Evaluating Nutritional Selection and Outcomes in Neonatal Abstinence Syndrome: A Retrospective Review

Abstract

Objective

The rising incidence of neonatal abstinence syndrome (NAS) has amplified the importance of nonpharmacological interventions in its management, which include the selection of feedings. With the goal of obtaining an accurate assessment of the effects of current feeding practices in NAS infants in our neonatal intensive care unit, we conducted a retrospective review of NAS infants at our hospital over a 3-year period to determine their nutritional selections and evaluate their length of stay (LOS), length of treatment (LOT), and growth outcomes.

Study Design

Retrospective chart review of term infants (≥37 weeks of gestation) with NAS. Maternal and infant demographics and characteristics were recorded. Infants were grouped based on majority (>50% of total feeding) nutritional selections and LOS, LOT, and growth parameters were evaluated. Linear regression was used to compare group outcomes. Significance was set at a p-value <0.05.

Results

A total of 70 infants were included and grouped based on majority feeds into maternal breast milk (MBM), standard term formula (STF), low lactose formula (LLF), and extensively hydrolyzed formula (EHF) groups. Feeding selections were provider-dependent and infants were placed on MBM or STF as an initial selection. In all infants included in our review, LLF was selected as the first choice following MBM or STF for increased gastrointestinal (GI) disturbance-related Finnegan Neonatal Abstinence Scoring scores and changed to EHF if LLF failed to improve the GI-related symptoms. The STF-fed infants had the shortest LOS, and none of these infants required pharmacological treatment. The LOT and LOS were similar in the MBM- and LLF-fed groups. Infants who were EHF fed had the longest LOT and LOS. All feeding groups demonstrated appropriate growth.

Conclusion

Nutritional selections in our NAS infants were modified for the severity of their withdrawal symptoms. All nutritional modifications driven by severity of withdrawal symptoms supported favorable growth outcomes in the infants.

Key Points

• Our NAS infants were fed with multiple types of nutrition.

• Infants with severe NAS required more elemental feeds.

• All formula selections supported favorable growth.

Keywords

neonatal abstinence syndrome - opioids - drug withdrawal - nutrition - formula - maternal breast milk - hydrolyzed formula

Neonatal abstinence syndrome (NAS) is a term used to describe a constellation of symptoms experienced by infants undergoing withdrawal from in utero exposure to maternal illicit or prescribed drugs.[1] These infants may experience significant central nervous system effects, gastrointestinal (GI) dysfunction, feeding difficulties, and poor growth, and as a result, may require pharmacological therapy, a prolonged hospitalization, and neonatal intensive care.[1] [2] [3] [4] The treatment for NAS has historically involved pharmacological management including medications such as morphine, clonidine, phenobarbital, and methadone, but recent efforts directed at limiting the need for these medications have led to a greater focus on nonpharmacological measures that include strategies to maintain the maternal–infant dyad, and interventions such as soothing techniques, care in a quiet dimly lit environment, and the management of feedings.[1] [2] [3] [4] [5] [6] [7]

Feeding and nutritional issues such as poor intake, regurgitation, emesis, and loose stools that are observed in NAS are significant and widespread despite which there is only limited information and evidence-based guidance regarding potential interventions for improving care in these areas.[6] [8] [9] [10] GI manifestations are among the major criteria evaluated by the Finnegan Neonatal Abstinence Scoring (FNAS) tool to assess withdrawal and FNAS scoring is heavily weighted on GI symptoms in directing the treatment for NAS.[11] [12] The emphasis on GI disturbances is additionally justified based on evidence indicating the presence of opioid receptor sites that are implicated in withdrawal in the GI tract.[13] [14] Infants with NAS are also frequently observed to have poor weight gain in the early neonatal period.[6] [15] [16] The neurologic excitability (including hypertonia and tremors), hyperthermia, and tachypnea experienced by these infants in addition to their GI symptoms lead to high energetic demands accompanied by inadequate caloric intake.[6] [16] Several studies have investigated the impact of breastfeeding on the improvement of NAS symptoms and on the length of treatment (LOT) and length of stay (LOS).[9] [17] However, studies evaluating the effects of other nutritional selections are currently limited and do not provide significant guidance for appropriate interventions in formula-fed infants or address their effects on growth.[6] [8] [9] [10] It is therefore evident that a greater emphasis on examining the effects of current nutritional interventions on growth and infant outcomes is needed to inform the development of effective strategies for the nutritional management of infants with NAS.

