Dopamine system genes associated with parenting in the context of daily hassles


*M. H. van IJzendoorn, Centre for Child and Family Studies, Leiden University, PO Box 9555, NL-2300 RB Leiden, The Netherlands. E-mail:


The current study examined the molecular genetic foundations of sensitive parenting in humans and is the first to test the interaction between genes and environment in modulating parental sensitive responses to children. In a community sample of 176 Caucasian, middle class mothers with their 23-month-old toddlers at risk for externalizing behavior problems, the association between daily hassles and sensitive parenting was investigated. We tested whether two dopamine-related genes, dopamine D4 receptor (DRD4) and catechol-O-methyltransferase (COMT) gene polymorphisms, modulate parents’ vulnerability to the negative influence of daily hassles on sensitive parenting behavior to their offspring. Sensitive parenting was observed in structured settings, and parents reported on their daily hassles through a standard questionnaire. In parents with the combination of genes leading to the least efficient dopaminergic system functioning (COMT val/val or val/met, DRD4-7Repeat), more daily hassles were associated with less sensitive parenting, and lower levels of daily hassles were associated with more sensitive parenting = 1.12. The other combinations of COMT and DRD4 polymorphisms did not show significant associations between daily hassles and maternal sensitivity, suggesting differential susceptibility to hassles depending on parents’ dopaminergic system genes. It is concluded that the study of (multiple) gene–environment interactions (in the current case: gene by gene by environment interaction, G × G × E) may explain why some parents are more and others less impacted by daily stresses in responding sensitively to their offspring’s signals.

Although parenting is crucial for child development, almost no studies have been conducted on the genetics of parental behavior in humans (Swain et al. 2007). In a series of programmatic studies on parenting in rodents, Meaney and colleagues provided evidence for genomic as well as non-genomic transmission of sensitive ‘licking/grooming and arched-back nursing’ in rats and showed that sensitive care enhanced the development of the dopaminergic system and stress regulation in the offspring (Meaney 2001, 2007).

In humans, sensitive parenting has frequently been investigated, in particular its role in the development of secure attachments that shape children’s responses to stress and their willingness to be compliant to parental requests (Ainsworth et al. 1978; Belsky 2005; De Wolff & Van IJzendoorn 1997). Parental sensitivity refers to the ability to accurately perceive children’s attachment signals and to respond to these signals in an adequate and prompt way (Ainsworth et al. 1978). Parental sensitivity has been shown to be affected by the strains and stresses implied in low socioeconomic status (Bakermans-Kranenburg et al. 2004), postnatal depression (Hoffman et al. 2006) and the experience of (parenting) daily hassles (Belsky et al. 1995, 1996a,b; Crnic & Low 2002; Phelps et al. 1998; Repetti & Wood 1997). However, much less is known about the genetic underpinnings of parental sensitivity and about the interaction between genes and environment in modulating sensitive responses to children. Previous studies focused on children’s genetic differences in relation to the parenting they retrospectively perceived (Lucht et al. 2006) or actually experienced (Bakermans-Kranenburg & Van IJzendoorn 2006, 2007; Gervai et al. 2007), but as yet no molecular genetic basis of observed human parenting has been found, and studies on genetic differences interacting with the environment related to parenting are lacking. In the current study, we examine two crucial dopamine-related genes and their interaction with parents’ daily hassles in addressing the question whether some parents are more and others are less vulnerable to the influence of stress in responding sensitively to their offspring.

The catechol-O-methyltransferase (COMT) gene is one of the candidate genes to be studied in the context of parenting as it influences the activity of dopamine in the prefrontal cortex through its catabolic effects on the neurotransmitter. The COMT gene, in humans localized on chromosome 22, band q11.2, plays a pivotal role in the regulation of dopaminergic activity, and the functional variation in the val158met polymorphism of this gene (Lotta et al. 1995) may contribute to differences in sensitive parenting under stress originating from daily hassles. The met allele codes for an enzyme with about one-fourth of the activity in degrading dopamine compared with the enzyme that is coded by the val allele (Chen et al. 2004). Smolka et al. (2005) suggested that for subjects with one or two COMT met alleles, increased limbic and prefrontal activation elicited by unpleasant stimuli might contribute to their lower emotional resilience against negative mood states, whereas in a comprehensive review, Bilder et al. (2004) hypothesized that subjects with one or more met158 alleles may better perform on tasks requiring a consistent behavioral program in the face of distraction. In a meta-analysis, Savitz et al. (2006) showed that the low-activity met allele is consistently associated with better performance on cognitive tasks, in particular those with a working memory component. The association with less distractibility may, however, also lead to decreased cognitive flexibility (Bilder et al. 2004; Drabant et al. 2006), thus both alleles seem to be associated with risks and benefits. As parental sensitivity requires continuous attention to the children’s signals even in stressful circumstances, less distractibility may help parents to remain focused on their child.

