 1; Holick 2007, Nigwekar et al. 2012), which is then excreted in the bile.
It is recognized that many tissues in the body, including the macrophages, brain, colon, prostate, breast and others, have the enzymatic machinery to produce 1,25(OH)2D locally (Holick & Chen 2008) and it has been discovered that human keratinocytes exhibit an autonomous vitamin D pathway also in vivo (Lehmann & Meurer 2010). Keratinocytes also possess vitamin D catabolic pathways. A five-step inactivation pathway from 1,25(OH)2D to calcitroic acid in epidermal keratinocytes is attributed to the multifunctional enzyme 25-hydroxyvitamin D-24-hydroxylase (CYP24A1), which is induced by 1,25(OH)2D. The hydroxylation of 25(OH)D to 1,25(OH)2D occurs in the kidneys under the influence of PTH, or in extrarenal cells and tissues under the influence of cytokines (Lips 2006).

5.2 Vitamin D insufficiency

5.2.1 Vitamin D insufficiency in general population

The best method for determining a person's vitamin D status is to measure the circulating 25(OH)D concentration, which is regarded as remaining fairy stable (Holick & Chen 2008, Pludowski et al. 2013). Most laboratories report the normal range of 25(OH)D to be 50 to 250 nmol/L (20 to 100 ng/mL) (Holick 2007). Circulating levels of 1,25(OH)2D are approximately 1/1000th that of 25(OH)D (K-DIGO 2009). The serum levels of 1,25(OH)2D range from 75 to 200 pmol/L (29 to 77 pg/mL) (Lehmann & Meurer 2010).
The desirable concentration of serum 25(OH)D is still under debate, but vitamin D deficiency can be diagnosed ifits serum levels of25(OH)D are below 50 nmol/L (Holick 2007, Querfeld & Mak 2010) and some authors have used the term severe deficiency for levels below 25 nmol/L. Relative vitamin D deficiency (vitamin D insufficiency) is considered at serum levels of25 to 75 nmol/L. Substitution with vitamin D preparations should aim at serum levels of at least 75 nmol/L or preferably 100 to 200 nmol/L. Highly elevated levels, above 200 nmol/L, or more typically above 150 nmol/L, may indicate vitamin D toxicity (Querfeld & Mak 2010).
Vitamin D deficiency is recognized as a pandemic and it has been estimated that 1 billion people worldwide have vitamin D deficiency or insufficiency (<75 nmol/L; Holick 2007). The prevalence of vitamin D deficiency in the general population has been described as ranging from 20% to 50%, with race, age, sunlight exposure, and comorbid conditions such as diabetes mellitus and obesity accounting for its wide variations (Nigwekar et al. 2012).
In one large-scale evaluation of trends in vitamin D insufficiency in the US population (Ginde et al. 2009) in which the serum 25(OH)D levels of 18 883 participants in the Third National Health and Nutrition Examination Survey (NHANES III) of 1988-1994 were compared with those of 13 369 participants in NHANES 2001-2004 the mean serum level had decreased markedly, from 75 nmol/L to 60 nmol/L between the two periods. Correspordingly, the prevalence of 25(OH)D levels of less than 25 nmol/L increased from 2% to 6%, and that levels of 75 nmol/L or more decreased from 45% to 23%.
In a European study, researchers measured 25(OH)D in 7437 whites from the 1958 British birth cohort when they were 45 years old (Hypponen & Power 2007) and found the prevalence of hypovitaminosis D to be highest during the winter and spring, when 25(OH)D concentrations <25, <40 and <75 nmol/L were found in 15.5%, 46.6% and 87.1% of participants, respectively. The corresponding proportions during the summer and autumn were 3.2%, 15.4%, and 60.9%, respectively.
Culture and habits affect serum 25(OH)D levels (Holick 2007). A Finnish study examining two groups of immigrant women (Bangladeshi and Somali) in Helsinki in winter pointed to a high prevalence of vitamin D insufficiency (25(OH)D <50 nmol/L) in the Somali group, 89.6% (Islam et al. 2012).

5.2.2 Health effects of vitamin D insufficiency

Serum levels of about 75 nmol/L are currently considered by many investigators to be optimal for health (Holick 2007, Holick & Chen 2008, Lehmann & Meurer 2010) while a concentration below 75 nmol/L is considered to be insufficient for bone fracture prevention (Dawson-Hughes et al. 2005, Bischoff-Ferrari et al. 2006). Cross-sectional studies have shown serum 25(OH)D concentrations to be related to serum PTH concentration and bone marrow density in adolescents (Lamberg-Allardt & Viljakainen 2008). When given in appropriate doses, calcium and vitamin D have been shown to be pharmacologically active, safe and effective for the prevention and treatment of osteoporotic fractures (Boonen et al. 2004) and a recent review states that the anti-fall and anti-fracture actions of vitamin D administration at a dose of at least 20 fig daily, achieving serum 25(OH)D levels of at least 60 nmol/L, appear to be effective and beneficial for the musculoskeletal machinery (Pludowski et al. 2013).
In addition to osteoporosis, low serum 25(OH)D concentrations have been associated with a risk of developing and dying from cancer (Holick & Chen 2008, Gandini et al. 2011, Pludowski et al. 2013). It has been suggested that adults with 25(OH)D below 50 nmol/L have a 30-50% increase in their risk of developing colorectal, breast and prostate cancers (Holick 2007). A meta-analysis showed that increasing the intake of vitamin D to 25 fig cholecalciferol daily is associated with a decrease of as much as 50% in the risk of colorectal and breast cancer (Holick & Chen 2008).
Although skin cancers are known to be associated with sun exposure, sunlight may provide protection from many cancers through the production of vitamin D. In a large cohort of 416 134 skin cancer cases and 3 776 501 non-skin cancer cases as the subject's first cancer extracted from 13 cancer registries (Tuohimaa et al. 2007) 10 886 of the melanoma and 35 620 of the non-melanoma skin cancer patients developed second cancers. The risk of a second primary cancer after a non-melanoma skin cancer was lower in sunny countries for most of the cancers except for the lip, mouth and non-Hodgkin lymphoma types. Vitamin D production in the skin seems to reduce the risk of many solid cancers, especially stomach, colorectal, liver and gallbladder, pancreas, lung, female breast, prostate, bladder and kidney cancers.
Vitamin D deficiency has also been linked to an increased incidence of many chronic illnesses, most notably type 1 and type 2 diabetes mellitus, hypertension, multiple sclerosis, Crohn's disease, rheumatoid arthritis, osteoarthritis, neurocognitive dysfunction, mental illness, infertility, adverse pregnancy and birth outcomes (Holick 2007, Holick & Chen 2008, Pludowski et al. 2013). It has also been thought to be associated with an increased risk of cardiovascular events and mortality (de Boer et al. 2009), although it is maintained in one review that the association between vitamin D status and cardiometabolic outcomes is uncertain (Pittas et al. 2010) as trials have shown no clinically significant effect ofvitamin D supplementation at the dosages given. Lower vitamin D status seems to be associated with an increased risk of hypertension and cardiovascular disease, but we do not yet know whether vitamin D supplementation will affect the clinical outcomes (Nigwekar et al. 2012). However, one large test of the association of low 25(OH)D levels with all-cause, cancer and cardiovascular disease mortality in 13 331 nationally representative adults aged 20 years or older as recorded in the Third National Health and Nutrition Examination Survey mortality files (Melamed et al. 2008) in which the participants' vitamin D levels were collected from 1988 to 1994 and the individuals were followed up passively for mortality until 2000 found the lowest quartile in terms of the 25(OH)D level (<45 nmol/L) to be independently associated with all-cause mortality in the general population. According to a more recent review, the all-cause mortality risk in the general population seems to be lowest at 25(OH)D levels ranging from 75 to 113 nmol/L (Pludowski et al. 2013).

