18b Vitamin D

Vitamin D (calciferol) is a generic term for a group of related fat-soluble sub­stances, includ­ing vitamin D₂ (ergo­cal­ciferol), vitamin D₃ (chole­cal­ciferol) and their meta­bolites (Figure 18b.1).

Vitamin D₂ is derived from the plant sterol, ergo­sterol, and is the form used to fortify vegan foods and in some high-dosage pharma­ceut­ical prep­ar­ations. Vitamin D₃ originates from the action of ultra­violet B (UVB) rays from sunlight or artificial sources on its precursor sterol, 7-dehydrochol­esterol. This is present in the skin and some animal-based diet­ary sources, including some fish liver oils, the flesh of fatty fish, eggs, and organ meats. Foods may be fort­ified with vitamin D (e.g., milk, cereals, bread, margarine) in some countries (Cashman and O’Dea, 2019).

18b.1 Functions of vitamin D

The active form of vitamin D, 1,25‑dihydroxy­vitamin D (calcitriol), primarily func­tions as a steroid hor­mone, ensuring ade­quate intest­inal absorption of calcium and phos­phorus and regul­ating bone mineral­ization. In this role, the pathway of syn­thesis of the active form 1,25‑dihydroxy­vitamin D (calcitriol), is endo­crine as it is made in a tissue only when it is required, then circulates to target tissues, where it is present for only a short period of time.

Vitamin D also has a critical role in a number of cellular func­tions in many tissues in the body, includ­ing bone, placenta, prostate, keratin­ocytes, macro­phages, T-lympho­cytes, dendritic cells, para­thyroid gland and some cancers (Wacker and Holick, 2013). The active meta­bolite 1,25‑dihydroxy­vitamin D (calcitriol) is involved in modulating the immune res­ponse and in regu­lating cell differ­entiation, pro­lifer­ation, and apoptosis (Norman, 2008). In these non-skel­etal roles, the pathway of syn­thesis of this active form is autocrine / paracrine, whereby the active form stays in the cell or is trans­ported locally.

18b.2 Metabolism of vitamin D

Vitamin D can be obtained in two ways: from skin synthesis of vitamin D3 (cholecalciferol) and from ingestion of the parent compounds (D2 (ergocalciferol) or D3) from foods or supplements. This situation creates problems in assessing vitamin D status as dietary intake alone is not sufficient to gauge risk for deficiency. Sun exposure along with behavioural and environmental factors affecting skin synthesis must also be taken into account.

18b.2.1 Skin syn­thesis of vitamin D₃ (cholecalciferol)

The require­ment for vitamin D can be met by skin syn­thesis alone provided UVB can reach the skin. The syn­thesis of vitamin D₃ in the skin involves two stages: the photochem­ical transform­ation of 7-dehydrochol­esterol to pre­vit­amin D₃ by UVB, followed by thermal isomerization of the pre­vit­amin to vitamin D₃ (cholecalciferol). Variables influ­enc­ing the form­ation of pre­vit­amin D₃ in the skin include skin pig­ment­ation, the intensity of the solar ultra­violet light, and envi­ron­mental factors such as clouds, smog, clothing and sunscreen use (Grant et al., 2016; Wacker and Holick, 2013). Over-expo­sure to UVB will not lead to excess skin syn­thesis of vitamin D, as pre­vit­amin D as well as vitamin D₃ are irreversibly con­verted to inactive meta­bolites (Wacker and Holick, 2013). In the absence of UVB expo­sure, the require­ment for vitamin D must be met from diet­ary sources. Mostly, the require­ment is met partially or fully by diet, even in equa­torial regions because people may embrace an urban or sun-avoiding lifestyle.

18b.2.2 Vitamin D₂ (ergocalciferol)

Vitamin D₂ (ergo­cal­ciferol) cannot be made by animals. The only source is ingestion of certain foods in the fungi king­dom. It is import­ant to note the fungi king­dom is separate from the other eukaryotic life king­doms of plants and animals and texts may erroneously refer to vitamin D₂ as a “plant” source. Sun-exposed mush­rooms naturally provide vitamin D₂, while UV-exposed yeast and mush­rooms added to the food supply enhance the vitamin D con­tent of foods (Wacker and Holick, 2013).

The major meta­bolic steps involved in the meta­bolism of vitamin D₂ are similar to those for the meta­bolism of vitamin D₃ (cholecalciferol). The evi­dence about efficacy of vitamin D₂ versus vitamin D₃ suggests that although vitamin D₃ is the more active (Logan et al., 2013), vitamin D₂ is a reasonable alter­nat­ive (Wacker and Holick, 2013). Hence, in the following dis­cussion and in figure 18b.2, the term “vitamin ” refers to either or both vitamin D₂ and vitamin D₃ and their meta­bolites.

18b.2.3 Production of 25‑hydroxy­vitamin D

Vitamin D enters the circu­lation from the skin or from the lymph via the thoracic duct, bound to a specific vitamin D-binding pro­tein. Vitamin D is trans­ported to adipose tissue where it is stored, or to the liver, where it is hydrox­yl­ated to 25‑hydroxy­vitamin D (25(OH)₂D [also called cal­cidiol when referring to 25(OH)₂D₃], the major circulating form of vitamin D (Figure 18b.2).

The plasma 25(OH)₂D meta­bolite has sev­eral fates. When there is a need for calcium (des­cribed below) then 25(OH)₂D is con­verted in the kidney by an en­zyme (25(OH)₂D-1-α-hy­droxy­lase) to pro­duce the bio­logi­cally active meta­bolite 1,25‑dihydroxy­vita­min D [1,25(OH)₂D] also called calcitriol. Circulating 25(OH)₂D can also be taken up by tissues and subsequently con­verted to 1,25(OH)₂D inside cells if the 1-hy­droxy­lase en­zyme has been activated. Finally, in the pathway for inactivation and excretion of vitamin D in bile, 25(OH)₂D is converted to 24,25‑dihydroxy­vitamin D by the enzyme 25‑hydroxy­vitamin D-24-hydroxy­lase (Wacker and Holick, 2013). The enzyme 25‑hydroxy­vitamin D-24-hydroxylase also inactivates 1,25(OH)₂D (calcitriol).

