The Association Between Feeding Mode, Growth and Developmental Outcomes in Infants with Complex Congenital Heart Disease at 6 and 12 Months of Age

Abstract

Objective

To assess the association between early anthropometric measurements, device assisted feeding and early neurodevelopment in infants with complex congenital heart defects (CHD). Study design Bayley Scales of Infant Development II, were used to assess cognitive and motor skills in 72 infants with CHD at 6 and 12 months of age. Linear regression models were used to assess the association between mode of feeding and anthropometric measurements with neurodevelopment at 6 and 12 months of age.

Results

Of the 72 infants enrolled in the study, 34 (47%) had single ventricle physiology. The mean Mental Developmental Index (MDI) and Psychomotor Developmental Index (PDI) scores at 6 months of age were 92 ± 10 and 81 ± 14, respectively. At 12 months of age the mean MDI and PDI scores were 94 ± 12 and 80 ±16, respectively. Lower length-for-age z-score (p<0.01) and head circumference-for-age z-score (p<0.05) were independently associated with lower MDI at 6 months, and both increased hospital length of stay (p<0.01) and lower length-for-age z-score (p=0.04) were independently associated with lower MDI at 12 months. Device assisted feeding at 3 months (p=0.04) and lower length-for-age z-score (p<0.05) were independently associated with lower PDI at 6 months. Both lower weight-for-age z-score (p=0.04) and lower length-forage z-score (p=0.04) were independently associated with PDI at 12 months.

Conclusion

Neonates with complex CHD who required device assisted feeding and those with lower weight, length and head circumference z scores at 3 months were at risk for neurodevelopmental delay at 6 and 12 months of age.

Congenital heart disease (CHD) is the most common congenital defect in neonates(1), affecting some 40,000 births each year. Advances in surgical, medical and nursing care for infants with complex CHD have resulted in an increase in survival and a growing population of infants with morbidities related to their underlying defect, surgical intervention and/or residual anatomic or hemodynamic abnormalities. Current literature suggests infants with complex CHD are at increased risk for worse neurodevelopmental outcomes in both cognitive and motor domains in late infancy and early childhood(2, 3). Various factors influencing neurodevelopmental outcomes in infants with complex CHD have been identified; these include genetic abnormalities and syndromes, prematurity, post-surgical physiology, length of hospital stay (LOS), history of cardiac arrest, utilization of extracorporeal membrane oxygenation or ventricular assist device, post-operative stroke, seizures or abnormalities on neuroimaging(4–6). The etiology of abnormal cognitive and motor neurodevelopment in these infants appears to be multifactorial and is not well understood(7). Most recently, there is new interest in nutritional factors such as feeding mode or growth status as predictors of neurodevelopmental outcomes (4, 8).

There are several reports of growth failure and feeding dysfunction in children with complex CHD (9–11). Early sucking and swallowing difficulties have been shown to be a significant predictor of worse neurodevelopmental outcomes in premature infants (12) and in children with brain injury(13). Although many infants with CHD are born full term, evidence suggests they may be neurologically immature and suffer neurologic insult early in life(14). We previously reported an association between poor oral feeding skills and growth failure at hospital discharge(15) in children with CHD. Despite a history of early growth faltering, most infants with complex CHD demonstrate adequate catch-up growth during the first year of life and few require supplemental tube feedings after 3 months of age(16). Given the importance of nutrition in assuring optimal brain growth during the first months of life, we hypothesized that poor early growth and device assisted feeding will be associated with worse neurodevelopmental outcomes at 6 and 12 months of age.

Methods

This study was a prospective cohort study of growth and development in the first year of life in infants who had undergone neonatal cardiac surgery during the first month of life. The Institutional Review Board at the Children’s Hospital of Philadelphia (CHOP) approved the project. Informed consent was obtained from a parent or legal guardian prior to enrollment of each infant. Neonates with CHD who underwent cardiac surgical intervention within 30 days of birth were screened for enrollment. Neonates were eligible if they had a post-menstrual age > 36 weeks at birth, and were discharged to home after surgical intervention.

