Association of lower vitamin a levels in neonates and their mothers with increased risk of neonatal late-onset sepsis: A case-control study

This study was a case-control study over a period of six months in the neonatal wards and neonatal intensive care units (NICUs). After obtaining informed consent from the parents of the infants, and ethics committee approval, 40 eligible infants were entered into the study according to the inclusion criteria. The case group included 20 term or near-term infants who had been hospitalized with late-onset neonatal sepsis diagnosis from three to seven days of age, while the control group consisted of 20 term or near-term icteric hospitalized neonates without any features of sepsis.

We treated all LOS patients with antibiotics according to our NICU protocol and existing literature. In the control group, we only used phototherapy for all patients with uncomplicated neonatal jaundice, without any medication [4].

This study was performed on two groups: the case group included full-term and near-term infants (above 35 weeks) with sepsis who had been hospitalized due to late-onset neonatal sepsis between the third and seventh day of life and had not yet received any vitamin supplements. LOS was diagnosed based on clinical and paraclinical signs. We did a complete sepsis work-up for each neonate in both groups which included CBC diff, CRP, and blood cultures to rule out sepsis, and when necessary, we also took CSF cultures in septic group neonates. Moreover, in both groups, we also took urine cultures, because a UTI can be one of the important causes of both neonatal prolonged jaundice and late-onset sepsis. We consider prolonged jaundice to be a jaundice lasting more than 2 weeks; regardless, we took a urine culture to rule out late-onset sepsis in the control group or identify it in the case (sepsis) group. In septic neonates without any positive culture, we used Rodwell criteria as a precise instrument for sepsis diagnosis [4, 8, 47].

The control group consisted of 20 term or near-term infants who had been diagnosed with uncomplicated jaundice for any reason, provided that their bilirubin level was not close to the blood transfusion level, and who had been hospitalized for jaundice within three to seven days of their life, matching the case group. The neonates in control group had no suspected clinical or laboratory symptoms of sepsis, and their clinical and paraclinical signs, as well as CBC diff, CRP, B/C and urine culture, were not in favor of infection in any way, so these signs and symptoms strongly and definitely ruled out the presence of sepsis. It should be noted that with the exception of jaundice, the control group infants were found to be be completely healthy in other respects, having no other health problems, and they should not have received vitamin A supplements [21].

Study exclusion criteria among the neonates in both case and control group consisted of having inherited metabolic disorders, IUGR infants, infants with asphyxia, malabsorption syndromes, major anomalies, genetic and immunodeficiency diseases, and malignant underlying diseases, as well as infants who had received any form of vitamin A supplementation during their pre-hospitalization period (which means the first week of their life). Infants whose parents did not consent to their participation in the study, and those who died during hospitalization or were discharged from the hospital with personal consent before completing the course of treatment, were also excluded [12, 13, 22].

The reason we did not consider early-onset sepsis was that in the first three days of life, there may be many other transient and non-septic conditions which may mimic sepsis; for example, TTN, or transient tachypnea of neonates, or pneumonia may be clinically mistaken as sepsis. These conditions (especially TTN) which are even more common than infection, if not complicated, are mostly recovered after the third day of life. As such, we limited the inclusion criteria to late-onset sepsis to reduce the overlap with transient and non-septic conditions in the first days of life.

Neonates in each group were selected among eligible patients which fulfilled the selection criteria. The two groups were matched in terms of gestational age, birth weight, day of admission, age, feeding and delivery type, maternal age, and Apgar score, to minimize the effect of confounding factors.

The case-to-case ratio in this study was one-to-one, and the main dependent variable examined was the level of vitamin A in maternal and neonatal serum. For infants in both groups, vitamin A levels were measured after the third day of life. Demographic data for each of the 40 infants, including Apgar score, gender, age, and weight based on the SECA 345 digital scale with a measurement accuracy of 3 grams, were collected.

Maternal and neonatal vitamin A levels in both groups were measured by HPLC during the first week after delivery in the Gholhak clinical laboratory, by the Agilent G6410B Triple Quadrupole LC/MS system, which is a liquid chromatograph triple quadruple mass spectrometer that performs MS/ MS using three sets of parallel rods. The Agilent G6410-B Triple Quadrupole LC/MS system features a mass range of 15 to 1650 m/z (mass to charge ratio), with mass accuracy of < 0.1 u over 24 hours with MS/MS sensitivity of 0.5 pg (pictogram = 0.000000000001 grams) of reserpine inject on column. The results were recorded and stored blindly, through codes, for each infant. 2 to 3 cc of blood were taken in a tube without anticoagulation and after complete blood clotting, the serum was separated through a centrifuge, poured in a separate tube, and then kept at -20°. Due to the sensitivity of vitamin A to light and heat, the sampling was performed in special containers and sent to the reference laboratory as soon as possible. The normal range of vitamin A levels, according to the Gholhak clinical laboratory, is 30-80 microgram/dl (μg/dl).

According to the WHO, vitamin A deficiency in any age is defined as vitamin A level less than 21 μg/dl. However, many references consider 30 μg/dl as the cutoff point of vitamin A deficiency in human being of any age, because when the level of vitamin A reaches 30 μg/dl, the hepatic storage of vitamin A is greatly affected, and the immunity system as well as tissue repair mechanisms and respiratory epithelium begin to be disturbed, leading to the patient being very prone to sepsis even though they may not have obvious clinical symptoms. According to studies done in developed countries, vitamin A concentration in cord blood immediately after the birth of normal baby is 30 μg/dl; this amount is 18-24 μg/dl in developing countries. Considering all of the above, we defined vitamin A deficiency as vitamin A level less than 30 μg/dl [12, 13, 18, 22, 46].

