Seasonal changes in vitamin D content

Seasonal variations in vitamin D content of milk are well documented, with concentrations found to be higher in the summer months than in the winter, most likely due to differences in both husbandry and feeding practices between the seasons. Reports dating back as far as the 1920s demonstrated that a single cow pasture-fed between May and July had a higher ‘anti-rachitic’ (vitamin D) content than the milk produced when the same cow was fed in-house and kept in the dark⁽ Reference Luce ^{16 )}. The same cow was then involved in another study, which collected milk samples for 18 months. In support of the initial findings, a 2–3-fold increase in the vitamin D content of the milk produced was observed when the cow was out to pasture, compared with the milk produced when the cow was housed in a dark stall⁽ Reference Chick and Roscoe ^{17 )}. Evidence suggests that this seasonal variation is the result of insufficient stores of vitamin D in the liver and fat tissues for mobilisation in times when dietary intake of the vitamin is low⁽ Reference Thompson ^{18 )}. Many subsequent studies have confirmed the seasonal variation of the vitamin D content of milk (approximate differences ranging between 0·004 and 0·0014 µg/g fat) across different countries and breeds of cattle (Table 1)⁽ Reference Thompson, Henry and Kon ^{19 –} Reference Lindmark-Månsson, Fondén and Pettersson ^{22 )}. Although seasonal variation in vitamin D content is widely reported in the literature, units of measurement are inconsistent, which makes it difficult to compare between studies. In the previous edition of the UK Food Composition Tables, no seasonal variation in the vitamin D content of milk was noted for whole, semi-skimmed and skimmed milk, but was observed in the whole milk samples from the Channel Islands, where mean concentrations for summer and winter were 0·04 µg/100 g and 0·03 µg/100 g, respectively⁽ Reference McCance and Widdowson ^{23 )}. In the most recent edition, a lack of seasonal variation is still apparent, with vitamin D only quantified for Channel Island whole milk, listed as 0·01 µg/100 g and trace for whole, semi-skimmed, skimmed and 1% milks⁽ Reference McCance and Widdowson ^{24 )}.

Table 1. Studies investigating the seasonal variation of vitamin D concentrations in cow's milk

* Seasonal means not reported.

While the seasonal variation in the vitamin D content of milk is established, not all studies or databases, such as the recent editions of the UK Food Composition Tables, report such variations, and a more comprehensive update of vitamin D in milk across the UK and Ireland is warranted.

UVB exposure

In a study by Hymøller et al., cows from two organic dairy farms in Denmark were selected to determine the effect of sunlight on the vitamin D status in March and April, and on each farm, cattle were allocated based on milk yield, parity and lactation stage to have daily outdoor access (from February to April) or to be confined indoors for the duration of the study (November–April)⁽ Reference Hymøller, Mikkelsen and Jensen ^{25 )}. Results from Farm 1 found no significant effect of treatment allocation on plasma 25-hydroxyvitamin D₃ (25(OH)D₃) concentration in March (P = 0·350) or April (P = 0·060), with mean plasma 25(OH)D₃ concentrations of 7·84 and 5·85 ng/ml for the outdoor and indoor groups, respectively⁽ Reference Hymøller, Mikkelsen and Jensen ^{25 )}. On Farm 2, the outdoor group had a significantly higher 25(OH)D₃ concentration in March, compared with the indoor group (5·71 v. 3·36 ng/ml; P < 0·05), but the same difference was not reported in April (P = 0·100)⁽ Reference Hymøller, Mikkelsen and Jensen ^{25 )}. Hymøller et al.⁽ Reference Hymøller, Mikkelsen and Jensen ^{25 )} concluded that the assumption was that supplemental vitamin D₃ may still be required in the spring as a means to maintain a healthy vitamin D status.