Neonates with multiple congenital, facial and/or complex gastrointestinal anomalies, chromosomal abnormalities, and/or congenital or acquired neurological insult were excluded because these factors are known to be associated with poor growth.

Data collection was performed prospectively throughout the hospitalization, and included: intraoperative support times, total hours of ventilation, history of infections, length of stay (LOS) and mode of feeding at discharge. Research procedures included anthropometric measurements at 3, 6, 9 and 12 months completed by research staff in the Growth and Nutrition Laboratory of the Clinical Translational Research Center (CTRC) at CHOP. Feeding mode (bottle or breast only, bottle or breast combined with tube assisted feeding with nasogastric tube or gastrostomy tube, and nasogastric tube or gastrostomy tube only) was recorded at discharge and at 3 months of age. Neurodevelopment was assessed at 6 and 12 months using the Bayley Scales of Infant Development- II (BSID-II) in the Behavioral Neuroscience Core of the CTRC by research psychometricians under the supervision of a licensed psychologist. The BSID-II yields two scores: the Psychomotor Developmental Index (PDI) which assesses gross motor and fine motor skills; and the Mental Developmental Index (MDI), which evaluates cognitive, memory, problem solving, generalization, vocalizations and social skills. The mean MDI and PDI score for the normal population is 100, with a standard deviation of 15(17).

Statistical Analyses

Measures of central tendency and variation (means and standard deviations, medians, and ranges) for continuous variables, as well as frequencies and percentages for categorical variables, were used to describe the study sample. Bivariate general linear regression models were created to assess the association between predictors [feeding mode at discharge and 3 months of age; 3-month anthropometric measurements (weight, length, head circumference)] and developmental outcomes (MDI and PDI at 6 and 12 months of age, separately). Multivariable models were generated for each outcome, separately, and included variables significant at the 0.05 level in bivariate models, along with post-operative physiology classified as Single Ventricle (SV) or Biventricular (2V), and LOS. To avoid multi-collinearity between feeding mode at discharge and 3 months, as well as between the three growth measures observed at 3 months, individual models were generated for the different combinations of these predictors. Finally, neurodevelopmental outcomes were described and compared by physiology group (SV vs. 2V) and feeding mode (oral vs. device assisted) using two-sample t-tests. Significance was set at 0.05 for all analyses. A post hoc analysis based upon our group of 72 participants achieves 80% power to detect an R² increase of 7% attributed to a single predictor of interest with a 0.05 level of significance, adjusting for 3 additional covariates that demonstrate a total R² of 0.30. The statistical analysis section was revised. All analyses were performed by the STATA software package, V.11.0 (STATA Corp., College Station, Texas, USA).

Results

The sample included 72 subjects with complex CHD, of whom 34 (47%) had SV physiology (Table I; available at www.jpeds.com). The mean gestational age for all subjects was 39±1 week, with a mean birth weight of 3406 ± 512 grams. Males comprised 69% of the study sample. Diagnoses included hypoplastic left heart syndrome in 24%, transposition of great arteries in 24%, coarctation of the aorta in 10%, double inlet left ventricle in 7%, tetralogy of Fallot in 7%, double outlet right ventricle in 6%, and tricuspid atresia in 6%. The median LOS was 14.5 (2 to 159) days. Thirty-eight (53%) subjects were exclusively orally fed at hospital discharge, whereas 34 (47%) required device-assisted feeding (i.e. nasogastric tube, gastric tube) combined with oral feeding. At 3 months of age, ten (14%) subjects continued to require device-assisted feeding.

TABLE 1.

(on line): Infants’ demographic and clinical characteristics, N=72.