18b.2.4 Renal production of circulating 1,25‑dihydroxy­vita­min D (calcitriol)

The active form of vitamin D, 1,25(OH)₂D (calcitriol) that circulates in plasma is made in the kidney in the Endocrine Pathway as shown in Figure 18b.2. This syn­thesis of 1,25(OH)₂D is homeo­statically con­trolled, mainly by the action of para­thyroid hor­mone (PTH) in res­ponse to serum calcium levels, and fibro­blast growth factor 23 (FGF-23) related to serum phos­phate levels (Wacker and Holick, 2013), that regulate the activ­ity of renal (25(OH)D-1-α-hy­droxy­lase). For exam­ple, a de­crease in plasma calcium prompts an inc­rease in para­thyroid hor­mone secretion from the para­thyroid gland that acts to mobilize calcium stores from the bone. Parathyroid hor­mone also promotes the syn­thesis of 1,25(OH)₂D in the kidney which, in turn, stimulates the mobilization of calcium from the bone and inc­reased in­test­inal calcium absorption (Figure 18b.2). Once plasma calcium levels are normal, the need for circulating 1,25(OH)₂D diminishes and there is no stimulation of the en­zyme 25(OH)D-1-α-hy­droxy­lase. Thus, circulating levels of 1,25(OH)₂D are not related to vitamin D status as in de­fi­ciency low levels of 1,25(OH)₂D may reflect lack of the precursor meta­bolite 25(OH)D. However, as vitamin D de­fi­ciency leads to sec­ond­ary hyper­para­thyroid­ism with PTH-enhanced 1,25(OH)₂D production, vitamin D deficiency can be assoc­iated with normal to high 1,25(OH)₂D (calcitriol) levels.

Another reason why the level of 1,25(OH)₂D is not useful in assess­ing vitamin D status is because 1,25(OH)₂D is a very short-lived meta­bolite causing its own destruc­tion by rapidly inducing synthesis of the enzyme 25‑hydroxy­vitamin D-24-hy­droxy­lase (Wacker and Holick, 2013).

18b.2.5 Extrarenal production of 1,25‑dihydroxy­vita­min D (calcitriol)

The extrarenal pathway of 1,25(OH)₂D (calcitriol) is locally pro­duced in almost every tissue in the body (Norman, 2008). As 1,25(OH)₂D acts locally, this syn­thesis pathway is called Paracrine / Autocrine. Activity of extra-renal 25(OH)D-1-α-hy­droxy­lase is not regul­ated by the hor­mones that con­trol renal 25(OH)D-1-α-hy­droxy­lase (i.e., PTH and FGF-23). The activity of the enzyme must be induced in the cell. The diverse actions of 1,25(OH)₂D, when acting locally as a trans­crip­tion factor in many differ­ent cell types, are called “non-calcemic” or “non-skel­etal” and include immuno-modulatory and cell-differ­entiating properties. It is these properties that have led re­searchers to inves­tigate vitamin D and its derivatives in the patho­gen­esis of cancer, respiratory dis­eases, and immune res­ponses. For further details of these noncalcemic func­tions see (Norman, 2008) and (Wacker and Holick, 2013).

18b.2.6 Serum 25(OH)D

Serum 25(OH)D is a biomarker of vitamin D exposure and status. Both meta­bolites 25(OH)D and 1,25(OH)₂D circulate in plasma. The former, 25(OH)D, reflects the sum of vitamin D from diet­ary intake and sunlight expo­sure, whereas plasma 1,25(OH)₂D con­cen­trations reflect the immediate physio­logical need and are under homeo­static con­trol in the kidney. Concentrations of 1,25(OH)₂D in plasma are about 0.1% of those of 25(OH)D. In vitamin D de­fi­ciency, serum 1,25(OH)₂D levels may be normal or even ele­vated, as a result of inc­reased renal production of 1,25(OH)₂D in res­ponse to the rise in serum para­thyroid levels (Wacker and Holick, 2013). In con­trast, plasma 25(OH)D con­cent­rations remain low until a reserve accumulates. As a result, the plasma 25(OH)D con­cent­ration reflects medium to long-term vitamin D availability from both diet­ary and endog­enous sources, thus making it the best biomarker of vitamin D exposure and status.

18b.3 Vitamin D deficiency in humans

Osteo­malacia may occur in severely vitamin D deficient adults, a con­di­tion charact­er­ized by a failure in the mineral­ization of the organic matrix of bone. This results in weak bones, diffuse skel­etal bone tender­ness, proximal muscle weakness, and an inc­reased frequency of fractures. Such dis­turb­ances are assoc­iated with serum 25(OH)D con­cent­rations below 7.5nmol/L (Haddad and Stamp, 1974). After treat­ment with vitamin D sup­ple­ments, serum 25(OH)D values rise and radio­logical lesions heal (Preece et al., 1975). Osteomalacia prevalence may be high globally due to lack of sun expo­sure, but remains largely undiagnosed due to the need to have radio­graphic data (Uday and Högler, 2019). It may occur in adults living in the tropics,who have no sun expo­sure, such as garment factory workers in Bangladesh (Islam et al., 2008). Low diet­ary intake may also play a role. A study in Germany of mainly white adults who died accidently, found the prevalence of osteo­malacia to be at least 25% based on osteoid volume/bone volume ratio (Priemel et al., 2010).

Some adult patients with chronic renal failure, gas­trect­omy, in­test­inal mal­absorp­tion and stea­torrhea arising from celiac dis­ease, inflam­matory bowel dis­ease, pan­creatic insufficiency, or mas­sive bowel re­sec­tion, may also develop osteo­malacia. Functional dis­turb­ances have been des­cribed in adults with low serum 25(OH)D con­cent­rations. These include sec­ond­ary hyper­para­thyroid­ism, an inc­reased bone turnover, and red­uced bone mass (Chapuy et al., 1997). In the elderly, sub­optimal vitamin D status decreases absorption of calcium, a factor assoc­iated with a lowering of the bone mineral con­tent during post­meno­pausal aging.

