Desfechos motores e cognitivos de recém-nascidos com baixo peso ao nascer no Brasil: uma revisão sistemática e metanálise

Radaelli, Graciane; Leal-Conceição, Eduardo; Kalil Neto, Felipe; Taurisano, Melissa Rogick Guzzi; Majolo, Fernanda; Bruzzo, Fernanda Thays Konat; Booij, Linda; Nunes, Magda Lahorgue

doi:10.1055/s-0042-1758866

Graciane Radaelli

conception, design, and interpretation of data

drafting the article

final approval of the version to be published

Pontifícia Universidade Católica do Rio Grande do Sul, Instituto do Cérebro do Rio Grande do Sul, Porto Alegre RS, Brazil

https://orcid.org/0000-0002-1208-362X

Eduardo Leal-Conceição

analysis and interpretation of data

final approval of the version to be published

Pontifícia Universidade Católica do Rio Grande do Sul, Instituto do Cérebro do Rio Grande do Sul, Porto Alegre RS, Brazil

https://orcid.org/0000-0001-9789-698X

Felipe Kalil Neto

article draft

final approval of the version to be published

Pontifícia Universidade Católica do Rio Grande do Sul, Instituto do Cérebro do Rio Grande do Sul, Porto Alegre RS, Brazil

https://orcid.org/0000-0002-7409-763X

Melissa Rogick Guzzi Taurisano

acquisition of data

final approval of the version to be published

Pontifícia Universidade Católica do Rio Grande do Sul, Instituto do Cérebro do Rio Grande do Sul, Porto Alegre RS, Brazil

https://orcid.org/0000-0002-6744-1500

Fernanda Majolo

interpretation of data

final approval of the version to be published

Universidade do Vale do Taquari, Programa de Pós-Graduação em Biotecnologia, Lajeado RS, Brazil

https://orcid.org/0000-0002-7955-078X

Fernanda Thays Konat Bruzzo

acquisition of data

final approval of the version to be published

Pontifícia Universidade Católica do Rio Grande do Sul, Instituto do Cérebro do Rio Grande do Sul, Porto Alegre RS, Brazil

https://orcid.org/0000-0003-2903-3618

Linda Booij

article draft and critical review for important intellectual content

final approval of the version to be published

Concordia University, Faculty of Arts and Science, Department of Psychology, Montreal QC, Canada

https://orcid.org/0000-0002-0863-8098

Magda Lahorgue Nunes Address for correspondence Magda Lahorgue Nunes (email: mlnunes@pucrs.br).

article draft and critical review for important intellectual content

final approval of the version to be published

Pontifícia Universidade Católica do Rio Grande do Sul, Escola de Medicina e InsCer, Disciplina de Neurologia, Porto Alegre RS, Brazil

https://orcid.org/0000-0002-3402-6810

Address for correspondence Magda Lahorgue Nunes (email: mlnunes@pucrs.br).

Conflict of Interest

The authors have no conflict of interests to declare.

