English (pdf)
Article in xml format
Article references
Automatic translation
Send this article by e-mail
Cited by SciELO 
Cited by Google 
Similars in
SciELO 
Similars in Google 
Permalink
INTRODUCTION
Vitamin D (VD) is a prohormone that plays a key role in humans, since it has to do with the balance of calcium, phosphorus, and bone structure1. VD deficiency has negative effects on body composition, bone metabolism2, cardiovascular, neurological, and respiratory systems; and the immune response3. Various studies have shown an association between lower VD levels and diseases such as asthma, multiple sclerosis, glucose intolerance, diabetes, arterial hypertension, acute respiratory infection, obesity, cancer, and cardiovascular diseases4-6.
It is estimated that nearly one billion people worldwide have VD deficiency or insufficiency7, representing a global public health problem, even in countries with adequate sun exposure throughout the year8. In the pediatric population, the prevalence of VD deficiency is between 31.4 % and 45.6 %9, and can reach 55 % in critically ill patients10.
In critically ill pediatric patients, VD deficiency and insufficiency have been associated with greater disease severity10, the need for ICU interventions, the need for ventilator support, vasoactive agents, and the risk of sepsis11-14. Additionally, it has been shown that VD deficiency during ICU stay increases the risk of dying by up to two times, compared to pediatric patients with normal VD levels15,16.
In Latin America, only one study has been conducted, in Chile, by Bustos et al. (2016). It showed an association between VD deficiency with the use of vasoactive drugs (RR 1.6; 95 % CI 1.2 - 2.3), mechanical ventilation (RR 2.2; 95 % CI 1.2 - 3.9), septic shock (RR 1.9; 95 % CI 1.3 - 2.9), and need for resuscitation fluids in the first 24 hours (RR 1.5; 95 % CI 1.1 - 2.1) 17. In Colombia, the prevalence of VD deficiency or insufficiency in pediatric patients admitted to intensive care units, and its association with disease severity, are unknown. The objective of this study was to evaluate the VD status and its association with adverse clinical outcomes in critically ill pediatric patients.
MATERIALS AND METHODS
Design
An observational follow-up study was conducted on a prospective cohort of pediatric patients admitted to the intensive care units (ICU) of two health institutions, between August 2021 and February 2022. The two participating health institutions were highly complex, and they care for patients from all socioeconomic strata in the southern region of Colombia. All admitted patients had their 25-OH-VD levels measured and were under followed-up until discharge from the ICU.
Participants
Pediatric patients between one-month old and 18 years, admitted to the pediatric ICU for medical or surgical reasons, and whose parents or legal guardian had signed the informed consent form. Patients with renal, hepatic or parathyroid endocrine disease (hyper/hypoparathyroidism) prior to ICU admission, with proven malabsorptive status, were excluded. Patients with known VD resistance or VD supplementation three months prior to admission were also excluded. The selection of the participants was carried out consecutively by including all the patients who met the inclusion criteria between August 2021 and February 2022.
Sample Size
The sample size calculation was carried out considering the prevalence of VD deficiency reported in the literature to be 54.8 %10, with a confidence level of 95 %, and an accuracy of 7.4 %, obtaining a sample size of 175 patients.
Variables
The exposure variable was VD level, which conformed three cohorts. Deficiency with a 25-OH-VD concentration less than 20 ng/ml; insufficiency, less than 30ng/ml; and normal VD levels, equal to or more than 30ng/ml. The outcome variable was disease severity, which included mortality, the need for mechanical ventilation, inotropes and renal replacement therapy in the ICU. The independent variables were sex, age, medical diagnosis, PRISM scale score, PELOD scale, and nutritional status, which was evaluated for children under and over 5 years of age. Paraclinical variables including phosphorus, calcium, and blood count, were also evaluated.
Data Source
A biological sample of VD (25-OH-VD) was collected on the date of admission to the ICU, through phlebotomy and extraction of 2 ml of blood, following the institutional protocol. The samples were protected from sunlight, centrifuged, and refrigerated. The analysis was performed using a Roche electrochemiluminescence immunoassay. The collection, processing, and analysis procedures were similar between the two institutions where the patients were recruited. Information on paraclinical tests, medical diagnosis, and nutritional status was collected during the first 24 hours of admission to the unit. Outcomes were identified every day through medical assessments, until the patient was discharged from the ICU.
Statistical Analysis
Categorical variables were analyzed using absolute and relative frequencies, whereas quantitative variables were described using measures of central tendency (median) and dispersion (interquartile range). The Wilcoxon rank test was used to identify significant difference in the level of VD, according to the occurrence of the evaluated outcomes. To calculate adjusted relative risk (95 % of confidence), four models of log-binomial regression were constructed for each outcome (mortality, mechanical ventilation, use of inotropes, and renal replacement therapy). Independent variables were deficient, insufficient, or suboptimal levels of VD, with covariables including age and severity of illness using the PRISM scale. These covariables were chosen based on clinical criteria. Due to convergence methods that might emerge with log-binomial regression, for this study, a generalized lineal model with binomial family (link="log") was used. Statistical analysis was performed using the RStudio 3.4 program.