The dopamine D4 receptor gene (DRD4) is another gene implicated in the dopaminergic system. The gene, located near the telomere of chromosome 11p, exhibits various polymorphisms, with variations in the number of 48-bp tandem repeats in exon 3. The three most common variants are 2R, 4R and 7R. The 7R gene variant (DRD4-7R) codes for a receptor that is less sensitive to endogenous dopamine compared with the receptors coded for by the shorter repeats. In previous studies, the DRD4-7R polymorphism showed associations with impulsive behavior in adults and attention-deficit hyperactivity disorder in children (Ebstein 2006; Swanson et al. 2007). Importantly, DRD4-7R (linked to attentional, motivational and reward mechanisms; Robbins & Everitt 1999) may make children more vulnerable to negative environmental influences in developing insecure, disorganized attachments (Lakatos et al. 2000, 2002, 2005; Gervai et al. 2005; but see Bakermans-Kranenburg & Van IJzendoorn 2007) and externalizing behavior problems (Bakermans-Kranenburg & Van IJzendoorn 2006, 2007), which are both related to stress dysregulation.

In the current gene by gene by environment interaction study, we investigate whether COMT val and DRD4-7R alleles (separately or in interaction) also make parents more vulnerable to the negative influence of daily hassles on sensitive parenting behavior to their offspring.

Materials and methods


The current paper is based on data obtained in the Screening and Intervention of Problem behavior in Toddlerhood study (SCRIPT), which investigated the effectiveness of an early intervention program aimed at reducing externalizing problems in 1- to 3-year-old children by enhancing maternal sensitivity and adequate discipline strategies (VIPP-SD, see Van Zeijl et al. 2006). It consisted of a screening phase in a general population sample and a pretest–posttest randomized case–control intervention in a subsample of children with scores above the 75th percentile on the Child Behavior Checklist (CBCL) Externalizing Problems scale. Data for the current paper were derived from the screening and pretest phase. Permission for the study was obtained from the Committee for Medical Ethics of Leiden University Medical Centre and the Ethics Committee of the Faculty of Social and Behavioral Sciences of Leiden University.

Participants were recruited from community records of several cities and towns in the western region of the Netherlands. Several exclusion criteria (e.g. twins, serious medical condition in child or mother, single parenthood and non-Dutch cultural background) resulted in a target selection sample of 1954 children. Children with scores above the 75th percentile on the CBCL Externalizing Problems scale (age 1 year: scores ≥13; age 2 years: scores ≥19; age 3 years: scores ≥20) were selected for the intervention study. Of the 438 selected families, parents of 237 children (54%) agreed to participate in the entire intervention study and were invited for a pretest. During a 1.5-h laboratory session, mother and child completed several tasks coded afterwards from videotapes with observational measures, by independent coders, unaware of other data concerning the participants. The average time between the screening and the laboratory session was 4 months. There were no significant differences between selected families who agreed to participate in the intervention phase and those who did not regarding maternal age, child age or gender, presence of siblings and level of child externalizing problems. About 2 years after the intervention, the sample was contacted to take part in the collection of DNA material. Cheek cells were collected from 176 mothers. Mean age of the mothers was 33 years (SD = 4.0), and mean age of the children was 23 months (SD = 10.1) at pretest. Fifty-six per cent of the children were boys, and 60% of the children had siblings. The majority of the parents were well educated.

As data concerned the screening and pretest phases, the intervention and control groups were combined in the analyses. Internal consistency of questionnaire data was assessed in the general population screening sample (= 2408).


