Advertisement

Associations between malaria in pregnancy and neonatal neurological outcomes

Open AccessPublished:July 17, 2021DOI:https://doi.org/10.1016/j.ijid.2021.07.037

      Highlights

      • A prospective observational study of in utero malaria exposure and neonatal neurological functioning was conducted.
      • In utero malaria exposure may increase the risk of suboptimal reflexes in term-born neonates.
      • The impact of in utero malaria exposure on child neurodevelopment must be established.

      Abstract

      Objective

      To compare neurological functioning of neonates born to mothers with and without malaria in pregnancy.

      Methods

      Pregnant women presenting at Korle Bu Teaching Hospital, Ghana were recruited into this prospective observational study. Malaria exposure was determined by clinically documented antenatal malaria infection; parasitemia in maternal, placental, or umbilical cord blood; or placental histology. Neurological functioning was assessed using the Hammersmith Neonatal Neurological Examination within 48 hours of birth. Performance was classified as "optimal" or "suboptimal" by subdomain and overall.

      Results

      Between November 21, 2018 and February 10, 2019, a total of 211 term-born neonates, of whom 27 (13%) were exposed to malaria in pregnancy, were included. In the reflexes subdomain, exposed neonates tended to score lower (adjusted mean difference -0.34, 95% confidence interval -0.70 to 0.03), with an increased risk (adjusted risk ratio 1.63, 95% confidence interval 1.09 to 2.44) of suboptimal performance compared with unexposed neonates. There were no significant between-group differences in scores or optimality classification for the remaining subdomains and overall.

      Conclusions

      Malaria-exposed neonates had similar neurological functioning relative to unexposed neonates, with differences confined to the reflexes subdomain, suggesting potential underlying neurological immaturity or injury. Further studies are needed to confirm these findings and determine the significance of malaria in pregnancy on long-term neurological outcomes.

      KEYWORDS

      1. Introduction

      Naturally acquired immunity to malaria is compromised during pregnancy, and pregnant women in endemic regions are at higher risk of malaria infection than their non-pregnant peers (
      • Doolan DL
      • Dobano C
      • Baird JK.
      Acquired immunity to malaria.
      ). It is well-established that malaria in pregnancy is associated with adverse pregnancy outcomes, including miscarriage and stillbirth (
      • Saito M
      • Briand V
      • Min AM
      • McGready R.
      Deleterious effects of malaria in pregnancy on the developing fetus: a review on prevention and treatment with antimalarial drugs.
      ), as well as maternal and fetal/neonatal complications such as malarial anemia, fetal growth restriction, preterm birth, and low birthweight (
      • Rogerson SJ.
      Management of malaria in pregnancy.
      ). Approximately 11 million pregnant women in sub-Saharan Africa were infected with malaria in 2018, resulting in 16% of all low birthweight deliveries in the region (). While the adverse neurodevelopmental outcomes of children who have suffered from cerebral malaria during childhood have been investigated extensively (
      • Carter JA
      • Neville BG
      • White S
      • Ross AJ
      • Otieno G
      • Mturi N
      • et al.
      Increased prevalence of epilepsy associated with severe falciparum malaria in children.
      ;
      • Idro R
      • Kakooza-Mwesige A
      • Balyejjussa S
      • Mirembe G
      • Mugasha C
      • Tugumisirize J
      • et al.
      Severe neurological sequelae and behaviour problems after cerebral malaria in Ugandan children.
      ), relatively little is known regarding the impact of malaria in pregnancy on neonatal neurological outcomes. Published reviews have theorized that malaria exposure can impair fetal neurological development and subsequent neurodevelopment (
      • Lawford HLS
      • Lee AC
      • Kumar S
      • Liley HG
      • Bora S.
      Establishing a conceptual framework of the impact of placental malaria on infant neurodevelopment.
      ;
      • McDonald CR
      • Elphinstone RE
      • Kain KC.
      The impact of placental malaria on neurodevelopment of exposed infants: a role for the complement system?.
      ); a number of socio-environmental and biological pathways are hypothesized to be involved, which we recently summarized in a conceptual framework (
      • Lawford HLS
      • Lee AC
      • Kumar S
      • Liley HG
      • Bora S.
      Establishing a conceptual framework of the impact of placental malaria on infant neurodevelopment.
      ).
      Human and animal studies suggest some neurological impact of malaria exposure in pregnancy. Cerebral blood flow redistribution (
      • Arbeille P
      • Carles G
      • Bousquet F
      • Body G
      • Lansac J.
      Fetal cerebral and umbilical artery blood flow changes during pregnancy complicated by malaria.
      ) and faster development in the cingulate gyrus (
      • Rijken MJ
      • de Wit MC
      • Mulder EJH
      • Kiricharoen S
      • Karunkonkowit N
      • Paw T
      • et al.
      Effect of malaria in pregnancy on foetal cortical brain development: a longitudinal observational study.
      ) have been documented in fetuses in response to maternal malaria infection, while neurocognitive deficits are evident in the offspring of malaria-infected mice relative to uninfected mice (
      • McDonald CR
      • Cahill LS
      • Ho KT
      • Yang J
      • Kim H
      • Silver KL
      • et al.
      Experimental malaria in pregnancy induces neurocognitive injury in uninfected offspring via a C5a-C5a receptor dependent pathway.
      ). However, only one study to date has reported the neurodevelopmental impact of malaria exposure in pregnancy among infants. This case report investigated neurodevelopmental outcomes at 12 and 24 months postpartum in dizygotic twins whose placentas were discordant for parasitemia; the placental malaria-exposed twin demonstrated consistently lower motor, cognitive, and language scores relative to the unexposed twin at both time points (
      • Conroy AL
      • Bangirana P
      • Muhindo MK
      • Kakuru A
      • Jagannathan P
      • Opoka RO
      • et al.
      Case report: birth outcome and neurodevelopment in placental malaria discordant twins.
      ). However, there was marked discordance in fetal growth, with the malaria-exposed twin exhibiting a lower birthweight (1320 g vs. 1920 g) and head circumference (27 cm vs. 32 cm). As neurodevelopmental disadvantage has previously been reported in the smaller twin of discordant twin pairs regardless of malaria status (
      • Halling C
      • Malone FD
      • Breathnach FM
      • Stewart MC
      • McAuliffe FM
      • Morrison JJ
      • et al.
      Neuro-developmental outcome of a large cohort of growth discordant twins.
      ), it is unclear whether the neurodevelopmental outcomes reported occurred as a component of the pathophysiology of malaria or whether this was an independent confounder.
      To date, no studies have reported neurological functioning of neonates exposed to malaria in utero. This prospective observational study was conducted to compare the neurological functioning of neonates ≤48 hours of age born to mothers with and without malaria in pregnancy. It was hypothesized that exposure to malaria in pregnancy would adversely affect neonatal neurological functioning.