The sample size is defined based on a study by Zhang et al [12], in which according to similar studies, the incidence of vitamin A deficiency was reported to be 58% in the sepsis group and 12% in the control group. Therefore, considering α=0.05, β=0.18, Z1-α/2= 1.96, Z1- B = 0.9, and also estimating a 10% drop out rate, the final sample size was calculated to be about 20 patients in each group.

Demographic data and maternal and neonatal serum levels of vitamin A in both sepsis and control groups were evaluated and analyzed with SPSS software version 25. Categorical variables were described as frequency rates and percentages, while continuous variables were described using mean and standard deviation (SD). Proportions for categorical variables were compared using the χ2 test, and the Fisher exact test was used when the data were limited. An independent sample t-test was used to compare the average level of vitamin A in the case and control groups. The chi-square test was utilized to check the compatibility of variables such as Apgar and gender in the two groups. A correlation test was used to evaluate the relationship between vitamin A levels in mothers and neonates. To evaluate the relationship between sepsis and neonatal or maternal factors, linear regression and logistic regression were used. A P-value of lower than 0.05 was considered significant.

No extra blood samples were taken from neonates; vitamin A levels were checked along with the blood that was sent for routine testing. Parents in both case and control groups were given the necessary information to obtain informed consent. This information included sampling and the amount of blood needed to test the level of vitamin A in mother and infant. If the parents agreed to their infant entering the study, this consent was recorded in written forms. All protocols were approved by the Ethics Committee of the Iran University of Medical Sciences. This study was carried out in compliance with the relevant guidelines. The Ethics Committee Approval code for this research is IR.IUMS.FMD.REC.1399.576. We ensure the quality and integrity of our research. We obtained informed consent from the parents of all infants participating in our study and also informed consent from the mothers who participated in our study. We respect the confidentiality and anonymity of our research respondents. We guarantee that the infants participated in the study voluntarily and according to their parents’ wishes. We also avoided any harm to our participants. There is no conflict of interest in this study and we did not have any financial support for this research.

Concerning neonates’ feeding and vitamins intake during the study period, as shown in Table 1, of all 40 neonates in this study, 17.5% were formula fed, 50% were exclusively breastfed, and 32.5% used a combination of formula and breast milk. There was no statistically significant difference between case and control group according to feeding type (p value: 0.4). Also, as it is shown in Table 2, there was no statistically significant difference between vitamin A level in neonates according to their feeding type (p value: 0.26); the reason could be the short duration of neonatal feeding before taking blood samples. Vitamin A level was checked in all neonates between their third and seventh day of life. The average day of admission in these 40 neonates was around the fourth day of their life, which not have been enough time to there to be significant impact on the serum vitamin A level. Based on the statistical results obtained from this study and information shown in Tables 1 and 2, maternal vitamin A status might have a greater effect than the neonatal feeding type on the neonatal vitamin A level and subsequently on the neonatal LOS rate. Regardless of the type of neonatal feeding, serum vitamin A level is significantly lower in the LOS group with any kind of feeding in comparison with the control group with any kind of feeding. Among the factors such as APGAR score, gestational age, birth weight, gender, feeding type and maternal serum vitamin A level, the only factors which were significantly different between the case and control group were maternal and neonatal serum vitamin A level. As there is no de novo fetal vitamin A synthesis, all vitamin A needed by the embryo is gained from maternal vitamin A storage, so it seems that paying attention to the maternal vitamin A level and storage, as well as maternal nutrition, might have a greater effect on the neonatal vitamin A level and consequently on the neonatal LOS rate [48].

Regarding taking vitamin A supplement drops, the average age of neonates who participated in our study was four days. Many infants in our country start taking supplement drops after the first few days of their life. As we mentioned in the exclusion criteria, we excluded any neonates who took supplement drops prior to hospitalization. After hospitalizing both groups of neonates, we also postponed prescribing vitamin A supplementation to the time after we took blood samples to check serum vitamin A level, which was on the first day of their admission.

Regarding using Total Parenteral Nutrition, according to our NICU ward protocol, we start TPN for all neonates as soon as they are admitted if their oral intake was not completely full. Our available TPN solution consists of intralipids, aminofusion and soluvit, none of which include any form of vitamin A. Unfortunately, appropriate intravenous vitamin A solution is not available in our ward. As a result, TPN and nutritional matters are not considered confounding factors in our study.

APGAR: Appearance, Pulse, Grimace, Activity, and Respiration; CRP: C-reactive Protein; LOS: Late onset sepsis; NICU: Neonatal intensive care unit; NVD: Natural Vaginal Delivery; PROM: Premature Rupture of Membranes; RDS: Respiratory Distress Syndrome; SD: Standard Deviation; WBC: white blood cells

This manuscript has been read and approved by all authors. All protocols were approved by the Ethics Committee of the University. Also, the study was carried out in compliance with the relative guidelines. The Ethics Committee Approval code for this research is IR.IUMS.FMD.REC.1399.576. We ensure quality and integrity of our research. We obtained informed consent from the parents of all infants participating in our study and also obtained informed consent from their mothers who participated in our study. We respect the confidentiality and anonymity of our research respondents. We guarantee that the infants participated in the study voluntarily and according to their parents’ wishes. We also avoided any harm to our participants. There is no conflict of interest in this study, and we did not have any financial support for this research.