Categorical variables	Frequency (%)
Infant sex
Male	50 (69)
Female	22 (31)
Ethnicity
Hispanic	7 (10)
Non-Hispanic	50 (69)
Unknown	15 (21)
Race
White	66 (92)
Native American	3 (4)
Black	2 (3)
Asian	1 (1)
Post-op cardiac physiology
Single ventricle	34(47)
Bi-ventricle	38(53)
Feeding mode at discharge
Oral	38 (53)
Tube assisted	34 (47)
Feeding mode at 3 months
Oral	62 (86)
Tube assisted	10 (14)
Continuous variables	Mean (SD*)	Median (Range)	IQR
Birth weight, grams	3406 (512)	3494 (2200–4910)	721
Gestational age, weeks (N=71)	39.0 (1.3)	39 (35–42)	2
Weight at 3 mo, z-score** (N=63)	−1.24(1.28)	−1.3 (−5.1 0.84)	1.92
Length at 3 mo, z-score (N=61)	−.89 (1.35)	−0.94 (−5.45 1.53)	1.72
Head circumference at 3 mo, z-score (N=59)	−.69 (1.26)	−0.58 (−3.99 2.46)	1.67
Hospital length of stay, days	22.67(25.37)	14.5 (2–159)	15
MDI Ϯ at 6 mo (n=50)	92 (10)	92 (63– 111)	12
PDI Ϯ at 6 mo (n=51)	81 (14)	82 (50–111)	22
MDI at 12 months	94(12)	92 (72–117)	18
PDI at 12 months	80(16)	77.5 (50–113)	21

Note.

^*Standard deviation,

^**Z-scores as are defined by WHO standardized growth scores,

^ϮBayley Mental Development Index scores,

^Ϯ Bayley Psychomotor Development Index scores.

At 3 months, mean weight-, length-, and head circumference-for-age z-scores were −1.24 ± 1.28, −0.89 ± 1.35 and −0.69 ± 1.26, respectively, on the World Health Organization standard growth chart. Thirty percent of the subjects had a weight-for-age z-score of ≤ −2.0. The mean MDI and PDI scores at 6 months of age were 92 ± 10 and 81 ± 14, respectively. At 12 months of age, the mean MDI and PDI scores were 94 ± 12 and 80 ± 16), respectively (Table I).

Table II depicts the effect of post-operative physiology, feeding mode at hospital discharge and 3 months of age on the neurodevelopmental outcomes of interest. Comparing the SV and 2V groups revealed no differences in MDI or PDI scores at 6 months of age. At 12 months of age, subjects with SV physiology demonstrated lower MDI scores (91.2 ± 10.3 vs. 97.1,± 12.6, p=0.04) along with lower PDI scores (73.9 ± 15.8 vs. 84.6,± 15.3, p<0.01) when compared with those with 2V physiology.

TABLE 2.

Descriptive statistics and comparisons for neurodevelopment by physiology, and feeding mode at discharge and 3 months

	MDI at 6 mo	MDI at 12 mo	PDI at 6 mo	PDI at 12 mo
	Mean (SD)	Mean (SD)	Mean (SD)	Mean (SD)
Physiology
Single ventricle	90.53 (8.78)	91.18 (10.33)	76.58 (14.77)	73.94 (15.78)
Bi-ventricle	92.81 (10.56)	97.05 (12.60)	83.97 (13.47)	84.58 (15.33)
Sub-sample P value	0.435	0.035	0.074	0.005
Feeding mode at discharge
Orally fed	94.62 (10.25)	96.16 (12.68)	85.70 (11.58)	80.47 (15.36)
Tube assisted	89.04 (8.79)	92.18 (10.70)	76.17 (15.54)	78.53 (17.54)
Sub-sample P value	0.045	0.157	0.016	0.618
Feeding mode at 3 months
Orally fed	93.18 (9.14)	94.98 (12.15)	83.43 (12.85)	81.84 (15.36)
Tube assisted	80.8 (10.33)	89.9 (9.35)	60.8 (10.71)	65.4 (15.75)
Sub-sample P value	0.007	0.211	<0.001	0.003

p-values from two samples t-test

When comparing subjects by mode of feeding, infants fed orally compared with infants fed by devise at hospital discharge had significantly higher MDI (94.6 ±10.3 vs. 89.0 ± 8.8, p=0.045) and PDI (85.7,± 11.6 vs. 76.2,±15.5, p=0.016) scores at 6 months of age. Similarly, infants fed orally at 3 months of age compared with infants fed by device at 3 months of age had significantly higher MDI (93.2,± 9.1 vs. 80.8 ± 10.3, p=0.007) and PDI (83.4,±12.9 vs. 60.8 ± 10.7, p<0.001) scores at 6 months of age. Additionally, infants fed orally at 3 months compared with infants infants fed by device at 3 months had a significantly higher PDI (81.8 ± 15.4 vs. 65.4 ± 15.8, p=0.003) score at 12 months of age.

