Seasonal Deficiency of Vitamin D in Children: A Potential Target for Osteoporosis-Preventing Strategies?


  • Parts of this study were presented at the Meeting of the Sociedad Española de Investigaciones Oseas y Metabolismo Mineral, Alicante, Spain, 1995.


Peak bone mass attained after skeletal growth is a major determinant of the risk of developing osteoporosis later in life, hence the importance of nutritional factors that contribute to bone mass gain during infancy and adolescence. An adequate supply of vitamin D is essential for normal bone homeostasis. This study was undertaken to determine what the levels are of 25-hydroxyvitamin D (25(OH)D) that may be considered desirable in children and to assess if normal children maintain these levels throughout the year. Vitamin D metabolites and parathyroid hormone (PTH) serum levels were measured in 21 children in March and October, prior to and after the administration of a daily supplement of 25(OH)D (40 μg for 7 consecutive days). There were inverse correlations between basal 25(OH)D levels and supplementation-induced changes in serum 1,25(OH)2D (r = 0.57, p < 0.05) and PTH (r = 0.41, p < 0.05). When basal levels of 25(OH)D were below 20 ng/ml, the supplement induced an increase in serum 1,25(OH)2D; with basal 25(OH)D under 10–12 ng/ml, the supplement also decreased serum PTH. The lowest serum level of 25(OH)D in 43 normal children studied in summer was 13 ng/ml. Those results suggested that the lowest limit for desirable levels of 25(OH)D in children was somewhere between 12 and 20 ng/ml. However, 31% of 51 normal children studied in winter had levels below 12 ng/ml, and 80% had levels lower than 20 ng/ml. Those children are likely to have suboptimal bioavailability of vitamin D, which might hamper their achievement of an adequate peak bone mass. Since cutaneous synthesis of vitamin D is rather limited in winter, oral vitamin D supplementation should be considered.


Osteoporosis may result from a reduced accumulation of bone tissue during skeletal growth and consolidation, or from an accelerated rate of bone loss later in life. The relative contribution of each factor varies among different subjects, but a low peak bone mass appears to be a major determinant of the subsequent risk of osteoporotic fractures.(1) Although family studies suggest that about 80% of the variability in the peak bone mass may be explained by genetic factors, it still leaves a significant proportion determined by environmental factors and thus potentially open to manipulation.(2) Until recently it was considered that peak bone mass was reached by the third or fourth decades of life. However, several studies have now shown that most of the skeletal mass is already accumulated by late adolescence.(3,4) Thus, in order to prevent the development of osteoporosis, it is important to pay attention to exogenous factors that contribute to bone mass gain during infancy and adolescence, including nutrition and physical activity.(5)

A marked deficiency of vitamin D results in rickets, but little is known about the effects of a mildly reduced vitamin D supply on the growing skeleton. Nevertheless, some indirect data suggest that it may impair bone mass accumulation. Among the nutrients required for bone growth, calcium plays an essential role. In fact, increasing calcium intake in children results in higher bone mass.(6) An insufficient availability of vitamin D impairs intestinal calcium absorption and increases bone resorption and turnover. Thus, low vitamin D levels are associated in adults with a low bone mass, whereas vitamin D supplementation increases bone mass and decreases fracture rate in populations at risk for vitamin D deficiency.(7–9) Although such a relationship between vitamin D and bone mass has not been demonstrated in children, a recent study suggests that reduced rates of skeletal remodeling during periods of bone growth may result in increased bone density.(10) In fact, in one study, vitamin D intake was positively related to peak bone mass,(11) and even an effect on fetal skeletal growth has been suggested.(12) Vitamin D stores can be assessed by measuring the levels of 25-hydroxyvitamin D (25(OH)D), which is the most abundant metabolite in serum. However, there is considerable debate about which levels should be considered as “healthy” or desirable in order to acquire and maintain bone mass.(13,14) Since vitamin D deficiency can be easily prevented and treated, it seems to be an attractive target for preventive programs aimed at potentiating the acquisition of an adequate peak bone mass. Therefore, we planned this study to determine what the levels are of 25(OH)D that may be considered desirable in children, and to assess if normal children maintain these levels throughout the year.