5.3 Vitamin D and antimicrobial peptides in innate immunity

1,25(OH)2D has been shown to inhibit cancer cell growth, induce cancer cell maturation, induce apoptosis and reduce angiogenesis and to have an immunomodulatory activity on monocytes and activated T and B lymphocytes (Holick & Chen 2008, Pludowski et al. 2013). It also has an antiproliferative effect and downregulates inflammatory markers. Extrarenal synthesis of 1,25(OH)2D, occuring under the influence of cytokines, is important for the paracrine regulation of cell differentiation and function and this may explain why vitamin D deficiency can play a role in the pathogenesis of autoimmune diseases and cancer (Lips 2006).
Our skin is constantly being challenged by microbes, but is rarely infected. Epidermal production of antimicrobial peptides (AMPs) is the primary protective system, and the expression of some AMPs increases further in response to microbial invasion (Schauber & Gallo 2008). AMPs act to form a chemical shield on the surface of the skin, and they are also thought to trigger and coordinate multiple components of the innate and adaptive immune systems. Many cell types that permanently reside in the skin produce AMPs, including keratinocytes, sebocytes, eccrine glands and mast cells while circulating cells recruited to the skin, such as neutrophils and natural killer cells, are also significant contributors to the total amount of AMPs present. Cathelicidins and (3-defensins are the best characterized of the AMPs found in the skin, but a list of the known cutaneous AMPs will identify more than 20 individual proteins that have shown antimicrobial activity (Schauber & Gallo 2008).
Cathelicidins are unique AMPs that protect the skin through both direct antimicrobial activity and the initiation of a host response resulting in cytokine release, inflammation, angiogenesis and re-epithelialization (Schauber & Gallo 2008). Human cathelicidin is often referred to by one of its peptide forms (LL-37) or by the nomenclature assigned to its precursor protein (hCAP18). Cathelicidin dysfunction emerges as a central factor in the pathogenesis of several cutaneous diseases, including atopic dermatitis, in which cathelicidin is suppressed, rosacea, in which cathelicidin peptides are abnormally processed to forms that induce inflammation, and psoriasis, in which a cathelicidin peptide converts self-DNA to a potent stimulus in an autoinflammatory cascade (Schauber & Gallo 2008).
Cathelicidin expression in the skin is known to be induced by 1,25(OH)2D, which is a potent immunomodulator (Pludowski et al. 2013). When a macrophage or monocyte is stimulated by an infectious agent through its toll-like receptor, the signal up-regulates the expression of the vitamin D receptor and 1a-hydroxylase (Holick 2007). A 25(OH)D level of 75 nmol/L or higher provides an adequate substrate for 1a-hydroxylase to convert 25(OH)D to 1,25(OH)2D which then travels to the nucleus, where it increases the expression of cathelicidin, a peptide capable of promoting innate immunity and inducing the destruction of infectious agents (Holick 2007).
UVB irradiation of human keratinocytes supplemented with 7-dehydrocholesterol has been shown to trigger the synthesis of hormonally active 1,25(OH)2D, which then affects expression of the AMPs cathelicidin and human (3-defensin 2 (Schauber & Gallo 2008, Peric et al. 2010). Recent work has demonstrated direct upregulation of AMPs in healthy skin by cutaneous UV exposure (Felton et al. 2013). The induction of antimicrobial peptides by UV irradiation appears to have been one of the first indications that UV irradiation may foster the innate immune response (Glaser et al. 2009). Since UV exposure can result in a disruption of the epidermal barrier, and may thus increase the risk of bacterial infections, the induction of antimicrobial peptides as a counterregulatory phenomenon may be beneficial. Suppression of the adaptive immune system by UV irradiation plays an important role in photocarcinogenesis and inhibits the induction of contact hypersensitivity (Schwartz 2010). Thus UV radiation exerts diverse effects on the immune system, suppressing the adaptive immune response but inducing the innate immune response. This seems to explain why T-cell-mediated immune reactions are suppressed upon UV exposure but host defence reactions to bacterial attacks are not (Glaser et al. 2009). Schwartz et al. (2012), in their recent investigation into whether 1,25(OH)2D might be involved in UV-induced immunosuppression, were able to show that Langerhans cells are required for the induction of regulatory T-cells by 1,25(OH)2D. This indicates that 1,25(OH)2D exerts similar immunosuppressive effects to UV but is dispensable for local UV-induced immunosuppression, as vitamin D receptor knockout mice were equally as susceptible to UV-induced immunosupppression as wild-type controls.

5.4 Ultraviolet B treatment for dermatological diseases

It has been known for more than 2000 years that several skin diseases improve upon exposure to the sun (heliotherapy). Systematic investigations into phototherapeutic modalities started at the beginning of the twentieth century (Berneburg et al. 2005) and in the 1920s Goeckerman showed the beneficial effect of natural sunlight in combination with tar for treating psoriasis. Broadband ultraviolet B (BB-UVB) therapy involves a wide spectrum of UVB wavelengths (280-320 nm) and has been available for more than 80 years for the treatment of psoriasis (Patel et al. 2009). In 1953, Ingram initiated a combination of UVB radiation, dithranol and tar bathing for psoriasis and later, Fischer and Alsins (1974) and Parrish and Jaenicke (1981) showed that wavelengths around 311 nm provoke the least erythema while being most effective for clearing psoriasis, as reviewed by Berneburg et al. (2005). As a consequence, a fluorescent bulb was developed (TL-01) that emitted a major peak at 311 (± 2 nm) and a minor peak at 305 nm (Karvonen et al. 1989) and this treatment, which later came to be called narrowband ultraviolet B (NB-UVB) was shown to be superior to broadband phototherapy for treating a number of dermatoses. Furthermore, NB-UVB was found to be 5 to 10 times less potent than BB-UVB in inducing erythema, hyperplasia, oedema and Langerhans cell depletion in the skin (Patel et al. 2009).
At present NB-UVB is a standard dermatological therapy, especially for psoriasis (Bandow & Koo 2004). It is mostly given three times a week (Dawe et al. 1998), usually for a total of approximately 20-25 exposures (Green et al. 1988, Bandow & Koo 2004). NB-UVB is also effective in the treatment of atopic eczema, polymorphic light eruption, early stage mycosis fungoides, pruritus, graft-versus-host disease and several other inflammatory dermatoses (Berneburg et al. 2005, Hearn et al. 2008). It suppresses the IFN-y and IL-17 signalling pathways in order to resolve psoriatic inflammation (Johnson-Huang et al. 2010, Racz et al. 2011) and its antipsoriatic effect may also, at least in part, be based on the antiproliferative and prodifferentiative action of newly synthesized 1,25(OH)2D on epidermal keratinocytes (Lehmann et al. 2007). NB-UVB is known to reduce systemic immune responsiveness via the induction of regulatory T cells (Milliken et al. 2012).
The UVB doses given to patients in phototherapy cabins are calculated in physical units J/m2 or J/cm2, but the physical dose can be weighted for the presence of erythema when measuring the UVB dose received by the skin. The minimal erythema dose (MED), which is an individual measure, is the smallest dose to cause erythema with well-defined borders in the phototest area 24 hours after irradiation of the skin while Diffey et al. (1997) proposed the term standard erythema dose (SED) when referring to erythema as a specific biological response. 1 SED is equivalent to an effective erythemal radiant exposure of 100 J/m2 and 10 mJ/cm2 (Commission Internationale de l'Eclairage (CIE) 1999). The MED in subjects with skin types I to IV lies between erythemal effective radiant exposures of 150-600 J/m2 CIE, which is equivalent to 1.5-6 SED. The ambient diurnal exposure on a clear summer's day in Europe is approximately 30-40 SED (Diffey et al. 1997). An action spectrum for vitamin D production in the skin has also been defined (Commission Internationale de l'Eclairage (CIE) 2006), but this has been criticised.
The early side effects ofNB-UVB include erythema and dryness ofthe skin. Erythema reaches its maximum 8-24 hours after irradiation (Berneburg et al. 2005). One notable risk after chronic UVB exposure is that of skin tumour induction, but phototherapy with NB-UVB seems to be a relatively safe treatment modality (Black & Gavin 2006). A German study of 195 psoriasis patients did not provide evidence for any increased skin cancer risk in patients treated with either BB-UVB or NB-UVB (Weischer et al. 2004). On the other hand, a small but significant increase in basal cell carcinomas was detected in 1908 Scottish patients treated with NB-UVB (Man et al. 2005). A larger British retrospective study that included 3867 NB-UVB treated patients (Hearn et al. 2008) failed to find any significant association between NB-UVB treatment and basal cell carcinomas, squamous cell carcinomas or melanomas, but one should be cautious about interpreting these results as the cohort contained relatively few patients who had a high number of treatments and there were also some patients who were followed up for only six months. A longer follow-up would be essential in order to determine the true carcinogenic risk attached to NB-UVB phototherapy (Hearn et al. 2008, Patel et al. 2009, Archier et al. 2012).

