Vitamin D Deficiency Associated With Increased Incidence of Gastrointestinal and Ear Infections in School-age Children

Kathryn A. Thornton, DMD, MPH; Constanza Marín, RD; Mercedes Mora-Plazas, MSc, RD; Eduardo Villamor, MD, DrPH

Pediatr Infect Dis J. 2013;32(5):585-593.

Results

The mean (± standard deviation) age of children at recruitment was 8.9 ± 1.6 years, and 52% of children were girls. The mean (± standard deviation) plasma concentration of 25(OH)D was 73.2 ± 19.8 nmol/L; 10.1% of children were vitamin D-deficient, whereas 46.7% of children were vitamin D-insufficient. Children contributed 62,642 days of observation with a median (interquartile range) 140 (91, 182) days per child; the distribution of total child-days did not vary significantly by vitamin D status. At baseline, vitamin D serostatus was inversely associated with female sex, age, body mass index-Z and single mother status (Table 1).

Children with VDD had higher rates of vomiting, diarrhea with vomiting and earache or ear discharge with fever than vitamin D-sufficient children (Table 2). Compared with children who were vitamin D-sufficient, those who were deficient had twice as many days with diarrhea and vomiting after adjustment for child's age, sex and household socioeconomic status (P = 0.009). In addition, vitamin D-deficient children had 2.4 times as many days with earache or ear discharge with fever compared with vitamin D-sufficient children (P = 0.008). This association was particularly strong among boys (adjusted IRR: 5.74; 95% CI: 2.32, 14.18) compared with girls (adjusted IRR: 1.09; 95% CI: 0.46, 2.60; P test for interaction with sex = 0.001).

We noted that vitamin D-insufficient (25(OH)D ≥50 and <75 nmol/L) children had lower rates of earache or ear discharge with fever and cough with fever, compared with children with 25(OH)D ≥75 nmol/L. The relation with cough and fever was mainly apparent in girls; vitamin D-insufficient girls reported 66% fewer days of cough with fever than girls who were vitamin D-sufficient (adjusted IRR: 0.34; 95% CI: 0.26, 0.43), whereas no significant association was observed in boys (adjusted IRR: 1.10; 95% CI: 0.81, 1.50; P test for interaction with sex < 0.0001).

1 2 3 4

Next Section

Table 1. Distribution of Sociodemographic Characteristics According to Vitamin D Status in School-age Children From Bogotá, Colombia
	Vitamin D-deficient	Vitamin D-insufficient	Vitamin D-sufficient	P*
	25(OH)D <50 nmol/L	25(OH)D ≥50 and <75 nmol/L	25(OH)D ≥75 nmol/L	P*
Total (n)	48	222	205
Child characteristics
Girls (%)	62.5	57.2	44.9	0.004
Age (yr)	9.2 ± 1.8	9.1 ± 1.5	8.6 ± 1.6	0.001
Stunted^† (%)	12.5	9.6	7.0	0.4
BMI-for-age Z score^‡	0.3 ± 1.2	0.2 ± 1.0	0.0 ± 1.0	0.02
Time spent playing outdoors (h/wk)	6.6 ± 8.7	7.3 ± 7.9	7.4 ± 8.4	0.6
Maternal characteristics
Single mother (%)	35.4	27.7	17.0	0.002
Education (yr)	8.5 ± 3.5	8.5 ± 3.3	8.7 ± 3.4	0.5
BMI (kg/m²)	24.6 ± 4.6	24.8 ± 4.0	24.0 ± 3.9	0.09
Lowest socioeconomic strata (%)^§	52.1	37.1	41.0	0.5

*For all binary variables except stunting, P value is from the Cochran-Armitage test for trend. For stunting, P value is estimated with Fisher exact test. For continuous characteristics, P value was from a test for linear trend for a variable representing 25(OH)D categories as continuous.
^†Height-for-age Z score less than −2.0 according to the World Health Organization (WHO) reference.³⁵
^‡According to the WHO reference.³⁵
^§Strata 1 or 2 of a maximum of 4 in the sample, according to the government's classification of households for tax and planning purposes.
BMI indicates body mass index.