Rickets occurs in infants and chil­dren with severe vitamin D de­fi­ciency. In rickets, abnormal softness of the skull (cranio­tabes) occurs. This may be accompanied by en­largement of the epiphyses of the long bones and of the costo­chond­ral junction (rachitic rosary). Bowlegs and knock knees may arise from these bone deform­ities. Rickets, arising from primary vitamin D de­fi­ciency may occur in infants in indus­trial­ized coun­tries who are breast-fed without vitamin D sup­ple­mentation. A supplement of 400 IU (10µg) per day to prevent rickets is often recom­mended for all infants from birth to 12mo of age, indepen­dent of their mode of feeding (Wagner and Greer, 2008; Munns et al., 2016). However, for breastfed infants whose mothers have an adequate vitamin D status, the content of vitamin D in breastmilk can be sufficient because the vitamin D content in breastmilk is dependent on maternal status (Stoutjesdijk et al., 2017). Nutritional rickets can occur in older chil­dren, particularly during the adol­es­cent growth spurt (Beck-Nielsen et al., 2009; Uday and Högler, 2019). However, some nutri­tional rickets is also caused by a lack of calcium, so both vitamin D and calcium should be monitored (Munns et al., 2016; Uday and Högler, 2019).

Metabolic defects also cause rickets, including both vitamin D-resistant rickets (familial hypophos­phatemia) and vitamin D-depen­dent rickets (VDDR type 1). The latter con­di­tion is a de­fi­ciency of the 25(OH)₂D-1-hy­droxy­lase en­zyme, while vitamin D-resistant rickets is a defect in proximal renal tubular resorp­tion of phos­phate. The yearly incid­ence of hypo­phos­phatemic rickets in infants 0–0.9y is about 3.9 per 100,000: vitamin  D-resistant rickets (VDDR type 1) is very rare (Beck-Nielsen et al., 2009).

18b.4 Food sources and diet­ary intakes

Natural food sources rich in vitamin D are restricted to fatty fish, organ meats, and UV-exposed mush­rooms. In some coun­tries — for exam­ple, the United States, Canada, and Finland — fluid milk is fort­ified with vitamin D. In the United Kingdom and Europe, low amounts of vitamin D can be added to some breakfast cereals, margarine, fat spreads, and vege­table oils, breakfast beverages, and breads (Calvo et al., 2005). In low and medium income coun­tries, fort­ifi­cation of wheat, edible plant-based oil or milk, could reduce vitamin D defi­ciencies (Cashman and O’Dea, 2019).

Intakes from national nutri­tion surveys provide data on vitamin D in differ­ent coun­tries. They show low (3.0µg/d) vitamin D intakes on average in coun­tries where there is little or no vitamin D fort­ifi­cation, such as the U.K. However in coun­tries such as Japan, fatty fish may be con­sumed in sufficient quantities to provide vitamin D intakes > 7µg/d (Calvo et al., 2005). Canada, where there is man­dat­ory fort­ifi­cation of milk and margarine, has an aver­age intake of 5µg/d which does not meet the adult US/Canadian Estimated Average Requirement (EAR) of 10µg/d. The Canadian government will double the level of man­dat­ory fort­ifi­cation by 2023 to try to reduce the prevalence of inade­quacy (Vatanparast et al., 2020).

High Intakes following self-dosing with excessive amounts of vitamin D supplements have been described, although the dose required to induce vitamin D toxicity is uncertain (Hathcock et al., 2007). Vitamin D intoxication does not arise from the consumption of conventional foods (including fortified foods), nor does excess exposure to UVB through sun or artificial lamps cause toxicity (Marcinowska-Suchowierska et al., 2019). Cases of toxicity have arisen, however, from accidental overfortification of milk with vitamin D₃, from uncontrolled use of vitamin D mega-doses, and from inappropriate use of vitamin D metabolites. An increased risk for vitamin D toxicity is also associated with certain diseases, including sarcoidosis, tuberculosis, and genetic disorders of rare polymorphisms of enzymes involved in vitamin D metabolism (e.g., idiopathic infantile hypercalcemia) (Wacker and Holick, 2013). Signs of vitamin D toxicity include hypercalcemia (i.e., elevated serum calcium concentrations) and hypercalciuria (elevated urine calcium levels). Hypercalcemia arises from hyperabsorption of intestinal calcium and, to a lesser degree, from the release of calcium from bone and can lead to calcification of soft tissues such as arteries (arteriosclerosis) and the kidney (nephrocalcinosis). The hypercalciuria associated with vitamin D toxicity reflects the presence of excess calcium in the serum arising from the release of calcium from bone.

18b.5 Nutrient reference values for vitamin D

Unlike other essential nutrients, measurements of vitamin D intake using traditional dietary assessment methods cannot be used to measure dietary exposure because some of the requirement for vitamin D can be met from skin synthesis, as noted earlier. Instead a surrogate biomarker is used to measure exposure to vitamin D; serum 25 hydroxy­vitamin D (25(OH)D) is the biomarker of choice. Serum 25(OH)D measures exposure to vitamin D from the effects of both diet and sunlight as noted earlier, and associations between concentrations of serum 25(OH)D and functional biomarkers of bone health have been considered when setting Nutrient Reference Values (NRVs) for vitamin D by several agencies. The evidence for extraskeletal outcomes has been considered inadequate, inconsistent, or insufficient to develop Nutrient Reference Values (NRVs). Nevertheless, in view of the variation in sunlight exposure and the variable response to that exposure, as well as concerns about skin cancer, several agencies, including the IOM (2011), EFSA (2016), and the UK (2016) have set their NRVs based on the assumption of minimal or no sunlight exposure. This means that in the presence of cutaneous vitamin D synthesis, the requirement for dietary vitamin D may be lower than that set, or may even be zero.

The US IOM (2011), as an example, set an Estimated Average requirement (EAR), the requirement to satisfy the need of half the population, of 10µg/d for all children and adults age 1–70y, as necessary to maintain bone health and achieve a serum 25(OH)D concen­tration of 40nmol/L (16ng/mL). The Recommended Dietary Allowance (RDA), designed to cover the requirements for 97.5% of the population, was set at 15µg/d, a level that corresponded to a serum 25(OH)D concen­tration of 50nmol/g (20ng/mL). A higher RDA level (i.e., 20µg/d) was recommended for those over 70y, due to age-related inefficiencies in vitamin D metabolism. For infants 0–12months, IOM set an Adequate Intake (AI) of 10µg/d because there was not enough information to establish an EAR for this group.