Vitamin D intake

A food-frequency questionnaire focusing on vitamin D– and calcium-rich foods was completed by interviewing the parents and institution managers. The average daily intake of vitamin D and calcium was computed using Spanish tables on nutrient contents (Instituto de Nutrición, Madrid, Spain).

Desirable vitamin D levels

We determined which serum 25(OH)D level is desirable by studying the response of serum 1,25(OH)2D and parathyroid hormone (PTH) to 25(OH)D supplementation. In the case of vitamin D deficiency, reduced 25(OH)D levels limit 1,25(OH)2D synthesis and may cause a compensatory increase in PTH secretion. Therefore, desirable 25(OH)D levels can be defined as those levels at which a further supplement of 25(OH)D neither increases serum 1,25(OH)2D nor decreases serum PTH.(13–16)

After obtaining parental consent, 21 children aged 9 ± 1 years (range 7–10 years), 12 male and 9 female, were studied successively in March and October, months which follow periods of low and high sun radiation, respectively. The group included 11 children with minor acute diseases (“normal children”) and 10 with mental deficiency. The latter subgroup was selected to study children within a wider range of vitamin D status, based on the assumption that they were more likely to present vitamin D deficiency. They lived in an institution during weekdays and went home for the weekends. All were able to walk. One was taking phenobarbital. During each period, serum levels of 25(OH)D, 1,25(OH)2D, PTH, calcium, phosphorus, and alkaline phosphatase were measured before and after the administration of 40 μg/day of 25(OH)D (Hidroferol, Faes Laboratories, Madrid, Spain) per os for 7 days.(15) Pre- and post-treatment samples of each child were run in the same assay.

Vitamin D status in normal children

In a cross-sectional study, serum levels of vitamin D metabolites and PTH were measured in 94 children (aged 8 ± 2 years); 51 were studied in January through April (“winter”), and 43 in August through October (“summer”). All lived in Cantabria, a region in northern Spain at 43° north latitude, with cloudy weather (annual means of sunshine time and global solar irradiation are 4.6 h/day and 3.4 kwh/m2, respectively). They had minor acute diseases or were being subjected to minor surgical procedures (inguinal hernia, strabismus, phimosis). All were reared at home by their parents, had no present or past chronic diseases, and were not receiving vitamin D supplements or other drugs known to interfere with calcium metabolism. For the purposes of this study, they were considered “normal children.”


Serum calcium, phosphorus, and ALP were measured by autoanalyzer (Hitachi 737, Hitachi, Tokyo, Japan). Serum 25(OH)D was measured by a competitive protein binding assay after purification by high performance liquid chromatography as previously reported in detail.(17) The interassay coefficient of variation (CV) was 10%. Serum 1,25(OH)2D was measured by a radioreceptor assay with calf thymus receptor (Nichols Institute, San Juan Capistrano, CA, U.S.A.). CV was 14%. Intact PTH was measured with a commercial kit (Nichols Institute); CV was 7%.


The results are expressed as means ± SD, unless otherwise indicated. Unpaired t-tests were used to compare the means of different groups. Changes within the same individual were analyzed by t-tests for paired data. Correlations were estimated as Pearson's coefficients. All tests were two-tailed, and p values less than 0.05 were considered as statistically significant.


Desirable serum 25(OH)D levels

Although we hypothesized that mentally retarded children were likely to have vitamin D deficiency, their serum 25(OH)D levels were not significantly different from those of normal children (19.2 ± 11.1 vs. 23.6 ± 11.2 ng/ml). Therefore, data from both subgroups were analyzed together. Baseline serum levels of 25(OH)D were significantly lower in March than in October (12.6 ± 5.5 vs. 29.9 ± 9.4 ng/ml, p < 0.001), but the average increment after the administration of exogenous 25(OH)D was similar in both months (Fig. 1). 1,25(OH)2D levels increased significantly after 25(OH)D supplementation in March, with the average increase being 12 pg/ml (p < 0.001). It was accompanied by a significant reduction in PTH levels (p < 0.002), without change in serum calcium or the other parameters. However, in October 25(OH)D supplementation resulted in a much lower increase in 1,25(OH)2D serum levels, the average being only 4 pg/ml (p < 0.05), and PTH did not change (Fig. 1).

Figure FIG. 1.