Since our facility lacked any instruction to address the nutritional aspects of NAS, the objective of our study was to provide an accurate assessment of the current practices in our Level III neonatal intensive care unit (NICU) regarding provider-driven feeding selection in the management of infants diagnosed with NAS. We conducted a retrospective chart review of NAS infants at our hospital over a 3-year period to determine nutritional selection, LOS, LOT, and growth outcomes during hospitalization. Maternal breast milk (MBM) and infant formulas predominantly utilized by providers were then compared with the same short-term outcomes to determine if nutritional selection had an effect on the LOS, LOT, and growth outcomes while in the hospital.

Materials and Methods

Study Design

We conducted a cross-sectional analysis of data utilizing a retrospective chart review of all term neonates (≥37 weeks of gestation) with a diagnosis of NAS admitted to the University of Louisville Hospital (ULH) between January 2018 and December 2020. Infants were monitored either in the mother/baby unit, where private rooms allowed for maternal rooming in, or in the NICU, where an open bay design limited maternal contact. The most common indications for transfer to the NICU were to monitor infants after maternal discharge or to initiate pharmacotherapy. Infants who were admitted to the NICU received input from registered dietitians and pharmacists as a part of their care. Our study was approved by the Institutional Review Board at the University of Louisville (#22.0007, January, 26, 2022).

Inclusion and Exclusion Criteria

Infants were defined as having NAS if they demonstrated symptoms of withdrawal and were exposed in utero to maternal illicit or prescribed substances. Subjects with a diagnosis of NAS were included for data collection and analysis if their charts indicated that they showed signs of withdrawal, regardless of a requirement for pharmacological treatment. Infants born at term (≥37 weeks of gestation) with a birth weight >1,800 g were included in the study. Infants who required evaluation for sepsis, respiratory support, or were found to have congenital abnormalities were excluded. The electronic medical record (EMR) was used for obtaining all pertinent information that was gathered for our review including information on demographics and maternal exposures. The outcomes evaluated for effects on NAS symptoms were LOT and LOS in days.

Growth Assessment

Growth parameters including infant weight, length, and head circumference at birth and at discharge were collected and used for calculating changes in z-score based on the World Health Organization (WHO) standard growth curves. The peak weight loss was calculated based on the lowest weight documented during the hospital stay. A return to birth weight was recorded when infants reached a weight that was ≥birth weight for 2 consecutive days. Infants who were discharged before a return to birth weight were not included in this measure.

Nutritional Data

Data regarding nutritional selection were gathered by a detailed review of the documentation of frequency, volume, and type of feeds in the infants' EMRs. Majority feeds were defined as feeds forming greater than 50% of the total feeds throughout the infant's hospitalization, based on the definition used in studies from Lembeck et al to determine the effects of breastfeeding and formula selection on outcomes.[9] MBM was not used for feeding if mothers showed evidence of polysubstance or illicit drug use or were not stabilized in a medically assisted treatment program. All the infants included in the study who were majority-fed MBM received expressed breast milk, which permitted an accurate estimation of the MBM intake. Formula types that were most commonly used by our providers and therefore included in this review were standard term formula (STF), low lactose formula (LLF), and extensively hydrolyzed formula (EHF), all of which had a caloric content of 19 to 20 kilocalories/ounce (Kcal/oz). Infants who were majority fed other infrequently used formula types were excluded. The caloric content of breast milk was assumed to be equivalent to STF, and calorimetry was not performed on breast milk samples. Data on infants who received feeds with a higher caloric content were also obtained and noted in [Table 1].

Table 1 Baseline maternal and infant characteristics

Maternal characteristics	n = 70
Maternal age, y, mean (± SD)	28.9 (4.3)
Race, n (%)
White	61 (87.1)
Black	9 (12.9)
Hispanic	0 (0.0)
Asian	0 (0.0)
Prenatal care, n (%)
Yes	41 (58.6)
No	29 (41.4)
Drugs used during pregnancy, n (%)
Opioids
Buprenorphine	19 (27.1)
Methadone	12 (17.1)
Oxycodone	6 (8.6)
Other opioids	25 (35.7)
Amphetamines	16 (22.9)
Cannabinoids	10 (14.3)
Benzodiazepines	6 (8.6)
Cocaine	5 (7.1)
Methamphetamines	2 (2.9)
Barbiturates	1 (1.4)
Gabapentin	1 (1.4)
Tobacco use during pregnancy, n (%)	36 (51.4)
Infant characteristics	n = 70
Infant sex, n (%)
Male	39 (55.7)
Female	31 (44.3)
Gestational age at delivery, wk, mean (± SD)	38.6 (1.7)
Breast milk feeds, n (%)
Any breast milk	21 (30.0)
Majority breast milk	5 (7.1)
Increased calorie feeds, n (%)
Any 22 Kcal/oz	16 (22.9)
Majority 22 Kcal/oz	4 (5.7)
Any 24 Kcal/oz	0 (0)
Majority 24 Kcal/oz	0 (0)
Infant drug exposure	n = 68[a]
Number of agents, n (%)
0	3 (4.3)
1	36 (51.4)
2	16 (22.8)
3	9 (12.8)
4	1 (1.4)
5	3 (4.3)

Abbreviation: SD, standard deviation.