Several other countries and regions have also restricted their recommendation to a single level — the AI (or equivalent) — because of insufficient evidence, including Australia and New Zealand (NHMRC, 2006), the European Union (EFSA, 2016), and most recently, the United Kingdom (SACN, 2016), although here only for infants and children less than 4y. However, the target serum 25(OH)D concentrations on which the AIs are based vary, ranging from at least 25nmol/L (10ng/mL) in Australia and New Zealand (NHMRC, 2006), and the UK (SACN, 2016), to 50nmol/L (20ng/mL) in the EU (EFSA, 2016). As a consequence, the AI levels range from about 10µg/d to 15µg/d for most age groups, similar to those for several other European countries (Spiro and Buttriss, 2014).

The UK , like the US has also set a Recommended Nutrient Intake (equivalent to the RDA) for all life-stage groups (including pregnant women, lactating women, and the elderly), except those < 4y. However, for the UK (SACN, 2016), the RNI is 10µg/d in contrast to the 15µg/d RDA set by IOM (2011). No EAR has been set by the UK. Cashman et al. (2008), have emphasized that dietary levels of about 10–15µg/d may not be adequate to keep most of the adult population in Europe above a higher target value of 50nmol/L in winter without adequate sun exposure in the summer season.`

Several agencies have established a Tolerable Upper Intake Level (UL) for vitamin D to discourage potentially dangerous self-medication. The UL is defined as the highest average intake that is likely to pose no risk. For infants 0–6mo, the IOM set an UL of 25µg/d, rising with age through childhood to 100µg/d for persons aged 9y and older. This upper limit corresponds to an average serum total 25(OH)D levels of 125nmol/L (50ng/mL) (2011).

The UL values for the European Food Safety Authority (EFSA) start for infants aged 7–11mo with an UL of 35µg/d which rises to 100µg/d for persons aged > 11y (EFSA, 2016). Data on the long-term adverse effects of high doses of vitamin D are limited, however, and caution is necessary when setting an UL for chronic intake (Aloia, 2011).

The Endocrine Society, focusing on clinical use for patients at high risk of deficiency with rickets, osteomalacia, osteoporosis, chronic kidney disease, etc., and some special populations (e.g., pregnant women) , have set a higher adult UL of 250µg/d (Holick et al., 2011) , a recommendation that is not appropriate for healthy individuals in the population. This distinction is important, and if not recognized, could result in inappropriate dietary recommendations for healthy individuals, for whom the NRVs should be used (Aloia, 2011).

18b.6 Biomarkers of vitamin D status

Historically, vitamin D status was assessed indirectly by meas­uring alkaline phos­phat­ase activ­ity, as well as calcium and phos­phorus con­cent­rations in serum: all very non­specific indices. Methods are now avail­able for the direct measure­ment of vitamin D meta­bolites in serum, and these are des­cribed below. If possible, these measure­ments should be performed in con­junc­tion with an assay of serum para­thyroid hor­mone and some func­tional assess­ment of skel­etal health. In adults, this assess­ment may include measure­ment of bone mineral con­tent or bone mineral density. In chil­dren, in extreme cases of rickets, bony deform­ities such as en­larged fontanelle, rachitic rosary, and swollen joints are clin­ical signs of rickets, whereas knock knees or bowed legs are clin­ical signs of the assoc­iated osteo­malacia noted in growing chil­dren (Uday and Högler, 2019).

18b.6.1 Serum 25‑hydroxy­vitamin D

Serum 25‑hydroxy­vitamin D is the circulating meta­bolite of vitamin D that is the most abundant and has the longest half-life of all the vitamin D derivatives. Concentrations of 25(OH)D in serum (or plasma) are also the most useful measure of vitamin D exposure and status in humans, as they reflect the total supply of vitamin D from both cutaneous syn­thesis and diet­ary intake of either vitamin D₂ or vitamin D₃ (Wacker and Holick, 2013). Moreover, they can be used to define vitamin D de­fi­ciency, insufficiency, hypovitaminosis, sufficiency, and toxicity (IOM, 2011). Concentrations in healthy adults vary from 30–130nmol/L, depend­ing in part on expo­sure to solar ultra­violet light.

18b.6.2 Factors affect­ing serum 25 hydroxy­vitamin D

Seasonal and latitude effects on serum 25 hydroxy­vitamin D status are marked in many areas of the world. Subjects living north of latitude 33°N and south of latitude 33°S (Wacker and Holick, 2013) have reduced dermal vitamin D syn­thesis during winter, and show the highest serum 25(OH)D levels in the late summer, and the lowest in late winter. In a U.K. nat­ional survey of people > 65y (Finch et al., 1998), mean plasma 25(OH)D con­cent­rations were sig­nif­icantly higher in free-living participants surveyed in the summer (July–September) than in the winter (Figure 18b.3).

Finch et al. (1998) further noted that in the summer, only 6% of a free-living group of elderly sub­jects had plasma 25(OH)D con­cent­rations < 25nmol/L com­pared to 35% of those who were insti­tutionalized.

Only UVB rays (290–315nm) elicit syn­thesis of vitamin D₃ (chole­cal­ciferol). When the incident angle of the sun is low, UVB does not reach the earth (Grant et al., 2016). The “Shadow Rule” states that if one's height is longer than the length of one's shadow, vitamin D syn­thesis is possible; if not, the sun's incident angle is too low to provide UVB. This also explains why the best time for vitamin D syn­thesis is bet­ween 10:00 and 14:00 in summer in temperate countries or year-round in low latitude regions. Other factors affect­ing dermal syn­thesis are des­cribed below.

Age-related changes from infancy to adulthood in the con­cent­rations of serum 25(OH)D can be marked. Newborn infants have serum con­cent­rations that correlate with maternal 25(OH)D con­cent­rations. Breastfed infants, as noted above, are at risk of poor vitamin D status unless mothers have an adequate vitamin D status and can pass on meta­bolites in their milk (Stoutjesdijk et al., 2017). Vitamin D meta­bolites are lower when meas­ured in cord blood. Serum 25(OH)D levels in chil­dren of both sexes dec­line with increasing age ( Gregory et al., 2000), a trend that may reflect both behavioral factors (diet and sun expo­sure) as well as bio­logi­cal effects (increasing requirements of vitamin D with growth).

Studies indicate that calcium absorption by premature infants is not only vitamin D depen­dent; how­ever, concern remains about the lack of data to outline the vitamin D needs for premature infants (Taylor et al., 2019).