5.5 Dialysis treatment and vitamin D insufficiency in chronic kidney diseases

Chronic kidney disease (CKD) is divided into five different stages according to the glomerular filtration rate (GFR) (Table 1; k/DOQI 2003). According to Report 2011 of the Finnish Registry for Kidney Diseases, the prevalence of renal replacement therapy, a term that includes both dialysis and kidney transplantation, is increasing, so that there were 2552 kidney transplantation patients and 1774 dialysis patients in Finland by the end of the year in question. Approximately 75% of renal replacement therapy patients recieve in-centre haemodialysis as their first modality, while 25% have peritoneal dialysis and 2% home haemodialysis (Finnish Registry of Kidney Diseases 2011).
Table 1. The five stages of chronic kidney disease (CKD) (k/DOQI 2003).
Uraemic patients may suffer from fatigue, nausea, loss of appetite and/or itching, but nowadays dialysis is most often started before the patient develops uraemic symptoms, at the latest when the GFR has decreased to 5-10 mL/min/1.73m2 (Korevaar et al. 2003). Conventional haemodialysis usually consists of a session of four to five hours three times a week. The most common cause of dialysis in Finland is type 2 diabetes (Finnish Registry of Kidney Diseases 2011), while the incidence of dialysis due to type 1 diabetes has not increased in recent years. Other common causes of dialysis are age-related ischaemic kidney disease, chronic glomerulonephritis, polycystic kidney disease and chronic interstitial nephritis.
Vitamin D insufficiency (25(OH)D <75nmol/L) is very common in CKD patients, affecting approximately 50 to 80%, in both pre-dialysis and dialysis populations (LaClair et al. 2005, Blair et al. 2008, Mehrotra et al. 2008, Clayton & Singer 2009, Bhan et al. 2010, Nigwekar et al. 2012). GFR is associated with vitamin D metabolism (Karhapaa et al. 2012). In advanced kidney disease the kidney is unable to produce 1,25(OH)2D from 25(OH)D due to the loss of renal 1a-hydroxylase (CYP27B1) activity and hyperphosphataemia (Pitts et al. 1988, Holick 2007, Nigwekar et al. 2012). The Kidney Disease Outcomes Quality Initiative guidelines published by the National Kidney Foundation have recommended that CKD patients should have a serum 25(OH)D concentration above 75 nmol/L (k/DOQI 2003). There is an ongoing debate about the definition of a sufficient vitamin D level for such this patients, however (K-DIGO 2009).
Many patients taking an active vitamin D analogue still do not have adequate vitamin D levels (Holick 2007). Levels of 25(OH)D are inversely associated with PTH levels regardless ofthe degree ofchronic renal failure. Patients with Stage 4 or 5 CKD and those requiring dialysis are unable to produce sufficient 1,25(OH)2D, and administration of 1,25(OH)2D or one of its less calcaemic analogues is often recommended to maintain calcium metabolism and to reduce PTH levels and the risk of renal bone disease. 25(OH)D levels below 37 nmol/L are associated with a greater severity of secondary hyperparathyroidism even in CKD patients on dialysis (k/DOQI 2003), while secondary hyperparathyreoidism and changes in mineral metabolism develop early in the course of CKD and worsen with its progression. The major factors responsible for stimulating parathyroid gland function are hypocalcaemia, hyperphosphataemia, increased fibroblast growth factor 23 and diminished 1,25(OH)2D levels (Holick 2007, Levin et al. 2007).
Vitamin D deficiency is associated with cardiovascular disease (de Boer et al. 2009), the most common cause of mortality in haemodialysis patients (Wolf et al. 2007). When the relationship between a low serum 25(OH)D level and death was determined among CKD patients in a large cohort of 3011 patients from the Third National Health and Nutrition Examination Survey who had CKD but were not on dialysis (Mehrotra et al. 2009), individuals with serum 25(OH)D levels less than 37 nmol/L had a higher risk of all-cause mortality than those with levels over 75 nmol/L. These results indicate there is a graded relationship between serum 25(OH)D and the risk of death among subjects with CKD. A meta-analysis of all-cause mortality in patients with CKD had similar results (Pilz et al. 2011), with higher 25(OH)D levels in patients with CKD being associated with significantly improved survival.
Both vitamin D deficiency and vitamin D toxicity are associated with cardiovascular complications of CKD. These associations are best illustrated by means of a biphasic, or U-shaped, curve (Querfeld & Mak 2010), but in clinical practice the therapeutic window is rather small, so that the avoidance of both vitamin D deficiency and toxicity presents a therapeutic challenge.
Quinibi et al. (2010) found that the recommended 25(OH)D level for CKD patients (>75 nmol/L; k/DOQI 2003) is difficult to achieve with a 6-month course of oral ergocalciferol supplementation. Only 57% of their haemodialysis patients achieved this with a supplementation of 500 jxg cholecalciferol weekly for 9 months (Tomak et al. 2008), although ergocalciferol supplementation of 1250 jxg weekly for 6 months was associated with significant improvements in serum 25(OH)D in CKD patients, from a baseline of 46 nmol/L to 105 nmol/L (Blair et al. 2008).
Treatment with vitamin D (ergocalciferol or cholecalciferol) to raise 25(OH)D levels was shown to increase not only serum 25(OH)D but also 1,25(OH)2D in both Stage 3 and Stage 4 CKD patients (Martin & Gonzalez 2001, Zisman et al. 2007). Thus, ergocalciferol or cholecalciferol should be used to correct vitamin D levels in patients with CKD either before or during active vitamin D therapy (Gal-Moscovici & Sprague 2010). In addition to the endocrine effects of the vitamin D axis on bone and mineral metabolism, there is also a certain amount of extrarenal conversion of 25(OH)D to 1,25(OH)2D in multiple cells leading to autocrine effects. Such local conversion has led to the speculation that CKD patients may also need ergocalciferol or cholecalciferol supplementation (Moorthi et al. 2011). Nutritional vitamin D may even be needed by patients with Stage 5 CKD (Melamed & Thadhani 2012).
Several papers have focussed attention on vitamin D supplementation in CKD patients (Kalantar-Zadeh & Kovesdy 2009, Gal-Moscovici & Sprague 2010, Matias et al. 2010), but although the available evidence shows that this supplementation improves biochemical end points, it has yet to be determined whether such improvements translate into clinically significant outcomes such as reduced cardiovascular mortality (Wolf et al. 2007, Kandula et al. 2011).

5.6 Effect of food fortification and oral supplementation on vitamin D balance

The occurrence of suboptimal vitamin D intake and vitamin D insufficiency in the general population led the authorities in many countries to take action to add vitamin D to various foods. Milk, some juice products, some breads, yogurts and cheeses are fortified with vitamin D in the United States (Holick & Chen 2008) and in Finland an expert meeting recommended the enhancement of milk, sour milk and yogurt with 0.5 fig/100 g of vitamin D and an increase in the fortification of margarine and cheese spreads from 7.5 fig to 10 fig/100 g. Thus one teaspoonful of margarine would yield 0.5 fig of vitamin D (Tylavsky et al. 2006). The Ministry of Social Affairs and Health agreed to these recommendations, and the fortifications took effect in February 2003. Even this fortification and the recommended daily dietary vitamin D intake of 7.5 fig, however, has proved inadequate to prevent vitamin D insufficiency among adolescent Finnish girls (Lehtonen-Veromaa et al. 2008) and two other Finnish studies have shown that the vitamin D in milk products brought only a slight and inadequate improvement in the poor vitamin D status of young Finnish men in winter (Laaksi et al. 2006, Valimaki et al. 2007).
To achieve the recently recommended serum 25(OH)D concentrations of above 75 nmol/L, a dietary intake of 17 to 20 fig vitamin D daily would be required (Bischoff-Ferrari et al. 2006, Holick 2007), and as much as 20 to 25 fig daily for older people (Dawson-Hughes et al. 2005, Viljakainen et al. 2006). This target is hard to reach without supplements, since for a mean daily intake of only 7.5 fig one needs to consume fish 2 to 3 times a week, for instance, together with drinking 6 dl of vitamin D fortified milk daily and eating 5 to 6 sandwiches containing vitaminized margarine.
Over-the-counter vitamin D supplements take the form of either vitamin D2, ergocalciferol, or D3, cholecalciferol (Holick 2007). Vitamin D2 is manufactured by the ultraviolet irradiation of ergosterol from yeast, and vitamin D3 by the ultraviolet irradiation of 7-dehydrocholesterol from lanolin. There has been much discussion about vitamin D2 being only 30-50% as effective as vitamin D3 in maintaining serum concentrations of 25(OH)D implying that vitamin D2 may need to be given in higher doses to raise the blood concentrations of 25(OH)D (Holick & Chen 2008).
At present the Finnish National Institute for Health and Welfare (National Institute for Health and Welfare 2011) recommends vitamin D supplementation as follows: 10 fig (400 international units, IU) daily for children below 2 years of age and for pregnant and breastfeeding women, 7.5 fg daily for children aged 2-18 years, 20 fg daily for over 60 years old and 10 fg daily from October to March for people aged 18-60 years if they do not regularly consume vitamin D fortified milk and eat fish.
It is well established that oral vitamin D supplementation markedly increases one's serum 25(OH)D concentration and two Finnish studies have shown this in elderly people. Viljakainen et al. (2006) observed that daily supplementation with 5, 10 and 20 fig of cholecalciferol increased serum 25(OH)D concentrations significantly and that an equilibrium was reached after 6 weeks while Kilpinen-Loisa et al. (2009) gave adults living in nursing homes 20 fig of oral vitamin D supplement for 6 months, and noted that 42% of them had a 25(OH)D concentration above 80 nmol/L. Previously Vieth et al. (2001) in Canada had given healthy adults 25 fig of oral cholecalciferol in winter and found that their serum 25(OH)D increased by 28.0 nmol/L while in Sweden Burgaz et al. (2007) observed that a vitamin D supplement intake in winter increased women's serum 25(OH)D concentration by 11.0 nmol/L and Bischoff-Ferrari et al. (2006) found that 100 fig vitamin D daily increased 25(OH)D concentrations in young men and women by 56 nmol/L to a mean of 125 nmol/L.
Acute vitamin D intoxication on account of ergocalciferol or cholecalciferol supplements is a rare event (Querfeld & Mak 2010). It is characterized by hypercalcaemia, hypercalciuria and nephrocalcinosis, and can be caused by inadvertent or intentional ingestion of excessively high doses. Doses of more than 1250 fig of vitamin D daily raise serum 25(OH)D over 375 nmol/L (Holick 2007). The actual serum 25(OH)D concentration that marks the threshold for vitamin D toxicity has not been established, however, although it is known that the lowest dose of vitamin D causing hypercalcaemia in some healthy adults is 1000 fig of ergocalciferol daily for many months (Vieth 2007). The Institute of Medicine has set the tolerated upper vitamin D intake level at 100 fig daily, defining this as "the highest level of daily nutrient intake that is likely to pose no risks of adverse health effects to almost all individuals in the general population" (Vieth 2007, Pludowski et al. 2013).