Table III summarizes bivariate regression model results for MDI and PDI at 6 and 12 months. The multivariate regression models for both MDI and PDI at 6 and 12 months of age are presented in Tables IV and V, respectively. When assessing 6-month MDI models with mode of feeding at discharge, lower length-for-age z-score (p<0.01) and lower head circumference-forage z-score (p<0.05) at 3 months of age were independently associated with lower MDI scores at 6 months of age. Similarly, length-for-age z-score (p<0.01) at 3 months of age was associated with lower MDI scores at 6 months of age when including feeding mode at 3 months of age. When assessing 12- month MDI models with mode of feeding at discharge, longer hospital LOS (p<0.01) was independently associated with lower MDI scores at 12 months for all 3 anthropometric outcomes models (weight, length, head-circumference, and mode of feeding at 3 months). The total variance for MDI ranged between 26 to 39% at 6 months, and between 26 to 32% at 12 months.

Table 3.

Bivariate Regression Models for MDI and PDI^*

	MDI at 6 months				MDI at 12 months				PDI at 6 months				PDI at 12 months
variables	β	SE Ϯ	95% CI**	P	β	SE	95% CI	P	β	SE	95% CI	P	β	SE	95% CI	P
Cardiac physiology	2.28	2.89	(−3.54 , 8.10)	0.435	5.88	2.73	(.42 , 11.33)	0.035	7.39	4.04	(−.74 ,15.51)	0.074	10.64	3.67	(3.32 , 17.96)	0.005
Length of staya	−4.36	2.02	(−8.42 , −.31)	0.036	−6.02	1.74	(−9.50 , −2.55)	0.001	−6.31	2.91	(−12.16 , −.46)	0.035	−7.40	2.44	(−12.27 , −2.54)	0.003
Feeding mode at discharge	−5.57	2.71	(−11.03 , −.12)	0.045	−3.98	2.78	(−9.53 , 1.57)	0.157	−9.54	3.81	(−17.20 , -1.88)	0.016	−1.94	3.88	(−9.68 , 5.79)	0.618
Feeding mode at 3 mo	−12.38	4.36	(−21.14 , −3.61)	0.007	−5.08	4.03	(−13.12 , 2.95)	0.211	−22.63	5.98	(−34.64 , −10.63)	0.000	−16.44	5.25	(−26.91 , −5.97)	0.003
Weight Z-score Ϯ at 3mo	2.26	1.04	(.16 , 4.36)	0.036	3.26	1.13	(1.01, 5.51)	0.005	3.97	1.59	(.77 , 7.17)	0.016	4.57	1.53	(1.50 , 7.64)	0.004
Length Z-score Ϯ at 3mo	3.64	1.01	(1.61 , 5.68)	0.001	3.35	1.09	(1.17 , 5.52)	0.003	4.54	1.62	(1.27 , 7.82)	0.008	4.14	1.47	(1.19 , 7.08)	0.007
Head circumference Z-score	3.57	1.28	(.98 , 6.17)	0.008	3.10	1.23	(.63 , 5.57)	0.015	3.18	2.01	(−.89 , 7.24)	0.122	4.899	1.56	(1.76 , 8.02)	0.003
Ϯ at 3mo

Note.

^*Bayley Mental and Psychomotor Development Index scores;

^ϮStandard Error;

^**95% Confidence intervals;

^ϮZ-scores as are defined by WHO standardized growth score;

^alog- transformed length of stay variable.

TABLE 4.