      2. Methods

      2.1 Sample

      The Impact of Malaria in Pregnancy on Infant Neurodevelopment (IMPRINT) study is a prospective observational study conducted at Korle Bu Teaching Hospital in Accra, Ghana. This is the largest tertiary teaching hospital in Ghana and the leading regional referral center, with additional referrals from primary and secondary health facilities in the southern region. It has a catchment population of more than three million in an area of 50 km radius (
      • Adu-Bonsaffoh K
      • Ntumy MY
      • Obed SA
      • Seffah JD.
      Perinatal outcomes of hypertensive disorders in pregnancy at a tertiary hospital in Ghana.
      ) and approximately 10,000 live births annually.
      Six physicians were recruited and trained to perform study assessments. Pregnant women presenting in the early stages of labor were approached for written informed consent. If granted and a member of the study team was available, neonates that met the eligibility criteria were enrolled. Women were not approached if they were <15 years of age, HIV-positive, or had sickle cell disease. A nested sample of singleton neonates was selected for this study by further excluding those who 1) were born preterm or post-term (<37 + 0 or >42 + 6, weeks + days of gestation), 2) had an Apgar score <7 at 5 minutes, 3) had any recorded admission to the neonatal intensive care unit, or 4) had any recorded diagnosis of congenital anomalies. Ethical approval was obtained from the institutional review boards of the University of Ghana and The University of Queensland, Australia.

      2.2 Malaria diagnosis

      Malaria infection during pregnancy was the primary exposure, measured as a binary variable. At Korle Bu Teaching Hospital, pregnant women are routinely tested for malaria at their antenatal visits. If a woman tests positive, she is treated as per the national malaria treatment guidelines for pregnant women. A neonate was classified to be in the "exposed" group if they met one or more of the following conditions: 1) medical records of antenatal malaria infection confirmed by rapid diagnostic test (RDT) or microscopy; 2) positive maternal, placental, or umbilical cord blood samples tested by RDT and/or microscopy; or 3) placental histology. Supplementary Material Appendix 1 further describes how malaria was diagnosed in the exposed group.

      2.3 Neurological evaluation

      The primary outcome was performance on the Hammersmith Neonatal Neurological Examination (HNNE). The HNNE can identify neonates at risk of neurological dysfunction and later neurodevelopmental impairment (
      • Dubowitz LM
      • Dubowitz V
      • Palmer PG
      • Miller G
      • Fawer CL
      • Levene MI.
      Correlation of neurologic assessment in the preterm newborn infant with outcome at 1 year.
      ;
      • Molteno C
      • Grosz P
      • Wallace P
      • Jones M.
      Neurological examination of the preterm and full-term infant at risk for developmental disabilities using the Dubowitz Neurological Assessment.
      ;
      • Molteno CD
      • Thompson MC
      • Buccimazza SS
      • Magasiner V
      • Hann FM.
      Evaluation of the infant at risk for neurodevelopmental disability.
      ;
      • Setanen S
      • Lehtonen L
      • Parkkola R
      • Aho K
      • Haataja L
      • Group PS.
      Prediction of neuromotor outcome in infants born preterm at 11 years of age using volumetric neonatal magnetic resonance imaging and neurological examinations.
      ;
      • Tuhkanen H
      • Pajulo M
      • Jussila H
      • Ekholm E.
      Infants born to women with substance use: exploring early neurobehavior with the Dubowitz neurological examination.
      ) and exhibits good sensitivity (88%) to identify significant neuropathology detected by magnetic resonance imaging (
      • Woodward LJ
      • Mogridge N
      • Wells SW
      • Inder TE.
      Can neurobehavioral examination predict the presence of cerebral injury in the very low birth weight infant?.
      ). The HNNE has a total of 34 items stratified into six subdomains: tone, tone patterns, reflexes, movements, abnormal signs/patterns, and orientation and behavior. A scoring system was developed in 1998 based on reference values from a low-risk, term-born sample of 224 British neonates (
      • Dubowitz L
      • Mercuri E
      • Dubowitz V.
      An optimality score for the neurologic examination of the term newborn.
      ). This scoring system allows the classification of neonate performance as "optimal" or "suboptimal" by each subdomain and overall. A score >10th centile of reference values is considered optimal. The HNNE administration and scoring are described in detail in the original publication (
      • Dubowitz L
      • Mercuri E
      • Dubowitz V.
      An optimality score for the neurologic examination of the term newborn.
      ).
      The HNNE was administered to all neonates in the IMPRINT study (irrespective of inclusion in this nested sample) at ≤48 hours after birth by trained physicians in the postnatal ward, using the standardized assessment proforma (
      • Dubowitz L
      • Mercuri E
      • Dubowitz V.
      An optimality score for the neurologic examination of the term newborn.
      ). Details regarding physician training for this study have been described in previous publications (
      • Lawford HLS
      • Nuamah MA
      • Liley HG
      • Lee AC
      • Kumar S
      • Adjei AA
      • et al.
      Neonatal neurological examination in a resource-limited setting: what defines normal?.
      ;
      • Lawford HLS
      • Nuamah MA
      • Liley HG
      • Lee AC
      • Botchway F
      • Kumar S
      • et al.
      Gestational age-specific distribution of the Hammersmith Neonatal Neurological Examination scores among low-risk neonates in Ghana.
      ). Examiners were not routinely blinded to gestational age at birth, but were blinded to malaria status.

      2.4 Sociodemographic, clinical, and placental characteristics

      Sociodemographic information was collected using a standardized questionnaire administered when participants were not in active labor and also following birth. Maternal and neonatal clinical data were extracted from medical records, and the placenta was characterized by examination. Further details are described in Supplementary Material Appendix 1.

