Participant Characteristics
A total of 55,925 participants from the 12 cities of Heilongjiang were included in this study, in which 30,683 (54.86%) were male. Table 1 presented the basic characteristics of the participants. The average serum level of 25(OH)D for all participants was 26.13±12.30 ng/mL. The median age of participants was 3.10 years old (interquartile range, 1.10–6.50 years old). Of the children’s outdoor time, 35.63%, 31.95%, and 32.42% were below 30min/d, 30-60 min/d and over 60min/d, respectively. Furthermore, 50.65% of the subjects in no vitamin D intervention group, 32.87% in supplementation group, and 16.48% in therapeutic group. The vitamin D level was higher in boys, younger children, children recruited in summer, with prolonged outdoor time, and those with vitamin D intervention (P < 0.001).
Table 1. 25(OH)D concentration in children aged 0–18 years old
|
N (%)
|
25(OH)D (ng/mL)
|
P value
|
Total
|
55,925
|
26.13±12.30
|
|
Gender
|
|
|
<0.001
|
Boys
|
30,683 (54.86)
|
26.42±12.07
|
|
Girls
|
25,241 (45.13)
|
25.78±12.56
|
|
Age (year)
|
|
|
<0.001
|
0< age <1
|
11,953 (21.37)
|
33.25±13.64
|
|
1≤ age <3
|
14,901 (26.64)
|
30.71±11.44
|
|
3≤ age <6
|
13,124 (23.47)
|
23.74±9.37
|
|
6≤ age <12
|
12,416 (22.20)
|
19.12±8.70
|
|
12≤ age <18
|
3531 (6.31)
|
16.24±8.30
|
|
Season
|
|
|
<0.001
|
Spring
|
15,795 (28.24)
|
25.98±11.77
|
|
Summer
|
17,421 (31.15)
|
26.58±13.07
|
|
Autumn
|
10,205 (18.25)
|
26.51±12.08
|
|
Winter
|
12,504 (22.36)
|
25.37±11.95
|
|
Ethnic
|
|
|
0.050
|
Han
|
54,967 (98.29)
|
26.12±12.29
|
|
Ethnic minorities in China
|
958 (1.71)
|
26.91±12.67
|
|
Outdoor time
|
|
|
<0.001
|
<30 min/d
|
19,929 (35.63)
|
21.61±11.81
|
|
30–60 min/d
|
17,867 (31.95)
|
27.65±11.47
|
|
>60 min/d
|
18,129 (32.42)
|
29.60±12.11
|
|
BMI for age
|
|
|
<0.001
|
Normal
|
51,294 (91.72)
|
26.58±12.26
|
|
Overweight
|
2851 (5.10)
|
21.84±12.43
|
|
Obesity
|
1780 (3.18)
|
20.07±9.90
|
|
Intervention
|
|
|
<0.001
|
None
|
28,328 (50.65)
|
20.60±8.65
|
|
Supplementation intervention
|
18,383 (32.87)
|
29.80±9.49
|
|
Therapeutic intervention
|
9214 (16.48)
|
35.80±17.14
|
|
NOTE: P < 0.05 shows a significant difference between the groups.
Characteristics of Children with Different Vitamin D Statuses
The overall rate of hypovitaminosis D was 65.60%. 6.57%, 25.51%, and 33.52% of the children included were found with severe deficiency, deficiency and insufficiency, respectively. Moreover, 18.11% children with severe deficiency were overweight or obese, compared with 11.34% and 8.17% in those with deficiency and insufficiency, respectively. However, the rate of overweight and obesity decreased to 4.21% in children with vitamin D sufficiency (P < 0.001). Children who have outdoor time of 30–60 min/d and over 60 min/d showed a percentage of 34.31% and 39.97% in the vitamin D sufficiency group, whereas it dropped to 25.72% in the < 30 min/d group (P < 0.001). In children with normal vitamin D status, 46.70% and 27.12% had supplementation and therapeutic doses of vitamin D, respectively; in contrast, 26.18% received no intervention, accounting for 90.31% patients with severe vitamin D deficiency (Table 2). Among the influential factors, outdoor activity time and intervention methods are possibly causal factors determine the vitamin D status in children, and also changeable factors for prevention of hypovitaminosis D.