Notes: Maternal and gestational age presented as mean with SD in parentheses. Rest of data presented as total number with the percentage of total number of included infants in parentheses.

^a Two infants did not have drug screens documented.

Neonatal Abstinence Scoring and Pharmacological Therapy

The medical team followed a standardized protocol for monitoring withdrawal symptoms, and for initiating, escalating, and weaning pharmacological therapy. Infants were assessed using the FNAS tool for NAS.[11] [12] If the infant had three consecutive scores ≥8 or two consecutive scores ≥12, morphine 0.05 mg/kg/dose every 3 hours was initiated and increased by 0.02 mg/kg/dose until the infant was stabilized, and there were no more than two scores > 8 in a 24-hour period. When approximately 50% of the recommended maximum morphine dose of 0.2 mg/kg/dose was reached or if morphine was not able to be weaned by day 7 of treatment, clonidine was initiated as adjunctive therapy starting at 1µg/kg/dose every 3 hours and increased as needed to a maximum dose of 2 µg/kg/dose. Phenobarbital is the third-line medication in our NAS protocol but was not required by any of these infants. Weaning was begun when scores remained <8 within a 24-hour time frame. All infants initiated on morphine were admitted to the NICU, if not already under observation in the NICU. Infants were not discharged on medication. Infants who did not require pharmacological treatment were monitored for at least 5 days in the mother/baby unit or NICU prior to discharge.

Data Analyses

The baseline characteristics of the mothers and infants were reported using the following descriptive statistics: mean and standard deviation (SD) for continuous variables and a measure of frequency with corresponding percentages for categorical data. The Shapiro-Wilk test, box plot and Q–Q plot were used to assess the normality of the continuous data; thus, we used the mean, SD, and parametric tests for normally distributed data. The sample size and power for the study were calculated based on Cohen's guidelines; f ² ≥ 0.02, f ² ≥ 0.15, and f ² ≥ 0.35 represent small, medium, and large effect sizes, respectively, for correlation models.[18] Since no previous studies have examined the effects of EHF on the LOS or LOT nor on growth outcomes in NAS infants, we estimated the sample size to detect a medium effect size. Setting the statistical significance level (α) at 5% and an adequate power of 80% (1-β), it was determined that the total sample size required was 55 participants to detect a medium effect size of (0.15) based on Cohen's guidelines.[18] G*Power software (v. 3.1.9.6; Heinrich-Heine-Universität Düsseldorf, Düsseldorf, Germany; http://www.gpower.hhu.de/) was used to perform the power analysis. For analyses of LOT, LOS and growth, given that the outcome variables (i.e., LOT, LOS and growth parameters) were continuous, we utilized univariate linear regression analysis to examine the relationship between nutritional selection and LOT, LOS and growth parameters (defined as changes in z-score for weight, length, and head circumference). Univariate linear regression analysis was used because the sample size was sufficient for one predictor, based on power analyses and the calculation of the minimum sample sizes needed to detect a power of 80% and a medium size of effect. No adjustment was performed. SAS software (SAS Institute, Cary, North Carolina, USA) version 9.4 was used to analyze the data. A two-tailed p-value of less than 0.05 was considered statistically significant. SDs and standard errors for all data are shown in the data tables and not restated in the written results.

Results

During the study period, 103 term infants born at ULH received a diagnosis of NAS. Of these infants, 33 were excluded for several clinical indications including respiratory distress, evaluation for sepsis, congenital anomalies, and significant hypoglycemia (which required prolonged intravenous fluids and delayed the commencement of feeds; [Fig. 1]). An additional three infants were excluded due to majority intake of formulas which are infrequently used at our hospital for which we were therefore unable to assess significance. Of the remaining 70 infants that were included in our study sample, 7% (n = 5) received majority feeds of MBM, 7% (n = 5) of STF, 61% (n = 43) of LLF, and 24% (n = 17) of EHF ([Tables 1] and [2]).

Fig. 1 Diagram of population of infants reviewed in study. EHF, extensively hydrolyzed formula; LLF, low lactose formula; MBM, maternal breast milk; NAS, neonatal abstinence syndrome; STF, standard term formula. *Group not evaluated for relationship with length of treatment, length of stay, or growth parameters.