      2.5 Statistical analysis

      Differences in sociodemographic, clinical, and placental characteristics between included and excluded neonates, and malaria-exposed and unexposed neonates were described using mean ± standard deviation (SD), median [interquartile range], or number (%), and were tested using Student's t-test or Mann–Whitney U test for continuous data and Pearson's chi-squared test of independence or Fisher's exact test for categorical data. The association between malaria exposure and mean raw scores for the six HNNE subdomains and overall was assessed using linear regression, and standardized effect sizes reported as Cohen's d values with 95% confidence intervals. The association between malaria exposure and the proportion of neonates classified as suboptimal for the HNNE subdomains and overall was assessed using a Poisson regression with robust error variance. Multivariable models were adjusted for covariates determined using our previously published conceptual framework (
      • Lawford HLS
      • Lee AC
      • Kumar S
      • Liley HG
      • Bora S.
      Establishing a conceptual framework of the impact of placental malaria on infant neurodevelopment.
      ), summarized in a qualitative causal model designed using www.dagitty.net (Supplementary Material Appendix 2). The selected covariates are shown in red (socioeconomic status, education, maternal age, and social risk). There was no adjustment for covariates on the causal pathway (shown in green). Measures of association were expressed as unadjusted and adjusted mean differences and risk ratios. The statistical analysis was conducted using Stata 16.0 (Stata Corp, College Station, TX, USA) and a significance level of 0.05 was used throughout inferential analys

      3. Results

      Figure 1 displays the sample recruitment. Between November 21, 2018 and February 10, 2019, a total of 302 mothers and 310 neonates (eight twin births) were recruited. In total, 36/310 neonates met the study criteria for exposure to malaria in pregnancy. The HNNE was administered to 296/310 neonates (34/36 exposed to malaria and 262/274 unexposed) within 48 hours of birth. Of the 14 neonates who were not administered the HNNE, eight were lost to follow-up, five were too unwell to be assessed, and there was one neonatal death.
      After the exclusion of seven exposed and 78 unexposed neonates who did not meet the criteria for this nested sample, the current study sample comprised 211 eligible neonates of whom 27 (13%) were exposed to malaria. Demographic and clinical characteristics of the included mother–neonate dyads (n = 211) and dyads that either did not have the HNNE administered (n = 14) or did not meet the inclusion criteria (n = 85) are compared in Supplementary Material Appendix 3 and Appendix 4, respectively.

      3.1 Characteristics of mother–neonate dyads

      Table 1 displays the sociodemographic, clinical, and placental characteristics of the 211 included mother–neonate dyads by malaria exposure. Compared with mothers of unexposed neonates, significantly more mothers of exposed neonates had no other children (P = 0.003) and lived in overcrowded dwellings with more than one person per room (P = 0.03). Mothers of exposed neonates had a smaller average middle-upper arm circumference compared with mothers of unexposed neonates (P = 0.03). Significant differences were evident in the timing of first intermittent preventative treatment in pregnancy using sulfadoxine–pyrimethamine (IPTp-SP); while the majority of mothers of exposed and unexposed neonates took their first IPTp-SP dose in the first/second trimester, fewer mothers of exposed neonates took no IPTp-SP but more took their first IPTp-SP in the third trimester relative to mothers of unexposed neonates (P = 0.03). Mothers of exposed and unexposed neonates did not differ significantly for the remaining sociodemographic or maternal clinical variables, and exposed and unexposed neonates did not differ significantly with regards to clinical or placental characteristics.
      Table 1Sociodemographic and clinical characteristics of neonates assessed using the Hammersmith Neonatal Neurological Examination according to malaria exposure
      Characteristics
      Data are presented as mean ± standard deviation, number (%), or median [interquartile range].
      All (N = 211)Malaria-exposed (N = 27)Malaria-unexposed (N = 184)P-value
      Maternal demographics
      Age, years31 ± 630 ± 732 ± 60.14
      Literate
       No46 (22.3)5 (18.5)41 (22.9)0.60
       Yes160 (77.7)22 (81.5)138 (77.1)
      Education
       None/primary/secondary137 (66.2)22 (81.5)115 (63.9)0.07
       Higher70 (33.8)5 (18.5)65 (36.1)
      Amount worked
       None/occasional/seasonal39 (18.8)5 (18.5)34 (18.9)0.96
       Full-time168 (81.2)22 (81.5)146 (81.1)
      Wealth quintile
       Poorest (1st–3rd)100 (47.4)15 (55.6)85 (46.2)0.36
       Richest (4th–5th)111 (52.6)12 (44.4)99 (53.8)
      Health insurance
       No6 (2.9)1 (3.7)5 (2.8)0.79
       Yes200 (97.1)26 (96.3)174 (97.2)
      Other children
       None42 (20.5)11 (42.3)31 (17.3)0.003
       ≥1163 (79.5)15 (57.7)148 (82.7)
      Overcrowding
       ≤1 person per room51 (24.6)2 (7.4)49 (27.2)0.03
       >1 person per room156 (75.4)25 (92.6)131 (72.8)
      Social risk
       Low risk (no risk factor)106 (50.2)15 (55.6)91 (49.5)0.55
       High risk (≥1 risk factor)105 (49.8)12 (44.4)93 (50.5)
      Maternal clinical
      Time of first antenatal visit
       Second/third trimester70 (34.3)10 (37.0)60 (33.9)0.75
       First trimester134 (65.7)17 (63.0)117 (66.1)
      Gravidity3.3 ± 1.72.7 ± 1.93.3 ± 1.70.09
      Middle upper arm circumference, cm31.3 ± 4.129.6 ± 3.631.5 ± 4.10.03
      Hemoglobin level, g/dl10.2 ± 1.510.0 ± 2.010.3 ± 1.40.29
      Anxiety2 [0, 4]2 [0, 6]2 [0, 4]0.41
      Depression2 [0, 5]2 [0, 5]2 [0, 5]0.40
      Clinical risk
       Low risk (no risk factor)152 (72.0)17 (63.0)135 (73.4)0.26
       High risk (≥1 risk factor)59 (28.0)10 (37.0)49 (26.6)
      Malaria prevention
      Insecticide-treated bed net use in pregnancy
       Did not use/no bed net133 (63.0)16 (59.3)117 (63.6)0.72
       Used in pregnancy78 (37.0)11 (40.7)67 (36.4)
      Total IPTp-SP doses2 [1, 3]2 [2, 3]2 [1, 3]0.28
      Trimester of first IPTp-SP
       No IPTp-SP/not specified36 (17.1)1 (3.7)35 (19.0)0.03
       Third trimester44 (20.9)10 (37.0)34 (18.5)
       First or second trimester131 (62.1)16 (59.3)115 (62.5)
      Neonatal clinical
      Mode of delivery
       Cesarean section134 (63.8)14 (51.9)120 (65.6)0.17
       Spontaneous vaginal/vacuum extraction76 (36.2)13 (48.1)63 (34.4)
      Gestational age, weeks39 [38, 40]39 [38, 40]39 [38, 40]0.81
      Birthweight, kg3.2 ± 0.43.2 ± 0.43.1 ± 0.40.40
      Birthweight Z-score-0.3 ± 0.9-0.2 ± 0.9-0.3 ± 0.90.44
      Low birthweight
       No201 (95.7)25 (92.6)176 (96.2)0.39
       Yes9 (4.3)2 (7.4)7 (3.8)
      Apgar score 1 minute8 [7, 8]8 [7, 8]8 [7, 8]0.60
      Apgar score 5 minutes9 [8, 9]9 [8, 9]9 [8, 9]0.54
      Length, cm50 [49, 52]51 [50, 52]50 [49, 52]0.13
      Chest circumference, cm33 [32, 34]33 [31, 34]33 [32, 34]0.18
      Head circumference, cm34 [33, 35]34 [33, 35]34 [33, 35]0.62
      Ponderal index2.6 ± 0.82.5 ± 0.32.6 ± 0.90.50
      Sex
       Female102 (48.6)15 (55.6)87 (47.5)0.44
       Male108 (51.4)12 (44.4)96 (52.5)
      Placental assessment
      Placental abnormality
       None50 (23.7)10 (37.0)40 (21.7)0.16
       1 abnormality82 (38.9)7 (25.9)75 (40.8)
       >1 abnormality79 (37.4)10 (37.0)69 (37.5)
      Cord length, cm52.0 [45.5, 60.0]55.5 [46.0, 60.5]51.6 [45.5, 59.2]0.37
      Cord diameter, cm1.2 [1, 1.5]1.3 [1, 1.5]1.2 [1, 1.5]0.69
      Umbilical coiling index0.08 ± 0.090.07 ± 0.080.08 ± 0.090.86
      Placental weight, g472.1 ± 100.6478.8 ± 105.4471.2 ± 100.30.73
      Placental thickness, cm1.9 ± 0.41.9 ± 0.31.8 ± 0.40.48
      IPTp-SP, intermittent preventative treatment in pregnancy using sulfadoxine–pyrimethamine.
      a Data are presented as mean ± standard deviation, number (%), or median [interquartile range].
      Of the 27 mothers who had malaria in pregnancy, 14 (52%) had an active malaria infection at birth (positive RDT and/or blood smear). The timing and type of antimalarial treatment for these cases were not recorded. There were five cases of past-chronic placental infection and one case of active-chronic placental infection. Eleven (41%) mothers had evidence of malaria infection from the medical records; of these, two were in the first trimester, two in the second trimester, and three in the third trimester. The timing of infection was not recorded for four infections.