Author	Year of publication	N	Birth weight (grams)	Birth weight of the control group (grams)	Gestational age (weeks)	Age at thr assessment (months)	Proportion of girls	Developmental outcomes
Mello et al.²⁹	1998	70	1,185	-	32.2	21	0.6	Cognitive and motor
Grantham-McGregor et al.²¹	1998	262	2,338	3,210	37	6 to 12	0.6	Cognitive and motor
Mello et al.³⁰	1999	83	1,176	-	32	12 to 30	-	Cognitive and motor
Eickman et al.¹⁵	2002	152	2,332	3,255	37	24	0.6	Cognitive and motor
Méio et al.²⁷	2003	94	1,903	-	32.04	48 to 60	0.6	Cognitive and motor
Meio et al.²⁸	2004	129	1,220	-	32	0 to 36	0.6	Cognitive and motor
Emond et al.¹⁶	2006	164	2,346	3,212	39	96	-	Cognitive and motor
Schirmer et al.³⁵	2006	69	1,552	-	32.06	36	0.5	Only motor
Carvalho et al.¹⁴	2008	36	1,058	-	30.44	12	0.5	Cognitive and motor
Manacero and Nunes²⁶	2008	88	1,753	-	32.8	4 to 8	0.3	Cognitive and motor
Espírito Santo et al.¹⁷	2009	80	1,788	-	32.3	48 to 60	0.5	Only motor
Mello et al.³¹	2009	100	1,126	-	29.6	12	0.5	Cognitive and motor
Bühler et al.¹³	2009	32	1,073	3,291	29.4	24	0.4	Only motor
Magalhães et al.²⁵	2009	70	1,171	3,215	30.23	84	0.6	Only motor
Procianoy et al.³³	2009	131	1,190	-	30.5	24	0.5	Only motor
Silva et al.³⁶	2011	69	1,236	-	31	24	0.4	Only cognitive
Oliveira et al.³²	2011	46	1,201	3,273	30	60 to 72	0.6	Only cognitive
Guimarães et al.²³	2011	92	1,424	3,158	28	36	0.5	Cognitive and motor
Reis et al.³⁴	2012	109	1,122	-	29	24	0.5	Cognitive and motor
Fernandes et al.¹⁹	2012	58	1,172	-	30	18 to 2	0.5	Cognitive and motor
Lemos et al.²⁴	2012	98	1,439	-	31	2 to 7	0.5	Cognitive and motor
Fan et al.¹⁸	2013	97	1,890	-	33,6	6 to 7	0.5	Cognitive and motor
Guerra et al.²²	2014	100	1,744	-	33.2	18 to 24	0.5	Cognitive and motor
Saccani et al.³⁷	2017	42	1,886	2,878	36	0 to 12	-	Only motor
Fuentefria et al.²⁰	2018	135	1,098	3,348	29.1	8 to 18	0.5	Cognitive and motor
Mean (±standard deviation)			1,497 (±427)	3,204 (±134)	32.73 (±3.5)
p < 0.0001