Final multivariate regression models for MDI^* at 6 and 12 months

	MDI at 6 months						MDI at 12 months
Final Model**	β	SE Ϯ	95% CI Ϯ	P	R2	β	SE	95% CI	P	R2
Physiologya	−1.37	3.02	(−7.49 , 4.74)	0.652	0.27	2.48	2.95	(−3.41 , 8.38)	0.403	0.29
Length of stayb	−3.94	2.45	(−8.90 , 1.01)	0.116		−6.22	2.04	(−10.30 , −2.13)	0.003
Tube Assisted Feeding at DCc	−5.09	2.89	(−10.92 , .79)	0.088		−.36	2.86	(−6.08 , 5.36)	0.900
Weight Z−scored	1.60	1.01	(−.44 , 3.64)	0.120		1.75	1.13	(−.52 , 4.02)	0.128
Physiology	−.70	3.03	(−6.84 , 5.44)	0.818	0.39	2.80	3.01	(−3.23 , 8.83)	0.356	0.31
Length of stay	−3.52	2.39	(−8.37 , 1.33)	0.150		−6.13	2.04	(−10.22 , −2.05)	0.004
Tube Assisted Feeding at DC	−3.86	2.90	(−9.74 , 2.02)	0.191		.175	2.99	(−5.81 , 6.16)	0.953
Length Z-score d	2.83	1.02	(.76 , 4.90)	0.009		2.14	1.11	(−.075 , 4.35)	0.058
Physiology	−1.28	3.25	(−7.87 , 5.32)	0.697	0.30	3.06	3.34	(−3.63 , 9.75)	0.363	0.26
Length of stay	−2.82	2.60	(−8.09 , 2.45)	0.284		−6.34	2.27	(−10.88 , −1.79)	0.007
Tube Assisted Feeding at DC	−5.34	2.99	(−11. 40 , .71)	0.082		−1.24	3.07	(−7.39 , 4.90)	0.687
Head circumference Z-score d	2.68	1.31	(.023 , 5.35)	0.048		.65	1.36	(−2.08 , 3.38)	0.636
Physiology	−.99	3.01	(−7.09 , 5.09)	0.744	0.26	3.00	3.00	(−3.01 , 9.02)	0.321	0.29
Length of stay	−3.28	2.65	(−8.65 , 2.08)	0.223		−6.34	2.00	(−10.35 , −2.34)	0.002
Tube Assisted Feeding at 3moc	−8.63	5.35	(−19.44 , 2.19)	0.115		2.71	4.33	(−5.95, 11.37)	0.534
Weight Z-score	1.10	1.08	(−1.09 , 3.28)	0.317		2.03	1.18	(−.342 , 4.39)	0.092
Physiology	−1.17	3.03	(−7.31 , 4.98)	0.703	0.37	3.39	3.12	(−2.85 , 9.64)	0.281	0.32
Length of stay	−2.67	2.52	(−7.79 , 2.44)	0.296		−6.22	1.99	(−10.21 , −2.23)	0.003
Tube Assisted Feeding at 3mo	−7.94	4.90	(−17.86 , 1.98)	0.113		2.90	4.25	(−5.60 , 11.41)	0.497
Length Z-score	2.80	1.01	(.75 , 4.84)	0.009		2.30	1.09	(.13 , 4.48)	0.038
Physiology	−1.14	3.31	(−7.86 , 5.57)	0.732	0.27	3.39	3.37	(−3.37 , 10.15)	0.319	0.27
Length of stay	−2.68	2.79	(−8.35 , 2.98)	0.343		−6.61	2.21	(−11.05 , −2.18)	0.004
Tube Assisted Feeding at 3mo	−8.16	6.04	(−20.43 , 4.08)	0.185		2.55	5.04	(−7.56 , 12.65)	0.615
Head circumference Z-score	2.11	1.42	(−.76 , 4.98)	0.145		.95	1.43	(−1.93 , 3.82)	0.511

Note.

^*Bayley’s Mental Development Index scores;

^**Models differed on the basis of various combinations of feeding mode and growth Z- scores to avoid multi-collinearity;

^ϮStandard Error;

^Ϯ95% Confidence intervals;

^aBi-ventricle cardiac physiology vs. single ventricle physiology;

^blog- transformed length of stay variable;

^cTube assisted vs. orally fed;

^dz- scores at 3 months of age as are defined by WHO standardized growth score; DC = Discharge.

TABLE 5.