      3.2 Neurological functioning of neonates

      Unadjusted and adjusted mean differences in raw scores for the six HNNE subdomains and overall were similar for exposed and unexposed neonates (Table 2). However, in both the unadjusted and adjusted models, exposed neonates tended to score lower on the reflexes subdomain (adjusted mean difference -0.34, 95% confidence interval -0.70 to 0.03).
      Table 2Unadjusted and adjusted mean differences in raw scores of the Hammersmith Neonatal Neurological Examination subdomains according to malaria exposure
      HNNE subdomainMalaria-exposedMalaria-unexposedMean differenceAdjusted mean difference
      Adjusted for socioeconomic status, education, maternal age, and social risk.
      NMean ± SDNMean ± SDMean difference (95% CI)P-valueCohen's d (95% CI)Mean difference (95% CI)P-valueCohen's d (95% CI)
      Tone277.1 ± 2.61827.3 ± 2.2-0.18 (-1.09, 0.73)0.70-0.08 (-0.48, 0.33)-0.00 (-0.93, 0.93)0.10-0.00 (-0.41, 0.40)
      Tone patterns274.2 ± 0.81844.1 ± 0.80.13 (-0.18, 0.44)0.410.17 (-0.24, 0.57)0.14 (-0.18, 0.46)0.400.18 (-0.23, 0.58)
      Reflexes274.6 ± 1.01785.0 ± 0.9-0.33 (-0.69, 0.03)0.07-0.37 (-0.78, 0.03)-0.34 (-0.70, 0.03)0.07-0.38 (-0.78, 0.03)
      Movements241.9 ± 0.91791.8 ± 0.80.06 (-0.30, 0.42)0.750.06 (-0.36, 0.49)0.03 (-0.33, 0.40)0.850.04 (-0.39, 0.47)
      Abnormal signs/ patterns272.4 ± 0.61812.3 ± 0.70.06 (-0.21, 0.34)0.660.09 (-0.31, 0.50)0.07 (-0.21, 0.34)0.640.10 (-0.31, 0.50)
      Orientation and behavior204.7 ± 1.51654.5 ± 1.60.19 (-0.54, 0.93)0.600.12 (-0.34, 0.59)0.28 (-0.47, 1.03)0.460.18 (-0.29, 0.64)
      Total HNNE score1925.2 ± 3.516025.0 ± 3.80.23 (-1.56, 2.02)0.800.06 (-0.41, 0.54)0.46 (-1.35, 2.27)0.620.12 (-0.35, 0.60)
      HNNE, Hammersmith Neonatal Neurological Examination; SD, standard deviation; CI, confidence interval.
      a Adjusted for socioeconomic status, education, maternal age, and social risk.
      As shown in Table 3, a large proportion of neonates were considered to be demonstrating "suboptimal" performance [using the original British scoring thresholds (
      • Dubowitz L
      • Mercuri E
      • Dubowitz V.
      An optimality score for the neurologic examination of the term newborn.
      )] by HNNE subdomain: 67% for tone, 67% tone patterns, 37% reflexes, 82% movements, 61% abnormal signs/patterns, 75% orientation and behavior, and 95% overall. In the reflexes subdomain, significantly more neonates exposed to malaria scored suboptimally than unexposed neonates (56% vs. 34%; adjusted risk ratio 1.63, 95% confidence interval 1.09 to 2.44). There were no significant differences between exposed and unexposed neonates in the risk of suboptimal scores for tone, tone patterns, movements, abnormal signs/patterns, or orientation and behavior. Finally, the association between scoring suboptimally by HNNE subdomain and overall was investigated separately for active (n = 14) and past (n = 13) malaria infection; however, no significant difference was evident.
      Table 3Risk of suboptimal scores in the Hammersmith Neonatal Neurological Examination subdomains according to malaria exposure
      HNNE subdomainNeonates with suboptimal scores, n/N (%)UnadjustedAdjusted
      Adjusted for socioeconomic status, education, maternal age, and social risk.
      Malaria-exposedMalaria-unexposedAllP-valueRisk ratio (95% CI)P-valueRisk Ratio (95% CI)
      Tone17/27 (63.0)123/182 (67.6)140/209 (67.0)0.650.93 (0.69, 1.27)0.520.90 (0.66, 1.23)
      Tone patterns16/27 (59.3)125/184 (67.9)141/211 (66.8)0.420.87 (0.63, 1.21)0.400.87 (0.62, 1.21)
      Reflexes15/27 (55.6)61/178 (34.3)76/205 (37.1)0.021.62 (1.09, 2.41)0.021.63 (1.09, 2.44)
      Movements18/24 (75.0)149/179 (83.2)167/203 (82.3)0.400.90 (0.71, 1.15)0.510.92 (0.71, 1.19)
      Abnormal signs/patterns17/27 (63.0)110/181 (60.8)127/208 (61.1)0.831.03 (0.76, 1.42)0.841.03 (0.75, 1.42)
      Orientation and behavior14/20 (70.0)125/165 (75.8)139/185 (75.1)0.610.92 (0.68, 1.25)0.530.91 (0.68, 1.22)
      Total HNNE score18/19 (94.7)152/160 (95.0)170/179 (95.0)0.961.00 (0.89, 1.12)0.951.00 (0.90, 1.11)
      HNNE, Hammersmith Neonatal Neurological Examination; CI, confidence interval.
      a Adjusted for socioeconomic status, education, maternal age, and social risk.