RESULTS
Clinical Aspects
During the study period, 175 pediatric patients admitted to the ICU from two health institutions were followed-up. 107 of them were male. Half of the patients were 48 months old (table 1).
Forty-eight percent of patients under 5 years of age had a normal weight-for-age. The prevalence of acute underweight (risk, moderate, and/or severe) was 44.44 % (table 1). In patients older than 5 years, 3 of every five patients had a normal weight (58.97 %), and the prevalence of overweight or obesity in patients older than 5 years was 15.37 % (table 1).
Table 1 Sociodemographic variables of pediatric patients admitted to the Intensive Care Unit in two health institutions
| Data | Total N= 175 |
|---|---|
|
Age (months) Median (IR) Mean (SD) |
48 (11.50 - 119.50) 70.28 (63.96) |
|
Age groups Minor infant Older infant Preschool School Adolescents |
44 (25.14) 13 (7.43) 45 (25.71) 34 (18.43) 39 (22.29) |
|
Sex Female Male |
68 (38.86) 107 (61.14) |
|
Stratum Low - low Low Middle - low |
135 (77.14) 35 (20.00) 5 (2.86) |
|
Nutritional status < 5 years Normal Acute underweight risk Moderate acute underweight Acute severe underweight Overweight risk Overweight |
48 (48.5) 21 (21.2) 10 (10.1) 13 (13.1) 4 (4.0) 3 (3.0) |
|
Nutritional status > 5 years Normal Thinness risk Thinness Overweight Obesity |
46 (59.0) 11 (14.1) 9 (11.5) 4 (5.1) 8 (10.2) |
Note. IR: Interquartile range; S.D: Standard deviation.
Source: own elaboration.
Regarding the phosphorus, calcium, and hematology parameters, no major alterations were observed (table 2). Half of the patients had a comorbidity of clinical importance, with neurological disease being the most frequent (18.29 %), and one out of every 10 patients had oncological disease (table 3). Patients admitted to the pediatric ICU had average PRISM and PELOD values of 7.0 and 3.0, respectively. The average number of days of mechanical ventilation and ICU stay were 6.60 and 5.0 days respectively (table 3).
Table 2 Level of phosphorus, calcium, leukocytes, neutrophils, lymphocytes, eosinophils, monocytes, hemoglobin and platelets in pediatric patients admitted to the ICU in two health institutions
| Data | Total N= 175 |
|---|---|
|
Phosphorus Median (IR) Mean (SD) |
4.10 (3.5 - 5.1) 4.39 (1.65) |
|
Calcium Median (IR) Mean (SD) |
1,3 (1,15 - 9,1) 4.56 (4.07) |
|
Leukocytes Median (IR) Mean (SD) |
11850 (6875 - 17815) 14661 (19275.8) |
|
Neutrophil Median (IR) Mean (SD) |
6860 (3345 - 12010) 8617.97 (6936.17) |
|
Lymphocytes Median (IR) Mean (SD) |
2230 (1375 - 4190) 4304.46 (12403.89) |
|
Eosinophils Median (IR) Mean (SD) |
50.0 (10.0 - 210.0) 244.8 (874.22) |
|
Monocytes Median (IR) Mean (SD) |
570 (330 - 1000) 750.88 (666.27) |
|
Hemoglobin Median (IR) Mean (SD) |
11.29 (9.7 - 12.60) 11.09 (2.29) |
|
Platelets Median (ir) Mean (sd) |
336000 (216500 - 432000) 333548.6 (183357.4) |
Note. IR: Interquartile range; S D: Standard deviation.
Source: own elaboration.
Table 3 PRISM, PELOD, ICU stay, days on MV and inotropic index in pediatric patients admitted to the ICU in two health institutions
| Data | Total N= 175 |
|---|---|
|
Comorbility Neurological Oncological Cardiac Respiratory Gastrointestinal Metabolic |
32 (18.29) 17 (9.71) 8 (4.57) 8 (4.57) 6 (3.43) 6 (3.43) |
|
Prism Median (ir) Mean (sd) |
7.0 (4.0 - 10.0) 6.78 (4.97) |
|
Pelod Median (ir) Mean (sd) |
3.0 (2.0 - 4.0) 3.25 (3.50) |
|
ICU stay Median (ir) Mean (sd) |
5.0 (3.0 - 10.0) 7.72 (7.44) |
|
Days of mv Mean (sd) |
6.60 (7.31) |
|
Inotropic index Mean (sd) |
8.39 (26.26) |
Note. IR: Interquartile range; SD: Standard deviation; MV: Mechanical ventilation. ICU: Intensive Care Unit.
Source: own elaboration.
Vitamin D Levels
The average VD was 25.43 mg/dl (SD = 10.17). The prevalence of alterations in the VD levels was 74.29 %. Of all patients, 46 (26.29 %) had VD deficiency, and 48 % had insufficiency.