Table 2. Characteristics of children with varying serum levels of 25(OH)D
|
Severe deficient
|
Deficient
|
Insufficient
|
Sufficient
|
P value
|
Vitamin D (ng/mL,`x±s)
|
7.45±2.19
|
15.35±2.87
|
25.04±2.81
|
38.74±10.57
|
<0.001
|
Percentage (%)
|
6.57
|
25.51
|
33.52
|
34.40
|
|
Male (%)
|
51.20
|
53.07
|
55.42
|
56.34
|
<0.001
|
Age (years,`x±s)
|
7.54±4.99
|
6.65±3.94
|
4.18±3.38
|
2.09±2.39
|
<0.001
|
Ethnic (Han, %)
|
98.31
|
98.49
|
98.28
|
98.14
|
0.063
|
Overweight and obesity (%)
|
18.11
|
11.34
|
8.17
|
4.21
|
<0.001
|
Season (%)
|
|
|
|
|
<0.001
|
Spring
|
30.50
|
29.82
|
31.12
|
32.29
|
|
Summer
|
26.53
|
29.33
|
28.19
|
27.83
|
|
Autumn
|
16.31
|
17.35
|
18.69
|
18.85
|
|
Winter
|
26.66
|
23.50
|
22.00
|
21.03
|
|
Outdoor activity time (%)
|
|
|
|
|
<0.001
|
<30 min/d
|
58.69
|
56.23
|
25.63
|
25.72
|
|
30–60 min/d
|
23.56
|
22.14
|
38.63
|
34.31
|
|
>60 min/d
|
17.75
|
21.63
|
35.74
|
39.97
|
|
Intervention methods (%)
|
|
|
|
|
<0.001
|
No intervention
|
90.31
|
74.10
|
50.16
|
26.18
|
|
Supplementation intervention
|
7.27
|
17.94
|
35.06
|
46.70
|
|
Therapeutic intervention
|
2.43
|
7.96
|
14.79
|
27.12
|
|
`x±s and percentage (%) were used to describe the continuous and categorical variables, respectively; P<0.05 was regarded as statistically significant.
Association of Intervention Methods with Vitamin D Status
After adjusting for age, sex, BMI for age, season, and outdoor activity time, an inverse association of vitamin D intervention with hypovitaminosis D was observed. The OR (95% CI) of hypovitaminosis D was 0.423 (0.404, 0.443) in children with supplementation doses and 0.191 (0.180, 0.202) in children with therapeutic doses compared with children with no supplements (P for trend <0.001, Table 3). After stratification by age group, the negative association of vitamin D interventions (including supplementation and therapeutic doses) with hypovitaminosis D existed in each age group (Supplement. Table.1).
Table 3. Association of vitamin D intervention with hypovitaminosis D analyzed by logistic regression
Intervention
|
Model 1
|
Model 2
|
Model 3
|
No intervention
|
1
|
1
|
1
|
Supplementation intervention
|
0.435 (0.415, 0.455)
|
0.435 (0.415, 0.455)
|
0.423 (0.404, 0.443)
|
Therapeutic intervention
|
0.202 (0.191, 0.214)
|
0.202 (0.191, 0.214)
|
0.191 (0.180, 0.202)
|
P for trend
|
<0.001
|
<0.001
|
<0.001
|
Associations were examined with multivariable logistic regression. Model 1: adjusted for age and sex. Model 2: adjusted for BMI for age on the basis of model 1. Model 3: adjusted for season and outdoor time on the basis of model 2.
Association of the Intervention Methods with Vitamin D Status in Children with Varying Outdoor Time
We conducted logistic regression to further eliminate the influence of outdoor activity time on the relationship between intervention methods and serum 25(OH)D level. Both therapeutic and supplementation intervention reduced the risk of hypovitaminosis D regardless of outdoor time compared to those with no vitamin D intervention. In children whose outdoor time was > 60 min/d, the OR was 0.196 (0.178, 0.216) in the former way, whereas 0.508 (0.470, 0.550) in the latter way, compared with those who were not given intervention (Table 4).