5.7 Effect of heliotherapy and UVB treatment on vitamin D balance

It has been shown during last years that heliotherapy and artificial UVB treatment have a positive effect on serum 25(OH)D levels. 25(OH)D is produced in the epidermis by UV radiation at wavelengths of 280-315 nm, and BB-UVB phototherapy in a similar range of 280-320 nm has been used successfully for decades to treat psoriasis. Now, in addition to standard BB-UVB, monochromatic UV 311 nm, i.e. NB-UVB, has become an important mode of treatment for psoriasis.
An experiment as conducted in Finland in which 23 patients with atopic dermatitis, 74% of whom had prior vitamin D insufficiency (25(OH)D <50 nmol/L), received a 2-week course ofheliotherapy in the Canary Islands (Vahavihu et al. 2008). The median personal UV dose received during the course was 60 SED in a group sent in January and 109 SED in a group sent in March. Serum 25(OH)D increased significantly in both groups, by 13.4 and 24.0 nmol/L, respectively. Thus only 17% of patients had vitamin D insufficiency after this 2-week course of heliotherapy, which significantly increased serum 25(OH)D concentration and caused a marked healing of the initial atopic dermatitis.
In a Swedish study 24 postmenopausal women with psoriasis were treated with BB-UVB two to three times per week for 8-12 weeks (Osmancevic et al. 2007), whereupon their levels of 25(OH)D increased from 92.0 to 149.0 nmol/L, serum PTH decreased, but the serum levels of 1,25(OH)2D, calcium, osteocalcin, thyroid hormones and creatinine were unaltered. The same team then compared the effect of BB-UVB on vitamin D synthesis in patients with psoriasis with that of NB-UVB therapy (Osmancevic et al. 2009) treating 26 patients with BB-UVB and 42 with NB-UVB two to three times per week for 8-12 weeks. Here 25(OH)D increased from 94.8 to 173.5 nmol/L and in the BB-UVB group and from 87.0 to 138.3 nmol/L in the NB-UVB group and PTH decreased in the BB-UVB group, but serum concentrations of 1,25(OH)2D, calcium and creatinine remained unaltered.
Most tanning beds emit 2 to 6% UVB radiation, and a group of tanners are reported to have had 25(OH)D levels of approximately 112 nmol/L at the end of the winter and higher bone density as compared with non-tanners with levels of approximately 45 nmol/L (Holick 2007). A Danish group investigated whether the use of tanning beds with sunlamps emitting mainly UVA and only 0.5% or 1.4% UVB would increase the level of serum 25(OH)D, giving healthy females sunbed radiation on eight days within a couple of weeks, and reported an average increases in serum 25(OH)D of 12 nmol/L in the 0.5% UVB group and 27 nmol/L in the 1.4% UVB group after four exposures (Thieden et al. 2008). In another sunbed study, repeated exposures to small doses from a commercial sunbed over five weeks raised the subjects' 25(OH)D levels from typical winter values to typical summer values, i.e. to 80 nmol/L. A mean increase of 15 nmol/L was seen after 2-4 weeks, followed by a decrease back to the pre-exposure level (Moan et al. 2009).
It has also been determined how different body sites respond to NB-UVB and affect serum 25(OH)D (Vahavihu et al. 2010). NB-UVB exposures were given on seven consecutive days either to the whole body, to the head and arms or to the abdomen. Similarly, seven solar simulator exposures were given to the face and arms. The cumulative UVB dose was 13 SED in all these regimens. Where 77% of the patients had baseline vitamin D insufficiency (25(OH)D <50 nmol/L) and 11% vitamin D deficiency (25(OH)D <25 nmol/L), their 25(OH)D concentration increased significantly, by a mean of 11.4 nmol/L, when the whole body was exposed to NB-UVB, by 11.0 nmol/L when only the head and arms were exposed and by 4.0 nmol/L when only the abdomen was exposed. After two months the serum 25(OH)D of the group who had received NB-UVB to the whole body was still higher than initially.
According to a Danish BB-UVB study, the increase in serum 25(OH)D is dependent mainly on the BB-UVB dose (Bogh et al. 2011b), although the area of body surface irradiated is also important at small UVB doses. When analysing each body surface area separately for any 25(OH)D increase, the team found a significant UVB response correlation for a 6% or 12% body surface area, but not for 24%. They also found a significant correlation of the body surface area response for a dose of 0.75 SED, but not for 1.5 or 3.0 SED. Factors other than the BB-UVB dose also affected the serum 25(OH)D concentration during UVB exposure. The increase in 25(OH)D level after UVB correlated negatively with the baseline 25(OH)D level and positively with the baseline total cholesterol level, but there were no significant correlations with constitutive or facultative skin pigmentation (Bogh et al. 2010). The increase in 25(OH)D after BB-UVB exposure is dependent on the dose but not on the timing of the dose (Bogh et al. 2011a), i.e. similar levels of 25(OH)D were achieved if the same total dose of BB-UVB was given over 1 minute or 20 minutes. It is also the case, however, that a significant increase in 25(OH)D can be achieved with a very low BB-UVB dose (four exposures of 0.375 SED). Sunscreens absorb UVB radiation, and vitamin D production is known to increase exponentially when thinner sunscreen layers than recommended (<2 mg/cm2) are applied (Faurschou et al. 2012).

6 AIMS OF THE RESEARCH

Vitamin D insufficiency is common, and this condition affects people in Finland especially during the winter. The main aims of the present thesis were to study the effect of NB-UVB phototherapy on vitamin D balance in dermatological patients in winter and to assess whether NB-UVB exposures improve the vitamin D balance of healthy subjects better than does oral vitamin D supplementation. Further aims were to study the effect of NB-UVB exposures on CKD patients receiving haemodialysis, since they are especially prone to vitamin D insufficiency, and to investigate the effects of NB-UVB on the enzymes hydroxylating vitamin D, antimicrobial peptides and cytokines in the skin.
The specific aims were:

1. to study the effects of NB-UVB phototherapy on the vitamin D balance of patients with psoriasis and atopic dermatitis (I),
2. to compare the effects of NB-UVB exposures and oral vitamin D supplementation on the vitamin D balance of healthy subjects in winter (II),
3. to study whether NB-UVB exposures improve the vitamin D balance of CKD patients receiving haemodialysis without (III) or with continuous oral cholecalciferol supplementation (IV), and
4. to examine whether NB-UVB exposures alter the levels of antimicrobial peptide and cytokine expression in the skin lesions of psoriasis and atopic dermatitis patients (I), and of the expression the enzymes hydroxylating vitamin D (CYP27A1, CYP27B1) and antimicrobial peptides in the skin of CKD patients receiving haemodialysis (III, IV).

7 MATERIALS AND METHODS

7.1 Patients and healthy subjects (I-IV)

Altogether 36 dermatological patients, 29 dialysis patients and 105 healthy subjects voluntarily participated in the four trials (I-IV) included in the present thesis (Table 2). The first (I) included 18 patients with psoriasis (mean age 46.9 years), 18 patients with atopic dermatitis (mean age 32.1 years) and 15 healthy hospital employees (mean age 41.8 years; Table 2).
In trial II, 99 healthy adult hospital employees and medical students were screened for 25(OH)D concentrations below 75 nmol/L and were randomly allocated to receive either NB-UVB or oral cholecalciferol (Table 2). The NB-UVB group consisted of 33 subjects (mean age 43.8, range 23-59 years) and the 20 fig oral cholecalciferol group of 30 subjects (mean age 40.2, range 20-62 years).
Trial III involved 15 Stage 5 CKD patients (mean age 48.3, range 33-65 years) receiving haemodialysis (Table 2), three of whom were taking synthetic vitamin D analogues (alphacalcidol in two cases and paricalcitol in one). Twelve healthy hospital employees (mean age 43.6, range 31-60 years) served as controls.
Fourteen Stage 5 CKD patients (mean age 53.6, range 34-69 years) receiving haemodialysis (Table 2) were enrolled for trial IV, in which seven of them received active vitamin D analogues (six received 5 fg intravenous paricalcitol twice weekly and one took 0.5 fig oral alphacalcidol daily). The patients had all been taking 20 fig oral cholecalciferol daily for a mean of 5.3 (range 1-16) months before the NB-UVB course. The 15 healthy hospital employees (mean age 46.1, range 19-62 years) who served as controls had similarly been taking 20 fig oral cholecalciferol daily before the NB-UVB course, for a mean of 3.4 (range 1-24) months.

Table 2. Demographic data on the dermatological and dialysis patients and healthy subjects participating in the four (I-IV) narrow-band UVB (NB-UVB) trials. Numbers of NB-UVB exposures, mean cumulative NB-UVB doses and oral vitamin D supplementation details are also given.
Inclusion criteria for the trials were Fitzpatrick skin type II-IV, indicating that the skin does not burn easily in the sun, no pregnancy, no history of skin cancer and no phototherapy, solarium visits, sun tanning or vitamin D supplementation within the two preceding months. The only exception to this was trial IV, where daily use of 20 fig oral cholecalciferol for at least one month were obligatory. The inclusion criteria also included age: over 18 years for trial I, 18-65 years for trials II and III and 18-70 years for trial IV.
Clinical improvement of dermatitis was measured using the psoriasis area and severity index (PASI; Fredriksson and Pettersson 1978) in the patients with psoriasis and by Severity Scoring of Atopic Dermatitis (SCORAD 1993) in those with atopic dermatitis. The indices were determined before and after the NB-UVB course in trial I.
The aetiology of the CKD in trials III and IV was glomerulonephritis in nine/eight patients, diabetic nephropathy in three/two, polycystic kidney disease in one/three, interstitial nephritis (in one/zero) and an unknown cause (in one/one). Four patients were the same in both of these trials. The patients had been on dialysis for a mean of 3.8 (range 0.5-16) years in trial III and 46.8 (range 9-117) months in trial IV. During the trials the patients received calcium carbonate (2 in trial III/14 in trial IV), non-calcium phosphate binder (7/12), cinacalcet (4/6) and active vitamin D analogues (paricalcitol 1/6 and alphacalcidol 1/1). The CKD patients in trial IV, but not in III, had been taking 20 fig oral cholecalciferol for a mean of 5.3 (range 1-16) months before the course of NB-UVB.
The protocols were approved by the ethics committee of Tampere University Hospital (code numbers R07149, R09186, R10112 and R11172) and all the subjects gave their informed consent. The authors followed the principles of the Declaration of Helsinki.
In trial I, 18 patients with psoriasis, 18 with atopic dermatitis and 15 healthy hospital employees, all volunteers, received a total of 15 NB-UVB exposures each, given three times a week, to the entire body area with a Waldmann UV 7001 cabin (Waldmann, Villingen-Schwenningen, Germany) equipped with 40 (not 20 as erroneously expressed in paper I) TL-01 tubes (Schulze & Bohm, Bruhl, Germany). The initial unweighted NB-UVB dose for all participants was 0.13 J/cm2, corresponding to one standard erythema dose (SED), which in turn is equivalent to 10 mJ/cm2 CIE (Commission Internationale de l'Eclairage) erythema-weighted irradiance. Thereafter the dose was gradually increased up to 1.19 J/cm2 (9.5 SED). The mean theoretical cumulative UVB dose after 15 NB-UVB exposures in the protocol was 8.88 J/cm2, which corresponds to 71.5 SED, but in practise the patients and healthy subjects received a somewhat lower cumulative UVB dose, from 62 to 66 SED (Table 2).