Older adults are part­ic­ularly vulnerable to low levels of serum 25(OH)D (McKenna et al., 1985). Lifestyle factors that reduce vitamin D status include low diet­ary intakes of vitamin D and limited sun expo­sure with a greater use of sunscreens and umbrellas. In addit­ion, there are bio­logi­cal reasons for the low status includ­ing a red­uced capacity of the skin to pro­duce vitamin D resulting from a reduc­tion of 7-dehydrochol­esterol in the skin, and impaired in­test­inal absorption of ingested vitamin D (Wacker and Holick, 2013). In coun­tries such as Canada where low vitamin D status is related to osteoporosis, it is recom­mended that older adults take vitamin D sup­ple­ments and age-related dec­lines in vitamin D status as meas­ured by serum 25(OH)D have not been observed in Canadian nat­ional survey data (Brooks et al., 2017).

Sex and Gender differ­ences in con­cent­rations of serum 25(OH)D have been noted, although no con­sis­tent pattern has emerged from nat­ional survey data in the USA (Wacker and Holick, 2013) and Canada (Brooks et al., 2017). Differ­entiating bio­logi­cal (sex) effects from lifestyle effects (gender)such as differences in food prefer­ences, clothing, and amounts of sun exposure during work or leisure may be difficult. In some coun­tries, consumption of fort­ified foods and sup­ple­ments as well as sunscreen use and sun avoidance practices, differ bet­ween males and females of all ages.

Skin pig­ment­ation differ­ences, often seen when comparing ethnic or racial groups, greatly influ­ences serum 25(OH)D con­cent­rations. Skin pig­ment­ation reflects the amount of the pigment melanin that absorbs UVB rays and lowers the amount of UVB acting on pre­vit­amin D. In the USA, for exam­ple, serum 25(OH)D con­cent­rations in African-Americans and Hispanics are much lower than in non-Hispanic whites (2011). In European coun­tries, African immigrants now living in northern latitudes have a greater risk of vitamin D de­fi­ciency com­pared to non-migrants. As an example, of the patients with nutri­tional rickets in Denmark, 74% were immigrant chil­dren (Beck-Nielsen et al., 2009). Not all differ­ences in vitamin D status bet­ween ethnic groups are due to skin pig­ment­ation. Behavioral differ­ences such as diet­ary intakes, clothing prefer­ences, and sun avoidance practices are also import­ant factors.

Melanin in skin does not block all chole­cal­ciferol syn­thesis, but it is slowed. Webb and Engelsen (2006) have cal­cu­lated the time needed for a person with each of the differ­ent Fitzpatrick categories of skin type to burn with sun expo­sure. Also shown in Table 18b.1,

Table 18b.1 Association of Fitzpatrick skin types with capacity of skin for vitamin D synthesis at 42°N on 21 June at 10:30 hours while exposing one-quarter of body surface area to sunlight. Time needed for vitamin D synthesis in spring or fall equinox or a higher latitude (62.5°N) would be double. Source Webb and Engelsen (2006) Photochemistry and Photobiology 82: 1697–1703.

Fitzpatrick Skin Type	How skin responds to sun exposure	Minutes to make 25µg (1000IU)
I	Always burn never tan	4
II	Burn slightly then tan slightly	6
III	Rarely burn tan moderatelu	7
IV	Never burn, tan moderately e.g. Mediterranean	10
V	Never burn, tan darkly Asian, Indigenous American, Pacific Islander	13
VI	Never burn, tan very darkly; Australian Aborigine, Tamil, West African	21

are the times needed to synthesize 1000 IU of vitamin D. The times are taken from a more com­plex analy­sis that depends on time of day and day of the year in areas that experience marked seasons, and the amount of skin exposed to sun. As shown, all skin types can synthesize vitamin D, but the time needed is longer as skin pig­ment­ation increases (higher Fitzpatrick numbers). The times in the southern hemisphere will be influenced by variation in theozone layer thickness.

Sunscreen lotions are used to deliberately block UV rays reaching the skin. Sunscreens are labeled with a sun pro­tect­ion factor (SPF) number which indicates the amount of UV blocked. An SPF blocks at 1/SPF, so that a product having an SPF of 8 would allow only 1/8 (12.5%) of the UV to penetrate the lotion and reach the skin. Theoretically sunscreens should reduce vitamin D syn­thesis. In a nat­ional survey in Canada, however, participants answering “yes” to using sunscreen had higher 25(OH)D levels (2.4 ±1.1nmol/L; P < 0.001) (Brooks et al., 2017). Several reasons may account for this seemingly abherrant finding. Users may apply sunscreens poorly or incompletely (for exam­ple answer “yes” to use but only apply to face). Alternatively, sunscreen users may spend more time outdoors. Other behavioral differ­ences may exist, for exam­ple in the Canadian survey, sunscreen users were more likely to take vitamin D sup­ple­ments than nonusers.

Smoking is assoc­iated with lower serum 25(OH)D con­cent­rations. This may partly explain the repor­ted inc­reased risk of osteoporosis among smokers. The mech­anism is unclear, but the relat­ion­ship does not appear to result exclu­sively from addit­ional con­founding lifestyle factors (Brot et al., 1999). In a large study of adults in Australia, both male and female nonsmokers, includ­ing ex-smokers, had higher mean levels of serum 25(OH)D com­pared to current smokers (Gill et al., 2017).

Obesity, prevalent in many pop­ul­ation groups worldwide, is assoc­iated with a trend towards lower serum 25(OH)D levels (Brooks et al., 2017; Wacker and Holick, 2013). Biologically, this trend can be attrib­uted to vitamin D, whether from cutaneous or diet­ary sources, being deposited in adipose tissue, where it is not bioavail­able (Wacker and Holick, 2013). Some re­searchers have found that obesity in men has less of an effect on reducing 25(OH)D than in women, perhaps because of gender differences in behavioral aspects such as sun avoidance (Rockell et al., 2006). Never­the­less, body weight must be con­sid­ered when evaluating vitamin D status

Disease con­di­tions affect­ing the gastro­intest­inal tract, the liver and kidneys may cause a sec­ond­ary de­fi­ciency of vitamin D (Table 18b.2).

Table 18b.2 Secondary causes of vitamin D deficiency. Source Holick (2007). Vitamin D deficiency. The New England Journal of Medicine, 357(3), 266–281.