Table 2. Morbidity According to Vitamin D Status in Colombian Schoolchildren
	Vitamin D-deficient	Vitamin D-insufficient	Vitamin D-sufficient
	<50 nmol/L	≥50 and <75 nmol/L	≥75 nmol/L
Total (n)	48	222	205
Child-days of observation	6707	29,723	26,212
Diarrhea
Days with symptom	78	342	275
Rate per child-year	4.24	4.20	3.83
Unadjusted rate ratio (95% CI)	1.11 (0.86, 1.43)	1.10 (0.94, 1.29)	1.0
Adjusted rate ratio (95% CI)*	1.17 (0.90, 1.52)	1.00 (0.85, 1.18)	1.0
Vomiting
Days with symptom	99	302	276
Rate per child-year	5.39	3.71	3.84
Unadjusted rate ratio (95% CI)	1.40 (1.11, 1.76)	0.97 (0.82, 1.14)	1.0
Adjusted rate ratio (95% CI)	1.30 (1.02, 1.66)	0.86 (0.72, 1.01)	1.0
Diarrhea with vomiting
Days with symptom	21	47	46
Rate per child-year	1.14	0.58	0.64
Unadjusted rate ratio (95% CI)	1.78 (1.06, 2.99)	0.90 (0.60, 1.35)	1.0
Adjusted rate ratio (95% CI)	2.05 (1.19, 3.53)	0.92 (0.61, 1.39)	1.0
Earache/discharge with fever
Days with symptom	16	16	35
Rate per child-year	0.87	0.20	0.49
Unadjusted rate ratio (95% CI)	1.79 (0.99, 3.23)	0.40 (0.22, 0.73)	1.0
Adjusted rate ratio (95% CI)	2.36 (1.26, 4.44)	0.35 (0.19, 0.65)	1.0
Cough with fever
Days with symptom	55	177	253
Rate per child-year	2.99	2.17	3.52
Unadjusted rate ratio (95% CI)	0.85 (0.63, 1.14)	0.62 (0.51, 0.75)	1.0
Adjusted rate ratio (95% CI)	0.77 (0.57, 1.04)	0.53 (0.44, 0.65)	1.0

*From Poisson regression models adjusted for child's age, sex and household socioeconomic stratum