      4. Discussion

      The objective of this study was to compare neurological functioning of malaria-exposed and unexposed neonates with a widely used, validated, structured neurological assessment tool. Examining neonates prior to hospital discharge allowed us to assess the impact of malaria in pregnancy without the risk of confounding from subsequent exposure to early childhood illnesses and family socioeconomic adversities that may affect studies of outcomes in infancy. Further, assessing neonates within the first 48 hours of life has the advantage of early detection of neurological abnormalities, which can lead to opportunities for targeted interventions.
      It was found that in utero malaria-exposed neonates ≤48 hours of age had similar total HNNE scores to their unexposed peers. Interestingly, in the reflexes subdomain only, a statistically significant higher risk of suboptimal scores was found (which persisted after adjusting for socioeconomic status, education, maternal age, and social risk), although the mean difference in raw scores was small and did not reach statistical significance. There were no significant associations between malaria exposure and mean raw scores or suboptimal functioning in the tone, tone patterns, movements, abnormal signs/patterns, or orientation and behavior subdomains of the HNNE. Finding a significant difference in only one of the six HNNE subdomains could signify a selective effect on specific neurological function, but also raises the possibility that the finding was due to chance alone, since no adjustment of statistical significance was made for multiple comparisons. These findings may also be a result of the study being underpowered due to the small sample size, thus the study may not be adequately powered to detect patterns of malaria-related abnormality but might support the finding with the reflexes subdomain only. Although all statistical analyses were predetermined according to a priori hypotheses, we recognize the limitations on the certainty of the current findings and as such, we emphasize the preliminary nature of the study findings and highlight that this study was designed for hypothesis generation.
      If there is a true differential impact on primitive reflexes over tone, movements, and behavior, the mechanism and implications are uncertain. HNNE reflexes scores have been strongly associated with motor and cognitive outcomes in preterm-born infants assessed at 32 weeks of postmenstrual age (
      • George JM
      • Colditz PB
      • Chatfield MD
      • Fiori S
      • Pannek K
      • Fripp J
      • et al.
      Early clinical and MRI biomarkers of cognitive and motor outcomes in very preterm born infants.
      ). Suboptimal reflexes subdomain scores have also predicted poor neurodevelopmental outcomes, including lower mental and psychomotor development indices (
      • Molteno C
      • Grosz P
      • Wallace P
      • Jones M.
      Neurological examination of the preterm and full-term infant at risk for developmental disabilities using the Dubowitz Neurological Assessment.
      ;
      • Sanchez K
      • Morgan AT
      • Slattery JM
      • Olsen JE
      • Lee KJ
      • Anderson PJ
      • et al.
      Neuropredictors of oromotor feeding impairment in 12 month-old children.
      ), and structural brain abnormalities, including reduced biparietal diameter, increasing severity of cerebral white and gray matter abnormalities, and cerebellar abnormalities (
      • Eeles AL
      • Walsh JM
      • Olsen JE
      • Cuzzilla R
      • Thompson DK
      • Anderson PJ
      • et al.
      Continuum of neurobehaviour and its associations with brain MRI in infants born preterm.
      ;
      • George JM
      • Fiori S
      • Fripp J
      • Pannek K
      • Guzzetta A
      • David M
      • et al.
      Relationship between very early brain structure and neuromotor, neurological and neurobehavioral function in infants born <31weeks gestational age.
      ;
      • Sanchez K
      • Morgan AT
      • Slattery JM
      • Olsen JE
      • Lee KJ
      • Anderson PJ
      • et al.
      Neuropredictors of oromotor feeding impairment in 12 month-old children.
      ;
      • Woodward LJ
      • Mogridge N
      • Wells SW
      • Inder TE.
      Can neurobehavioral examination predict the presence of cerebral injury in the very low birth weight infant?.
      ). A recent study in Brazil reported reduced head circumference in neonates born to malaria-infected mothers (
      • Dombrowski JG
      • de Souza RM
      • Lima FA
      • Bandeira CL
      • Murillo O
      • de Sousa Costa D
      • et al.
      Association of malaria infection during pregnancy with head circumference of newborns in the Brazilian Amazon.
      ); however there was no intergroup difference in head circumference in the current study. We can speculate that exposure to malaria in pregnancy results in adverse neurodevelopmental outcomes and/or subtle alterations in brain development. Unlike changes in gross brain structure, subtle changes would not manifest as differences in HNNE scores across all domains. However, without incorporating neurodevelopmental follow-up of exposed neonates, or including neuroimaging in the study, we cannot determine whether any such brain pathology or long-term neurological adversities exist in malaria-exposed infants.
      An alternative explanation for why so little difference was found between malaria-exposed and unexposed neonates is the heterogeneity of malaria exposure in the sample and the lack of a dense placental inflammatory response with pigmented monocytes that may be mitigating the possible effects of malaria infection. An important pathway identified in our previously published conceptual framework was the role of maternal immune-inflammatory dysfunction and the downstream effects of inflammatory factors and the immune system on fetal brain development (
      • Lawford HLS
      • Lee AC
      • Kumar S
      • Liley HG
      • Bora S.
      Establishing a conceptual framework of the impact of placental malaria on infant neurodevelopment.
      ). However, if there was only clinically mild malaria in the sample with little acute or chronic placental malaria infection, it is unlikely that heightened maternal immuno-inflammatory responses would occur, which would be responsible for impaired fetal brain development and subsequent neonatal neurological functioning. It is possible that replicating this study in a population with denser placental parasitization would find different results. However, this approach presents serious ethical challenges common to other studies of "natural history" of disease, in that a duty of care would be owed to mothers participating in research to provide them with optimal treatment if malaria is diagnosed early in pregnancy. While there are no major ethical challenges around recruiting women with intense placental inflammation in the labor ward, as in this study, this does increase the challenge of determining the importance of timing of malaria infection on neurological outcomes.
      A limitation of this study is that exposure to antimalarial treatment among women with active malaria at birth was not recorded. According to the standard treatment guidelines for malaria in Ghana, pregnant women are administered either artesunate + amodiaquine, artemether + lumefantrine, or oral quinine for uncomplicated malaria in the second or third trimester (