7.2 Methods

7.2.1 Narrow-band UVB exposures (I-IV)

The 67 healthy subjects in trial II had serum 25(OH)D concentrations below 75.0 nmol/L and were randomly allocated to receive NB-UVB or oral cholecalciferol. The NB-UVB group, comprising 33 subjects, received 12 exposures given three times a week over four weeks to the entire body area with a Waldmann UV 7001 cabin. The spectral irradiance of the cabin was measured with a stray light-corrected single-monochromator spectroradiometer. The initial physical NB-UVB dose was 0.21J/cm2 of CIE erythema-weighted irradiance (1.25 SED), and this was increased gradually according to a fixed protocol up to 1.45 J/cm2 CIE (8.55 SED). If the subject experienced itching or slight erythema the NB-UVB dose was either lowered or the same dose was repeated before any further increase. The mean cumulative amount of NB-UVB given to the subjects was 8.23 (range 6.14-9.34) J/cm2 CIE, which is equivalent to 48.4 (range 36.1-54.9) SED. The oral cholecalciferol group received 20 jxg (800 IU) daily for 4 weeks. This group consisted of 30 subjects because four subjects had to be excluded due to excessively high baseline serum 25(OH)D concentrations (>75 nmol/L).
The 15 dialysis patients and 12 healthy subjects participating in trial III received NB-UVB exposure with a Medisun 700 UVB-311 apparatus equipped with TL01 tubes three times a week. A total of nine NB-UVB exposures were given during three weeks, to the face, arms, chest and abdomen, accounting for approximately 25% of the total body area. The cumulative dose of NB-UVB was 15 SED. The exposure time ranged from 48 to 96 seconds, and the exposure for the patients took place just before their dialysis session.
The 14 dialysis patients and 15 healthy subjects in trial IV received nine NB-UVB exposures given three times a week for three weeks to the entire body area with a Waldmann UV 7001 cabin. Again these were given to the patients just prior to their dialysis. The first NB-UVB dose was 0.19 J/cm2 CIE (1.11 SED), and this was increased incrementally according to a fixed gradual protocol up to 0.97 J/cm2 (5.70 SED). The highest NB-UVB dose took approximately 90 seconds. If the subjects experienced mild itching or erythema, the dose was not increased or was reduced on the next occation. The mean cumulative dose of NB-UVB given to the CKD patients was 4.53 (range 2.97-4.75) J/cm2 CIE, which is equivalent to 26.6 (range 17.5-27.9) SED, while that given to the healthy subjects was 4.37 (range 3.21-4.75) J/cm2 CIE, which is equivalent to 25.7 (range 18.9-27.9) SED. The cumulative NB-UVB doses did not differ between the dialysis patients and healthy subjects (p = 0.46).

7.2.2 Measurement of serum 25(OH)D and 1,25(OH)2D concentrations and dietary intake of vitamin D (I-IV)

Blood samples were taken before, during and after the NB-UVB exposures, as described in papers I-IV. The samples were protected from light, centrifuged and the serum was frozen at -20°C. Serum 25(OH)D concentrations were measured as in duplicate by radioimmunoassay (I-IV), and serum 1,25(OH)2D concentrations were analysed in duplicate by immunoextraction followed by radioimmunoassay (III). The methods included the measurement of the hydroxylated metabolites of both vitamin D2 and D3. The dietary intake of vitamin D was assessed with a Food Frequency Questionnaire (II).

7.2.3 Skin biopsies and mRNA expression of enzymes hydroxylating vitamin D, antimicrobial peptides and cytokines (I, III, IV)

Punch biopsies of same representative skin lesions were taken from seven patients with psoriasis and eight patients with atopic dermatitis before treatment and after six NB-UVB exposures, with seven healthy subjects serving as controls. The samples were frozen immediately and stored at -70°C. Total messenger ribonucleic acid (mRNA) was extracted from the biopsies, and the transcript levels of two antimicrobial peptides, cathelicidin and human (3-defensin 2 (HBD2), and the mRNA expression levels of various cytokines in the psoriasis and atopic dermatitis lesions were analysed by the real-time quantitative polymerase chain reaction (PCR) technique. The methods are described in detail in paper I.
For trial III, punch biopsies were taken from the abdominal skin of 12 dialysis patients before the first and ninth NB-UVB exposure, with skin biopsies from 12 additional healthy subjects not treated with NB-UVB serving as controls. The biopsies were frozen immediately and stored at -70°C. Total RNA was isolated from them and 1g of RNA was reverse transcribed to complementary deoxyribonucleic acid (cDNA). The mRNA expressions of CYP24A1, CYP27B1 and cathelicidin were evaluated, and fold induction relative to that in healthy volunteers was calculated.
For trial IV, punch biopsies were taken from the skin of the buttocks of 10 dialysis patients and 13 healthy subjects before the first and ninth NB-UVB exposure. The mRNA expression levels of the enzymes CYP27A1 and CYP27B1 and of the antimicrobial peptide cathelicidin were evaluated.

7.2.4 Statistical analyses (I-IV)

The changes in serum 25(OH)D concentration in the three groups and in PASI and SCORAD where applicable were analysed in paper I by means of a Monte-Carlo p-value using the permutation test. Confidence intervals for the changes were obtained by bootstrapping (5000 replicates). Statistical analyses of cathelicidin and HBD2 expression in the psoriasis and atopic dermatitis skin lesions before and after NB-UVB treatment were compared with results for healthy controls using the Mann-Whitney test. Cathelicidin and HBD2 expression levels were compared between the untreated and NB-UVB treated skin lesions with the Wilcoxon matched pairs test, and cytokine expression in the psoriasis skin vs. the atopic dermatitis skin was analysed with the non-parametric Mann-Whitney test.
In paper II, 67 subjects with a serum 25(OH)D concentration below 75.0 nmol/L were randomly allocated to receive NB-UVB or oral cholecalciferol. The sample size needed to discover significant differences in 25(OH)D concentrations, i.e. 30 subjects per group, was calculated with a Power and Sample Size Calculations program. The results were expressed as means ± SD and 95% confidence intervals. Statistically significant differences between the groups were tested using the permutation test. Repeated measures were analysed using generalizing estimating equations models with an unstructured correlation structure. No adjustment was made for multiple testing.
The difference in serum 25(OH)D concentration between the dialysis patients and healthy subjects and the differences between the serum 25(OH)D and 1,25(OH)2D concentrations measured before and after the NB-UVB exposures were analysed in paper III with a Monte-Carlo p-value by means of the permutation test. Confidence intervals were obtained by bias-corrected bootstrapping (5000 replications). The differences in the mRNA expression levels of CYP24A1, CYP27B1 and cathelicidin between the dialysis patients and healthy subjects were analysed with an unpaired t-test and the differences between the levels before and after NB-UVB exposure with a paired t-test.
Statistical comparisons between the groups in paper IV were performed using the t-test, permutation test or Chi-square test. Repeated measures were analysed using generalizing estimating equations models with an unstructured correlation structure using bootstrap-type standard error. The changes within the group of dialysis patients were analysed by applying a t-test and a permutation test to the relevant samples.

8 RESULTS

The 25(OH)D baseline levels and responses to NB-UVB exposures in the four trials are presented in Table 3.
Table 3. Mean serum 25-hydroxyvitamin D (25(OH)D) concentrations at baseline and at the end of the narrow-band UVB (NB-UVB) exposures in the four trials included in the present thesis.
The mean 25(OH)D concentrations in trial I were low at baseline in the patients with atopic dermatitis (32.2 nmol/L) and psoriasis (36.8 nmol/L) but markedly higher (60.5 nmol/L) in the healthy subjects, who were doctors, nurses and other employees at our university hospital (Table 3). Overall, 89% of the psoriasis patients, 94% of the atopic dermatitis patients and 57% of the healthy subjects had vitamin D insufficiency (25(OH)D <50.0 nmol/L). The inclusion criterion for the healthy subjects in trial II was a serum 25(OH)D concentration below 75 nmol/L, after which they were allocated to two groups. The mean baseline serum 25(OH)D concentration was 52.9 nmol/L in the NB-UVB group and 53.5 nmol/L in the cholecalciferol group (Table 3).
As expected, the mean baseline 25(OH)D concentration for the dialysis patients in trial III was as low as 32.5 nmol/L (Table 3), with five of them (33%) having a concentration below 25.0 nmol/L and 14 (93%) below 50.0 nmol/L. The corresponding figure for the dialysis patients receiving a supplement of 20 fig oral cholecalciferol daily in trial IV was markedly higher, i.e. 57.6 nmol/L (Table 3), but the level was still below 50.0 nmol/L in six patients (43%) and between 50.0 nmol/L and 70.0 nmol/L in another six (43%). The mean baseline serum 25(OH)D concentration in the healthy subjects was 74.3 nmol/L in trial III and 60.2 nmol/L in trial IV (Table 3).