Changes in serum 25(OH)D (left axis), 1,25(OH)2D and PTH (right axis) induced by the administration of a 25(OH)D supplement (40 μg/day for 7 days) in March (black bars) and October (dashed bars). Mean and SEM. *p < 0.05 (postsupplementation vs. baseline levels); **p < 0.01; ***p < 0001.

When the changes in serum 1,25(OH)2D and PTH after 25(OH)D supplementation were related to the individual basal levels of 25(OH)D, an inverse relationship became evident. With low basal 25(OH)D levels, 1,25(OH)2D increased and PTH decreased after 25(OH)D administration. The correlation coefficients between basal serum 25(OH)D and postsupplementation changes in serum 1,25(OH)2D and PTH were r = 0.57 (p < 0.05) and r = 0.41 (p < 0.05), respectively. If serum 25(OH)D was above 10–12 ng/ml, the administration of an exogenous supplement of 25(OH)D was not followed by significant changes in serum PTH. When basal 25(OH)D levels were higher than 20 ng/ml, the supplement did not induce changes in either serum PTH or 1,25(OH)2D (Fig. 2).

Figure FIG. 2.

Average changes in serum 1,25(OH)2D (black bars) and PTH (dashed bars) induced by a 25(OH)D supplement, according to different intervals of basal 25(OH)D levels. *p < 0.05; **p < 0001.

Vitamin D intake

The average daily intake of calcium was 790 ± 156 mg. Vitamin D intake was 160 ± 80 IU/day in normal children and 72 ± 24 IU/day in those with mental deficiency. There was a positive correlation between dietary vitamin D intake and serum 25(OH)D in March (r = 0.57, p < 0.05), but not in October (r = 0.26).

Vitamin D status of normal children

As expected, 25(OH)D levels were higher in summer than in winter (29.1 ± 9.9 vs. 15.4 ± 5.4 ng/ml, p < 0.001). Among 43 children studied in summer, the lowest 25(OH)D level was 13 ng/ml, and 5 (12%) children had levels below 20 ng/ml (Fig. 3). Sixteen out of 51 children studied in winter (31%) had serum 25(OH)D levels below 12 ng/ml, and 40 out of 51 (80%) had levels lower than 20 ng/ml. We did not find significant seasonal differences in either serum 1,25(OH)2D, PTH, calcium, phosphate, or ALP. Nevertheless, in winter (but not in summer) there was a positive correlation between serum 25(OH)D and 1,25(OH)2D (r = 0.31, p < 0.05). Thus, children with serum 25(OH)D below 12 ng/ml also had lower 1,25(OH)2D levels than those with higher 25(OH)D (39 ± 11 vs. 47 ± 11 pg/ml, p < 0.05).

Figure FIG. 3.

Serum 25(OH)D levels in normal children studied in winter and summer months. Horizontal lines are drawn through 20 and 12 ng/ml, which would be the minimum desirable levels according to 1,25(OH)2D and PTH responses, respectively, to 25(OH)D supplementation (see text for details).


Severe vitamin deficiency results in rickets and osteomalacia, commonly associated with serum levels of 25(OH)D below 5 ng/ml.(18) However, during the last 15 years, the concept has emerged that the main consequences of mild vitamin D deficiency are secondary hyperparathyroidism and osteoporosis.(18–20)

The skin is considered the main source of vitamin D in humans. Accordingly, many studies in different countries have shown that serum 25(OH)D levels experience marked seasonal changes, both in adults and children,(17,21,22) related to the changes in ultraviolet irradiation. Dawson-Hughes et al.(23) reported that bone mineral density (BMD) decreases in winter months in healthy postmenopausal females, in association with a seasonal increase in PTH levels. Increasing vitamin D intake prevents both the rise in PTH and the decrease in BMD.(23,24) Thus, assuming that the decrease in BMD is an undesirable effect, we should consider that vitamin D availability in winter is insufficient.