      Ministry of Health & Ghana Health Services. Guidelines for Case Management of Malaria in Ghana. 3rd Edition ed 2014.

      ), all of which have a good safety profile. Antenatal maternal treatment could have reduced the impact of exposure to malaria on the neonate, biasing the study towards finding no difference between the groups (whereas a study of women without access to treatment might have shown differences). However, if antimalarial drugs adversely affected the neurological functioning of neonates, we would have expected this to have exaggerated differences between the malaria-exposed and unexposed groups.
      It is important to acknowledge that, despite being the first study published to date investigating the impact of malaria in pregnancy on neonatal neurological functioning, this study may have been underpowered because of the small sample size (particularly the sample of neonates exposed to malaria in pregnancy). Given the small sample size and the multiple comparisons in this study, we advise caution in interpreting statistical significance. It is possible that the finding of a difference in neonates meeting the threshold for suboptimal performance in only one of six subdomains is the result of a type 2 error. It is also possible that the increased risk of suboptimal reflexes seen in neonates exposed to malaria could be due to chance (a type 1 error), subtle biases, or unmeasured confounders. The adjusted mean difference between groups for raw scores for the reflexes subdomain was only about a third of a standard deviation and was not statistically significant. The difference found may or may not be clinically significant, and a much larger sample size might find subtle (and yet clinically significant) differences in other subdomains or in total HNNE scores that this study was too small to detect. Ultimately, longitudinal studies are needed to determine the significance of malaria exposure during pregnancy on childhood neurodevelopment, and to distinguish the effects of maternal malaria infection from concomitant comorbid conditions. This will allow an understanding of both the childhood impact of malaria in pregnancy and the specificity and predictive value of neurological assessments at birth in this context.
      The HNNE was selected as the most appropriate neurological assessment tool for this study; it assesses neurological functioning at birth, has been used widely both in clinical and research contexts, has excellent (>96%) inter-rater reliability (
      • Dubowitz L
      • Mercuri E
      • Dubowitz V.
      An optimality score for the neurologic examination of the term newborn.
      ), and has high predictive validity to identify structural brain abnormalities and later neurological dysfunction. However, the HNNE has only been used infrequently for research in low- and middle-income countries and has not been validated or standardized in Ghana. Therefore, we are hesitant to interpret Ghanaian neonates as performing "suboptimally" using this (original British) scoring system without more extensive validation of the HNNE in Ghana or follow-up of the sample to determine long-term neurological functioning. As we are unsure of the reasons why such a high proportion of Ghanaian neonates in the comparison group scored suboptimally, HNNE raw scores were also compared, but still little difference was found between the groups.
      Based on the original HNNE scoring system (
      • Dubowitz L
      • Mercuri E
      • Dubowitz V.
      An optimality score for the neurologic examination of the term newborn.
      ), we would expect that approximately 10% of the malaria-unexposed comparison group would be scoring suboptimally. However, it was found that a much higher proportion of unexposed neonates scored below the 10th centile when the British scoring system was applied, although we are uncertain about whether this indicates a much higher baseline of adverse neurological functioning in term-born, malaria-unexposed neonates specific to the study site. The possible reasons for these findings have been discussed in both the IMPRINT study (
      • Lawford HLS
      • Nuamah MA
      • Liley HG
      • Lee AC
      • Kumar S
      • Adjei AA
      • et al.
      Neonatal neurological examination in a resource-limited setting: what defines normal?.
      ) and studies conducted in other low- and middle-income countries [Thailand, Myanmar (
      • McGready R
      • Simpson J
      • Panyavudhikrai S
      • Loo S
      • Mercuri E
      • Haataja L
      • et al.
      Neonatal neurological testing in resource-poor settings.
      ), Vietnam (
      • Hieu NT
      • Gainsborough M
      • Simpson JA
      • Thuy NT
      • Hang NN
      • Taylor AM
      • et al.
      Neurological status of low-risk Vietnamese newborns: a comparison with a British newborn cohort.
      ), and Uganda (
      • Hagmann CF
      • Chan D
      • Robertson NJ
      • Acolet D
      • Nyombi N
      • Nakakeeto M
      • et al.
      Neonatal neurological examination in well newborn term Ugandan infants.
      )], which have also reported differences from the original British norms.
      An important characteristic of the current study sample that should be noted is the mode of delivery. Overall, 64% of deliveries were by cesarean section. As discussed in our previous work, reasons for this could include the study setting (Korle Bu Teaching Hospital is a tertiary referral hospital), or it could be a reflection of higher maternal socioeconomic status (
      • Lawford HLS
      • Nuamah MA
      • Liley HG
      • Lee AC
      • Botchway F
      • Kumar S
      • et al.
      Gestational age-specific distribution of the Hammersmith Neonatal Neurological Examination scores among low-risk neonates in Ghana.
      ). It is important to note that when neonates were stratified by mode of delivery, there was no difference in total HNNE score between neonates delivered by cesarean section versus vaginally (25.3 ± 3.7 vs. 25.0 ± 3.7; P = 0.52). Therefore, it is unlikely that cesarean section or the use of postpartum analgesia impacted HNNE scores in this study. Nevertheless, we consider that any confounding or bias in HNNE results in the Ghanaian setting caused by unmeasured comorbidities, test conditions or conduct, or uncertainties in gestational age estimation should have applied equally to both arms of the current study, not just to the malaria-exposed group.
      In conclusion, given the high burden of malaria infection in pregnancy, understanding whether in utero exposure to malaria adversely impacts neurological development is important. The study results suggest that a group of term-born neonates exposed to malaria in pregnancy (and whose mothers had generally received treatment) had HNNE scores similar to an unexposed comparison group born in the same hospital. However, a higher risk of suboptimal functioning in only the reflexes subdomain was found, which could be a result of malaria exposure in pregnancy.