Causes of Secondary vitamin D Deficiency
Pathology	Diseases
Malabsorption of fat reduces absorption of dietary vitamin D	Cystic fibrosis Celiac disease, Whipple's disease, Crohn's disease, Bypass surgery.
Liver failure prevents production of 25(OH)D	Cirrhosis Hepatitis
Inability to produce 1,25(OH)2D in kidney	Chronic kidney disease
Medications
Drugs reducing Vitamin D absorption	Cholesterol-lowering agents: cholestyramine Weight loss drug orlistat and food additive olestra
Drugs reducing 25(OH)D levels due to increased catabolism	Anticonvulsant medications such as carbamazepine, phenobarbital, and phenytoin, gabapentin. Antiretrovirals agents such as ritonavir and efavirenz, valproic acid (AIDS treatment) Histamine H2 receptor antagonist cimetidine
Drugs Impairing vitamin D metabolism	Oral corticosteroids such as glucocorticoids

Diseases causing fat mal­absorp­tion will reduce the absorp­tion of dietary vitamin D, which could be sig­nif­icant in winter. Liver dis­ease will prevent the con­vers­ion of vitamin D to 25(OH)D, whereas kidney dis­ease prevents the con­vers­ion of 25(OH)D to the active form (i.e. 1,25(OH)₂D, calcitriol) in the endo­crine pathway. Disease states such as in­test­inal mal­absorp­tion and stea­torrhea caused by pan­creatic insufficiency, inflam­matory bowel dis­ease, celiac dis­ease, or mas­sive bowel re­sec­tion have also been associated with lower serum 25(OH)D con­cent­rations. Here, vitamin D deple­tion arises from mal­absorp­tion of diet­ary vitamin D.

Medication use can affect vitamin D status. Any drug which affects liver or kidney cyto­chrome en­zymes will likely affect con­vers­ions of vitamin D meta­bolites. Table 18b.2 provides a list of drugs known to impact vitamin D status. While this list does not cover all possible sec­ond­ary causes of de­fi­ciency, it emphasizes the need to monitor dis­ease states and medi­cation use as possible reasons for vitamin D de­fi­ciency. As is des­cribed below, the require­ment for vitamin D may be ele­vated in persons who have chronic con­di­tions or for whom medi­cation use is required.

Magnesium status may impact 25(OH)D levels and there­fore vitamin D status through the require­ment for two en­zymes of vitamin D meta­bolism: 25(OH)D-1-α-hy­droxy­lase and 25(OH)D-24-hy­droxy­lase. In mag­nes­ium de­fi­ciency, there is a red­uction in the active form 1,25(OH)₂D associated with “Mg-depen­dent vitamin-D-resistant rickets” (Dai et al., 2017). More re­search is needed to deter­mine the intake of mag­nes­ium that affects vitamin D status.

Analytical methods have a marked effect on serum 25(OH)D con­cent­rations. To overcome inter-assay differ­ences, and establish the accuracy and precision of the assay, verified standards should be run with every batch using the Vitamin D Standardization Program. See Section 18b.9.

18b.8 Interpretive cri­teria

Table 18b.3

shows a lack of consensus in defining cutoff points for serum total 25(OH)D (nmol/L) (i.e., the sum of serum 25(OH)D₂ and 25(OH)D₃ concen­trations). The uncertainty is attributed in part to the lack of standardization of the laboratory measurement of serum total 25(OH)D and the lack of potential functional biomarkers of vitamin D status. Only three of the numerous meta-analyses completed to date are based on standardized serum total 25(OH)D levels (Sempos and Binkley, 2020).

Agreement exists Table 18b.3 between two groups of nations (i.e., United Kingdom, the Netherlands with Australia/New Zealand, European Union, USA) for the cutoff point to define severe deficiency (i.e., 25–30nmol/L, 10–12ng/mL), but there is much less agreement for cutoff points to define vitamin D insufficiency, vitamin D sufficiency, and vitamin D toxicity. Conflicting recommendations have been compiled by some non-governmental medical societies and organizations.

In 2011, the US Institute of Medicine (IOM) published cut-off points for serum total 25(OH)D that represent public health guidelines for generally healthy non-diseased populations Table 18b.4

Table 18b.4. Comparison of Institute of Medicine and Endocrine Society cut-points for serum total 25‑hydroxy­vitamin D (ng/mL) [nmol/L = ng/mL × 2.5]

Interpretation	IOM (ng/mL)	Endocine Society (ng/mL)
Deficient	<12	<20
Insufficient	12–20	21–29
Sufficient	20–30	30–100
No added benefit	30–50
Possible harm	>50	>100

(IOM, 2011). A summary report is available in Ross et al. (2011). The lower limit of adequacy, interpreted as deficiency (i.e.,< 12ng/mL; 30nmol/L), was based on relationships to biomarkers of bone health. Cutoff points for insufficiency, sufficiency, no added benefit, and possible harm were also defined by IOM, and have been adopted by several other organizations, including the Global Consensus Recommendations on the Prevention and Management of Nutritional Rickets (Munns et al., 2016, Sempos and Binkley, 2020).

Note the levels set by IOM differ from those set for clinical practice by the Endocrine Society in 2011, and presented in Table 18b.4 (Holick et al., 2011). However, these are intended as guidance for clinicians for the evaluation, treatment, and prevention of vitamin D deficiency, with emphasis on the care of patients who are at high risk of deficiency (e.g., those with rickets, osteomalacia, osteoporosis, chronic kidney disease, etc). Such practice guidelines to treat disease should not be applied to the apparently healthy population. Clearly, more research is urgently required using assays that meet the standardized criteria developed by the Vitamin D Standard­ization Program (VDSP) (Section 18b.9) to develop a consensus on a single set of rigorous interpretive criteria for serum total 25(OH)D levels relative to vitamin D status and interventions that are appropriate for both public health and clinical needs. In the meantime, readers are advised to use the IOM (2011) cutoffs proposed and listed in Table 18b.4 for public health use.

18b.9 Measurement of serum 25‑hydroxy­vitamin D

There are three main methods for meas­uring serum 25(OH)D: immuno­assay, high performance liquid chromatography (HPLC), and liquid chromatography with tandem mass spectrometry (LC-MS/MS) However, there are two main steps that must be undertaken before selecting a method for assaying total 25(OH)D in serum. The first step is verification of “fit-for-purpose” . This step is only necessary when an immunoassay is the intended method and is necessary because some immunoassays are not appropriate for certain patient population groups (e.g., those with certain disease states, pregnant women, and vegetarians such as vegans) (Sempos and Binkley, 2020). In addition, the chosen immunoassay should have an appropriate measurement range for the intended study population (e.g., measure 25(OH)D levels in persons who are deficient). The “fit for purpose” step can be accomplished by testing the immunoassay against a Vitamin D Standardization Program (VDSP) standardized LC-MS/MS assay.