Daily hassles

In the screening phase, the mothers were asked to rate 25 indices of potentially stressful events (Kanner et al. 1981). They rated the intensity of hassles they experienced on a 5-point scale for each event (0 = no hassle to 4 = big hassle). Items concerned daily hassles related to life in general (e.g. money problems or trouble at work). Reliability and validity of this scale were shown by Kanner et al. (1981); in our sample alpha reliability was 0.87. The average level of hassles reported in the current sample was similar to that in a non-select community sample of families with 4-year-old children (Pannebakker 2007).

Maternal sensitivity

Mothers’ sensitivity was observed in the laboratory session during a series of problem-solving tasks. Dyads were given three tasks during a total time of 15 min; they were asked to solve puzzles that were too difficult, considering the age of the child (different puzzles were used in each age group), and mothers were instructed to help their child in the way they usually did. Mothers’ supportive presence, intrusiveness and clarity of instruction were rated on 7-point scales described by B. Egeland, M. F. Erickson, J. C. Clemenhagen-Moon, M. K. Hiester, J. Korfmacher, unpublished data, University of Minnesota, Minneapolis. These scales include but also extend the original scales for ‘sensitive responsiveness’ of Ainsworth et al. (1978), developed for parent–infant interaction in the first year after birth. The scales take the wider age range of the current sample into account and measure an age-appropriate concept of sensitivity that also pertains to the developmental domain of coping with cognitive challenges. The average intraclass correlation (single rater, absolute agreement) for intercoder reliability (for all separate pairs of seven coders) was 0.75 (= 30). For overall sensitivity, ratings of the separate tasks were averaged, the intrusiveness scores were reversed and the standardized subscale scores were added.

DNA isolation

Buccal swabs from the mothers were collected in lysis buffer (100 mm NaCl, 10 mm ethylenediaminetetraacetic acid, 10 mm Tris, pH 8, 0.1 mg/ml proteinase K and 0.5% w/v sodium dodecyl sulfate) until further processing. Genomic DNA was isolated from the samples using the Chemagic buccal swab kit on a Chemagen Module I workstation (Chemagen Biopolymer-Technologie AG, Baesweiler, Germany). DNA concentrations were measured using the Quant-iT DNA Assay kit (Invitrogen, Breda, the Netherlands). The average yield was approximately 4 μg of genomic DNA per buccal swab sample.

Polymerase chain reaction amplification and genotyping


The region flanking amino acid position 158 from the COMT gene (Genbank accession number Z26491) was amplified from genomic DNA using primers COMT-F1 (5′-TGGACGCCGTGATTCAGGAG-3′) and COMT-R1 (5′-GGTGGGGAGGACAAAGTGCG-3′). Typical polymerase chain reaction (PCR) reactions were set up using 1–10 ng genomic template DNA, 10 pmol per primer, 5% dimethyl sulfoxide and BioThermAB polymerase (GeneCraft, Munster, Germany). Cycling conditions were as follows: initial denaturation step of 5 min at 94°C, followed by 36 cycles of 30 seconds at 94°C, 60 seconds at 62°C, 60 seconds at 72°C and a final extension step of 4 min at 72°C. Polymerase chain reaction fragments were sequenced using primer COMT-F (5′-ACTGTGGCTACTCAGCTGTG-3′; Eisenberg et al. 1999) and dye terminator chemistry. Sequence reactions were run on an ABI-3730 automated sequencer, and homozygous or heterozygous met/val variants were determined either by manual inspection of chromatograms or by automated sequence analysis using SeqScape software (v.2.1, Applied Biosystems, Foster City, CA, USA). Genotypes (= 48 met/met, = 94 met/val, = 34 val/val) were in Hardy–Weinberg equilibrium, χ2 (2, = 176) = 1.01, = 0.61.

Dopamine D4 receptor

For amplification, primers 5′-GCGACTACGTGGTCTACTCG-3′ and 5′-AGGACCCTCATGGCCTTG-3′ were used (Lichter et al. 1993). For genotyping of the DRD4 gene (Genbank accession number AC021663), exon 3 fragments were amplified by an initial denaturation step of 5 min at 95°C, followed by 38 cycles of 45 seconds at 95°C, 30 seconds at 60°C, 1 min at 72°C and a final extension step of 5 min at 72°C. The number of repeats for each sample was determined by size fractionating the exon 3 PCR products on a 2% agarose gel. The 100-bp DNA ladder (Fermentas UAB, Vilnius, Lithuania) was used for size determination of PCR products. For three mothers, genotyping of DRD4 was not successful. The main genotypes in the sample (2/4, 4/4, 4/7 and 7/7) were in Hardy–Weinberg equilibrium, χ2 (3, = 140) = 3.07, = 0.38. Sixty mothers carried at least one DRD4 7R allele. The presence of DRD4 7repeat alleles was independent of COMT genotype, χ2 (2, = 173) = 3.58, = 0.17.


