Diastolic Alterations in Infants Exposed to Intrauterine Cocaine: A Follow-up Study by Color Kinesis

Abstract

Background:

During the first 48 hours of life, newborn infants exposed to cocaine in utero have left ventricular diastolic segmental abnormalities. It is unknown whether these abnormalities are transient because of short-term effects or persist in older infants, possibly reflecting a teratogenic effect of cocaine.

Methods:

This study prospectively evaluated global and segmental systolic and diastolic cardiac parameters by color kinesis. The patients were 2- to 6-month-old infants who were exposed to cocaine in utero (N = 56). Their data were compared with normal control patients with no intrauterine drug exposure (N = 60) and newborns exposed to drugs other than cocaine (N = 72).

Results:

At the age of 2 to 6 months, there was no significant difference in the measured color kinesis parameters among the cocaine-exposed and the 2 control groups (infants prenatally exposed to other drugs and no drugs). Infants exposed to heavy cocaine prenatally, as compared with the noncocaine-exposed group, had a significant (P = .007) increase in septal fractional area change during left ventricular filling.

Conclusions:

At 2 to 6 months of age, infants have recovered from initial left ventricular diastolic segmental alterations seen in the first 48 hours of life except for the septal wall in the heavily cocaine-exposed group.

Infants with prenatal exposure to cocaine have structural cardiovascular malformations and sustained arrhythmias that may lead to congestive heart failure, cardiorespiratory arrest, and death.^1,2 They may have poor left ventricular (LV) systolic function and a higher blood pressure during the first day of life.³ By causing coronary artery spasm,⁴ cocaine also affects LV segments. Despite the usefulness of color kinesis in the objective assessment of LV segmental function,^5,6 studies related to cocaine are limited. Within 48 hours of birth, asymptomatic cocaine-exposed neonates admitted to the normal newborn nursery had altered segmental diastolic function of the LV.^7,8 The long-term effects of cocaine on the newborn’s cardiovascular system and the natural history of the above diastolic alterations are unknown. The current study reports the follow-up of these infants who were cocaine-exposed in utero at the age of 2 to 6 months.

METHODS

Infants who were asymptomatic, from the normal newborn nursery, less than 48 hours old, weighing more than 1500 g, and between 33 and 42 weeks gestation were enrolled in the study between 1997 and 2000. Infants were ineligible for the study if they or their mothers had significant medical or surgical problems, or if they had in utero exposure to illicit drugs other than alcohol, marijuana, cocaine, and nicotine.⁸ The exposed group included infants who were exposed to cocaine and other drugs in utero such as alcohol, marijuana, and nicotine. Of the 2 control groups, the first group included infants with exposure to alcohol, marijuana, nicotine, or a combination of these, but no cocaine, and a second control group had no intrauterine drug exposure. Within 48 hours of life, in utero drug exposure was determined by a structured interview questionnaire;^9,10 laboratory analysis of maternal and infant urine; and meconium testing for cocaine and its metabolites, barbiturates, benzodiazepines, cannabanoids, opiates, phencycladine, amphetamines, and cotinine.⁸ The frequency of drug use multiplied by the amount used per day computed the severity of use score for the month before pregnancy and for each trimester. Averaged use score provided a total score for the prenatal exposure for each drug.¹¹ A positive exposure included exposure of a drug by maternal self-report or toxicology studies. Heavy cocaine use was defined a priori as “the amount of cocaine used during the pregnancy that exceeded the 70th percentile of cocaine usage in our prior study.”¹² The remaining infants who were cocaine-exposed were defined as “light cocaine.” Their data were compared with infants not exposed to cocaine.

The above infants were studied again between the ages of 2 to 6 months at the General Clinical Research Center of MetroHealth Medical Center. At the time of follow-up, all infants were asymptomatic, on no medications, and had no prior significant illnesses. After the initial history (maternal postpartum questionnaire) and physical examination, these infants underwent color kinesis echocardiography. The study was approved by the institutional review board for human investigation. Informed written consent was obtained from the legal guardians or parents of all participants.