The second step required is standardization, a process whereby the intended assay is calibrated to meet the VDSP performance criteria (i.e., total coefficient of variation (CV) < 10% and a mean bias with the range of –5% to +5%). Some assays, even when based on a similar methodology, are less accurate and precise, making comparison of data across assays or laboratories difficult. The gold standard analytical method is LC-MS/MS and that should always be the method of choice for national surveys. However, even the LC-MS/MS method must meet the assay standardization criteria of the VDSP. Serum total 25(OH)D measurements can be 'prospectively' standardized or 'retrospectively' standardized, using methods developed by VDSP; see Sempos et al., 2017 and Durazo-Arivizu et al., 2017 for more details. A standardized laboratory measurement is defined as one that provides the 'true' serum total serum 25(OH)D concentrations. Several NIST Standard Reference Materials — SRM 972a, 2973, and 1949 — which provide target values for 25(OH)D₂ and 25(OH)D₃ are available to check on the accuracy and precision of the chosen assay. NIST SRM 1949 also provides target values for serum total 25(OH)D values for the first, second, and third trimester of pregnancy and these should be used in studies of pregnancy.

18b.10 Measurement of serum 1,25‑dihyroxyvitamin D

The active form of vitamin D is 1,25(OH)₂D (also called calcitriol) interacts with its nuclear receptor in the intest­ine, bone, and kidney for most of its functions to regul­ate calcium and bone meta­bolism. 1,25(OH)₂D also has many other noncalcemic cellular actions that reside in the Paracrine/Autocrine pathway (Norman, 2008; Wacker and Holick, 2013).

1,25(OH)₂D in serum is not a useful marker of vitamin D status because it has a short half-life (4–6h) and levels are under stringent homeo­static regu­lation by factors at the site of 1,25(OH)₂D syn­thesis in the kidney. Hence, it is not surp­rising that no seasonal vari­ation in serum 1,25(OH)₂D con­cent­rations has been repor­ted (Landin-Wilhelmsen et al.,1995).

Synthesis of 1,25(OH)₂D in the kidney is stimulated by low serum con­cent­rations of calcium or phos­phorus and is inhibited by excess 1,25(OH)₂D. In cases of vitamin D sufficiency, a positive relat­ion­ship exists bet­ween serum 1,25(OH)₂D and 25(OH)D con­cent­rations (Need et al., 2000), presumably because 25(OH)D is the substrate for 1,25(OH)₂D. In vitamin D de­fi­ciency, however, this relat­ion­ship is reversed because with a fall in serum 25(OH)D con­cent­rations there is a rise in the con­cent­ration of para­thyroid hor­mone resulting in an inc­rease in the renal production of 1,25(OH)₂D ( Figure 18b.2) (Wacker and Holick, 2013). As a result, the circulating con­cent­rations of 1,25(OH)₂D often become normal or even ele­vated. In con­trast, serum 1,25(OH)₂D levels de­crease in renal dis­ease (which affects the en­zyme, 25(OH)D-1-α-hy­droxy­lase); levels are very low in anephric patients (i.e., those lacking a func­tional kidney), and in patients on hemodialysis. Even in normal healthy sub­jects, con­cent­rations of serum 1,25(OH)₂D are in the picomolar range, making analy­sis dif­fi­cult (Zittermann, 2003). Serum 1,25(OH)₂D con­cent­rations ranging from 60–100pmol/L are normal in adults; concentrations in chil­dren tend to be higher. Methods of analy­sis include HPLC, LC-MS/MS and radioimmuno­assay. (Zittermann et al., 2016).

18b.11 Other vitamin D meta­bolites

The 24 hydroxy meta­bolites 24,25‑dihydroxy­vita­min D and 1,24,25‑trihydoxyvitamin D can be meas­ured to assess the con­vers­ion of 25(OH)D and 1,25(OH)₂D to their respective inactive meta­bolites. Recently, measure­ment of meta­bolite-to-parent compound ratios were used to estim­ate meta­bolite hy­droxy­lase activ­ity in a study comparing vitamin D₂ and vitamin D₃ bolus doses (Martineau et al., 2019). In the future, additional vitamin D meta­bolites will be meas­ured as the vitamin D field continues to expand. For exam­ple, measurement of 3-epi-25(OH)D₃ and vitamin D-binding pro­tein (VDBP) may help in understanding aspects of vitamin D status (Sempos and Binkley, 2020). Hence, investigators should store serum sam­ples app­rop­riately for possible future analy­ses using LC-MS/MS together with the available refer­ence materials for standardization.

18b.12 Serum alkaline phos­phat­ase

Alkaline phos­phat­ase (EC 3.1.1.1) activ­ity in serum can be used as an indirect measure of vitamin D status. Activity inc­reases in osteo­malacia in adults and child­hood rickets but is gener­ally normal in osteoporosis. Increases in the activ­ity of alkaline phos­phat­ase are usually pro­por­tion­al to the severity of vitamin D deple­tion. For exam­ple, elderly Irish indiv­iduals with serum 25(OH)D levels indic­ative of severe or marg­inal vitamin D deple­tion had slightly higher serum alkaline phos­phat­ase activ­ity than those with 25(OH)D levels classified as replete (Figure 18b.4).

Seasonal changes in serum alkaline phos­phat­ase activ­ity have also been obser­ved in cross-sectional studies, with levels decreasing with seasonal rises in serum 25(OH)D levels (McKenna et al., 1985).

Serum alkaline phos­phat­ase activ­ity is also affected by sex and age, again in the opposite direction to changes in the levels of serum 25(OH)D. Serum alkaline phos­phat­ase activ­ity is sig­nif­icantly higher in females relative to males and in older versus younger adults; activ­ity is also higher in growing chil­dren and pregnant women, espe­cially during the third trimester (McKenna, 1992). In older surveys such as the U.K. nat­ional survey of young people 4–18y, mean alkaline phosphatase activ­ity was meas­ured. The lowest level was found in the oldest adol­es­cents (15–18y), espe­cially among the girls (Gregory et al., 2000).