Bivariate correlations

In Table 1, the means, SDs and bivariate correlations for the main variables are presented. Mothers with a higher educational level appeared significantly more sensitive to their children, but educational level was not associated with daily hassles. Higher scores on parental sensitivity were associated with fewer daily hassles. The presence of the DRD4-7R allele was not associated with any of the main variables (computed as point biserial correlation), whereas the number of met alleles of COMT showed a significant association with daily hassles: mothers with val/met or met/met reported more daily hassles (see Table 1).

Table 1.  Means, SDs and correlations for the main variables
  • *

    < 0.01.

  • **

    < 0.05.

  • = 173, 0 = DRD4-7R absent, 1 = DRD4-7R present.

  • 0 = val/val, 1 = val/met, 2 = met/met.

Child age (months)23.4110.08−0.100.12−0.06−0.130.11
Maternal education3.681.04 0.37**0.000.13−0.08
Maternal sensitivity0.122.26 −0.16*0.090.07
Hassles20.7212.65 0.050.17*
DRD4-7R 0.01

Multivariate regressions predicting maternal sensitivity

In Table 2, the results of the regression analysis on maternal sensitivity are presented. In the first step, child age and maternal educational level were entered. In the second step, daily hassles, DRD4 (DRD4-7R or no DRD4-7R) and COMT (val/val, val/met or met/met) were entered. The predictors of the second step added significantly to the explained variance of maternal sensitivity, Fchange(3, 167) = 2.79, = 0.04. In the third step, the two-way interactions between DRD4, COMT and hassles (centered) were entered (Aiken & West 1991). The three two-way interactions did not contribute significantly to the prediction, Fchange(3, 164) = 1.60, = 0.19. In the final step, the three-way interaction of hassles, DRD4 and COMT was entered. The interaction term contributed significantly to the equation, Fchange(1, 163) = 4.27, = 0.04.

Table 2.  Regression analysis predicting maternal sensitivity from DRD4, COMT and daily hassles, controlling for child age and maternal education
  • = 173.

  • **

    < 0.01.

  • The statistics are derived from the final block of the regression model.

  • 0 = DRD4-7R absent, 1 = DRD4-7R present.

  • §

    0 = val/val, 1 = val/met, 2 = met/met.

Step 1 0.1616.57**
 Age child0. 
 Education mother0.820.150.385.47<0.01 
Step 2 0.208.51**
Step 3 0.235.97**
 DRD4-7R × COMT−0.110.24−0.03−0.470.64 
 DRD4-7R × hassles−0.020.01−0.10−1.270.17 
 COMT × hassles0. 
Step 4 0.255.89**
 DRD4-7R × COMT × hassles0. 

The interpretation of the three-way interaction was facilitated by predicting maternal sensitivity from daily hassles, controlling for child age and maternal educational level, for separate groups with the various combinations of DRD4 and COMT. In the subgroup of parents with a combination of COMT val/val or val/met and the DRD4-7R allele (= 40), a strong association between hassles and maternal sensitivity was found: more daily hassles were associated with less sensitivity of the mothers to their child, and lower levels of daily hassles were associated with more sensitivity, β = −0.49, < 0.01. The effect size for this association was = 1.12, which is large. The association was not significant for any of the other groups (see Fig. 1): in the subgroup with COMT met/met and the shorter variants of DRD4 (= 28), the association between hassles and sensitivity (controlling for child age and maternal educational level) was β = −0.01 (= 0.97); in the subgroup with COMT met/met and the DRD4-7R allele (= 20), β = 0.16 (= 0.41) and in the subgroup with COMT val/val or val/met and the shorter variants of DRD4 (= 85), β = −0.13 (= 0.23). The association between daily hassles and maternal sensitivity (controlling for child age and maternal educational level) in the group of parents with COMT val and DRD4-7R was significantly different from this association in each of the other groups, with values for zdiff ranging from 2.03 to 2.38 (0.01 < < 0.04). Parents with a combination of COMT val and the DRD4-7R allele were more sensitive when experiencing fewer hassles and less sensitive when experiencing more hassles.