Table 2 Length of treatment and stay and growth outcomes in all feeding groups

Medications and growth outcomes	Overall	MBM	STF	LLF	EHF
Study population, n (%)	70 (100)	5 (7.1)	5 (7.1)	43 (61.4)	17 (24.3)
Use of pharmacological therapy, n (%)
None	11 (15.7)	1 (20.0)	5 (100.0)	5 (11.6)	0 (0.0)
Morphine only	28 (40.0)	2 (40.0)	0 (0.0)	24 (55.8)	2 (11.8)
Morphine + clonidine	30 (42.9)	2 (40.0)	0 (0.0)	13 (30.2)	15 (88.2)
Clonidine only	1 (1.4)	0 (0.0)	0 (0.0)	1 (2.3)	0 (0)
Phenobarbital	0 (0.0)	0 (0.0)	0 (0.0)	0 (0.0)	0 (0.0)
Total morphine (mg), mean (± SD)	20.0 (22.5)	15.6 (12.7)	0.0 (0.0)	16.0 (17.4)	37.5 (29.1)
Length of treatment (d), mean (± SD)	22.5 (15.6)	19.4 (12.9)	0.0 (0.0)	18.8 (11.3)	39.7 (12.6)
Length of stay (d), mean (± SD)	28.0 (15.8)	26.8 (12.3)	5.4 (1.1)	24.1 (12.0)	44.7 (12.7)
Growth outcomes, mean (± SD)
Peak weight loss %	6.0 (2.5)	6.5 (3.7)	5.8 (1.9)	6.3 (2.6)	5.0 (2.6)
Days to return to birth weight	10.2 (3.5)	10.5 (0.6)	–	10.8 (3.8)	8.9 (3.1)
Change in weight z-score	−0.52 (0.44)	−0.62 (0.39)	−0.50 (0.29)	−0.57 (0.39)	−0.35 (0.59)
Change in head circumference z-score	−0.16 (0.82)	−0.16 (0.77)	0.77 (0.88)	−0.31 (0.88)	−0.06 (0.44)
Change in length z-score	−0.29 (0.96)	−0.12 (0.28)	−0.22 (0.38)	−0.34 (1.02)	−0.20 (1.03)

Abbreviations: EHF, extensively hydrolyzed formula; LLF, low lactose formula; MBM, maternal breast milk; SD, standard deviation; STF, standard term formula.

Notes: Total study population and numbers of infants on pharmacological therapies presented as total number with percentage of total number of included infants in parentheses. Rest of data presented as mean with SD in parentheses.

Maternal and Infant Characteristics

Demographic characteristics for mothers and infants are presented in [Table 1]. A total of 89% of mothers had a urine drug screen that was positive for opioids prior to delivery, including buprenorphine, methadone, oxycodone, or other opioid-containing medications. Amphetamines were the second most common substance identified. Infant toxicology screens also were obtained and the numbers of infants exposed to 1 or greater than 1 drug have been listed in [Table 1]. In our review of these data, we found that only 28.5% (20/70) of the infant screens showed concordance with paired maternal toxicology screens in terms of substances identified, which corresponds with findings in previous studies.[19] [20]

Infants for whom mothers were unable or unwilling to breastfeed were fed with formula, which was selected at the discretion of the provider wherein the standard practice followed by providers in our NICU is to initiate feeds with STF. In all the infants included in our review, we found that providers selected LLF as the first formula choice following STF for an increase in FNAS scores for GI disturbances and then changed to EHF if LLF failed to improve the GI-related symptoms. In addition to increased GI-related FNAS scores, additional GI signs and symptoms were documented as indications for changing nutritional selections from either STF to LLF or from LLF to EHF in 22 infants. These included loose or watery stools with/without emesis (15 infants, 68.3%), poor feeding (5 infants, 22.8%), abdominal distention (2 infants, 9.1%), and bloody stools (2 infants, 9.1%). Four infants (5.7%) were fed a majority formula that contained 22 Kcal/oz. All the infants in the study were discharged on 19 to 20 Kcal/oz feeds.

Length of Treatment

We next determined the LOT in the majority feeding groups in all of the infants ([Table 2]). Of the 70 NAS infants who met inclusion criteria for our study, 84% (n = 59) required pharmacological treatment. Of these infants, 83% (n = 58) required morphine and 43% (n = 30) required clonidine therapy in addition to morphine with one additional infant requiring only clonidine ([Table 2]). Infants on MBM and LLF had a similar LOT at 19.4 and 18.8 days respectively, whereas infants on EHF had a longer LOT at 39.7 days. Notably, none of the infants who received STF as >50% of the total feeding (n = 5) required pharmacological treatment.