      Acknowledgements

      We acknowledge the contributions of Dr. Akomah Kennedy, Dr. Godwin A. Awuni, Dr. Newton E. Ofosu, Dr. Oyeronke S. Oyawoye, Dr. Temitope Akinyemi, Dr. Vida Akrasi-Boateng, and Mr. Hanson G. Nuamah at the University of Ghana for assistance with the data collection. Most importantly, we would like to express our sincere gratitude to the children and their families who participated in the IMPRINT study.

      Declarations

      Funding: This study was supported by a Mater Foundation Principal Research Fellowship to Dr. Samudragupta Bora (2016-21), a National Institutes of Health (NIH) Eunice Kennedy Shriver National Institute of Child Health and Human Development grant to Dr. Anne CC Lee (5K23HD091390), and The University of Queensland Research Training Program and Frank Clair scholarships to Dr. Harriet L.S. Lawford. The funding sources had no role in the writing of the manuscript or the decision to submit it for publication.
      Ethical approval: The study protocol was approved by the Institutional Review Board/Human Research Ethics Committee of the University of Ghana and The University of Queensland, Australia.
      Conflict of interest: The authors have no conflict of interest relevant to this study to disclose.

      Author contributions

      The corresponding author, Samudragupta Bora had full access to all of the study data and is primarily accountable for all aspects of the work, including the decision to submit for publication. The corresponding author, first author, and the statistical advisor, Ms. Alison Griffin, verified all of the reported data analysis.
      Harriet L.S. Lawford conceptualized and designed the study protocol, coordinated data acquisition, performed data analyses, interpreted the results, drafted and revised the initial manuscript, and approved the final manuscript as submitted.
      Mercy A. Nuamah designed the study protocol, coordinated and supervised data acquisition, interpreted the results, critically reviewed and revised the initial manuscript, and approved the final manuscript as submitted.
      Helen G. Liley conceptualized the study, supervised data analyses, interpreted the results, critically reviewed and revised the initial manuscript, and approved the final manuscript as submitted.
      Alison Griffin developed the statistical analysis plan, supervised preliminary data analyses, performed data analyses, interpreted the results, critically reviewed and revised the initial manuscript, and approved the final manuscript as submitted.
      Cecilia E. Lekpor designed the study protocol, acquired data, interpreted the results, critically reviewed and revised the initial manuscript, and approved the final manuscript as submitted.
      Felix Botchway designed the study protocol, acquired data, interpreted the results, critically reviewed and revised the initial manuscript, and approved the final manuscript as submitted.
      Samuel A. Oppong supervised the designing of the study protocol, coordinated data acquisition, interpreted the results, critically reviewed and revised the initial manuscript, and approved the final manuscript as submitted.
      Ali Samba supervised the designing of the study protocol, coordinated data acquisition, interpreted the results, critically reviewed and revised the initial manuscript, and approved the final manuscript as submitted.
      Ebenezer V. Badoe supervised the designing of the study protocol, coordinated data acquisition, interpreted the results, critically reviewed and revised the initial manuscript, and approved the final manuscript as submitted.
      Sailesh Kumar conceptualized the study, interpreted the results, critically reviewed and revised the initial manuscript, and approved the final manuscript as submitted.
      Anne CC Lee conceptualized the study, interpreted the results, critically reviewed and revised the initial manuscript, and approved the final manuscript as submitted.
      Richard K. Gyasi designed the study protocol, acquired data, interpreted the results, critically reviewed and revised the initial manuscript, and approved the final manuscript as submitted.
      Andrew A. Adjei supervised the designing of the study protocol, coordinated data acquisition, interpreted the results, critically reviewed and revised the initial manuscript, and approved the final manuscript as submitted.
      Samudragupta Bora acquired funds and resources, conceptualized the study, designed the study protocol, supervised data acquisition and data analyses, interpreted the results, critically reviewed and revised the initial manuscript, and approved the final manuscript as submitted.