Study	Design	Age N Sex	BW(g), GA	Domains	Measures	Data analysis	Results
Bühler et al.¹³	Cohort (incl control group)	2 years N = 32 n₁ = 12 PT n₂ = 20 FT Sex 18 boys 14 girls	≤1500 Mean BW: 1073 GA: 29.4 Vs ≥2500 BW:3291 GA: 39.1	Motor and cognitive	PELCDO	Analysis of variance, (Mann-Whitney test), Kruskal-Wallis test Spearman correlations tests	VLBW preterm infants displayed poorer cognitive development scores than term infants starting at 6 months of age. Expressive language among VLBW preterm infants was also delayed.
Carvalho et al.¹⁴	Cohort (no control group)	N = 36 Mothers and their children	GA < 37 BW < 2500	Motor and cognitive	Bayley-II Scale	The data from the instruments was analyzed according to their respective standards	The majority of the infants exhibited normal development on Bayley-II at 12 months CCA; however, 25% of the infants displayed cognitive problems and 40% motor problems.
Eickman et al.¹⁵	Cross-sectional (incl control group)	2 years N = 152 64 boys 88 girls n₁ = 76 LBW n₂ = 76 ABW	1500-2499 BW: 2332 ± 160 Vs. 3000-3499 BW: 2332 ± 160	Motor and cognitive	Bayley Scale HOME	Chi-squared, Analysis of variance (T-test), Multiple linear regression.	LBW term infants showed poorer motor development (95.1 ± 19.2, p < .001) than ABW term infants (1 05.3 ± 13.6) after controlling for SES factors. Mental development followed the same pattern (89.8 ± 15.8 for LBW; 98.9 ± 14.8, p < .001 ). BW was a risk factor for impaired motor and mental development (p = .010; p = .005). Socio-economic level and home stimulation were predictors of both motor and cognitive development (p <.01).
Emond et al.¹⁵	Cohort (incl control group)	8 years N = 164 n₁ = 83 LBW-T n₂ = 81 ABW-T Sex unknown	1500-2499. BW: 2346 ± 148 GA: 39 ± 1.4; Vs. 3000-3499 BW: 3212 ± 147 GA:40 ± 1.3 AII GA > 37	Motor and cognitive	WISC III WISC TEAch Bayley Scale M-ABC Strengths and Difficulties questionnaires Educational assessments HOME	Comparison of means, Chisquared, Analysis of variance (T-test, Mann-Whitney test), Spearman correlations tests, Multiple linear regression.	LBW term infants were more vulnerable to have motor and cognitive impairments. Home factors (e.g. stimulation) as significant predictors on motor and cognitive scores. Low developmental levels among children with least-educated mothers. Growth in head size was strongly associated with IQ as a better predictor of cognitive outcomes than BW.
Espírito Santo et al.¹⁷	Cross-sectional (no control group)	4-5 years N = 80 40 boys 40 girls	<2500+ Born preterm Mean BW: 1788 ± 502 GA: 32.3 ± 3 Subdivision <1000 vs. <1500 vs. 1500–2500	Motor and cognitive	WPPSI, CPRS-R Denver Test Bayley Scale Neurological examination	Chi-square Analysis of variance	Significant positive correlations between Bayley cognitive scale (p <.05) the Denver test (p < .01 ), and ADHD symptoms. Lower IQ was associated with the Bayley cognitive scale (p < .001 ) the Denver test (p < .001 ), and neurobiological examination (p < .01).
Fan et al.¹⁸	Cross-sectional (no control group)	6-7 years N = 97 49 girls 48 boys	GA < 37 BW < 1500	Motor and cognitive	WISC-III Bayley-II Scale Denver-II Test	Z-test to compare the mean values obtained in the sample with the WISC-III reference.	The results of the Denver and Bayley tests were associated with the cognitive performance (p,0.001).
Fernandes et al.¹⁹	Cross-sectional (no control group)	18-24 months N = 58 30 girls 28 boys	GA < 37 VLBW (< 1500)	Motor and cognitive	Bayley-III Scale	Numerical variables were compared by Mann-Whitney or Student t test and categorical variables by chi-square or Fisher's exact test. Factors associated with developmental scores were analyzed by linear regression, and statistical significance level was established at p < 0.05.	Out of the 58 children included, four presented cognitive delay, four motor, 17 language, 16 social-emotional and 22 adaptive-behavior delay. The female sex was associated with higher motor, language and social-emotional developmental scores.
Fuentefria et al.²⁰	Cohort (incl control group)	N= 135 Preterm infants (N = 83) Control group (N = 52)	Preterm infants GA: 29.1 BW:1098 Control group GA: 39.1 BW: 3348	Motor	AIMS BLS	SPSS, version 21.0. To compare the means between the groups, the Student's t-test was applied. For the control of confounding factors, the analysis of covariance (logistic regression) was utilized. In the comparison of proportions, Pearson's chi-square test was used. The associations between numerical variables in each group were evaluated by Pearson's correlation coefficient.	At 8 months corrected age, preterm infants scored significantly lower in total AIMS score (p V4 0.001). At 18 months, they scored significantly lower on the stand subscale from AIMS (p V4 0.040) and exhibited poor psychomotor development in the BLS (p V4 0.006). The nutritional status showed significant differences between the groups, in both age groups (p < 0.001 ). There were positive correlations between nutritional status and AIMS (r 1/4 0.420; p < 0.001 ) and BLS (r 1/4 0.456; p < 0.001 ) at 8 months, and between head circumference and BLS (r 1/4 0.235; p < 0.05) at 8 months and AIMS (r 1/4 0.258; p< 0.05) at 18 months. Conclusion Very low BW preterm infants at 8 and 18 months corrected age showed significant differences in the neurodevelopment and growth pattern when compared with their full-term peers.