Table 3. Epidemiologic Studies Evaluating the Relation of Vitamin D Serostatus With Infection Outcomes in Infants, Children and Adolescents
Author, Publication Year, Reference	Type of Study	Location	Sample Size	Age of Children	Health Status of Children	Vitamin D Dosage/Exposure	Outcome	Results	Variables Adjusted for
Morcos et al, 1998⁵³	RCT	Egypt	24	1.5–13 yr	Tuberculosis (TB)	Vitamin D3 (VD3) 1000 IU/d + TB treatment vs. TB treatment only for 2 mo	Improvement in TB signs and symptoms after 2 mo	VD3 group: more evident improvement clinically and on abdominal sonography compared with controls
Urashima et al, 2010¹⁶	RCT	Japan	430	6–15 yr	With or without underlying diseases	VD3 1200 IU/d vs. placebo daily for 4 mo	Influenza A, influenza B, influenza-like illness, pneumonia, gastroenteritis over 4 mo	VD3 group: ↓ influenza A (RR: 0.58; 95% CI: 0.34, 0.99; P = 0.04); ↔ influenza B, influenza-like illness, pneumonia, gastroenteritis
Manaseki- Holland et al, 2010¹⁸	RCT	Afghanistan	453	1–36 mo	Pneumonia	VD3 100,000 IU vs. placebo single dose	Pneumonia over 3 mo	↔ Duration of initial pneumonia episode (P = 0.2); VD3 group: ↓ risk of (RR: 0.78; 95% CI: 0.64, 0.94; P = 0.01) and ↑ time to (HR: 0.71; 95% CI: 0.53, 0.95; P = 0.02) repeat episode of pneumonia
Camargo et al, 2012¹⁷	RCT	Mongolia	247	9–11 yr	Presumably healthy	VD3 300 IU/d in Mongolian milk vs. regular Mongolian milk daily for 7 wk	Parent-reported ARI over 3 mo	VD3 group: ↓ ARI (RR: 0.52; 95% CI: 0.31, 0.89)
Manaseki-Holland et al, 2012¹⁹	RCT	Afghanistan	3046	1–11 mo	Presumably healthy	VD3 100,000 IU vs. placebo every 3 mo for 18 mo	Pneumonia, hospital admissions, all-cause mortality over 18 mo	↔ Incidence (IRR: 1.07; 95% CI: 0.90, 1.27; P = 0.48) or severity (IRR: 1.13; 95% CI: 0.74, 1.73; P = 0.58) of pneumonia, hospital admissions, or all-cause mortality; VD3 group: ↑ repeat pneumonia episodes (IRR: 1.69; 95% CI: 1.28, 2.21; P < 0.0001)
Mehta et al, 2009⁵⁴	Prospective cohort	Tanzania	884	Neonates	Born to HIV-infected women	Maternal serum 25(OH)D <80 nmol/L	HIV infection, mortality up to 2 yr of age	↑ HIV infection (RR: 1.46; 95% CI: 1.11, 1.91; P < 0.01) and mortality (RR: 1.58; 95% CI: 1.26, 1.97; P < 0.01)	Maternal age, HIV disease stage, CD4 cell counts at baseline, multivitamin regimen
Belderbos et al, 2011¹⁵	Prospective cohort	The Netherlands	156	Neonates	Healthy	Mean cord blood 25(OH)D; VDD*	RSV lower respiratory tract infection (LRTI) by 1 yr of age	Mean cord blood 25(OH)D: ↓ RSV LRTI (P = 0.02); VDD:↑ RSV LRTI (RR: 6.2; 95% CI: 1.6, 24.9; P = 0.01)	Birth month, birth weight, maternal ethnicity
Camargo et al, 2011⁵⁵	Prospective cohort	New Zealand	922	Neonates	Apparently healthy	Cord blood 25(OH) D <25 nmol/L	Respiratory infection, any infection by 3 mo of age	↑ respiratory (OR: 2.16; 95% CI: 1.35, 3.46) and any (OR: 2.21; 95% CI: 1.26, 3.90) infection	Season of birth
Finkelstein et al, 2012⁵⁶	Prospective cohort	Tanzania	609	Neonates	Born to HIV-infected women	Maternal serum 25(OH)D <80 nmol/L	Respiratory and gastrointestinal morbidity over median of 58 mo	↑ cough (RR: 1.11; 95% CI: 1.02, 1.21; P = 0.01); ↔ cough and fever (RR: 1.08; 95% CI: 0.93, 1.26; P = 0.29), other respiratory symptoms, or diarrhea (RR: 1.02; 95% CI: 0.87, 1.21; P = 0.79)	Maternal and child factors†
Morales et al, 2012⁵⁷	Prospective cohort	Spain	1693	Infants	Presumably healthy	Maternal plasma 25(OH)D in pregnancy	LRTI by 1 yr of age	↓ trend for LRTI across cohort and season-specific quartiles of maternal 25(OH)D (Q4 vs. Q1: OR: 0.67; 95% CI: 0.50, 0.90; test for trend, P = 0.02)	Sex, number of siblings, breastfeeding duration, day-care attendance; maternal education, smoking, prepregnancy BMI and asthma
Wayse et al, 2004¹²	Case-control	India	150	<5 yr	Severe acute lower respiratory tract infection (ALRI) vs. healthy controls	Serum 25(OH)D >22.5 nmol/L	Severe ALRI	↓ severe ALRI (OR: 0.09; 95% CI: 0.03, 0.24; P < 0.001)	Age, height-for-age Z score, breastfeeding history, hemoglobin level, sun exposure, cooking fuel