Final multivariable regression models for PDI^* at 6 and 12 months

	PDI at 6 months						PDI at 12 months
Final Model**	β	SE Ϯ	95% CI Ϯ	P- value	R2	β	SE	95% CI	P- value	R2
Physiology a	4.33	4.60	(−4.97 , 13.63)	0.352	0.29	7.20	4.20	(−1.21 , 15.61)	0.092	0.23
Length of stay b	−2.26	3.74	(−9.82 , 5.30)	0.549		−4.49	2.91	(−10.31 , 1.34)	0.128
Tube Assisted Feeding at DC c	−8.10	4.38	(−16.96 , .76)	0.072		4.22	4.07	(−3.94 , 12.37)	0.305
Weight Z-score d	2.95	1.54	(−.16 , 6.05)	0.062		3.38	1.62	(.14 , 6.62)	0.041
Physiology	6.61	4.80	(−3.11 , 16.33)	0.176	0.32	8.34	4.23	(−.13 , 16.81)	0.053	0.24
Length of stay	−1.47	3.80	(−9.16 , 6.23)	0.702		−4.27	2.86	(−10.00 , 1.47)	0.142
Tube Assisted Feeding at DC m	−7.76	4.54	(−16.96 , 1.43)	0.096		4.44	4.19	(−3.9 , 12.84)	0.295
Length Z-score d	3.27	1.59	(.06 , 6.49)	0.046		3.26	1.55	(.15, 6.37)	0.040
Physiology	4.87	5.02	(−5.30 , 15.03)	0.339	0.22	6.22	4.50	(−2.80 , 15.23)	0.172	0.21
Length of stay	−1.57	4.02	(−9.71 , 6.58)	0.699		−3.71	3.06	(−9.84 , 2.41)	0.230
Tube Assisted Feeding at DC	−8.74	4.58	(−18.03 , .55)	0.064		3.63	4.13	(−4.65 , 11.91)	0.383
Head circumference Z-score d	1.75	2.03	(−2.37, 5.86)	0.395		3.16	1.84	(−.52 , 6.85)	0.091
Physiology	4.81	4.50	(−4.29 , 13.91)	0.292	0.30	4.97	4.25	(−3.54 , 13.48)	0.247	0.24
Length of stay	−.53	3.99	(−8. 60 , 7.53)	0.894		−3.64	2.83	(−9.30 , 2.03)	0.204
Tube Assisted Feeding at 3moc	−16.65	8.04	(−32.89 , −.41)	0.045		−9.34	6.12	(−21.60 , 2.92)	0.133
Weight Z-score	1.92	1.62	(−1.36 , 5.21)	0.243		2.23	1.68	(−1.12 , 5.58)	0.189
Physiology	5.68	4.75	(−3.94 , 15.29)	0.240	0.35	6.07	4.35	(−2.64 , 14.78)	0.168	0.26
Length of stay	.39	3.97	(−7.64 , 8.42)	0.922		−3.28	2.78	(−8.84 , 2.29)	0.243
Tube Assisted Feeding at 3mo	−16.66	7.68	(−32.22 , −1.10)	0.037		−9.17	5.92	(−21.04 , 2.69)	0.127
Length Z-score	3.13	1.56	(−.017 , 6.29)	0.051		2.19	1.51	(−.84 , 5.23)	0.153
Physiology	5.03	5.06	(−5.22 , 15.28)	0.327	0.21	5.28	4.53	(−3.80 , 14.37)	0.249	0.22
Length of stay	−.83	4.29	(−9.51 , 7.86)	0.848		−2.91	2.97	(−8.87 , 3.05)	0.332
Tube Assisted Feeding at 3mo	−15.82	9.27	(−34.61 , 2.97)	0.096		−7.19	6.77	(−20. 80, 6.39)	0.293
Head circumference Z-score	.61	2.17	(−3.79 , 5.01)	0.781		2.32	1.93	(−1.54 , 6.18)	0.234

Note.