In the field of bio-fortification/bio-addition, a recent Danish study⁽ Reference Jakobsen, Jensen and Hymøller ^{9 )} investigated the potential impact of supplemental UVB light on vitamin D₃ synthesis in sixteen housed Holstein cattle, a common dairy breed, which had been severely depleted of their vitamin D stores. The cows were randomised to receive artificial UVB light 30, 90 or 120 min daily for 24 d or 60 min for 73 d; the length of UVB exposure was designed to be equivalent to 1, 2, 3 and 4 h of sunlight at pasture at 56°N, respectively⁽ Reference Jakobsen, Jensen and Hymøller ^{9 )}. After 24 d, the exposure to supplemental UVB light significantly increased the vitamin D₃ and 25(OH)D₃ concentrations in the milk in a dose-dependent manner over 30, 90 and 120 min⁽ Reference Jakobsen, Jensen and Hymøller ^{9 )}. In the cattle allocated to receive 60 min daily, a significant increase (P = 0·029) in the vitamin D₃ (but not the 25(OH)D₃) concentration of the milk produced between days 0 and 24 was noted, but this did not increase further up to day 73 (P = 0·400)⁽ Reference Jakobsen, Jensen and Hymøller ^{9 )}.

This important preliminary evidence, albeit from a limited number of studies, suggests that vitamin D bio-fortification of cow's milk does, at least in theory seem probable. Future studies therefore should investigate this novel on-farm method as a means of minimising the seasonal variation in cow's vitamin D status and the milk produced.

Diet

The seasonal changes in the vitamin D content of milk, have long been associated with the change in UV intensity and a reduction in the time spent outdoors, rather than as a result of the change in feed⁽ Reference Lindmark-Månsson, Fondén and Pettersson ^{22 ,} Reference Bartlett, Cotton and Henry ^{26 )}. That being stated, in the UK cattle are solely reliant on dietary vitamin D during the winter, obtained through grass stores (hay, silage or haylage) or feed concentrates. Prior to 2010, both vitamins D₂ and D₃ were authorised by the European Commission as sources of vitamin D, which could be added to feeds intended for cattle; however, in November 2010 no submission was made for the re-authorisation of a vitamin D₂ dossier, and as a result cattle can now only obtain vitamin D₂ from the consumption of fungi growing among the vegetation (fresh grass, hay, silage or haylage) used as roughage in the diet⁽ Reference Richardson and Logendra ^{5 )} and not from concentrates. Within the EU, vitamin D₃ is now the only authorised source of supplemental vitamin D for cattle^{( 27 )}, with the maximum permitted levels set at 4000 IU (100 µg)/kg of feed^{( 28 )}.

Although cattle are reliant on dietary vitamin D during the winter months, it has been suggested that fat-soluble vitamins from such dietary sources are destroyed once they enter the rumen, owing to the fermentative environment⁽ Reference Rode, McAllister and Cheng ^{29 ,} Reference Bourne, Wathes and McGowan ^{30 )}. Research using a fistula model was designed to test this hypothesis in vitamin D⁽ Reference Hymøller and Jensen ^{4 )}. A maximum of 15 kg ruminal contents were removed and mixed with a vitamin D₂ and D₃ (both 250 mg) and vitamin E pre-mix⁽ Reference Hymøller and Jensen ^{4 )}. The contents were then returned to the rumen; ruminal and blood samples were then collected over the subsequent 30 h period⁽ Reference Hymøller and Jensen ^{4 )}. Once collected, ruminal samples were freeze-dried (in vivo samples), additional ruminal samples were collected at the 1 h time-point, and stored in plastic bottles, which were then placed in a water-bath (37 °C; in vitro samples). The concentrations of both vitamins D₂ and D₃ declined over the study period in the in vivo samples, with concentrations remaining stable in the in vitro samples, suggesting no degradation in the intact ruminal sample⁽ Reference Hymøller and Jensen ^{4 )}. Results showed that the plasma concentrations of both vitamins D₂ and D₃ increased over the first few hours, from levels below the limit of detection, and reached a maximum concentration after 24 h (99(15) and 163(16) ng/ml, respectively), with vitamin D₃ concentrations significantly higher than those for vitamin D₂ ⁽ Reference Hymøller and Jensen ^{4 )}. It has previously been hypothesised that vitamin D degradation in the rumen may be a natural protective detoxification process when large quantities of the vitamin are consumed⁽ Reference Horst, Reinhardt and Reddy ^{31 )}, and this may also be a possible reason for the rapid conversion to 25-hydroxyvitamin D observed by Hymøller and Jensen⁽ Reference Hymøller and Jensen ^{4 )}.