8.1 Effect of narrow-band UVB treatment on serum 25(OH)D concentrations in patients with psoriasis and atopic dermatitis (I)

NB-UVB exposures given three times a week for a total of 15 times markedly increased mean serum 25(OH)D concentrations by 68.2 nmol/L in the patients with atopic dermatitis, 59.9 nmol/L in the patients with psoriasis and 90.7 nmol/L in the healthy subjects (Table 3, Fig. 2). The clinical improvements in psoriasis and atopic dermatitis were statistically significant (PASI from 8.0 to 3.8. p <0.001; SCORAD from 37.1 to 14.1. p <0.001), but they did not correlate with the increases in serum 25(OH)D.

Figure 2. Effect of narrow-band ultraviolet B (NB-UVB) treatment on serum 25-hydroxyvitamin D (25(OH)D, calcidiol) concentration. Fifteen NB-UVB exposures significantly increased the serum 25(OH)D concentration (mean; 95% confidence intervals) in 18 patients with psoriasis, 18 patients with atopic dermatitis and 15 healthy subjects (p <0.001). The serum 25(OH)D level was still elevated in the patients and healthy subjects 1 month after the last NB-UVB exposure. (Original figure in paper I, Fig. 1)
The serum 25(OH)D concentration remained at an elevated level for at least one month in the patients with psoriasis and atopic dermatitis but showed some decrease in the healthy subjects (Fig. 2, Table 4).
Table 4. Mean serum 25-hydroxyvitamin D (25(OH)D) concentrations 1 month and 2 months after the narrow-band UVB (NB-UVB) exposures in the four trials included in the present thesis.

8.2 Comparison of the effects of narrow-band UVB and oral vitamin D supplementation on serum 25(OH)D concentrations in healthy subjects (II)

Twelve NB-UVB exposures increased the mean serum 25(OH)D by 41.0 nmol/L (95% CI 34.8-47.2, p <0.001; Table 3, Fig. 3), while oral cholecalciferol achieved an increase of 20.2 nmol/L (95% CI 14.6-26.0, p <0.001; Table 3, Fig. 3). The difference between the two treatments was therefore a significant one: 20.7 nmol/L (95% CI 12.2-29.2, p <0.001). In fact this was already evident after two weeks of treatment (p = 0.033). Altogether 28 (85%) subjects in the NB-UVB group and 13 (43%) subjects in the oral cholecalciferol group had an the active treatment regime.
Figure 3. Mean (95% CI) increase in serum 25-hydroxyvitamin D (25(OH)D) from the baseline after 12 narrow-band UVB (NB-UVB) exposures or daily oral cholecalciferol (20 pg) supplementation, both administered over 4 weeks, in healthy subjects. The response to NB-UVB is significantly higher than that to oral cholecalciferol at all time points. The 25(OH)D concentrations had continued to increase by one month after the treatments, and although they had decreased somewhat after 2 months, they were clearly higher than at the baseline. (Original figure in paper II, Fig. 1)
25(OH)D concentration above 75.0 nmol/L at the end of ne month after the treatments the 25(OH)D concentrations had increased further in both groups, being 97.2 ± 26.2 nmol/L and 84.9 ± 17.2 nmol/L, respectively (p <0.016; Table 4). Twenty-seven (84%) subjects in the NB-UVB group and 14 (56%) subjects in the oral cholecalciferol group had a 25(OH)D concentration above 75.0 nmol/L. Two months after the treatments the 25(OH)D concentrations had decreased in both groups, to 81.4 ± 21.0 nmol/L and 67.7 ± 17.2 nmol/L (p <0.003, Table 4), respectively, but these concentrations were still clearly higher than at the baseline. The daily dietary intake of vitamin D before the treatment had been 8.14 ± 3.02 ftg (mean ± SD) in the NB-UVB group and 8.17 ± 3.35 daily in the cholecalciferol group, while two months after the treatments the values were 7.90 ± 2.93 and 8.01 ± 3.28 ftg, respectively.

8.3 Effect of narrow-band UVB exposures on serum 25(OH)D concentrations in dialysis patients with (IV) or without (III) continuous oral cholecalciferol supplementation

The serum 25(OH)D concentrations of the dialysis patients had increased to 46.3 ± 12.0 nmol/L after eight NB-UVB exposures (Table 3, Fig. 4), a statistically significant change (p <0.001). After intervention, none of these patients had serum 25(OH)D below 25.0 nmol/L, but the concentration was still below 50.0 nmol/L in concentration from 60.2 ± 18.0 to 69.2 ± 17.7 nmol/L (p = 0.032; Table 3, Fig. 4).
Mean serum 25(OH)D levels in the dialysis patients at follow-up one and two months after NB-UVB exposure had decreased, but were still 10% higher than initially in paper III (Table 4), while those in the healthy subjects were still significantly increased one month after NB-UVB exposure but had decreased to close to the pre-treatment level after two months (Table 4). Plasma intact parathyroid hormone levels in the dialysis patients did not change during the NB-UVB exposures, and the levels of blood haemoglobin, plasma total calcium, serum ionized calcium and plasma phosphorus all remained unchanged.
The NB-UVB course increased serum 25(OH)D by 14.0 nmol/L (95% CI 8.7 to 19.5, p <0.001), or 24.2% (Table 3, Fig. 5). Only one dialysis patient (7%) had a serum 25(OH)D concentration below 50.0 nmol/L after treatment. The increase in 25(OH)D did not differ between the seven patients taking active vitamin D analogues (mean 13.2 nmol/L) and those not doing so (mean 15.0 nmol/L, p = 0.72). In the healthy subjects the NB-UVB course increased serum 25(OH)D by 17.0 nmol/L (CI 13.7 to 20.2, p <0.001), or 22.8% (Table 3, Fig. 5).

Figure 4. Serum 25-hydroxyvitamin D (25(OH)D) concentrations in 15 chronic kidney disease (CKD) patients on haemodialysis and 12 healthy subjects without oral cholecalciferol supplementation before (pre) and after (post) eight narrow-band ultraviolet B (NB-UVB) exposures. The increase was significant in both the CKD patients (p <0.001) and the healthy subjects (p = 0.032). (Original figure in paper III, Fig. 1) subjects before and after the narrowband ultraviolet B (NB-UVB) course. Both groups had been receiving oral cholecalciferol supplementation at a daily dose of 20 pg. The increase was significant (p <0.001) in both groups. (Original figure in paper IV, Fig. 1)
chronic kidney disease (CKD) patients on haemodialysis and 15 healthy

Figure 5. Serum 25-hydroxyvitamin D (25(OH)D) concentrations in 14
As reported in paper IV, follow-up serum 25(OH)D levels one and two months after the NB-UVB course (Table 4) were still significantly higher than at the baseline in both the dialysis patients (mean 69.8 ± 18.1 and 64.5 ± 22.3 nmol/L; p <0.001 and p = 0.031) and the healthy subjects (88.3 ± 19.9 and 91.8 ± 19.7 nmol/L; p = 0.002 and p <0.001).
Excluding the serum 25(OH)D levels, NB-UVB treatment had only marginal effects on the laboratory findings in the case of the dialysis patients (IV). Serum ionized calcium decreased, whereas intact parathyroid hormone, phosphorous and haemoglobin did not show any significant changes.

Cathelicidin and human $-defensin expression (I, III)

Analysis of the mRNA expression levels of cathelicidin before NB-U VB treatment showed significantly elevated levels in psoriasis lesions as compared with the normal skin of healthy subjects (I: Fig. 2a), while those in the atopic dermatitis lesions were also increased, but the difference was not significant (I: Fig. 3a). After six NB-UVB treatments cathelicidin expression had increased markedly in the psoriasis skin lesions and slightly in the atopic dermatitis ones, but the increases did not reach statistical significance. HBD2 mRNA expression levels were also significantly higher in the psoriasis and atopic dermatitis lesions than in the healthy control skin before NB-UVB

8.4 Effect of narrow-band UVB exposures on cutaneous antimicrobial peptides, cytokines and vitamin D hydroxylating enzymes (I, III, IV)

treatment (I: Fig. 2b and Fig. 3b), and this treatment significantly reduced the levels in both types of lesions, although the change did not reach statistical significance.
The baseline cathelicidin mRNA expression in the dialysis patients in trial III was somewhat lower than in the healthy subjects, and the nine NB-UVB exposures further reduced it to a significant extent (III: Fig. 4).

Cytokine and chemokine expression in psoriasis and atopic dermatitis (I)

The psoriasis lesions expressed significantly higher amounts of interleukin (IL)-1p, IL-17A and interferon (IFN)-g before NB-UVB treatment than did the atopic dermatitis lesions (I: Fig. 4), whereas CCL17 (thymus and activation regulated chemokine, TARC) was significantly higher in the atopic dermatitis lesions and IL-10, transforming growth factor (TGF)-b1, tumour necrosis factor (TNF)-a and CCL20 (macrophage inflammatory protein-3a, MIP-3a) were expressed similarly in both diseases. No IL-4 expression was found in the psoriasis and atopic dermatitis lesions. The expression of IL-1P and IL-17A was markedly, but not significantly, reduced in the psoriasis lesions after six NB-UVB treatments (I: Fig. 4), but no changes were observed in the expression of IL-4, IL-10, IFN-g, TGF-P1, TNF-a, CCL17 or CCL20.

