The Influence of Maternal Smoking During Pregnancy on the Toddler's Negativity FREE

IN RECENT YEARS, externalizing behavior (ie, aggression) in young children has been found to be stable during the toddler period of development and associated with later childhood aggression and problem behavior such as conduct disorder and drug use.^{1 - 2} Consequently, it is important to examine the etiology and prevention of negativity and its sequelae in young children.

Three psychosocial risk factors for problem behavior have been identified. These psychosocial risk factors include demographic factors, parental intrapersonal difficulty, and parenting difficulties.^{1 ,3 - 4} Thus, certain background and family factors may result in negative behavior in young children. On the other hand, there is the suggestion that problem behavior, including negativity, is related to disturbances in neurophysiological functioning.^{1 ,5} It may be, as several investigators have suggested, that genetic or other biological factors may also be involved in the development of negative behavior.⁵

One such exogenous biological factor is the occurrence of various antenatal and perinatal complications. Raine et al⁶ have noted that birth complications, combined with early maternal rejection at age 1 year, predispose to violent crime at age 18 years. In recent years, there has been the suggestion that one perinatal risk factor in particular, the use of cigarettes by mothers, has an adverse impact on the physical and psychological health of children. Maternal smoking during pregnancy may have long-lasting effects on the offspring's behavior, predicting, for example, later conduct disorder in boys.⁷ A recent study of 4169 men found a dose-response relationship between the amount of maternal prenatal smoking and persistent criminal behavior over time.⁸ Maternal smoking has also been shown to have adverse effects on the child's physical health. The most frequently mentioned results of maternal smoking during pregnancy are preterm delivery, low birth weight, increase in respiratory illnesses, delayed fetal growth, and sudden infant death syndrome.^{1 ,3 ,9} Another consequence is delay in neuropsychological development.⁵

The precursors of deviant behavior such as that noted above may be seen early in childhood, and their effects may be long-lasting. Early childhood negativity may be such a precursor. Although making a connection between such early childhood factors and adult behavior is beyond the scope of this study, it is of interest to assess the effects of maternal smoking during pregnancy on early childhood negativity to further trace its deleterious effects throughout the life span.

The present study was designed to examine the impact of maternal smoking during pregnancy (a perinatal factor) on the risk for negativity in the toddler. Although maternal smoking has been found to increase the probability of low birth weight and have negative consequences, a sizeable number of mothers still continue to smoke.¹⁰ One goal of the present study was to examine the impact of maternal smoking during pregnancy on the angry and negative behavior of 2-year-old children.

Maternal smoking during pregnancy has been found to be associated with some impairment in the child's cognitive ability and language skills.^{11 - 15} Related to this, several investigators have reported an association between maternal smoking and behavior problems in children, such as impulsivity, attentional behavior, and hyperactivity.^{14 ,16 - 17} Maternal smoking during pregnancy has also been found to be related to delinquency in the offspring.¹⁸

Using a large sample with control on social class and sex, Bagley¹⁹ found that mothers who smoked heavily from the fourth month of pregnancy were likely to have children who had behavioral problems at age 16 years.²⁰ Kandel and Wu²¹ and Kandel et al²² reported that parental smoking is related to the offspring's smoking behavior. Moreover, the influence of maternal smoking on the youngster was greater than paternal smoking.

This research builds on and extends the findings of earlier investigations by examining the link between maternal smoking during pregnancy and problem behavior in a community sample of young children, controlling for several significant factors. As noted previously by Brook and Brook,²³ the relationship between maternal perinatal smoking and behavior problems in the offspring might merely reflect the relationship between maternal characteristics and behavior problems in the offspring. For instance, mothers who experience intrapersonal distress may be more likely to smoke, and may be more likely to have children who demonstrate negative behavior. In such a case, it may be that the negative behavior noted in the children may be due to the causal link between the mother's intrapersonal distress and the child's behavior rather than the relationship between maternal smoking and the child's behavior. To make certain that the relationship between maternal smoking and the child's behavior is not due to other factors, 4 types of risk factors will be controlled: demographic, maternal intrapersonal distress, other pregnancy risk factors, and parenting factors.

The present study also examines the impact of 4 major psychosocial risk factors on the child's behavior. Such risk factors include demographic factors, perinatal factors, parental intrapersonal factors, and parenting difficulties.

The sample came from the third generation of our longitudinal study of childhood etiology.¹ Based on a sample of 154 two-year-old toddlers and their mothers, we selected the mothers who are smokers. To control for the effect of maternal smoking during pregnancy, we focused on the final sample of 99 mothers who had smoked when the data were collected. Among these 99 mothers, 52 smoked during pregnancy and 47 either stopped smoking when they were pregnant or started smoking after pregnancy. Eighty-six percent of the toddlers were white and 12% were black. There were 51 boys and 48 girls in the final sample. However, the sex of the toddlers was not considered in the study because this variable did not interact with the maternal variables to affect the toddler's negativity in our preliminary analysis. The mean (SD) of the mother's age when the target child was born was 21.73 (4.52) years. Sixteen percent of the mothers finished grade 13 or higher in school and 43% finished grade 12. The median yearly family income was $20 000. The school of medicine's institutional review board approved this project, and informed consent was obtained from all of the mothers and 2 legal guardians.