Since our review demonstrated that a substantial number of our infants with NAS were transitioned to and majority fed with either LLF or EHF, we sought to compare the LOT between infants on these formulas. We evaluated the relationship between nutritional selection and LOT in these formula groups by linear regression and found that the LOT in the EHF-fed group was significantly longer than the LLF-fed group ([Table 3]), as also indicated by the data in [Table 2].

Table 3 Relationship of length of treatment, length of stay, and growth outcomes with nutritional selection

Formula type	β-coefficient	Standard error	p-Value
Length of treatment
Comparison of alternate formula groups
LLF	−20.88	3.34	<0.0001
EHF	Reference
Length of stay
Comparison of alternate formula groups
LLF	−20.61	3.50	<0.0001
EHF	Reference
Growth outcomes
Change in weight z-score
LLF	−0.23	0.14	0.107
EHF	Reference
Change in length z-score
LLF	−0.24	0.30	0.435
EHF	Reference
Change in head circumference z-score
LLF	−0.31	0.20	0.119
EHF	Reference

Abbreviations: EHF, extensively hydrolyzed formula; LLF, low lactose formula.

Notes: Univariate linear regression analysis used to examine the relationships between nutritional selection and length of treatment, length of stay, and growth outcomes. A p-value of less than 0.05 was considered statistically significant.

Length of Stay

We also evaluated the LOS in infants who were majority fed either MBM or different formulas ([Table 2]). We found that infants on STF had the shortest LOS of all the groups (5.4 days), which was anticipated since none of the STF-fed infants during the time period examined in our review required pharmacological treatment and were therefore discharged after observation. Infants on MBM and LLF had a similar LOS at 26.8 and 24.1 days, respectively. The infants on EHF had the longest LOS at 44.7 days. Similar to the LOT, we also compared LOS between infants who were majority-fed LLF and EHF using linear regression and found that the EHF-fed infants showed a significantly longer LOS ([Table 3]), as also indicated by the data in [Table 2].

Growth Outcomes

We examined weight, length, and head circumference to evaluate growth in all the infants. We found that all the nutritional selection groups showed a <7% weight loss, a return to birth weight by day 11, and a <0.80 reduction in z-scores in all growth parameters. Importantly, we also did not find statistically significant differences in growth outcomes between the LLF- and EHF-fed groups ([Table 3]).

Discussion

The incidence of NAS has increased significantly over the past several decades in the United States,[5] which has elevated the relevance and urgency of determining optimal interventions for its management. Apart from the pharmacological management of NAS, nonpharmacological measures including environmental interventions, soothing, and feeding selection have been found to play an increasingly vital role in alleviating withdrawal symptoms, improving outcomes, and optimizing health care resources.[21] [22] [23] The goal of our study was to review the nutritional selections made in NAS infants at our hospital and determine if these selections influenced their LOS, LOT, and growth outcomes. In our hospital, the first nutritional choice for infants with NAS is breast milk (or STF if they are unable to feed breast milk). We observed that in all reviewed infants, subsequent nutritional selections were based on the severity of their NAS-related GI symptoms. Nutritional selections were thus tailored to the specific treatment needs of the infants and not preselected based simply on the infant's diagnosis of NAS. Our results showed that, as anticipated based on our observation of the nutritional choices made in response to NAS-related GI intolerance, NAS infants who were transitioned to an EHF were infants who demonstrated higher FNAS scores and therefore the longest LOS and LOT. Importantly, our data demonstrated adequate and comparable growth outcomes with no statistically significant differences in growth parameters (weight, length, and head circumference) between all the nutritional selection groups (MBM, STF, LLF, and EHF). We therefore found that nutritional selection based on withdrawal severity appeared to support and maintain adequate growth.

Previous studies have demonstrated benefits from MBM and breastfeeding in reducing the severity of NAS symptoms and decreasing the length of pharmacological treatment and hospitalization in infants with NAS.[24] [25] Possible reasons for these effects include improved mother–infant bonding, the soothing effect of breastfeeding, and the potential passage of opioids into MBM.[1] The benefits of breastfeeding are potentially even more profound since mothers who are able to provide breast milk are more likely to be in stable recovery programs, less likely to have polysubstance use, and more likely to exhibit greater attachment to their infants.[24] In a recent study by Lembeck et al, LOS, LOT, and weight change were compared among NAS infants who were breastfeeding, standard formula fed, and low-lactose formula fed and showed that infants who were majority-fed breastmilk had a reduction in LOT and LOS when compared with standard formula or LLF.[9] Due to the limited number of infants who were majority fed with MBM in the period covered by our review (n = 5), we were unable to analyze our results in these infants relative to formula-fed infants. We also were unable to compare results between our MBM-fed group and STF-fed infants due to the limited number of infants majority fed on STF (n = 5) in our review period.