Outcomes
Mechanical ventilation was the most frequent, with an incidence of 21.71 %, followed by the use of inotropes with 20.57 % (table 4). Patients with VD deficiency or low values had a significantly higher risk of requiring inotropes during their ICU stay (table 4). No association was observed between VD levels and requirement for mechanical ventilation, renal replacement therapy, or in-hospital mortality (table 4).
Table 4 Association between vitamin D levels with mechanical ventilation, use of inotropes, RRT and mortality in pediatric patients admitted to the ICU in two health institutions
| Data | MV N = 39 (21.71%) | p | Crude RR (CI95%) | Adjusted RR (CI95%)* |
|---|---|---|---|---|
|
Vitamin D levels Deficiency Insufficiency Alteration |
15 (32.61) 15 (17.86) 30 (23.08) |
0.06 0.81 0.83 |
1.75 (1.01 - 3.04) 0.89 (0.42 - 1.88) 1.15 (0.59 - 0.84) |
1.56 (0.84 - 2.91) 0.83 (0.39 - 1.78) 1.56 (0.84 - 2.91) |
| Data |
Inotropes N= 58 (20.57%) |
|||
|
Vitamin D levels Deficiency Insufficiency Alteration |
17 (36.96) 16 (19.05) 33 (25.38) |
<0.01 0.07 <0.01 |
2.51 (1.43 - 4.40) 2.86 (0.88 - 9.29) 3.81 (1.23 - 11.81) |
2.07 (1.12 - 3.82) 2.57 (0.83 - 7.97) 2.07 (1.12 - 3.82) |
| Data |
Death N= 10 (5.71%) |
|||
|
Vitamin D levels Deficiency Insufficiency Alteration |
5 (10.87) 4 (4.76) 9 (6.92) |
0.13 0.66 0.46 |
2.80 (0.85 - 9.25) 2.14 (0.25 - 18.60) 3.11 (0.41 - 23.91) |
2.33 (0.68 - 8.01) 1.68 (0.17 - 16.93) 2.34 (0.68 - 8.01) |
| Data |
RRT N= 6 (3.43%) |
|||
|
Vitamin D levels Deficiency Insufficiency Alteration |
4 (8.70) 1 (1.19) 5 (3.85) |
0.04 1.0 1.0 |
5.61 (1.06 - 29.61) 0.54 (0.03 - 8.36) 1.73 (0.21 - 14.42) |
4.02 (0.73 - 21.98) 0.45 (0.02 - 7.55) 4.02 (0.73 - 21.98) |
Note. * Adjusted RR by age and PRISM scale; MV: Need of mechanical ventilation; RRT: Renal replacement therapy; RR: Relative risk; CI: Confidence intervals.
Source: own elaboration.
VD levels were lower in patients who required inotropes (median 24.91 versus 20.72, p<0.01), renal replacement therapy (median 24.32 versus 11.10, p=0.08), and in deceased patients (median 24.32 versus 24.35, p = 0.15) (figure).
DISCUSSION
This is the first published cohort study carried out in Colombia, and one with the largest sample size in Latin America. It evaluated VD deficiency and insufficiency, and their association with disease severity in critically ill pediatric patients. The study found that critically ill pediatric patients have decreased levels of VD. Additionally, patients with VD deficiency had a higher risk of requiring inotropes during their stay in the ICU.
The results are consistent with preliminary reports. In patients admitted to pediatric ICU, studies have found VD deficiency between 40 % and 80 % of patients18-21, with an average VD less than 25 ng/ml7. Although it is unknown if this subnormal level of VD is due to the critical condition of the patient, as it has been found that critically ill patients have lower levels compared to healthy-patient controls20,22. Therefore, it is necessary to assess the level of VD in critically ill patients, since it has been shown that acute decrease could even be more dangerous than chronic deficiency3.
Studies in the critically ill adult population have shown an association between the level of DV and disease severity. In this regard, VD deficiency represents a risk for clinical outcomes such as the need for mechanical ventilation, longer hospital stay23, sepsis, in-hospital mortality24, and vasoactive requirement3. On the other hand, in the pediatric population, the studies have not been conclusive due to the high heterogeneity in the methodology and in their results. However, cardiovascular support has been one of the most found severe outcomes3.
This study found that suboptimal levels and VD deficiency increased the risk of needing inotropes, similar to the study report by Bustos et al17. This could be related to the behavior of VD that influences the structure and function of myocytes25, and the functioning of the respiratory system3. In this regard, analytical studies with a larger sample size are needed to analyze the association between VD and disease severity. Additionally, clinical trials are needed to evaluate the effect and safety of VD supplementation in critically ill pediatric patients.
This study has some limitations that should be taken into account. The first is the lack of knowledge of each patient's VD level prior to admission to the ICU, so it is not possible to assure that the deficiency is acute or chronic. As a second limitation, in this study, the change in the level of VD during the ICU stay was not evaluated, we report an information bias that could influence the disease severity. Thirdly, despite finding a significant association between deficiency and the need for inotropes, it is necessary to assess other criteria to establish causality, since these are multiple-risk factor patients. Finally, a selection bias is reported in this study, because it was conduct in only two health institutions, therefore results cannot be generalizable to all pediatric critically ill patients.