^*Bayley’s Mental Development Index scores;

^**Models differed on the basis of various combinations of feeding mode and growth Z- scores to avoid multi-collinearity;

^ϮStandard Error;

^Ϯ95% Confidence intervals;

^aBi-ventricle cardiac physiology vs. single ventricle physiology;

^blog- transformed length of stay variable;

^cDevice assisted vs. orally fed;

^dZ-scores at 3 months of age as are defined by WHO standardized growth score; DC = Discharge.

When assessing 6-month PDI models with device assisted feeding at discharge, 3 month length-for-age z score (p=0.04) was associated with lower PDI scores at 6 months. Similarly, device assisted feeding at 3 months (p=0.04) was significantly associated with lower PDI scores at 6 months. At 12 months, lower PDI scores were associated with 3 month weight-for-age z score(-p=0.04) and length-for-age z score (p=0.04) in separate models. The total variance for PDI ranged between 21% to 35% at 6 months and between 21% to 26% at 12 months.

Discussion

In this prospective cohort study, we describe neurodevelopmental outcomes at 6 and 12 months of age in neonates requiring surgical intervention for complex CHD before 30 days of age. Consistent with previous reports, PDI was significantly impaired in subjects in this study, with MDI being less severely affected(4). Unique to this study was the exploration of the association of growth measures of early mode of feeding (hospital discharge and at 3 months of age) with neurodevelopmental outcomes in the first year of life.

Growth measurements at 3 months of age proved to be significantly associated with both cognitive and motor outcomes at both 6 and 12 months of age. This was consistent with the report from the Pediatric Heart Network Infant Single Ventricle Trial(8) where linear growth in the first year of life was associated with neurodevelopmental outcomes at 12 months of age. In addition to length-for-age z score, we found an association between weight-for-age z score and head circumference-for-age z score at 3 months with both PDI and MDI scores. Malnutrition in fetal and early neonatal animal models results in a reduction in brain size. The rapid brain growth and developmental changes that occur in late gestation make brain growth vulnerable to an inadequate diet(18).

Another finding of this study was the association between mode of feeding and neurodevelopment. It is not surprising that those subjects who required device-assisted feeding at hospital discharge and at 3 months of age were at the greatest risk for lower MDI and PDI scores at 6 months. Infants requiring tube or device-assisted feeding are unable to ingest optimal calories by oral feeding for several potential reasons such as congestive heart failure, gastroesophageal reflux, swallowing dysfunction and other endocrine and genetic factors. The need for device-assisted supplemental feeding may serve as a predictive proxy for the infant’s overall health status. In addition, inability to feed may be an indicator of immature feeding skills. Maturity of feeding organization parallels neurologic integrity(19) which suggests that the need for early device-assisted feeding may be indicative of sub-optimal brain development. Successful feeding depends on integration of both sensory and motor functions. Mussatto et al (4) suggested that the ability to achieve full oral feeding is one of the most significant factors associated with developmental progress.

Our study has several limitations. It is a single-center study making the results specific to our population and possibly not generalizable to other care settings. We were unable to assess caloric intake due to inconsistency in parental reporting of feeding at home. Given the small number of subjects in the fully tube-fed group we did not perform a three group analysis of oral, partially tube-fed and fully tube-fed infants. The lack of consistent airway/vocal cord evaluation and its potential impact on feeding mode and growth also remains a study limitation. Infants were not assessed for the presence of residual lesions and therefore we were unable to explore its potential impact on mode of feeding or MDI/PDI scores. Although we excluded infants with known genetic syndromes, it is possible that infants with an undiagnosed and unrecognized genetic abnormality or syndrome were included in the study. Our association between growth status and mode of feeding with neurodevelopment does not prove causality.

Infants with complex CHD with worse growth and those who require device-assisted feeding early in life are at increased risk for neurodevelopmental disability. Close monitoring of feeding skills and growth trajectories are necessary to identify those infants at increased risk for developmental delays.

Abbreviations

SV

Single Ventricle

2V

Biventricular

BSID-II

Bayley Scales of Infant Development-II

CHD

Congenital Heart Disease

CTRC

Clinical Translational Research Center

DA

Device assisted

LOS

Length of Stay

MDI

Mental Developmental Index

PDI

Psychomotor Developmental Index