Our standard practice for infants with NAS who are unable to receive breastmilk is to initiate feeds with STF. As noted in the results section, we found that none of the five infants fed with STF had FNAS scores that reached the threshold required for pharmacological therapy based on our institutional guidelines. To gain a better understanding of the course of these infants, we analyzed their exposures in greater detail and found that only one of these infants tested positive for opioids in their urine drug screen at birth, whereas the rest of the group was exposed to nonopioid substances including cocaine, amphetamines, and benzodiazepines. Although infants exposed to cocaine and amphetamines in utero may experience symptoms of withdrawal including hypertonia, tremors, poor feeding, and abnormal sleep patterns,[26] their symptoms are of lower intensity and shorter duration than those seen in opioid-dependent infants,[27] which may account for the shorter LOS and lack of need for treatment in infants who fed STF in our study.

We additionally sought to assess and gain an understanding for the prolonged LOS in the MBM-fed group relative to the STF-fed group and therefore also conducted a detailed review of their exposures. Our examination revealed that all the infants in the MBM-fed group were exposed to longer-acting opioid agents in utero (buprenorphine in three infants and methadone in two infants), and we postulate that this likely explains the longer LOT and LOS in the MBM group infants relative to the STF group.

Of the 21 criteria scored in the FNAS tool, 5 criteria assess GI disturbances including excessive sucking, poor feeding, regurgitation and projectile vomiting, and stool consistency.[12] LLFs are often used as an alternative to STFs in infants with feeding intolerance and have therefore become a frequent choice in infants with GI disturbances as part of their withdrawal symptoms. Recent data, however, have shown that low lactose feeds in infants treated for NAS did not reduce duration of treatment or LOS when compared with standard feeds.[8] [28] The lack of symptomatic relief from LLFs may be explained by data indicating that the GI effects of opioid withdrawal are likely driven by pathways activated by the 5HT3 receptor.[29] Our review did not find that LLF was an initial formula choice for any of the infants but did find that 43 infants were transitioned to and majority-fed LLF for increased GI symptoms of withdrawal. Due to the limited number of infants on either MBM or STF in our review, we were unable to statistically analyze the LOS or LOT in our LLF group relative to MBM-fed or STF-fed groups, but similar to recent studies, we did not find a trend toward decreased LOT or LOS in LLF-fed infants relative to either group.

Notably, our study did find that EHF-fed infants had a statistically significant longer LOT and LOS than LLF-fed infants. To our knowledge, our study is the first to evaluate outcomes in infants with NAS who were fed with EHF. In our NICU, provider preference has dictated the use of EHF for infants with NAS when use of less hydrolyzed formulas has failed to improve the feeding intolerance and GI symptoms of NAS. We presume that this practice arose from the success observed from the use of EHF for feeding intolerance associated with non-NAS-related GI conditions in the NICU.[30] We found that infants majority-fed EHF were the infants observed to have the greatest degree of withdrawal as demonstrated by the highest FNAS scores, greater morphine requirement, and the longest LOT. Our results suggest that a requirement for EHF in NAS infants may indicate the possibility for a prolonged need for pharmacological treatment and a resultant longer LOS.

An important objective of our study was to determine whether different nutritional selections resulted in differences in growth. Our data found no differences in all three examined growth parameters among the feeding groups. In the early neonatal period, weight loss and a return to birth weight are assessed as indicators of adequate nutrition with an average weight loss of approximately 8 to 10% cited as a starting threshold for closer nutritional evaluation and an average return to birth weight by 7 to 14 days of life.[31] [32] As the use of z-scores is becoming more prevalent for the accurate identification of acute malnutrition in neonates,[32] we used changes in z-score from birth to discharge on the WHO standard growth curves as our primary growth outcomes. Within all the feeding groups, the infants had lower z-scores at the time of discharge, which is similar to other studies that display decreased growth parameters between birth and discharge from the NICU.[25] Notably, none of the infants on any feeding selection met criteria for malnutrition, defined as a decline of 0.8 in z-score.[32] As with our other outcomes, our results reflect the responses of the providers to the growth trajectory of infants during hospitalization. Similar to our providers' practice of changing feeds based on GI symptoms, infants who were noted to have slower weight gain had their feeds and caloric content adjusted as required to optimize their growth.