Table 4. Logistic regression analysis of the association of vitamin D intervention methods with hypovitaminosis D in children with varying outdoor time
Intervention methods
|
Model 1
|
Model 2
|
Model 3
|
>60 min/d
|
|
|
|
No intervention
|
1
|
1
|
1
|
Supplementation intervention
|
0.508 (0.470, 0.549)
|
0.508 (0.469, 0.549)
|
0.508 (0.470, 0.550)
|
Therapeutic intervention
|
0.196 (0.177, 0.216)
|
0.196 (0.177, 0.216)
|
0.196 (0.178, 0.216)
|
P for trend
|
<0.001
|
<0.001
|
<0.001
|
30–60 min/d
|
|
|
|
No intervention
|
1
|
1
|
1
|
Supplementation intervention
|
0.362 (0.332, 0.394)
|
0.362 (0.332, 0.394)
|
0.361 (0.332, 0.394)
|
Therapeutic intervention
|
0.183 (0.164, 0.205)
|
0.184 (0.164, 0.205)
|
0.183 (0.164, 0.205)
|
P for trend
|
<0.001
|
<0.001
|
<0.001
|
<30 min/d
|
|
|
|
No intervention
|
1
|
1
|
1
|
Supplementation intervention
|
0.376 (0.345, 0.411)
|
0.377 (0.346, 0.411)
|
0.377 (0.346, 0.411)
|
Therapeutic intervention
|
0.176 (0.160, 0.194)
|
0.174 (0.159, 0.192)
|
0.174 (0.159, 0.192)
|
P for trend
|
<0.001
|
<0.001
|
<0.001
|
Associations were examined using multivariable logistic regression. Model 1: adjusted for age and sex. Model 2: adjusted for BMI for age on the basis of model 1. Model 3: adjusted for season on the basis of model 2.
Association of Outdoor Time with Vitamin D Status
We found another inverse association of prolonged outdoor time with hypovitaminosis D after age, sex, BMI for age, season, and intervention methods were adjusted. The OR (95% CI) of hypovitaminosis D was 0.737 (0.701, 0.776) in children with outdoor time of 30–60 min/d and 0.479 (0.456, 0.504) in those with 60 min/d compared to those with less than 30 minutes (P for trend < 0.001, Table 5). Besides, after age stratification, prolonged outdoor time was negatively associated with hypovitaminosis D in all age groups (Supplement. Table.2).
Table 5. Logistic regression analysis of the association of outdoor time with hypovitaminosis D
Outdoor time (min/d)
|
Model 1
|
Model 2
|
Model 3
|
<30
|
1
|
1
|
1
|
30–60
|
0.696 (0.663, 0.731)
|
0.697 (0.663, 0.732)
|
0.737 (0.701, 0.776)
|
>60
|
0.533 (0.508, 0.560)
|
0.534 (0.508, 0.560)
|
0.479 (0.456, 0.504)
|
P for trend
|
<0.001
|
<0.001
|
<0.001
|
Associations were examined using multivariable logistic regression. Model 1: adjusted for age and sex. Model 2: adjusted for BMI for age on the basis of model 1. Model 3: adjusted for season and intervention methods on the basis of model 2.
Association of Outdoor Time with Vitamin D Status in Various Intervention Methods
The results of logistic regression models suggested that there was a decline in the risk of hypovitaminosis D with increased outdoor time whether vitamin D intervention was used [0.858 (0.769, 0.957) for 30–60 min/d and 0.469 (0.420, 0.525) for over 60 min/d compared with below 30 min/d in children with therapeutic vitamin D intervention, P for trend<0.001, Table 6].
Table 6. Logistic regression analysis of the effect of outdoor time on hypovitaminosis D in children with vitamin D intervention methods
Outdoor time (min/d)
|
Model 1
|
Model 2
|
Model 3
|
Therapeutic intervention
|
|
|
|
<30
|
1
|
1
|
1
|
30–60
|
0.849 (0.762, 0.947)
|
0.857 (0.768, 0.956)
|
0.858 (0.769, 0.957)
|
>60
|
0.464 (0.415, 0.519)
|
0.469 (0.420, 0.525)
|
0.469 (0.420, 0.525)
|
P for trend
|
<0.001
|
<0.001
|
<0.001
|
Supplementation intervention
|
|
|
|
<30
|
1
|
1
|
1
|
30–60
|
0.704 (0.650, 0.761)
|
0.704 (0.650, 0.762)
|
0.704 (0.650, 0.762)
|
>60
|
0.608 (0.557, 0.663)
|
0.608 (0.558, 0.663)
|
0.608 (0.558, 0.663)
|
P for trend
|
<0.001
|
<0.001
|
<0.001
|
No intervention
|
|
|
|
<30
|
1
|
1
|
1
|
30–60
|
0.798 (0.728, 0.874)
|
0.797 (0.728, 0.874)
|
0.796 (0.727, 0.873)
|
>60
|
0.422 (0.390, 0.455)
|
0.421 (0.390, 0.455)
|
0.427 (0.390, 0.455)
|
P for trend
|
<0.001
|
<0.001
|
<0.001
|
Associations were examined using multivariable logistic regression. Model 1: adjusted for age and sex. Model 2: adjusted for BMI for age on the basis of model 1. Model 3: adjusted for season on the basis of model 2.