However, it is unclear which are the desirable serum levels of 25(OH)D. Some investigators considered that the level separating vitamin deficiency from vitamin D–replete states would be the lower limit shown by healthy people in summer, which is about 12 ng/ml in our region (as shown in the present study) as well as in other European countries and the United States.(17,25,26) Peacock et al. addressed the issue by examining the response of serum 1,25(OH)2D to supplementation with 25(OH)D.(15) They found an inverse relationship between the change in serum 1,25(OH)2D induced by oral 25(OH)D and the basal serum 25(OH)D. At basal 25(OH)D levels above 20 ng/ml, no increase in serum 1,25(OH)2D was seen. The results of our study in children are remarkably similar. A limited availability of vitamin D may increase PTH secretion. Hence, the decrease in serum PTH levels following sunlight exposure or 25(OH)D supplementation suggests a suboptimal supply of vitamin D.(13,14,16,24,25) In the present study, only in children with serum 25(OH)D lower than 12 ng/ml did serum PTH decrease after 25(OH)D supplementation. Thus, our data also suggest that the threshold between vitamin D–deficient and vitamin D–replete states in children may be somewhere between 12 and 20 ng/ml (30–50 nmol/l).

Most children in our region have serum levels of 25(OH)D below the cut-off point of 20 ng/ml during the winter, and in roughly one-third of children the levels are lower than 12 ng/ml, indicating that they have an insufficient supply of vitamin D. Both an impaired cutaneous synthesis of vitamin D and an inadequate dietary supply seem to be responsible for this vitamin D insufficiency. At 42° north latitude, there is no ongoing conversion of 7-dehydrocholesterol to previtamin D3 in skin from November through February.(27) Thus, in winter vitamin D availability depends on the body stores accumulated in summer and on diet. However, vitamin D intake in our population was about 160 IU/day, well below the current recommended daily allowance of 400 IU/day.

Does mild vitamin D deficiency have any deleterious effect on bones of growing children? Our study does not permit an answer, but data from adult populations suggest that the answer should be positive. In postmenopausal women with a low intake of vitamin D, bone mass experiments a significant decrease in winter, which can be prevented by vitamin D supplementation.(23) Serum levels of 25(OH)D in the range that suggests subclinical vitamin D deficiency are associated with a reduced bone mass.(20,26,28) Furthermore, supplementation with vitamin D and calcium may decrease the incidence of bone fragility fractures.(9,29) However, recent data also suggest an inverse relationship between remodeling rate and bone density during skeletal growth.(10) Based on these studies, it seems likely that children with an insufficient vitamin D supply may be at risk of not reaching the peak bone mass allowed by their genetic background.

Mechanisms implicated in impairing bone homeostasis in individuals with vitamin D insufficiency might include an increased bone turnover and reduced calcium availability. Children with low 25(OH)D levels also had lower levels of 1,25(OH)2D, which may impair their ability to actively absorb dietary calcium. Moreover, although 1,25(OH)2D is the most potent vitamin D metabolite, 25(OH)D itself may have a stimulatory effect on intestinal calcium absorption.(30,31) When calcium intake is high, the nonvitamin D–dependent passive absorption increases. However, in the case of low intake, the vitamin D–dependent active transport becomes the major route for calcium absorption. Therefore, the effect of vitamin D deficiency on calcium bioavailability may be particularly notorious when it coexists with a low dietary intake of calcium.(14,31) In growing children the influence of calcium availability on the gain of bone mineral density has been demonstrated.(5,6) Thus, a reduced supply of both vitamin D and calcium, common in individuals who do not consume adequate amounts of dairy products, may have a particularly deleterious effect on bone tissue accumulation.

In conclusion, based on the response of 1,25(OH)2D serum levels to 25(OH)D supplementation, 20 ng/ml appears to be the minimum desirable level of 25(OH)D. A level of 12 ng/ml would be the minimum acceptable on the basis of PTH response and normal levels found in summer months. Most children in our region have 25(OH)D levels lower than 20 ng/ml in winter, and about one-third have levels below 12 ng/ml. Lacking a long prospective study comparing vitamin D levels during the growing years and the peak bone mass attained after adolescence, the relationship between such subclinical vitamin D deficiency and peak bone mass remains speculative. Nevertheless, based on the present knowledge about bone physiology, efforts to maintain an optimal supply of both calcium and vitamin D throughout the year seem warranted in order to facilitate bone accumulation and prevent osteoporosis later in life. The administration of a single oral dose of 150,000 IU of vitamin D early in winter may be a convenient way to maintain 25(OH)D levels within the desirable range.(21,22)