Our review was a retrospective observational study, and we acknowledge several limitations. Most notably, only limited numbers of infants received majority feeds with either MBM or STF in the reviewed period, and we will need to evaluate a larger number of infants before we are able to arrive at any conclusions regarding the relationship of these nutritional selections with outcomes in our population. Additionally, formula selections were entirely provider-dependent and not based on guidelines. An added limitation for our data analysis was our inability to accurately correct for polydrug use due to poor concordance between drugs identified by paired maternal and infant toxicology screens (as has been previously described[19] [20]). Using a definition of polydrug use as >1 drug documented in maternal and/or infant screens (based on observations of signs of withdrawal and the need for pharmacological therapy in infants with evidence of either maternal or infant drug exposure), we found no statistically significant correlation between polydrug use and LOS (p = 0.83) and LOT (p = 0.84) and therefore did not correct for it. Further, since this was a retrospective review, the data gathered relied on the completeness and accuracy of the medical records. Data regarding potential confounding factors including details of overall maternal health, socioeconomic status, level of antenatal care, degree of medication use, or severity of substance use and other infant medical conditions, which could have altered outcomes were not documented. In addition, other confounding variables that might have changed and potentially improved nonpharmacological outcomes, e.g., parental involvement, skin-to-skin care, and environmental modifications were not documented or controlled for in this review. Despite these potential drawbacks, however, we were encouraged to pursue our study based on previous successful retrospective studies that have yielded important data to guide the development of feeding protocols for infants with NAS.[8] [9] In addition to these issues, we recognize that our study focuses on immediate outcomes of nutritional selection in NAS. An evaluation of the long-term neurodevelopmental, growth, and health outcomes from different nutritional selections would provide valuable information and guidance for the development of optimal feeding strategies for these infants, and we propose to evaluate these outcomes in future prospective analyses.

Conclusion

In conclusion, our results indicate that infants diagnosed with NAS who were majority-fed EHF had significantly longer LOS and LOT than infants fed less elemental formula. The severity of NAS GI symptoms appeared to drive nutritional selection; therefore, the observed correlation of a longer LOS and LOT with the use of hydrolyzed formula was expected. Notably, the growth outcomes in all the NAS infants were adequate and comparable between all nutritional selections including MBM and all examined formula types. Most importantly, our data underscored the key importance of provider responsiveness in tailoring clinical management to support appropriate growth in these infants. Our results suggest that modifications in feedings for individual needs serves as an effective strategy in the management of infants with NAS. In addition, although not the focus of our study, our results additionally suggest that, utilizing this strategy, a need for elemental formula may be predictive of a greater severity of withdrawal that may require prolonged pharmacological treatment and hospitalization. Identification and knowledge of such predictive markers may alert providers to enhance other individual nonpharmacological interventions to improve outcomes.

Conflict of Interest

None declared.