CYP27A1 and CYP27B1 expression in dialysis patients (III, IV)

The normal skin of the dialysis patients in trial IV showed a significantly lower baseline level of CYP27A1 mRNA expression than that of the healthy subjects, whereas their expression of CYP27B1 mRNA was somewhat higher in trial III and significantly higher in trial IV when these patients received oral cholecalciferol supplementation. The nine NB-UVB exposures caused a significant decrease in the CYP27A1 and CYP27B1 mRNA expression levels of the dialysis patients (IV: Fig. 3A and B-) , and a similar result was found in the healthy subjects, but they did not cause any change in CYP24A1 mRNA expression level in the dialysis patients without cholecalciferol supplementation (III: Fig. 3B), showing that there was no major degradation of 25(OH)D and 1,25(OH)2D. CYP24A1 was erroneously interpreted as a synthesizing enzyme in the original paper (III).

9 DISCUSSION

9.1 Narrow-band UVB treatment increases serum 25(OH)D concentrations and affects cutaneous antimicrobial peptides and cytokines in patients with psoriasis and atopic dermatitis (I)

The mean serum 25(OH)D increase during the course of NB-UVB treatment was highest in the patients with atopic dermatitis (212%) and psoriasis (163%; Table 3), probably due to the fact that these patients received more NB-UVB exposures and larger cumulative NB-UVB doses than the others considered in the present thesis (Table 2). It should be noted, however, that these cumulative doses were somewhat lower than those we usually give to patients with psoriasis in our department during a course of 15 NB-UVB exposures, only 84% of the total dose, and that the increase in 25(OH)D after UVB exposure is dependent on the dose given (Bogh et al. 2011a). Moreover, the patients with atopic dermatitis and psoriasis had very low mean serum 25(OH)D concentrations, and the increase is known to be greatest when baseline levels are low (Bogh et al. 2010, Roman! et al. 2012). An analogous situation has been noted with oral vitamin D supplementation, in that subjects with a lower baseline 25(OH)D concentration were found to respond more than those with a higher baseline concentration (Viljakainen et al. 2006).
Several studies have been published in recent years on the effect of NB-UVB treatment on serum 25(OH)D concentrations in psoriasis patients in winter (Osmancevic et al. 2009, Ryan et al. 2010, Lesiak et al. 2011, Roman! et al. 2012), and although the total dose of NB-UVB given has varied greatly, the increases in serum 25(OH)D concentration have been significant in all these reports, as also in the present instance (I). In a more recent examination of how 12 psoriasis patients receiving 20 jxg supplementary oral cholecalciferol daily responded to a course of NB-UVB treatment (Ala-Houhala et al. 2013) the mean baseline serum 25(OH)D concentration, 74 nmol/L, was much higher than in trial I, when no supplementation was given, but in spite of this, serum 25(OH)D had increased by 13 nmol/L at the 9th NB-UVB exposure and by 49 nmol/L at the 18th exposure and the PASI scores improved as expected. It is also important to note that serum 25(OH)D concentrations remained far from the toxicity level in these psoriasis patients who received both vitamin D supplementation and NB-UVB treatment.
Analysis of the baseline mRNA expression levels of the antimicrobial peptides cathelicidin and HBD2 in paper I showed significantly elevated levels in the psoriasis lesions and elevated levels in atopic dermatitis relative to the normal skin of healthy subjects. These results are in agreement with earlier reports concerning psoriasis and atopic dermatitis (Gambichler et al. 2006, Hollox et al. 2008). It is of interest that cathelicidin and HBD2 can act as proinflammatory mediators, or "alarmins", and seem to have a role in the pathogenesis of skin inflammation in psoriasis and atopic dermatitis (Ong et al. 2002, Hollox et al. 2008). After six NB-UVB treatments cathelicidin expression had increased markedly in the psoriasis skin lesions and slightly in the atopic dermatitis lesions. In our more recent NB-UVB treatment trial, however (Ala-Houhala et al. 2013), where the psoriasis patients received supplementary oral cholecalciferol, we did not find any change in cathelidicin expression, while NB-UVB treatment significantly reduced HBD2 expression in healing psoriasis lesions both in the present patient series (I) and in our more recent one (Ala-Houhala et al. 2013) and slight decrease of HBD2 was found in atopic dermatitis lesions, as in previous reports (Gambichler et al. 2006, Ballardini et al. 2009, Peric et al. 2009). Alongside this decrease in HBD2
we saw a significant increase in serum 25(OH)D, the best indicator of the hormonally active form of vitamin D, i.e. 1,25(OH)2D. This exerts potent anti-inflammatory action through the inhibition of NfccB activation and inhibits IL-17A-induced HBD2 in keratinocytes in vitro (Peric et al. 2008). As NB-UVB is known to induce local synthesis of 25(OH)D and 1,25(OH)2D in keratinocytes (Lehmann et al. 2007), we propose that the effects of NB-UVB-triggered vitamin D production may outweigh the HBD2-inducing effect of short-duration UVB in our patient series.
We were also able to show that NB-UVB treatment reduced the expression of IL-1(3 and IL-17A in psoriatic skin lesions. IL-1(3 is a proinflammatory cytokine and is thought to be a critical mediator of the differentiation of human T cells that produce IL-17 (McKenzie et al. 2006, Wilson et al. 2007). In turn, IL-17 has been shown to enhance vitamin D-induced expression of the antimicrobial peptide cathelicidin (Peric et al. 2008). Although the expression of cathelicidin did not immediately decrease along with IL-17A in our material (I), lowered IL-17 levels may later reduce inflammation and neutrophil activation at lesional skin sites (Nickoloff 2007, Di Cesare et al. 2009). Our results suggest that the interplay between cytokines and vitamin D in the regulation of antimicrobial peptides in keratinocytes is complex and is still only partly understood, so that further research would be warranted.

9.2 Comparison of the effects of narrow-band UVB course and oral vitamin D supplementation on serum 25(OH)D concentrations in healthy subjects (II)

Paper II contains a comparison of the effects of NB-UVB exposures and oral vitamin D supplementation on healthy adult hospital employees in winter. Subjects with 25(OH) D below 75 nmol/L were randomly given either a course of 12 NB-UVB exposures or 20 fig of oral cholecalciferol daily for 4 weeks. The results showed that the NB-UVB exposures increased the mean serum 25(OH)D concentration by 41 nmol/L (Table 2) and oral cholecalciferol by 20 nmol/L, a significant difference. One month after the treatments the 25(OH)D concentrations in both groups were still higher than the baseline values.
Soon after our paper was published, a Swedish group reported similar results (Bogh et al. 2012a). They had given total body NB-UVB exposures three times a week for six weeks and were also able to show that NB-UVB treatment was more effective than a daily oral intake of 40 fig cholecalciferol, as mean serum 25(OH)D increased from 19 to 75 nmol/L in the NB-UVB group and from 23 to 61 nmol/L in the oral cholecalciferol group. They treated 73 participants with vitamin D deficiency (25(OH)D <25 nmol/L), but only 32 completed the treatment. Moreover, the participants were not described in any detail.
The main differences between the Swedish report and the present paper II concerned the subjects, their baseline serum 25(OH)D concentrations (<25 nmol/L vs. <75 nmol/L) and the oral cholecalciferol dose (40 fig vs. 20 fig daily). The total cumulative NB-UVB dose was similar in both cases, however, 9 J/cm2 vs. 8 J/cm2. It should be noted that the NB-UVB doses used in both trials were smaller than those conventionally used when treating psoriasis patients. The findings reported in these two papers, that 12-18 whole-body NB-UVB exposures given over 4-6 weeks were more efficient in both treating vitamin D deficiency and improving the vitamin D status of healthy subjects in winter than a daily oral vitamin D intake of 20-40 fig, were noted in a British Journal of Dermatology editorial (Diffey 2012), where it was also pointed out that it is controversial as to whether someone exhibiting a "normal" 25(OH)D concentration, i.e. from 50 to 75 nmol/L, should be subjected to a medical intervention by NB-UVB treatment.

9.3 Narrow-band UVB exposures increase serum 25(OH)D concentrations in dialysis patients and affect cutaneous enzymes hydroxylating vitamin D (III, IV)