All infants underwent 2-dimensional, Doppler, and color kinesis echocardiographic examination with a 5-MHz transducer (Sonos 5500 Ultrasonograph, Hewlett-Packard Co, Andover, Mass) by an experienced pediatric-trained sonographer. The color kinesis studies were obtained in the parasternal short-axis view. A cardiologist trained in color kinesis who was blinded to the infant’s drug exposure analyzed the color kinesis images offline. The analysis was performed using automated software (Quick Color Kinesis, EchoSoft Co, Wilmington, Del). The LV end-systolic and diastolic cavity was divided into 6 60-degree, wedge-shaped segments (clockwise: anterior, lateral, posterior, inferior, septal, and anteroseptal).⁸ Incremental area change was normalized from the end-diastolic area (EDA) of the corresponding segment resulting in regional fractional area change (percent) during systole and diastole. We evaluated global and regional, systolic and diastolic parameters including EDA, global fractional area changes (GFAC) during filling and systole, and index of asynchrony (IA). IA for LV filling represents the SD of the mean percent of filling of all segments at 50% filling time. The GFAC is an incremental area change divided by the EDA. Incremental area change of a segment divided by its EDA provides the percent segmental fractional area change for that segment.

Data were analyzed using the statistical software SAS (version 8.1, SAS Institute, Cary, NC). Hypothesis testing regarding group differences for demographic data was performed using analysis of variance. Distributional assumptions were assessed to test for appropriate underlying assumptions. Analysis and comparison among the 3 groups (within 48 hours from birth and 2- to 6-month follow-up) were done by the method of maximum likelihood, assuming a repeated measures analysis for unbalanced data. For each analysis it was assumed that a model with separate means at each time point for each group fits the data. The covariance structure of the repeated measures was modeled using an unstructured model. This makes no assumption about the correlation structure of the repeated measures. In addition to estimating the mean responses by group across time, under this model, we also tested specific hypotheses about the difference among groups. Two specific a priori hypotheses that were tested are the following: hypothesis 1, the means of the groups are not different at birth; and hypothesis 2, the means of the groups are not different at the 2- to 6-month follow-up assessment.

For each outcome variable, this study reports the least squares mean and SEM. Normality for IA was achieved with the square-root transformation. Statistical significance was defined a priori as “a P value < .05, 2-tail.” For the post hoc comparisons among the groups, the Bonferroni correction for reducing test-wise error was performed.

RESULTS

Of the initially enrolled newborns (N = 325), optimum color kinesis measurements were obtained in 256 infants. At follow-up, measurements were available on 188 infants. Ninety-five percent of the infants were between the ages of 2 and 6 months at the follow up assessment. The number of infants with cocaine exposure at birth and at follow-up were 81 and 56, respectively; with exposure to other drugs but not cocaine, 93 and 72, respectively; and with exposure to no drugs in utero, 82 and 60, respectively (Table 1).

Table 1.

Patient characteristics of infants at birth and follow-up

	Group 1: cocaine	Group 2: other drugs	Group 3: no drugs	P value
At birth Sample size (N)	N = 81	N = 93	N = 82
Male	42%	55%	54%	.181
Gestational age (wks)	38.45 ± 1.78	39.09 ± 1.54	39.40 ± 1.29	<.005a, b
Weight (g)	2888 ± 518	3129 ± 448	3259 ± 438	<.0001a, b
Length (cm)	48.08 ± 2.84	49.89 ± 2.50	49.89 ± 2.34	.0001a, b
Head circumference (cm)	33.22 ± 1.55	33.88 ± 1.55	34.07 ± 1.32	<.0008a, b
Body surface area (m2)	0.19 ± 0.02	0.21 ± 0.02	0.21 ± 0.02	<.0001a, b
Caucasians (%)	31	34	19	.08
At follow-up Sample size (N)	N = 56	N = 72	N = 60
Age (mos)*	2.44 ± 1.18	2.44 ± 1.59	2.50 ± 1.94	.97
Body surface area (m2)	0.20 ± 0.03	0.21 ± 0.03	0.21 ± 0.03	.32

^*Ninety-five percent of subjects were between 2 and 6 months.

Post hoc significant comparisons:

^acocaine versus no drugs at birth

^bcocaine versus other drugs at birth.