• References

• 1 Kocherlakota P. Neonatal abstinence syndrome. Pediatrics 2014; 134 (02) e547-e561
• 2 Stover MW, Davis JM. Opioids in pregnancy and neonatal abstinence syndrome. Semin Perinatol 2015; 39 (07) 561-565
• 3 Tolia VN, Patrick SW, Bennett MM. et al. Increasing incidence of the neonatal abstinence syndrome in U.S. neonatal ICUs. N Engl J Med 2015; 372 (22) 2118-2126
• 4 Patrick SW, Schumacher RE, Benneyworth BD, Krans EE, McAllister JM, Davis MM. Neonatal abstinence syndrome and associated health care expenditures: United States, 2000-2009. JAMA 2012; 307 (18) 1934-1940
• 5 Hirai AH, Ko JY, Owens PL, Stocks C, Patrick SW. Neonatal abstinence syndrome and maternal opioid-related diagnoses in the US, 2010-2017. JAMA 2021; 325 (02) 146-155
• 6 Hudak ML, Tan RC. Committee on Drugs, Committee on Fetus and Newborn, American Academy of Pediatrics. Neonatal drug withdrawal. Pediatrics 2012; 129 (02) e540-e560
• 7 Ryan G, Dooley J, Gerber Finn L, Kelly L. Nonpharmacological management of neonatal abstinence syndrome: a review of the literature. J Matern Fetal Neonatal Med 2019; 32 (10) 1735-1740
• 8 Alsaleem M, Berkelhamer SK, Wilding GE, Miller LM, Reynolds AM. Effects of partially hydrolyzed formula on severity and outcomes of neonatal abstinence syndrome. Am J Perinatol 2020; 37 (11) 1177-1182
• 9 Lembeck AL, Tuttle D, Locke R. et al. Breastfeeding and formula selection in neonatal abstinence syndrome. Am J Perinatol 2021; 38 (14) 1488-1493
• 10 Casper T, Arbour M. Evidence-based nurse-driven interventions for the care of newborns with neonatal abstinence syndrome. Adv Neonatal Care 2014; 14 (06) 376-380
• 11 Finnegan LP, Kron RE, Connaughton JF, Emich JP. Assessment and treatment of abstinence in the infant of the drug-dependent mother. Int J Clin Pharmacol Biopharm 1975; 12 (1–2): 19-32
• 12 Finnegan LP, Connaughton Jr JF, Kron RE, Emich JP. Neonatal abstinence syndrome: assessment and management. Addict Dis 1975; 2 (1–2): 141-158
• 13 Mosińska P, Zielińska M, Fichna J. Expression and physiology of opioid receptors in the gastrointestinal tract. Curr Opin Endocrinol Diabetes Obes 2016; 23 (01) 3-10
• 14 Patrick SW, Barfield WD, Poindexter BB. Committee on Fetus and Newborn, Committee on Substance Use and Prevention. Neonatal opioid withdrawal syndrome. Pediatrics 2020; 146 (05) e2020029074
• 15 Dryden C, Young D, Campbell N, Mactier H. Postnatal weight loss in substitute methadone-exposed infants: implications for the management of breast feeding. Arch Dis Child Fetal Neonatal Ed 2012; 97 (03) F214-F216
• 16 Weinberger SM, Kandall SR, Doberczak TM, Thornton JC, Bernstein J. Early weight-change patterns in neonatal abstinence. Am J Dis Child 1986; 140 (08) 829-832
• 17 Taylor K, Maguire D. A review of feeding practices in infants with neonatal abstinence syndrome. Adv Neonatal Care 2020; 20 (06) 430-439
• 18 Cohen J. Statistical Power Analysis for the Behavioral Sciences. 2nd ed.. Hillsdale, NJ: Lawrence Erlbaum Associates; 1988
• 19 Alghamdi NRT, Sharp C, Jortani S. Interpretation of discordant results in maternal/newborn dyad drug screening. Am J Pediatr 2022; 8 (04) 244-251
• 20 Colby JM. Comparison of umbilical cord tissue and meconium for the confirmation of in utero drug exposure. Clin Biochem 2017; 50 (13–14): 784-790
• 21 Young LW, Ounpraseuth ST, Merhar SL. et al; ACT NOW Collaborative. Eat, sleep, console approach or usual care for neonatal opioid withdrawal. N Engl J Med 2023; 388 (25) 2326-2337
• 22 Grossman M, Seashore C, Holmes AV. Neonatal abstinence syndrome management: a review of recent evidence. Rev Recent Clin Trials 2017; 12 (04) 226-232
• 23 Grossman MR, Berkwitt AK, Osborn RR. et al. An initiative to improve the quality of care of infants with neonatal abstinence syndrome. Pediatrics 2017; 139 (06) e20163360
• 24 Abdel-Latif ME, Pinner J, Clews S, Cooke F, Lui K, Oei J. Effects of breast milk on the severity and outcome of neonatal abstinence syndrome among infants of drug-dependent mothers. Pediatrics 2006; 117 (06) e1163-e1169
• 25 Favara MT, Carola D, Jensen E. et al. Maternal breast milk feeding and length of treatment in infants with neonatal abstinence syndrome. J Perinatol 2019; 39 (06) 876-882
• 26 Dixon SD. Effects of transplacental exposure to cocaine and methamphetamine on the neonate. West J Med 1989; 150 (04) 436-442
• 27 Oro AS, Dixon SD. Perinatal cocaine and methamphetamine exposure: maternal and neonatal correlates. J Pediatr 1987; 111 (04) 571-578
• 28 Pandey R, Kanike N, Ibrahim M. et al. Lactose-free infant formula does not change outcomes of neonatal abstinence syndrome (NAS): a randomized clinical trial. J Perinatol 2021; 41 (03) 598-605
• 29 Chu LF, Liang DY, Li X. et al. From mouse to man: the 5-HT3 receptor modulates physical dependence on opioid narcotics. Pharmacogenet Genomics 2009; 19 (03) 193-205
• 30 Mihatsch WA, Franz AR, Högel J, Pohlandt F. Hydrolyzed protein accelerates feeding advancement in very low birth weight infants. Pediatrics 2002; 110 (06) 1199-1203
• 31 Kellams A, Harrel C, Omage S, Gregory C, Rosen-Carole C. ABM clinical protocol #3: supplementary feedings in the healthy term breastfed neonate, revised 2017. Breastfeed Med 2017; 12: 188-198
• 32 Goldberg DL, Becker PJ, Brigham K. et al. Identifying malnutrition in preterm and neonatal populations: recommended indicators. J Acad Nutr Diet 2018; 118 (09) 1571-1582

Publication History

Received: 19 December 2023

Accepted: 17 September 2024

Article published online:
10 October 2024