The activ­ity of serum alkaline phos­phat­ase is also altered by various dis­ease states such as hyper­para­thyroid­ism, Paget's dis­ease, sec­ond­ary bone cancer, and cholestasis (Sauberlich, 1999). Serum alkaline phos­phat­ase activ­ity may de­crease in zinc de­fi­ciency but, as noted earlier, appears to be close to normal in osteoporosis in con­trast to osteo­malacia where it is high (Uday and Högler, 2019).

In general, measure­ment of alkaline phos­phat­ase activ­ity in serum is best used to con­firm a clin­ical diag­nosis of vitamin D de­fi­ciency, or as a screen­ing tool, but it is not very useful for detecting subclin­ical vitamin D de­fi­ciency. The diag­nosis of osteo­malacia can be made in the presence of high alkaline phos­phat­ase activ­ity accompanied by high PTH, low diet­ary calcium intake (< 300mg/d) and/or low serum 25(OH)D (< 30nmol/L) (Uday and Högler, 2019).

18b.12.1 Interpretive cri­teria

Serum alkaline phos­phat­ase activ­ity is normally expressed as U/L. The refer­ence range for normal adults is 30–135U/L (Gregory et al., 2000). Total plasma alkaline phos­phat­ase activ­ity was meas­ured in the past U.K. nat­ional surveys (Gregory et al., 2000, Finch et al., 1998) , with the excep­tion of the survey on pre-school chil­dren. Mean, median, and lower and upper 2.5 or 5th per­cen­tiles by age and sex are presented.

18b.12.2 Measurement of alkaline phos­phat­ase

Several methods are avail­able for the assay of serum or plasma alkaline phos­phat­ase (ALP); for plasma, heparin­ized blood sam­ples should be used (Bessey et al., 1946). Total alkaline phosphatase meas­ures all sources of en­zyme activ­ity includ­ing liver, while bone-specific alkaline phosphatase (BALP) meas­ures activ­ity derived from osteoblast activ­ity. Higher total or bone alkaline phos­phat­ase levels accompanied by low urin­ary calcium could prompt inves­tigation of vitamin D de­fi­ciency (Kennel et al., 2010). The assay of total alkaline phosphatase activity is sufficient evi­dence as long as other liver en­zymes are normal. The activ­ity of serum or plasma alkaline phos­phat­ase should be expressed as U/L. The within-sub­ject coef­fic­ient of vari­ation for serum/plasma alkaline phos­phat­ase activ­ity ranges from 4.6% to 9.2%, depend­ing on the time frame and diet­ary regimen (Gallagher et al., 1989). The en­zyme is reasonably stable in frozen serum or plasma.

18b.13 Serum para­thyroid hor­mone

Parathyroid hor­mone (PTH) levels in serum or plasma are con­sid­ered to be a func­tional biomarker of vitamin D status in the normocalcemic state. In vitamin D de­fi­ciency, when calcium absorption is red­uced, serum PTH levels rise to induce calcium mobilization from the bone and inc­rease tubular reabsorption of calcium in the kidney. In this way, serum calcium is maintained at a physio­logically optimum level. As a result, serum PTH con­cent­rations are inversely related to serum 25(OH)D levels (Chapuy et al., 1997; Need et al., 2000; Wacker and Holick, 2013).

Threshold values for serum 25(OH)D con­cent­rations that induce an inc­rease in serum para­thyroid secretion vary. In a study of Australian post­meno­pausal women, the threshold value was 40nmol/L (Need et al., 2000), whereas in French adults it was 78nmol/L (Chapuy et al., 1997). In the 2011 IOM report, studies exam­ined showed that PTH is inversely assoc­iated with serum 25(OH)D con­cent­rations at lower 25(OH)D con­cent­rations but there was incon­sis­tent evi­dence for a threshold above 27nmol/L. Such variable evi­dence for a threshold may result from differ­ent assays being used, both to measure serum PTH and serum 25(OH)D. Even small inc­reases in PTH levels may have a negative influ­ence on bone mass and inc­rease the risk of non-vertebral fracture (Chapuy et al., 1997). Hence, a com­bin­ation of serum PTH and serum 25(OH)D con­cent­rations is often recom­mended as an indicator of vitamin D status.

Serum PTH levels inc­rease with age, indepen­dent of 25(OH)D, ionized calcium, phos­phate, and renal func­tion (Carrivick et al., 2015 ). Further re­search is required to explore the underlying mech­anisms and clin­ical relevance. Season also influ­ences PTH con­cent­rations, as might be expected from seasonal vari­ations in 25(OH)D. Parathyroid hor­mone can be meas­ured by radioimmuno­assay using commercial kits. (Zittermann et al., 2016).

18b.14 Calcium and phos­phorus in serum and urine

Many studies of vitamin D status have included the measure­ment of calcium and phos­phorus con­cent­rations in serum or urine. The measure­ments are most useful if combined with the measure­ments of serum 25(OH)D and PTH con­cent­rations. In vitamin D de­fi­ciency in infants and chil­dren, the serum calcium and phos­phorus levels are usually red­uced. For exam­ple, sig­nif­icantly lower mean serum calcium con­cent­rations were repor­ted in French neonates with serum 25(OH)D con­cent­rations < 30nmol/L and ele­vated PTH con­cent­rations, in com­par­ison with individuals with normal values for these bio­chem­ical parameters (Zeghoud et al., 1997). Unlike total serum calcium alone, the exist­ence of such a triad of bio­chem­ical dis­turb­ances strongly indicates vitamin D de­fi­ciency.

Serum calcium is also used to identify possible vitamin D intoxication. In such cases, con­cent­rations of serum 25(OH)D and serum calcium are ele­vated and provide addit­ional evi­dence for hypervitaminosis D.

The res­ponse of urin­ary calcium and phos­phorus levels to changes in vitamin D status varies, as ex­cretion of these minerals is also affected by diet­ary intakes. Early osteo­malacia can be detected as a de­crease in urin­ary calcium to creat­in­ine ratio (Uday and Högler, 2019). Changes in urin­ary calcium and phos­phorus con­cent­rations, however, are not specific for vitamin D status. Details of the measure­ment of calcium and phos­phorus in serum and urine are given in Chapter 23.

Acknowledgements

SJW would like to thank past and present vitamin D collaborators. Both authors are grateful to Michael Jory for the HTML design and his tireless work in directing the translation to this HTML version.