Color kinesis filling parameters that were found to be significantly higher in newborns exposed to in utero cocaine (GFAC [perecnt], regional fractional area changes [percent] for the anterior, septal, inferior, and lateral wall, and IA) were not significantly different among the groups at 2 to 6 months (Table 2). For the IA during diastole, the group × age interaction was not significant (P = .053). For the anterior wall fractional area change, the group × age interaction and the test of group means at 2 to 6 months were not significant (P = .512 and P = .055, respectively). Similarly, for the septal wall fractional area change, the group × age interaction and the test of group means at 2 to 6 months were also not significant (P = .910 and P = .083, respectively).

Table 2.

Selective color kinesis data in infants exposed to cocaine and control groups: least squares means and SE from the repeated measures analysis

	Cocaine N mean (SE)	Other drugs N mean (SE)	No drugs N mean (SE)	P value
Sample size (N)
Birth	81	93	82
2–6 mos	56	72	60
Heart rate
Birth	128.80 (1.44)	126.21 (1.29)	125.37 (1.39)	.20
2–6 mos	143.97 (2.19)	135.54 (1.67)	138.91 (2.13)	.01b
GFAC-Systole
Birth	78.09 (0.87)	77.49 (0.77)	76.59 (0.86)	.465
2–6 mos	77.34 (0.99)	75.82 (0.84)	75.80 (0.91)	.421
GFAC-Diastole
Birth	76.17 (1.06)	72.28 (1.00)	72.05 (1.06)	.008a, b
2–6 mos	74.06 (1.28)	72.45 (1.12)	72.19 (1.22)	.518
Index of asynchrony-systole*
Birth	9.60 (0.07)	9.46 (0.06)	9.22 (0.07)	.829
2–6 mos	8.98 (0.07)	8.99 (0.06)	8.86 (0.07)	.968
Index ofasynchrony-diastole*
Birth	12.59 (0.07)	10.93 (0.07)	11.19 (0.07)	.043e
2–6 mos	11.86 (0.10)	12.99 (0.08)	11.73 (0.09)	.283
RFAC-diastole
Lateral wall
Birth	80.95 (1.36)	74.63 (1.26)	74.27 (1.35)	.001a, b
2–6 mos	77.54 (1.44)	76.21 (1.27)	77.56 (1.41)	.710
Inferior wall
Birth	77.99 (1.35)	74.14 (1.26)	73.39 (1.35)	.036e
2–6 mos	80.53 (1.53)	78.87 (1.34)	77.54 (1.48)	.376
Posterior wall
Birth	75.49 (1.43)	70.44 (1.33)	72.06 (1.42)	.034b
2–6 mos	76.78 (1.54)	76.48 (1.37)	75.01 (1.50)	.669
Anterior wall
Birth	72.82 (1.45)	67.69 (1.37)	65.43 (1.46)	.001a, b
2–6 mos	71.07 (1.85)	65.37 (1.64)	66.29 (1.78)	.056
Septal wall
Birth	76.44 (1.25)	72.82 (1.17)	71.99 (1.25)	.028a
2–6 mos	73.40 (1.63)	70.34 (1.43)	68.37 (1.56)	.083
Anteroseptal wall
Birth	75.11 (1.34)	70.90 (1.25)	73.17 (1.35)	.073
2–6 mos	63.99 (1.75)	63.53 (1.53)	64.92 (1.69)	.829

^*Square-root transformation used for analysis, means reported on original scale, SE reported from transformed scale. Post hoc significant comparisons with Bonferroni correction were:

^acocaine versus none at birth

^bcocaine versus other drugs

^eno post hoc pairwise differences were statistically significant at the adjusted a level of 0.017 ( = 0.05/3).

GFAC, global fractional area change; RFAC, regional fractional area change.

The color kinesis measurements on the basis of heavy cocaine, light cocaine, and noncocaine are summarized in Table 3. The septal wall fractional area change during filling at 2 to 6 months was significantly different between the heavy cocaine versus noncocaine group (P = .007). The anterior wall was also different at follow-up among the 3 groups (P = .043), however, none of the pairwise comparisons were significant at the Bonferroni adjusted α-level of 0.017. The group × age interaction for both the septal and anterior wall was not significant (P = .660 and P = .915, respectively).

Table 3.