Stephensen et al, 2006⁵⁸	Case-control	United States	359	14–23 yr	HIV-infected vs. HIV-uninfected controls	Mean plasma 25(OH)D; plasma 25(OH)D <37.5 nmol/L	HIV infection	↔ mean plasma 25(OH)D (P = 0.62) or prevalence of plasma 25(OH)D <37.5 nmol/L (P = 0.91)	Sex, sex × HIV status
Karatekin et al, 2009¹⁴	Case-control	Turkey	40	Neonates	ALRI vs. healthy controls	Mean serum 25(OH) D; serum 25(OH) D <25 nmol/L or <50 nmol/L	ALRI	Mean serum 25(OH)D: lower in cases (P = 0.01); serum 25(OH) D <25 nmol/L: ↑ ALRI (OR: 4.25; 95% CI: 1.06, 17.07; P = 0.04); serum 25(OH)D <50 nmol/L: ↔ ALRI (P = 0.50)
McNally et al, 2009⁵⁹	Case-control	Canada	197	<5 yr	ALRI vs. hospital controls without respiratory symptoms	Mean serum 25(OH) D; VDD	ALRI	Mean serum 25(OH)D: ↔ (P = 0.71); VDD:↑ admission to pediatric intensive care unit for ALRI (OR: 8.23; 95% CI: 1.4, 48.0; P = 0.02)	Age, sex, prematurity status
Roth et al, 2009⁶⁰	Case-control	Canada	129	1–25 mo	ALRI vs. controls undergoing elective surgery with no history of ALRI	Mean serum 25(OH) D; 10 nmol/L increase in serum 25(OH)D	ALRI	Mean serum 25(OH)D: ↔ (P = 0.96); ↑ in serum 25(OH)D of 10 nmol/L: ↔ ALRI (OR: 1.00; 95% CI: 0.998, 1.002)
Oduwole et al, 2010⁶¹	Case-control	Nigeria	34	2–60 mo	Pneumonia vs. hospital controls without pneumonia	Mean plasma 25(OH)D	Pneumonia	↔ (P = 0.50)
Roth et al, 2010¹¹	Case-control	Bangladesh	50	1–18 mo	ALRI vs. controls matched on age, sex and village	Mean serum 25(OH) D; 10 nmol/L increase in serum 25(OH)D	ALRI	Mean serum 25(OH)D: 10.0 nmol/L lower in cases (95% CI: 6.32, 13.68; P = 0.01); ↑ in serum 25(OH)D of 10 nmol/L: ↓ ALRI (OR: 0.23; 95% CI: 0.06, 0.81)	Underweight status, socioeconomic status
Elemraid et al, 2011⁶²	Case-control	Yemen	149	0.6–15 yr	Chronic suppurative otitis media (CSOM) vs. healthy controls	Mean serum 25(OH) D; VDD	CSOM	Mean serum 25(OH)D: 2.0 nmol/L higher in cases than controls (95% CI: 1.78, 2.28; P = 0.001); VDD: ↓ CSOM (OR: 0.22; 95% CI: 0.06, 0.82; P = 0.02)	Age
Rutstein et al, 2011⁶³	Case-control	United States	453	5–23 yr	Perinatally acquired HIV vs. healthy controls	Mean serum 25(OH) D; serum 25(OH) D <27.5 nmol/L	HIV infection	Mean serum 25(OH)D: lower in cases (P < 0.0001); serum 25(OH) D <27.5 nmol/L: ↔ HIV infection (OR: 0.6; 95% CI: 0.2, 1.8; P = 0.33)	Age, sex, BMI Z score, season, season × group, race
Williams et al, 2008⁶⁴	Cross-sectional	United Kingdom	64	0.1–17 yr	Active or latent TB		Prevalence of serum 25(OH)D <20 nmol/L and <75 nmol/L	25(OH)D <20 nmol/L: 37.5% of children 25(OH)D <75 nmol/L: 86% of children
Bener et al, 2009⁶⁵	Cross-sectional	Qatar	458	<16 yr	Healthy	VDD	Gastroenteritis	↑ Gastroenteritis (P = 0.02)
Katikaneni et al, 2009⁶⁶	Cross-sectional	United States	315	<3 mo	Hospital and pediatric clinic sample	Vitamin D supplementation	Urinary tract infection (UTI)	↑ UTI in infants exclusively formula-fed (RR: 2.24; 95% CI: 1.29, 3.90)
Gray et al, 2012⁶⁷	Cross-sectional	Australia (refugees)	328	0.5–17.5 yr	Active, latent, or no TB	Mean serum 25(OH) D; VDD	TB	Mean serum 25(OH)D: lower in children with active (P = 0.05) or latent TB (P = 0.03);VDD: ↔ (P = 0.1)
Inamo et al, 2011¹³	Case series	Japan	28	1–48 mo	ALRI	Serum 25(OH)D <25 nmol/L	Severity of ALRI	↑ Need for supplementary oxygen or ventilator management (P < 0.01)

Direction of the arrow refers to direction of association between exposure and outcome:↑ indicates statistically significant positive association with outcome;↓, statistically significant inverse association with outcome; ↔, no association with outcome.
*VDD defined as serum or plasma 25(OH)D <50 nmol/L.
†Maternal factors adjusted for included age, education, occupation, WHO HIV disease stage, CD4 T-cell counts, short stature, anemia, mid-upper arm circumference and multivitamin regimen. Child factors adjusted for included sex, gestational age, low birth weight, mid-upper arm circumference, breast-feeding status and HIV status.
RCT indicates randomized controlled trial; RR, relative risk; OR, odds ratio; BMI, body mass index; HR, hazard ratio; ARI, acute respiratory infection.