Previously the potential of intravenous supplements to improve the vitamin D status of the cow and the milk produced have also been considered. Thompson and Hidiroglou⁽ Reference Thompson and Hidiroglou ^{32 )} orally administered 1 000 000 IU (25 000 µg) vitamin D₂ and 1 000 000 IU (25 000 µg) vitamin D₃ mixed in maize oil to two dairy cows, collecting milk and blood samples for 10 d after. The results showed that the maximum plasma vitamin D concentrations were observed after 2–3 d, with the maximum concentrations in the milk 1–3 d after⁽ Reference Thompson and Hidiroglou ^{32 )}. At the same time twelve additional cows were allocated to be orally or intravenously administered with vitamin D₃ in doses of 5 000 000 IU (125 000 µg) or 10 000 000 IU (250 000 µg). Increases in the vitamin D content of the milk produced varied between animals, with the maximum levels reached between 3 and 7 d for the oral doses and up to 10 d for the intravenous doses, with the maximum observed ranges between 8 IU (0·2 µg) and 92 IU (2·3 µg)/100 ml⁽ Reference Thompson and Hidiroglou ^{32 )}. It is important to interpret these results with caution as the doses administered in this trial are extreme and would not be feasible to incorporate into the daily management of a dairy herd. Furthermore, little is also known on the safety, efficacy and longer-term effects of prolonged usage ‘mega-doses’, other than the data available for acute doses used in the treatment of milk fever⁽ Reference Olsen, Jorgensen and Bringe ^{33 ,} Reference Weiss, Azem and Steinberg ^{34 )}.

A research team led by Hollis collected milk samples from two groups of cows (4000 IU (100 µg) v. 40 000 IU (1000 µg) daily), and found concentrations of vitamin D, 25-hydroxyvitamin D and 1,25-dihydroxyvitamin D in the milk to be greater in those cattle receiving a higher daily dose of vitamin⁽ Reference Hollis, Roos and Draper ^{8 )}. Similar results were noted for 24,25-dihydroxyvitamin D and 25,26-dihydroxyvitamin D⁽ Reference Hollis, Roos and Draper ^{8 )}. This research indicates that the intake of sufficient quantities of dietary vitamin D is enough to increase the vitamin D content of the milk produced.

A cross-over study randomised fourteen Danish Holstein cows based on parity and milk yield to receive a one-off 250 mg dose of vitamin D₂, followed by the same dose of vitamin D₃ in capsule-form or vice versa⁽ Reference Hymøller and Jensen ^{6 )}. Plasma samples were obtained and area under the curve was used to determine the impact of the two different doses on the plasma 25-hydroxyvitamin D status. Results found that the concentrations of plasma 25-hydroxyvitamin D₂ when D₂ was administered first was less than half that of 25(OH)D₃ when the vitamin D₃ dose was given first (P ≤ 0·001)⁽ Reference Hymøller and Jensen ^{6 )}, suggesting that vitamin D₂ may impair the utilisation of vitamin D₃.

McDermott et al. assigned twenty Holstein cows to receive 0 IU, 10 000 IU (250 µg), 50 000 IU (1250 µg) or 250 000 IU (6250 µg) vitamin D₃ daily, for 14 weeks starting at 2 weeks pre-partum⁽ Reference McDermott, Beitz and Littledike ^{35 )}. Vitamin D₃ concentrations in the colostrum were significantly higher (P < 0·05) in cows receiving 250 000 IU/d compared with the other groups, although this dropped during the transition to normal milk from colostrum, about 1 week post-partum. At the end of the study the vitamin D₃ content of the milk was approximately 0·075 ng/ml, 0·16 ng/ml, 20 ng/ml and 22 ng/ml for 0, 10 000, 50 000 and 250 000 IU, respectively⁽ Reference McDermott, Beitz and Littledike ^{35 )}. A mean concentration of 0·15 ng/ml for 25(OH)D₃ was observed in normal milk⁽ Reference McDermott, Beitz and Littledike ^{35 )}.