Grantham-McGregor et al.²¹	Cohort (incl control group)	6 and 12 months N = 262 Sex 157 girls 105 boys n₁ = 131 LBW n₂ = 131 ABW	<2500 Mean BW: 2338 ± 152 vs. ≥2500 Mean BW: 3210 ± 142	Motor and cognitive	Bayley Scale HOME	Chi-squared, Analysis of variance (T-test), Multiple linear regression	LBW term infants showed poorer motor and cognitive development than term infants. LBW term infants with literate mothers had higher scores than those with illiterate mothers (p = .003). Home factors as significant predictors on motor and cognitive scores.
Guerra et al.²²	Cross-sectional (no control group)	With delay (N = 55) Without delay (N = 45)	With delay GA: 33.2 BW: 1727 Without delay GA: 33.3 BW: 1762	Motor and cognitive	Bayley-III Scale	The numerical variables were compared by the Mann-Whitney-U test or t-test and the categorical variables by the chi-square or Fisher's exact-test.	The percentages and 95% confidence intervals of those children with developmental delays were as follows: cognitive (2.0%; 0.6-7.0%), language (5.0%;2.2-11.2%), motor (3.0%; 1.0-8.5%), socio-emotional (13.0%; 7.8-20.1%), general adaptive (26.0%; 18.4-35.4%), conceptual (1 7.0%; 10.9-25.6%), social (46.0%; 36.6-55.7%) and practical (21.0%; 14.2-30.0%). Factors associated with delay in at least one developmental domain were as follows: caesarean delivery, low per capita income and peri-intraventricular haemorrhage. Factors associated with a reduction in developmental scores were as follows: non-white ethnicity, lower social class, caesarean delivery, male gender, peri-intraventricular haemorrhage, mechanical ventilation and length of hospitalisation.
Guimarães et al.²³	Cross-sectional (incl control group)	3 years N=92 n₁ = 46 PT n₂= 46 FT Sex: 46 girls 46 boys	PT GA:28-33 31.1 ± 1.5 BW: 1424 ± 321.1 Vs. FT GA: 38-40 38.6 ± 0.5 BW:3158 ± 565.4	Motor	TIMP	Homogeneity of variance between groups, Analysis of variance (T-test, Mann-Whitney), Spearman and Pearson correlations, F-test.	PT infants showed poorer motor development (58 ± 7.9; p < .001 ) than FT infants (67.9 ± 5.3) after controlling for SES factors. 100% of PT were showed atypical motor development (p < .001 ) whereas 100% of FT infants showed normal scores. Prematurity was associated with impaired motor development.
Lemos et al.²⁴	Cross-sectional (no control group)	Preschool (2-7 years) N = 98	GA < 37 BW < 2500	Motor and cognitive	PEDI	Chi-square and the variance analysis.	There was found a delay of 10,2%, 12,2% and 14,3% in the functional abilities in the areas of self-care, mobility and social function, respectively, and of 11,2%, 19,4% and 15,3% in the assistance level received from the caregivers (independence), in the same areas. It was not found statistically significant differences or associations between groups of different degrees of prematurity or birth weight and the PEDI performance.
Magalhães et al.²⁵	Cross-sectional (incl control group)	7 years	1.GA</=34 BW</=1500 2. GA > 37	Motor	M-ABC test	Wilcoxon's test	Significant difference between the two groups in terms of the overall score (Z = -4,866, p < 0,001 ) and the score for specific sub-sections of the M-ABC, the preterm group performing less well than the term group.
Manacero and Nunes²⁶	Cross-sectional (no control group)	<1 year N = 44 21 boys 23 girls n₁ = 14 < 1750 n₂ = 30 ≥ 1750	<2500 Subdivision <1750 BW: 1417 ± 292 GA:32.4 ± 0.7 vs. ≥1750 BW: 2090 ± 278 GA:33.2 ± 0.8	Motor Motor	AIMS Neurological examination Denver Test	Analysis of variance (ANOVA)	Despite BW differences, motor ability acquisition exhibited a normal progress at 40 weeks, 4 months, and 8 months according to mean percentile of normality on the AIMS (43.2% to 45.7%).
Meio et al.²⁷	Cohort (no control group)	N = 94 (64,6% girls)	GA < 37 BW < 1500	Cognitive	WPPSI-R	T-student Chi-squared with Yates correction	Seventy-nine children aged 4 and 5 years were studied. The mean full WIPPSI-R score was 75.6 (±11.9). The incidence of abnormal 1 and 2 SD full score was 77.2% and 32.9%, respectively. After adjusting for the method of delivery, small for gestational age (OR = 6.19, 95% CI 1.60-23.86), abnormal cerebral ultrasound exam (OR = 5.90, 95% CI 1.04-9.83) and male sex (OR = 3.20, 95% CI 1.32-26.35) were predictors of full score <70.
Meio et al.²⁸	Cohort (no control group)	N = 79/129 Age pre-school	GA < 37 BW < 1500	Cognitive	WPPSI-R	Epi-Info 6.0 SPSS 6.1 Chi-square T-student Fischer's test	No significant statistical difference was found between the groups (study and loss). Children who entered this study showed to have a borderline intellectual functioning at the moment of the evaluation. Results indicate they may face learning difficulties at school, thus requiring adequate stimuli that should be provided by the family and the school.