Figure 1.

Association between daily hassles and sensitive parenting controlling for child age and maternal educational level. The graph shows a strong negative association between hassles and maternal sensitivity for parents with COMT val/val or val/met and the DRD4-7R allele (β = −0.49, < 0.01) and no significant associations for parents with COMT val/val or val/met without DRD4-7R (β = −0.13, = 0.23), for parents with COMT met/met and the DRD4-7R allele (β = 0.16, = 0.41) and for parents with COMT met/met without DRD4-7R (β = −0.01, = 0.97).


Daily hassles lead to less sensitive parenting, depending on the genetic makeup of the parents involved. We found that parents display less sensitive behavior to their children’s attachment needs when they have to deal with more daily hassles, but only in the group of parents who have a DRD4-7R as well as a COMT val allele. At the same time, in the case of fewer daily hassles, this group shows higher levels of sensitive parenting compared with other parents. Both the DRD4-7R allele and the COMT val allele are associated with a less effectively functioning dopaminergic system. In the other groups of parents, the association between sensitive parenting and hassles is absent. Thus, the combination of COMT val and DRD4-7R appears to imply increased susceptibility to daily stresses in parents, for the better and the worse (Belsky 1997; Belsky et al. in press), leading to more sensitive interaction with their toddlers when they experience relatively few hassles and lower sensitivity when they are faced with more hassles. The fact that sensitive parenting is a well–documented, crucial determinant of young children’s socio-emotional development (Cassidy & Shaver 1999) adds to the relevance of the current findings.

Daily hassles have been conceptualized as the irritating and annoying demands that to some degree are involved in any everyday transaction with the environment (Kanner et al. 1981). Kanner et al. (1981) and Lazarus (1984) found that the accumulation of relatively minor daily hassles as appraised by individuals may have greater impact on their sense of competence, well being and somatic health than major life events. Daily hassles interfere with parents’ sensitivity to their children (Crnic & Low 2002) – although not in all parents. Some parents are more vulnerable to daily hassles than others, but the causes of this differential susceptibility are unknown (Phelps et al. 1998). Here we present the first molecular genetic study addressing this issue. We found 9% of the variance in parental sensitivity explained by COMT, DRD4 and daily hassles. Of this variance, 2% was explained by the interaction between COMT, DRD4 and daily hassles (G × G × E). Evans (1985) states that even moderator effects explaining as little as 1% of the total variance should be considered important.

The greater vulnerability of the DRD4-7R/COMT val group for negative effects of daily hassles on parental sensitivity seems to fit into the so-called diathesis-stress model (Clark et al. 1992; Phelps et al. 1998). In this model, negative outcomes such as insensitive parenting emerge when an underlying vulnerability or diathesis is activated by environmental stressors; without stressful environment, the diathesis remains ineffective and the negative consequences absent. Originally, the diathesis was understood to be some traumatic event or experience with harsh or abusive parenting in childhood, but the model can easily be extended in the direction of biological or genetic vulnerabilities (Belsky 1997, 2005; Boyce & Ellis 2005a,b; Meehl 1962; Paris 2000) related to neurotransmitter systems such as the dopaminergic system.

From inspection of the scatterplot of the G × G × E interaction effect (see Fig. 1), however, it appears that the interaction term requires an explanation that is not covered by the stress-vulnerability model alone. The interaction term between DRD4, COMT and daily hassles appears to consist of two components. The first has already been discussed and relates to parents with the less efficient allele combination being more affected by the stress of daily hassles. The second component pertains to parents with the same genetic combination who are more sensitive to their children than comparison groups when daily hassles are relatively low. The interaction effect thus points to ‘differential susceptibility’ (Belsky 1997; Belsky et al. in press) instead of a stress-vulnerability model in that the genetic ‘risk’ combination may confer greater reactivity to the environment, a reactivity that can be either positive or negative, depending on environmental conditions. Conversely, the combination of ‘non-risk’ polymorphisms may be less responsive both to environmental stressors in the form of high hassles and to environmental support in the form of low hassles, so that the expected relation between hassles and parenting is absent in this group. (We thank one of the anonymous reviewers who suggested this interpretation.)