The need to supplement cattle over the summer months with vitamin D₃ was investigated in Swedish Holsteins, assigned to receive a mineral feed containing vitamin D₃ concentrations in accordance with Swedish recommendations (control) or the same feed providing approximately 20 000 IU (500 µg) vitamin D₃ daily⁽ Reference Hymøller, Jensen and Lindqvist ^{2 )}. Plasma samples collected over the 2-year period showed a significant effect of treatment on the cattle's circulating 25(OH)D₃ concentrations compared with control (P ≤ 0·001) and moreover, the 25(OH)D₃ concentrations in both the supplemented and unsupplemented cows increased when the cattle were out at pasture over the summer months⁽ Reference Hymøller, Jensen and Lindqvist ^{2 )}. The authors concluded that cattle obtain adequate vitamin D₃ from dermal synthesis over the summer, but that stores were not adequate to maintain the status and they had to rely on supplemental vitamin D over the winter⁽ Reference Hymøller, Jensen and Lindqvist ^{2 )}.

Overall the results of the earlier studies provide evidence to suggest that dietary vitamin D₃ is adequate to improve the vitamin D content of the milk produced and to help maintain the status in times where dermal synthesis is not feasible, despite the fermentative environment of the rumen. These findings are of particular importance in relation to recent changes in husbandry practices, which have seen a growing shift to the year-round housing of cattle.

Breed

The variation in the vitamin D content of milk produced by different cattle breeds is supported by evidence conducted across the world (Table 2). The Holstein–Friesian cross has become the most common breed of dairy cow, used for milk production across the world, due to the high production rates⁽ Reference Ramhola, Santos and Casal ^{36 )} and also remains a popular choice within the majority of British herds^{( 37 )}.

Table 2. Studies investigating the impact of cattle breed on the vitamin D concentrations of milk

* Summer vitamin D concentration. Winter concentrations for all three breeds were 0·002 µg/g fat.

† Provitamin D₃ also measured – mean concentrations higher in the milk of Holstein–Friesian than Minhota cattle (0·77 and 0·45 µg/g fat).

The average vitamin D content of milk produced in the UK is currently documented as ‘trace’ for whole, semi-skimmed and skimmed milk, albeit breed is not specified, with the exception of whole milk from the Channels Islands where Jersey cows are the dominant breed (0·1 µg/100 g)⁽ Reference McCance and Widdowson ^{24 )}. The differences in vitamin D reported in the current Food Composition Tables support the results of Wallis⁽ Reference Wallis ^{38 )} who compared the vitamin D content of the milk from Holsteins and Jerseys in the 1940s. Results from this early work showed that although Holsteins produced vastly greater quantities of milk, the vitamin D content of the Jersey cows was on average 3-fold higher owing to higher butterfat concentrations⁽ Reference Wallis ^{38 )}. Bechtel and Hoppert noted that not only was the vitamin D content of the milk higher in the summer months, but also that the milk fat produced from the Guernsey cattle was higher than the milk of the Holstein cattle⁽ Reference Bechtel and Hoppert ^{39 )}. A British study involving three cattle breeds (Friesian, Jersey and Ayrshire) commonly milked in the UK observed differences across the three breeds in summer milk, with little difference apparent in winter milk⁽ Reference Thompson, Henry and Kon ^{19 )}. In Portugal, two studies have noted a higher vitamin D content of milk from indigenous dairy breeds (Barrosã and Minhota) when compared that from with Friesians and Holstein–Friesians⁽ Reference Ramhola, Santos and Casal ^{36 ,} Reference Pires, Fernandes and Vilarinho ^{40 )}.

Hair coverage and dominant colour

To determine if cattle could synthesis vitamin D₃ regardless of hair coverage, Hymøller and Jensen⁽ Reference Hymøller and Jensen ^{41 )} designed a study involving sixteen Danish Holstein cattle, which had been depleted of their vitamin D stores, and randomised based on parity and milk yield to one of the four groups. The treatment groups consisted of different levels of body coverage with a fabric, which prevented vitamin D synthesis for 28 d: a horse blanket; an udder cover; a horse blanket and an udder cover; no coverage⁽ Reference Hymøller and Jensen ^{41 )}. The cattle were on pasture for 5 h each day and inside for the remainder of the day, and were fed a vitamin D₃-free diet throughout the study⁽ Reference Hymøller and Jensen ^{41 )}. Mean plasma 25(OH)D₃ concentrations increased from 2·8  (0·2) ng/ml at baseline in all treatment groups, in a dose-dependent manner with the increasing level of body coverage⁽ Reference Hymøller and Jensen ^{41 )} (Table 3).

Table 3. Studies investigating the impact of hair coverage and dominant hair colour on the vitamin D synthesis

25(OH)D, 25-hydroxyvitamin D.