Mello et al.²⁹	Cohort (no control group)	N = 70; Mean GA:32.2 weeks	GA < 37 BW < 1500	Motor and cognitive	Dubowitz & Dubowitz method and brain US	Analysis of variance through the F-test; the differences between the proportions were tested by chi-square. The statistical significance, prevalence, sensitivity, specificities, predictive values and confidence intervals were calculated.	25.7% of the children had neuromotor impairment, and 20.3% had cognitive impairment. Neonatal neurological examination was more sensitive than neuromotor change (sensitivity: 77.7%, specificity: 57.6%), and cognitive (sensitivity: 78.5%, specificity: 56.4%). Low predictive value for neuromotor change (38.9%) and cognitive (31.4%). Ultrasonography was discharged specificity for neuromotor (92.3%) and cognitive development (89.1%). The predictive value of ultrasonography was high for neuromotor abnormalities (69.2%) and low for changes cognitive (50.0%).
Mello et al.³⁰	Cohort (no control group)	12-30 months N = 83	GA < 37 BW < 1500	Motor and cognitive	Bayley Scale	Comparison of means (Chisquared, F-test.)	Cerebral ultrasonography (US) was normal in 68 babies (81.9%) and abnormal in 15 (18.8%). With a mean age of 21 months, 63 children (75.9%) had normal motor development and 20 (24.0%) had motor abnormalities. The cognitive development was normal in 68 children (81.9%). The negative predictive value of the cerebral US for motor development was 85.3%, and for cognitive development, 86.8%. The positive predictive value of the cerebral US for motor development was 66.7% and for cognitive development, 42.9%. The probability for children with normal neonatal ultrasonography to have normal motor and cognitive development is greater than 85%.
Mello et al.³¹	Cohort (no control group)	N = 100	BW: 1126 GA: 29.6	Motor and cognitive	Bayley Scale	A multivariate logistic regression model was constructed. Neonatal variables and neuromotor abnormalities up to 6 months of corrected age were selected by bivariate analysis.	Mean birth weight was 1126g (SD: 240). Abnormal neuromotor development was presented in 60 children at 12 months corrected age.
Oliveira et al.³²	Case-control (incl control group)	5-6 years N = 46 n¹ = 23 ≤ 1500 9 boys 14 girls n₂ = 23 ≥ 2500 Sex unknown	≤ 1500 Mean BW: 1201 ± 177 GA: 30 ± 2 Vs. ≥2500 BW:3273 ± 348 GA:39 ± 0.4	Motor and cognitive	M-ABC WISC DCDQ HOME	Comparisonof means (Mann-Whitney test), Spearman correlations tests	LBW children were more vulnerable to have motor and cognitive impairments. Home factors as significant predictors on motor and cognitive scores.
Procianoy et al.³³	Cohort (incl control grup)	2 years N = 96 49 boys 47 girls n₁ = 55 SGA n₂= 41 AGE	SGA <1500 CA: 31.7 ± 2 BW:1130 ± 250 vs. AGE <1500 GA: 29.3 ± 1.6 BW:1250 ± 218	Motor and cognitive	Bayley Scale	Comparison of means (Chisquared, F-test), ANOVA, ANCOVA	Mental and motor development were similar between both groups of term infants at 8, 12, 18, and 24 months corrected age. Both groups had similar neurodevelopment outcome.
Reis et al.³⁴	Cohort (no control group)	N = 109 < 2 years (6 m, 12 m, 18–24 m)	GA < 37 BW < 1500	Motor and cognitive	Bayley-II Scale MDI	The stability of the scores between assessments was verified by the analysis of variance for repeated measures. The association of the major social and neonatal characteristics with mental development was confirmed using multivariate analysis by linear regression,	The association of the major social and neonatal characteristics with mental development was confirmed using multivariate analysis by linear regression, considering the following outcomes: mental development indices at 6 months, 12 months and between 18-24 months of corrected age. The cognitive development index did not show stability during the first two years, except for children with neonatal pneumonia.
Schirmer et al.³⁵	Cross-sectional (no control group)	3 years N = 69 n₁ = 30 <1500 n₂ = 39 1500–2500 Sex unknown	≤2500 Subdivision <1500 vs. 1500–2500 AII GA > 37	Motor and cognitive	Denver-II Test Bayley-II Scale Language Assessment	Chi-squared, Analysis of variance (T-test), multivariate regression, Risks estimates (OR, Wald)	GA correlates with language development; GA predicts language acquisition. Gestational age < 32 weeks increases 3 times the risk for delay in language acquisition.
Silva et al.³⁶	Cross-sectional (no control group)	4–24 months N = 69	GA < 37 BW < 1500	Cognitive Hand-eye coordination Language Posture Sociability	Brunet Scale Lèzine's Scale	Data were analyzed using descriptive and inferential statistics.	85% of scores within the normal range in the third assessment. The specific areas of hand-eye coordination and language had the worst initial results, while posture had the best scores. Correlation was found between birth weight and posture, language and social areas at the first assessment and between birth weight and social and hand-eye coordination at the third assessment.
Saccani et al.³⁷	Cross-sectional (incl control group)	0–12 months N = 42 n1 = 21 n2 = 21	<2500 Mean BW: 1886 ± 402 GA: 36 Vs ≥2500 BW:2878 ± 348 GA: 36	Motor	Alberta Infant Motor Scale	The independent t-test, the chi-square test of Pearson and the Eta2 test (strong association> 0.60).	Fifteen (71.42%) children with low birth weight were classified as small for gestational age. The mean motor development score percentile was 17.90 ± 17.74 for the LBW group and 34.57 ± 25.80 for the ABW group, indicating a better motor development of the second group (p = 0.02). There was a greater number of children with developmental delay in the LBW group (52.4%), whereas in the ABW group most were within the normal range (47.6%).