The study is designed to investigate the impact of maternal smoking during pregnancy on the toddler. We used the measurement of the toddler's negativity as the dependent variable.

The scale of negativity was a composite of 3 subscales of toddlers' troublesome behaviors assumed to be precursors of later unconventionality and lessened emotional control: impulsivity, risk taking, and rebelliousness. The Cronbach α for the todder's scale was .85.

Maternal smoking during pregnancy was measured by a single-item question, "How often did you smoke during your pregnancy?" Based on the evidence from our previous studies that maternal personality and drug use were linked to the offspring's problem behavior, we included the mother's intolerance of deviance, low ego integration, depression, and drug use in the study. The measurement of maternal drug use is a single-item question about the frequency of use of tobacco, alcohol, marijuana, and other illicit drugs by the individual.

Regarding the maternal child-rearing domain, we included affection toward the child, nonconflictual mother-child relationship, and power-assertive disciplinary methods. Table 1 presents the reliabilities, sample questions, and option panels for the maternal variables used in the study.

We also took into account the respondents' socioeconomic status (SES), which includes both mothers' and fathers' levels of highest education, occupation, and the household income.

To clarify the relative effects of the independent variables, a logarithm of the odds ratio statistics was used. All the independent variables were dichotomized. The cutoff point for each variable was chosen to obtain sufficient frequency in each category.

Children's negativity was related to the level of maternal smoking during pregnancy (χ²₁=5.60, P<.05). The toddler's negativity was also related to a conflictual mother-child relationship (χ²₁=24.46, P<.001) and power-assertive techniques of discipline (χ²₁=7.37, P<.01). There was a trend for maternal age at birth of the child (χ²₁=2.80, P<.10) and low maternal affection (χ²₁=2.90, P<.10) to be linked to the child's level of negativity. Toddlers who scored high and low on negativity were also compared on maternal demographic variables (ie, SES and the mother's marital status), maternal personality/behavior/attitudes (intolerance of deviance, ego integration, depression, and drug use), and perinatal variables (maternal alcohol use during pregnancy and maternal age at birth of the child). These demographic and personality variables were not related to the toddler's negativity.

Two logistic regressions were conducted to examine the relationship between maternal smoking during pregnancy and toddler negativity. The first controlled for parental personality/behavior/attitudes, perinatal variables, and demographic variables. A second logistic regression controlled for parenting behavior as well as the demographic variables.

The first logistic regression analysis assessed the relationship of maternal smoking, background factors, and maternal personality attributes and the child's negativity. Only maternal smoking and age of mother at birth of the child were related to toddlers' negativity when the demographic and maternal variables were entered into the analysis (Table 2).

A second logistic regression was conducted, regressing the toddler's negativity on maternal smoking during pregnancy, 3 aspects of child rearing (affection toward the child, mother-child nonconflictual relationship, and power-assertive discipline methods), SES, and the mother's age at birth of the child. As in the previous regression, each variable was examined with control on the other variables in the logistic model (Table 3).

Table 3 indicates that the toddler's negativity was related to maternal smoking during pregnancy, as well as to the mother's age at the child's birth (>23 years of age), maternal techniques of power assertion, and a conflictual mother-child relationship. The findings indicated that maternal smoking during pregnancy increased the odds of the toddler's negativity, even after controlling for the other variables in the model. Thus, the effect of maternal smoking during pregnancy on the toddler's negativity was independent of demographic, perinatal, and maternal personality attributes, as well as child-rearing techniques. In addition, a conflictual mother-child relationship was the variable most strongly predictive of the toddler's negativity (odds ratio=16.12).

Mothers who smoked during pregnancy were far more likely to have toddlers who displayed negativity than mothers who did not smoke during pregnancy. The relationship between maternal smoking and toddler negativity was maintained despite controlling for a number of psychosocial risk factors that have previously been found to be associated with negativity in toddlers. These psychosocial risk factors include demographic variables (ie, intolerance of deviance and depression), maternal use of alcohol during pregnancy, maternal age at the birth of the toddler, the mother-child bond, and maternal power-assertive disciplinary methods.

These results are in accord with previous research, which has demonstrated that maternal smoking during pregnancy is related to increased rates of problem behavior in the child.^{15 - 17 ,30 - 31} The findings of the present study extend previous research by examining the effect of maternal smoking during pregnancy on the toddler's behavior while controlling for important maternal factors associated with negativity in the toddler, and by using a community-based sample. Thus, the influence of maternal smoking during pregnancy on later negativity in the toddler cannot be explained by these other maternal risk factors.