In advanced kidney disease the kidney is unable to produce 1,25(OH)2D from 25(OH)D due to the loss of renal CYP27B1 activity (Pitts et al. 1988, Nigwekar et al. 2012). It has been shown in UVB-treated skin cultures that keratinocytes are able to hydroxylate 25(OH)D to 1,25(OH)2D (Lehmann et al. 2007). The finding that the CYP27B1 enzyme also exists outside the kidney is of interest with regard to oral vitamin D treatment for CKD patients (Melamed & Thadhani 2012).
In trial III, nine NB-UVB exposures given to dialysis patients increased serum 25(OH)D by 42% (Table 3) and serum 1,25(OH)2D by 27%. The NB-UVB exposures were given only to a body surface area of approximately 25%, whereas in the other trials (I, II, IV) whole-body exposures were given. If the exposed body area had been larger, e.g. the whole body, the serum 25(OH)D response might have been even more substantial. In the healthy subjects the NB-UVB course increased serum 25(OH)D by 15% (III; Table 3).
In trial IV the subjects were given an oral cholecalciferol supplement of 20 fig daily before, during and after the NB-UVB course, so that and the baseline serum 25(OH) D concentrations were clearly higher than in Study III (Table 2). In spite of this, nine NB-UVB whole-body exposures given to the dialysis patients increased their serum 25(OH)D significantly, i.e. by 24% (Table 3), in spite of the fact that they had been taking 20 fig oral cholecalciferol daily for a mean of 5 months beforehand. The NB-UVB course also increased serum 25(OH)D in the healthy subjects, by 23% (Table 3).
It should be noted that though the NB-UVB courses given to the dialysis patients in trials III and IV were quite short, the increase in serum 25(OH)D was significant. NB-UVB exposures have been used to relieve uraemic pruritus (Ko et al. 2011), but to our knowledge, these present trials are the first to show that NB-UVB exposures can also be used to improve the vitamin D balance in CKD patients undergoing dialysis.
Although the mean serum 25(OH)D concentration in the dialysis patients in trial III had decreased one month after the NB-UVB exposures, it was still 10% higher than at the baseline (Table 3). This kind of rapid decrease in serum 25(OH)D was not seen in the healthy subjects in either the present trial or in the psoriasis and atopic dermatitis patients and healthy subjects in our previous ones (I, II). Similarly, 25(OH)D concentrations began to decrease in the dialysis patients in trial IV during the follow-up of two months, in contrast to the situation in the NB-UVB-treated healthy subjects. The more profound decrease in 25(OH)D may be due to the higher body mass index (BMI) of the dialysis patients than in the healthy subjects, i.e. linked to active metabolism of vitamin D precursors in the fat tissue (Lehmann & Meurer 2010, Forsythe et al. 2012).
Although our dialysis patient series were small and the NB-UVB courses given short, the relatively rapid decrease in serum 25(OH)D suggests that dialysis patients may need a longer course of NB-UVB exposures or cyclic NB-UVB exposures to maintain their increased serum 25(OH)D concentrations. It has been shown in healthy subjects that BB-UVB exposures every second week are sufficient to maintain summer 25(OH)D levels during the winter (Bogh et al. 2012b). It would be worth trying continuous NB-UVB exposures also of this kind with dialysis patients in order to maintain their increased 25(OH)D concentrations.
Since as few as nine NB-UVB exposures were shown to cause significant increases in serum 25(OH)D concentrations in dialysis patients, it was of interest to study the effect of NB-UVB exposures on the enzymes hydrolyzing vitamin D in the skin. mRNA expression of the CYP27B1 enzyme was already elevated at the baseline, not significantly in trial III but significantly in trial IV, in which the dialysis patients had received cholecalciferol supplementation. This important observation suggests that the loss of renal CYP27B1 activity in the dialysis patients leads to activation of this enzyme in the skin, and possibly also in other organs, this activity in the skin being further intensified by the NB-UVB exposures in our patients. Interestingly, the NB-UVB exposures themselves caused a significant decrease in the mRNA expression of CYP27B1 and CYP27A1 in both the dialysis patients and the healthy subjects in trial IV, where they all received 20 fg of supplementary oral cholecalciferol. Such a decrease can be expected, because there is a very sensitive natural feedback controlling mechanism caused by the UVB-induced increase in cutaneous vitamin D synthesis (Holick 2007, Schauber et al. 2007, Lehmann & Meurer 2010).

10 CONCLUSIONS AND FUTURE PROSPECTS

The conclusions to be drawn from the present trials concerning the effects of NB-UVB treatment on the vitamin D balance of dermatological and dialysis patients and healthy subjects performed at Tampere University Hospital in 2008-2012 and the future prospects for research in this field may be summarized as follows:

Narrow-band UVB treatment increases serum 25(OH)D concentrations and affects cutaneous antimicrobial peptides and cytokines in patients with psoriasis and atopic dermatitis (I)

NB-UVB treatment is widely used for psoriasis and atopic dermatitis, but its effects on patients' vitamin D balance, i.e. serum 25(OH)D concentrations, had previously been examined only occasionally. In the present series 89% of the patients with psoriasis, 94% of those with atopic dermatitis and 53% of the healthy subjects were found to have baseline vitamin D insufficiency (serum 25(OH)D <50 nmol/L). This finding is in agreement with recent information on the frequent occurrence of vitamin D insufficiency in the general population (Holick 2007, Hypponen & Power 2007). Fifteen whole-body NB-UVB exposures significantly increased serum 25(OH)D (p <0.001), by 59.9 nmol/L in the psoriasis cases, 68.2 nmol/L in atopic dermatitis and 90.7 nmol/L in the healthy subjects. It was concluded that, in addition to a significant improvement in the status of psoriasis and atopic dermatitis, NB-UVB treatment effectively corrects vitamin D insufficiency. This finding is in agreement with several recent studies of psoriasis in winter (Osmancevic et al. 2009, Ryan et al. 2010, Lesiak et al. 2011, Roman! et al. 2012, Ala-Houhala et al. 2013).
It would be of interest in the future to study systemically how much the serum 25(OH)D concentration would increase, e.g. after the 25-30 NB-UVB exposures which are often needed for complete clearance of moderate or severe psoriasis. The fact that the PASI score denoting the status of psoriasis, did not show any correlation with the increase in serum 25(OH)D as reported here and previously (Roman! et al. 2012), suggests that vitamin D balance is not directly linked to an improvement in psoriasis, but further investigations, especially with regard to the skin lesions themselves, would be needed to confirm or exclude this possibility.
Antimicrobial peptides such as cathelicidin and HBD2 seem to have a role in the pathogenesis of skin inflammation in psoriasis (Schauber & Gallo 2008), and as expected, we found increased expression of these in the untreated psoriasis lesions and we also able to show that NB-UVB exposure further increased cathelidicin and reduced HBD2 in healing lesions. Similar effects of NB-UVB were not seen in our more recent trial (Ala-Houhala et al. 2013), in which the psoriasis patients received supplementary oral vitamin D. Whether these different effects of NB-UVB could be mediated by the improved vitamin D balance in the psoriasis lesions should be examined in more detail.

Comparison of the effects of narrow-band UVB exposures and oral vitamin D supplementation on serum 25(OH)D concentrations in healthy subjects (II)

Since there had been no direct comparisons on the effects of NB-UVB and oral vitamin D supplementation on serum 25(OH)D concentrations, we assigned healthy adult subjects with serum 25(OH)D below 75 nmol/L to receive randomly either a course of 12 whole-body NB-UVB exposures or 20 fig of oral cholecalciferol daily for 4 weeks. The increase in serum 25(OH)D was significantly greater in the NB-UVB group than in the cholecalciferol group (mean 41.0 nmol/L vs. 20.2 nmol/L). We concluded that a short course of low-dose NB-UVB is an effective way of improving vitamin D balance in winter, and toted that the response was still in evidence after 2 months. Soon after this trial, a Swedish group published similar results concerning subjects who had an initial serum 25(OH)D concentration below 25 nmol/L (Bogh 2012a). It would be of interest in the future to compare NB-UVB exposures with higher oral cholecalciferol doses than 40 fg daily. In addition, it would be of interest to know how high serum 25(OH)D concentrations would rise if NB-UVB courses longer than 4-6 weeks were given. This is important, because it is known that exposure of the skin to UVB triggers a feedback mechanism that controls vitamin D synthesis in such a way that an overdose is not possible (Holick 2007).

Narrow-band UVB exposures increase serum 25(OH)D concentrations and affect cutaneous enzymes hydroxylating vitamin D in dialysis patients (III, IV)

Most chronic kidney patients on dialysis are known to have vitamin D insufficiency because the loss of renal CYP27B1 activity means that the kidney is unable to convert 25(OH)D to 1,25(OH)2D (LaClair et al. 2005, Bhan et al. 2010, Nigwekar et al. 2012). In agreement with this, we found low serum 25(OH)D concentrations in our dialysis patients who did not receive supplementary oral cholecalciferol, 32.5 nmol/L, and in those who received such supplements, 57.6 nmol/L. Eight NB-UVB exposures significantly increased serum 25(OH)D by 43%, in our first dialysis trial (III) and by 24% in the second (IV), and serum 1,25(OH)2D also increased significantly, by 27% (III). To our knowledge, these trials are the first to show that NB-UVB exposures can be used to improve the vitamin D balance of dialysis patients.
Although the serum 25(OH)D concentrations had started to decrease one and two months after the NB-UVB exposures, they were still higher than initially (Table 4). Thus although our dialysis patient series were small and the NB-UVB courses were short, the relatively rapid decrease in serum 25(OH)D suggests that dialysis patients would need a longer course of NB-UVB treatment or cyclic NB-UVB exposures to maintain their increased serum 25(OH)D concentrations. It would also be worthwhile trying to give dialysis patients regular NB-UVB exposures in order to maintain higher 25(OH)D concentrations.
We also examined the mRNA expression of the enzymes CYP27A1 and CYP27B1 that hydroxylate vitamin D in skin biopsy samples taken from dialysis patients before and after a course of eight NB-UVB exposures. In first dialysis trial (III) the NB-UVB exposures significantly increased the mRNA expression of CYP27B1, whereas in the second (IV), baseline CYP27B1 was already higher. The increased cutaneous CYP27B1 levels in the dialysis patients suggest that loss of the renal activity of this enzyme is at least partially compensated for by the skin, but further trials with more NB-UVB exposures would be needed to confirm this observation.
In conclusion, NB-UVB exposures were shown to be an efficient way of increasing serum 25(OH)D concentrations in dermatological and dialysis patients, and in healthy subjects in winter. A short NB-UVB course was shown to increase serum 25(OH)D in healthy subjects significantly more than does daily supplementation with 20 fig oral cholecalciferol. NB-UVB exposures offer a new possibility for improving the vitamin D balance dialysis patients with vitamin D insufficiency. NB-UVB treatment is currently considered safe, but a longer follow-up would be needed.

ID	Name	Uploaded	Size	Downloads
3265	F5.jpg	admin 12 Nov, 2013	46.86 Kb	1115
3264	F4.jpg	admin 12 Nov, 2013	31.04 Kb	1155
3263	F2.jpg	admin 12 Nov, 2013	32.90 Kb	1280