No.	Author	Obtained score/ maximum score	Relative frequency (%)
01	Bühler et al.¹³	19/22^b	86
02	Carvalho et al.¹⁴	21/22^b	95
03	Eickman et al.¹⁵	21/22^a	95
04	Emond et al.¹⁶	22/22^b	100
05	Espírito Santo et al.¹⁷	21/22^a	95
06	Fan et al.¹⁸	20/22^a	91
07	Fernandes et al.¹⁹	20/22^a	91
08	Fuentefria et al.²⁰	20/22^b	91
09	Grantham-McGregor et al.²¹	19/22^b	86
10	Guerra et al.²²	20/22^a	91
11	Guimarães et al.²³	21/22^a	95
12	Lemos et al.²⁴	20/22^a	91
13	Magalhães et al.²⁵	20/22^a	91
14	Manacero and Nunes²⁶	21/22^a	95
15	Méio et al.²⁷	20/22^b	91
16	Meio et al.²⁸	20/22^b	91
17	Mello et al.²⁹	22/22^b	100
18	Mello et al.³⁰	21/22^b	95
19	Mello et al.³¹	20/22^b	91
20	Oliveira et al.³²	22/22^c	100
21	Procianoy et al.³³	22/22^b	95
22	Reis et al.³⁴	22/22^b	100
23	Schirmer et al.³⁵	20/22^a	91
24	Silva et al.³⁶	20/22^a	91
25	Saccani et al.³⁷	20/22^a	91

[TFN1] Notes: *Gestational age at delivery is presented as mean ± standard deviation values; p < 0.001 (significant); birthweight in grams of also in weeks.

Desfechos motores e cognitivos de recém-nascidos com baixo peso ao nascer no Brasil: uma revisão sistemática e metanálise

Resumo