The results indicate that the influence of maternal smoking during pregnancy on the toddler's negativity cannot be fully explained by variables that are strongly related to negativity, such as mother-child conflict.^{18 ,32} One possible mechanism through which maternal smoking during pregnancy influences toddlers' negativity is by affecting the neural structure and functioning in the brain of the offspring which, in turn, affects negativity in the youngster. It may be that maternal smoking during pregnancy is a marker for stress that causes physiological effects, rather than the direct cause of these physiological effects. Another possible mechanism has to do with the adverse effects on the child's physical health such as low birth weight and increased respiratory illnesses. Assuming these statements are accurate, the benefit of prevention programs designed to forestall maternal smoking during pregnancy is clear.

The findings indicated that there are 3 major types of maternal risk factors that are related to negativity in the child, namely, maternal smoking during pregnancy, a conflictual mother-child relationship, and maternal use of power-assertive discipline. These findings are in accord with those of others.^{4 ,33}

As noted previously, a conflictual mother-child bond has a particularly strong relationship with the child's negativity. It may be that this measure reflects the mutual frustrations of the mother's actions and the child's reactions that reciprocally reinforce one another.

This study extends previous results in demonstrating the independence of these risk factors, thereby showing that the child's negativity is vulnerable to 3 distinct types of risks. The implication is that interventional procedures need to focus separately on the alleviation of these distinct risk factors. Moreover, reduction of one of these risk factors should result in a decrease in the toddler's negativity, despite the continued presence of the other 2 risk factors. Alternatively, in any particular family, the accumulation of these risk factors may expose the child to greatly increased conflict. What remains to be determined is whether these risk factors, as well as the toddler's negativity itself, forecast future problem behavior.

A plausible connecting link between the toddler's negativity and later problem behavior is childhood aggression. There is evidence from longitudinal analyses conducted over 20 years with children as young as 5 years that childhood aggression is related to later delinquent behavior and drug use.²⁵

As noted earlier, mothers who were older than 23 years at the birth of their children reported greater negativity in their offspring. (The mother's maturity may be related to increased perception of negativity in her offspring.) The finding is somewhat puzzling, as further analyses indicated that the mother's age at the birth of her child was not related to the major maternal predictors of her child's negativity reported in Table 2 and Table 3 or the number of children in the family. As expected, low maternal SES was associated with greater child negativity, which may be related to the increased stress in the lives of mothers of low SES.

The findings of the present study warrant some qualifications. First, the findings are based on a relatively small sample. A larger sample would have enabled the analyses to be conducted separately for boys and girls. It is conceivable that the effects of maternal smoking on boys are different from those on girls, since negativity tends to be greater in male toddlers than in female toddlers. In our own data, males did score higher than females on our measure of negativity (t=2.16; P<.05). Second, the data on maternal smoking were based on maternal reports. It is quite possible that some mothers would be reluctant to admit to smoking during pregnancy because of the social stigma surrounding the act. In addition to maternal self-reports of smoking behavior, direct physiological measures of maternal smoking during pregnancy would be beneficial.

Despite the significant findings of the study, it is conceivable that the influence of maternal smoking during pregnancy on negativity in the offspring is the result of some other factors not included in the present study that affect both maternal smoking and infant negativity. Such attributes might include maternal personality disorders that are transmitted to the child through genetic transmission or other aspects of the parent-child bond not assessed in this study. Even if the relationship between maternal smoking and toddler negativity was due to a third factor (eg, maternal personality disorder), it would still suggest that maternal smoking is a marker for that underlying third factor. In such a case, it would signal the need for early intervention to lessen negativity in the toddler.

Overall, smoking during pregnancy is not only related to the health of the mother and the child,^{34 - 35} but also to the behavioral traits of the child. Our research has attempted to distinguish between those who smoke during pregnancy and those who smoke but abstain during pregnancy. We were therefore able to highlight the importance of the relationship between maternal smoking during pregnancy and the development of negativity among 2-year-old toddlers.

To the extent that maternal smoking during pregnancy, or the stress associated with it, has an effect on negativity, a reduction in maternal smoking should lead to decreased negativity in the toddler. A decrease in the toddler's negativity is likely to result in decreased problem behavior in childhood and adolescence. Future research needs to be directed at the antecedents of maternal smoking during pregnancy, as well as the mechanism through which maternal smoking during pregnancy affects the child's development.

Accepted for publication September 18, 1999.

This research was supported in part by grant DA 03188 and by Research Scientist Award K05 DA 00178 from the National Institute on Drug Abuse, Bethesda, Md (Dr J. Brook).

We gratefully acknowledge the suggestions of Jacques Normand, PhD. We are greatly indebted to Li-Jung Tseng, PhD, for her invaluable comments and advice and for performing the statistical analyses.

Reprints: Judith S. Brook, EdD, Mount Sinai School of Medicine, Box 1044A, 1 Gustave L. Levy Pl, New York, NY 10029.

Editor's Note: So how do we know that cigarette smoking is not a marker of stress or whatever effects negativity in the child? On the other hand, how can you be positive if someone/the entire house smells of smoke all the time?—Catherine D. DeAngelis, MD