Patient characteristics of infants at birth and follow-up

	No cocaine mean (SE)	Light cocaine mean (SE)	Heavy cocaine mean (SE)	P value
Sample size (N)
Birth	175	43	38
2–6 mos	132	27	29
Heart rate
Birth	125.82 (0.92)	123.91 (1.88)	134.81 (2.08)	.0001a, b
2–6 mos	137.02 (1.39)	140.96 (2.94)	147.53 (3.23)	.009d
GFAC-systole
Birth	77.09 (0.57)	78.43 (1.19)	77.69 (1.26)	.577
2–6 mos	75.81 (0.62)	77.77 (1.41)	76.93 (1.37)	.388
GFAC-diastole
Birth	72.17 (0.73)	75.38 (1.45)	77.10 (1.54)	.006a
2–6 mos	72.33 (0.82)	72.42 (1.80)	75.74 (1.80)	.222
Index of asynchrony-systole*
Birth	9.35 (0.05)	8.57 (0.10)	10.78 (0.10)	.042b
2–6 mos	8.93 (0.05)	8.69 (0.10)	9.25 (0.10)	.801
Index ofasynchrony-diastole*
Birth	11.05 (0.05)	11.51 (0.10)	13.87 (0.11)	.004a
2–6 mos	12.41 (0.06)	12.74 (0.13)	10.96 (0.13)	.323
RFAC-diastole
Lateral wall
Birth	74.46 (0.92)	81.01 (1.88)	80.92 (1.95)	<.001a, c
2–6 mos	76.82 (0.93)	74.50 (2.04)	80.49 (2.00)	.104
Inferior wall
Birth	73.79 (0.92)	77.33 (1.87)	78.78 (1.97)	.032e
2–6 mos	78.28 (0.99)	78.50 (2.20)	82.46 (2.13)	.202
Posterior wall
Birth	71.19 (0.97)	75.02 (1.99)	76.00 (2.07)	.045a
2–6 mos	75.82 (1.01)	76.31 (2.20)	77.25 (2.16)	.830
Anterior wall
Birth	66.63 (1.00)	71.24 (1.99)	74.62 (2.12)	.001a
2–6 mos	65.79 (1.20)	69.74 (2.66)	72.43 (2.58)	.043e
Septal wall
Birth	72.43 (0.85)	75.47 (1.72)	77.54 (1.81)	.022a
2–6 mos	69.44 (1.05)	70.47 (2.33)	76.15 (2.25)	.027d
Anteroseptal wall
Birth	71.95 (0.92)	75.64 (1.87)	74.55 (1.94)	.142
2–6 mos	64.17 (1.14)	63.43 (2.52)	64.48 (2.43)	.952

^*Square-root transformation used for analysis, means reported on original scale, SE reported from transformed scale. Post hoc significant comparisons with Bonferroni correction were

^aheavy cocaine versus no cocaine at birth

^bheavy cocaine versus light cocaine at birth

^clight cocaine versus no cocaine at birth

^dheavy cocaine versus no cocaine at 2–6 months

^eno post-hoc pairwise differences were statistically significant at the adjusted a level of 0.017 (=0.05/3).

GFAC, global fractional area change; RFAC, regional fractional area change.

From the first 48 hours of life to 2 to 6 months of age, fractional area changes during filling significantly (P < .001) decreased in anteroseptal wall among all 3 groups. Similar changes over time were not noted in the other 5 LV segments during filling.

Certain color kinesis parameters were associated with heart rate (bpm), gestational age (weeks), body surface area at birth, and birth weight. At both birth and follow-up, these associations were: (1) heart rate with the regional fractional area change during filling for inferior wall (P = .037, slope = −0.093); and (2) gestational age with the IA during systole (P = .012, slope = −0.050) and the regional fractional area change during filling for lateral wall only in the group exposed to other drugs (P = .008, slope = −1.59). Only at birth was there an association between body surface area (m²) with GFAC during systole (P = .002; slope = 80.008) and the regional fractional area change during filling for septal wall (P = .002, slope = −111.47). In addition, at 2 to 6 months of age, birth weight (grams) was only associated with the regional fractional area change during filling for anteroseptal wall (P = .024, slope = −0.004).