      Appendix. Supplementary materials

      References

        • Adu-Bonsaffoh K
        • Ntumy MY
        • Obed SA
        • Seffah JD.
        Perinatal outcomes of hypertensive disorders in pregnancy at a tertiary hospital in Ghana.
        BMC pregnancy and childbirth. 2017; 17: 388
        • Arbeille P
        • Carles G
        • Bousquet F
        • Body G
        • Lansac J.
        Fetal cerebral and umbilical artery blood flow changes during pregnancy complicated by malaria.
        Journal of ultrasound in medicine. 1998; 17: 223-229
        • Carter JA
        • Neville BG
        • White S
        • Ross AJ
        • Otieno G
        • Mturi N
        • et al.
        Increased prevalence of epilepsy associated with severe falciparum malaria in children.
        Epilepsia. 2004; 45: 978-981
        • Conroy AL
        • Bangirana P
        • Muhindo MK
        • Kakuru A
        • Jagannathan P
        • Opoka RO
        • et al.
        Case report: birth outcome and neurodevelopment in placental malaria discordant twins.
        The American journal of tropical medicine and hygiene. 2019; 100: 552-555
        • Dombrowski JG
        • de Souza RM
        • Lima FA
        • Bandeira CL
        • Murillo O
        • de Sousa Costa D
        • et al.
        Association of malaria infection during pregnancy with head circumference of newborns in the Brazilian Amazon.
        JAMA network open. 2019; 2e193300
        • Doolan DL
        • Dobano C
        • Baird JK.
        Acquired immunity to malaria.
        Clinical microbiology reviews. 2009; 22: 13-36
        • Dubowitz L
        • Mercuri E
        • Dubowitz V.
        An optimality score for the neurologic examination of the term newborn.
        The journal of pediatrics. 1998; 133: 406-416
        • Dubowitz LM
        • Dubowitz V
        • Palmer PG
        • Miller G
        • Fawer CL
        • Levene MI.
        Correlation of neurologic assessment in the preterm newborn infant with outcome at 1 year.
        The journal of pediatrics. 1984; 105: 452-456
        • Eeles AL
        • Walsh JM
        • Olsen JE
        • Cuzzilla R
        • Thompson DK
        • Anderson PJ
        • et al.
        Continuum of neurobehaviour and its associations with brain MRI in infants born preterm.
        BMJ paediatrics open. 2017; 1e000136
        • George JM
        • Colditz PB
        • Chatfield MD
        • Fiori S
        • Pannek K
        • Fripp J
        • et al.
        Early clinical and MRI biomarkers of cognitive and motor outcomes in very preterm born infants.
        Pediatric research. 2021; https://doi.org/10.1038/s41390-021-01399-5
        • George JM
        • Fiori S
        • Fripp J
        • Pannek K
        • Guzzetta A
        • David M
        • et al.
        Relationship between very early brain structure and neuromotor, neurological and neurobehavioral function in infants born <31weeks gestational age.
        Early human development. 2018; 117: 74-82
        • Hagmann CF
        • Chan D
        • Robertson NJ
        • Acolet D
        • Nyombi N
        • Nakakeeto M
        • et al.
        Neonatal neurological examination in well newborn term Ugandan infants.
        Early human development. 2015; 91: 739-749
        • Halling C
        • Malone FD
        • Breathnach FM
        • Stewart MC
        • McAuliffe FM
        • Morrison JJ
        • et al.
        Neuro-developmental outcome of a large cohort of growth discordant twins.
        European journal of pediatrics. 2016; 175: 381-389
        • Hieu NT
        • Gainsborough M
        • Simpson JA
        • Thuy NT
        • Hang NN
        • Taylor AM
        • et al.
        Neurological status of low-risk Vietnamese newborns: a comparison with a British newborn cohort.
        Journal of health, population and nutrition. 2006; 24: 57-63
        • Idro R
        • Kakooza-Mwesige A
        • Balyejjussa S
        • Mirembe G
        • Mugasha C
        • Tugumisirize J
        • et al.
        Severe neurological sequelae and behaviour problems after cerebral malaria in Ugandan children.
        BMC research notes. 2010; 3: 104
        • Lawford HLS
        • Nuamah MA
        • Liley HG
        • Lee AC
        • Kumar S
        • Adjei AA
        • et al.
        Neonatal neurological examination in a resource-limited setting: what defines normal?.
        European journal of paediatric neurology. 2020; 29: 71-80
        • Lawford HLS
        • Lee AC
        • Kumar S
        • Liley HG
        • Bora S.
        Establishing a conceptual framework of the impact of placental malaria on infant neurodevelopment.
        International journal of infectious diseases. 2019; 84: 54-65
        • Lawford HLS
        • Nuamah MA
        • Liley HG
        • Lee AC
        • Botchway F
        • Kumar S
        • et al.
        Gestational age-specific distribution of the Hammersmith Neonatal Neurological Examination scores among low-risk neonates in Ghana.
        Early human development. 2020; 152105133
        • McDonald CR
        • Cahill LS
        • Ho KT
        • Yang J
        • Kim H
        • Silver KL
        • et al.
        Experimental malaria in pregnancy induces neurocognitive injury in uninfected offspring via a C5a-C5a receptor dependent pathway.
        PLoS pathogens. 2015; 11e1005140
        • McDonald CR
        • Elphinstone RE
        • Kain KC.
        The impact of placental malaria on neurodevelopment of exposed infants: a role for the complement system?.
        Trends in parasitology. 2013; 29: 213-219
        • McGready R
        • Simpson J
        • Panyavudhikrai S
        • Loo S
        • Mercuri E
        • Haataja L
        • et al.
        Neonatal neurological testing in resource-poor settings.
        Annals of tropical paediatrics. 2000; 20: 323-336
      1. World malaria report 2019. World Health Organization, 2019
      2. Ministry of Health & Ghana Health Services. Guidelines for Case Management of Malaria in Ghana. 3rd Edition ed 2014.

        • Molteno C
        • Grosz P
        • Wallace P
        • Jones M.
        Neurological examination of the preterm and full-term infant at risk for developmental disabilities using the Dubowitz Neurological Assessment.
        Early human development. 1995; 41: 167-176
        • Molteno CD
        • Thompson MC
        • Buccimazza SS
        • Magasiner V
        • Hann FM.
        Evaluation of the infant at risk for neurodevelopmental disability.
        South African medical journal. 1999; 89: 1084-1087
        • Rijken MJ
        • de Wit MC
        • Mulder EJH
        • Kiricharoen S
        • Karunkonkowit N
        • Paw T
        • et al.
        Effect of malaria in pregnancy on foetal cortical brain development: a longitudinal observational study.
        Malaria journal. 2012; 11: 222
        • Rogerson SJ.
        Management of malaria in pregnancy.
        Indian journal of medical research. 2017; 146: 328-333
        • Saito M
        • Briand V
        • Min AM
        • McGready R.
        Deleterious effects of malaria in pregnancy on the developing fetus: a review on prevention and treatment with antimalarial drugs.
        The lancet child & adolescent health. 2020; 4: 761-774
        • Sanchez K
        • Morgan AT
        • Slattery JM
        • Olsen JE
        • Lee KJ
        • Anderson PJ
        • et al.
        Neuropredictors of oromotor feeding impairment in 12 month-old children.
        Early human development. 2017; 111: 49-55
        • Setanen S
        • Lehtonen L
        • Parkkola R
        • Aho K
        • Haataja L
        • Group PS.
        Prediction of neuromotor outcome in infants born preterm at 11 years of age using volumetric neonatal magnetic resonance imaging and neurological examinations.
        Developmental medicine and child neurology. 2016; 58: 721-727
        • Tuhkanen H
        • Pajulo M
        • Jussila H
        • Ekholm E.
        Infants born to women with substance use: exploring early neurobehavior with the Dubowitz neurological examination.
        Early human development. 2019; 130: 51-56
        • Woodward LJ
        • Mogridge N
        • Wells SW
        • Inder TE.
        Can neurobehavioral examination predict the presence of cerebral injury in the very low birth weight infant?.
        Journal of developmental and behavioral pediatrics. 2004; 25: 326-334