* While no figures were reported it was also determined that dominant hair colour (black or white) had no impact on plasma 25(H)D concentrations.

More recently, Hymøller and Jensen⁽ Reference Hymøller and Jensen ^{42 )} randomised twenty Danish Holstein heifers based on milk yield and dominant hair colour (black or white) to five different groups, allocated to an increasing length of time on pasture per day (0, 15, 30, 75, 150 or 300 min)⁽ Reference Hymøller and Jensen ^{42 )}. At baseline, the mean plasma 25(OH)D₃ concentration for all the heifers was 44·9 (2·4) nm/l. Over 28 d, the cattle on pasture for 15, 30 or 75 min were unable to maintain their 25(OH)D₃ concentrations from that at baseline. A significant increase in mean 25(OH)D₃ concentration was observed however in those outside on pasture for 150 or 300 min⁽ Reference Hymøller and Jensen ^{42 )}. In addition, they found that the dominant coat colour (black or white) had no significant effect on the plasma concentrations of 25(OH)D₃, illustrating that prominent coat colour does not influence the dermal synthesis of vitamin D₃ in such cattle⁽ Reference Hymøller and Jensen ^{42 )} (Table 3).

The results of these two unique studies eloquently demonstrate that cattle can synthesis vitamin D₃ through all areas of their skin and not just in the udders or muzzle, where hair coverage is scarce. The work by Hymøller et al. also illustrates that, unlike human subjects, pigmentation has no effect on the synthesis of vitamin D₃ following UVB exposure⁽ Reference Libon, Cavalier and Nikkels ^{43 )}. Further work in other cattle breeds is required to further investigate the variance in vitamin D levels in the milk produced. In addition, it may be beneficial to further explore the research by Hymøller and Jensen in other breeds to determine other factors that may prevent the dermal synthesis of vitamin D, such as long haired cattle breeds.

Age

A German two-series study investigated the impact age has on the metabolism of 25(OH)D₃ ⁽ Reference Wilkens, Cohrs and Lifschitz ^{44 )}. In the first series, fourteen multiparous cows were supplemented orally with 3 mg 25(OH)D₃ daily from day 270 of gestation until parturition, with blood samples collected every other day⁽ Reference Wilkens, Cohrs and Lifschitz ^{44 )}. Ninety cows were allocated in the second series to receive 0, 4, 6 mg 25(OH)D₃ daily through mineral feed additives for the last 8–10 d of gestation, with blood samples also taken every other day until parturition, and at several intervals thereafter⁽ Reference Wilkens, Cohrs and Lifschitz ^{44 )}. Calculated slopes found the difference in 25(OH)D₃ between cattle in their second and third lactation to be significantly higher in the second lactation (P < 0·001), suggesting that younger cattle are more efficient at absorbing 25(OH)D₃ or that in older cattle the rate of 25(OH)D₃ elimination is faster, with 1,25-dihydroxyvitamin D₃ increased in cattle in the third lactation or higher⁽ Reference Wilkens, Cohrs and Lifschitz ^{44 )}.

Stage of lactation

A Japanese study collected milk samples from three Holstein cows at stage points post-partum: 1 d after, colostrum; 2–4 d after, early milk and 15 d after, later milk⁽ Reference Okano, Yokoshima and Kobayashi ^{45 )}. Similar concentrations of vitamin D were recorded across the three points for two of the cows (33·2 IU/l (0·83 µg/l), 30·9 IU/l (0·77 µg/l), 35·6 IU/l (0·89 µg/l); and 47·0 IU/l (1·18 µg/l), 47·0 IU/l (1·18 µg/l), 55·7 IU/l (1·39 µg/l), respectively), with no trend noted in the third (77·0 IU/l (1·93 µg/l), 88·9 IU/l (2·22 µg/l) and 47·4 IU/l (1·19 µg/l))⁽ Reference Okano, Yokoshima and Kobayashi ^{45 )}.

Further work required to fully elucidate the impact of age and lactation on the vitamin D content of milk, as this has previously been established for other nutrients such as fatty acids⁽ Reference Kelsey, Corl and Collier ^{46 ,} Reference Bainbridge, Cersosimo and Wright ^{47 )}, this is of importance as the cattle milked on a farm will be at various stages of lactation depending on calving dates and parity.