After adjusting for the above potential confounders, the results were similar to the unadjusted analysis at birth and at follow-up for the IA during systole and the regional fractional area change during filling for lateral wall. However, different conclusions at birth, on the basis of group tests, were found for GFAC during systole (P = .028, post hoc tests significant for cocaine vs no drugs), and the regional fractional area change during filling for the following segments: inferior wall (P = .008, cocaine vs both no drugs and other drugs), anteroseptal wall (P = .044, cocaine vs other drugs), and septal wall (P = .001, cocaine vs both no drugs and other drugs). Similar associations were found in the models that categorized exposure as heavy cocaine, light cocaine, and no cocaine. Conclusions only changed for regional fractional area change during filling for inferior wall becoming significant at birth (P = .002, post hoc tests significant for heavy cocaine vs no cocaine) and at follow-up (P = .019, post hoc tests significant for heavy cocaine vs light and no cocaine groups), however, IA during systole at birth became nonsignificant (P = .096).

DISCUSSION

Prenatal cocaine exposure in newborns causes uneven LV segmental filling within 48 hours of birth. The frequency and severity of these alterations increases with severity of cocaine-exposure, with the heavy cocaine-exposed group being the most severely affected. At the age of 2 to 6 months, however, most infants showed apparent recovery of the LV muscle, except the diastolic filling of the septal wall that remained affected in the heavily cocaine-exposed group. These findings have not been previously reported.

LV dysfunction occurs in 7% of apparently healthy adults who abuse cocaine, whereas regional wall motion abnormalities were noted in 2 of 84 (2.3%) by radionuclide angiography.¹³ Hypokinesia of the anterior segment occurred in the first patient and anterior, posterior, and inferior segments were involved in the second patient. LV segmental involvement was similar in newborns in the current study. However, although patients with impaired diastolic function of the LV myocardium have lower filling fractions, infants who were cocaine-exposed had increased anterior, septal, and lateral filling fractions at birth and septal filling fractions at 2 to 6 months of age. It could be speculated that chronic cocaine exposure produced a “training effect” such as is seen in athletes. For example, older athletes show higher diastolic velocity of the inferior wall.¹⁴ As aerobic training has segmental affinity, normal segmental functions at rest may not be reflective of segmental functions during exercise. Because these infants who were asymptomatic did not have alterations in the cardiac output or systolic function, the clinical significance of the reported segmental and diastolic findings or their response to exercise remains unclear.

Altered segmental filling fractions indicate a regional involvement of the myocardium, either as a result of coronary artery involvement or as part of global changes that represent a unique reaction of the developing myocardium. It is of interest that from the first 48 hours of life to 2 to 6 months of age, fractional area changes during filling decreased significantly in the anteroseptal wall among all 3 groups. The documentation of this transition has not been reported. Our findings support the notion that growth of the LV myocardium, at least in terms of physiological properties, is nonhomogenous. The reported segmental differences may be a result of the difference in the growth rate of the myocardial cells, coronary blood supply, pressure changes, or a combination of these.

Other authors have reported no associations of the color kinesis measurements with heart rate, body surface area, and other medical or demographic variables at young age.¹⁵ This is also the case for our data in general, however, we did find some significant relationships. Because of the exploratory nature of this analysis, our findings of associations should be interpreted with some caution. Although certain clinical parameters had significant associations with color kinesis parameters, their influence did not alter final conclusions at 2 to 6 months of age on the basis of adjusted versus unadjusted analysis, except regional fractional area change for inferior wall became significantly different for heavy-exposed cocaine group after adjustment. Because the associations were not uniform across all the LV segments and the degree of change was relatively minor, our conclusions were on the basis of the unadjusted analysis.

The current study has potential limitations. The fractional area changes were measured only in short-axis and do not necessarily reflect volume changes or cardiac output, and they may not be reflective of changes in the long-axis or in other LV dimensions. The possible contribution of LV rotational changes to our measurements of GFAC or regional fractional area changes during the relaxation period are expected to be minimal because similar rotational changes probably occurred in the control population. Furthermore, apical walls do the most rotation and septal walls at each short-axis level undergo the least rotational wall motion.¹⁶

In summary, at 2 to 6 months of age, although most infants showed apparent recovery of the LV segmental alterations, the heavy cocaine-exposed group continued to show septal wall filling abnormality. Long-term implications of the above filling alterations secondary to in utero cocaine exposure remain unknown.