Animal research suggests that the DRD4 receptor plays an important role in the dopaminergic modulation of the perception of stimuli (Falzone et al. 2002). Biochemical studies showed that the shorter exon III repeats code for a more efficient gene in terms of transcription, translation and second messenger generation compared with the DRD4-7R (Ebstein 2006). Seeger et al. (2004) suggest that the DRD4-7R allele is associated with a reward deficiency syndrome that may lead to sensation and novelty seeking as well as impulsivity. The short variants of DRD4 are related to rigidity and inhibition (Ebstein 2006; Ebstein et al. 2002) or, to put it more moderately, to less distraction by irrelevant stimuli. Parenting in mildly stressful environments might be more effective if interference and distraction by child-independent stimuli from daily hassles is blocked or at least decreased.

Catechol-O-methyltransferase affects prefrontal functioning and working memory and has also been associated with emotional dysregulation. Greater connectivity of the amygdala and the hippocampus with the orbitofrontal cortex has been shown for met/met subjects (Drabant et al. 2006). This enhanced connectivity is associated with low levels of novelty seeking, with some inflexibility in processing affective information and with (affective) perseverance. Here we suggest that in normally functioning individuals with the COMT met/met genotype, the predisposition to focus their attention on relevant stimuli and to inhibit interference from other stimuli (Bilder et al. 2004; Drabant et al. 2006) may help to keep their attention focused on the child’s attachment signals and needs in a mildly stressful context of daily hassles. At the same time, more COMT met alleles were also associated with the parent’s report of more daily hassles. Thus, the finding that in the groups with COMT met the influence of hassles on parenting was absent cannot be ascribed to their experiencing fewer hassles. It also points to the possibility that COMT met may not always be functioning as a protective factor and COMT val as a risk or vulnerability factor (Drabant et al. 2006).

It should be noted that DRD4 and COMT did not predict parental sensitivity separately but only in interaction. In his seminal review on the molecular genetic architecture of human personality, Ebstein (2006) proposed to go beyond single-gene association studies of personality and to consider various combinations of genetic polymorphisms (e.g. see Ebstein 2006). Here we focused on two genes implicated in the dopaminergic system, rather than an exclusive focus on only one dopamine-related gene that in itself appeared to be unrelated to parenting (Gervai et al. 2007). Ebstein (2006) also argued that molecular genetic studies should go beyond the self-report of personality characteristics and that a focus on observed or experimentally manipulated features of a person’s functioning is necessary to decrease error variance and to enhance the validity and precision of the phenotypic measures. In the current study, we addressed the question whether observed differences in sensitive parenting were related to genetic differences in situations of daily stress. Although we were not able to observe the daily hassles in the lives of the parents and had to rely on their own reports, we confirmed Ebstein’s (2006) proposition that the environment interacting with genetic makeup should be taken into account when explaining human (parenting) behavior (e.g. Bakermans-Kranenburg & Van IJzendoorn 2007; Caspi & Moffitt 2006; Caspi et al. 2002; Rutter 2006). It should be noted that according to Lazarus (1984), the subjective appraisal of daily hassles is more important than the objective count of daily strains and stresses for the way in which hassles impact on the individual.

Increasingly, the study of gene–environment interaction, the interplay between ‘nature’ and ‘nurture’ deriving from genetic effects on susceptibility to environmental risks has been moving to the forefront (e.g. Bennett et al. 2002; Caspi et al. 2002, 2003; Fox et al. 2005; Kaufman et al. 2004). As shown by Suomi (1999) for primates and by Caspi et al. (2002, 2003) for humans, an individual’s response to environmental influences may be moderated by his or her genetic makeup. In our own series of studies on children, a moderating role of the DRD4 gene was found. Maternal unresolved loss or trauma was associated with infant attachment disorganization, but only for children with the DRD4-7R allele (Van IJzendoorn & Bakermans-Kranenburg 2006). In a related study on toddlers’ externalizing behavior, maternal insensitivity was associated with externalizing (oppositional or aggressive) behaviors, but only in the presence of the DRD4-7R polymorphism in the child (Bakermans-Kranenburg & Van IJzendoorn 2006). In the current study, we focused on parents and found a similarly important role for the parental DRD4 gene, in combination with COMT, also pointing at the greater susceptibility for environmental stresses of carriers of the DRD4-7R variant.

Our findings on the role of DRD4 and COMT should be independently replicated. Multiple gene effects are based on an increasingly large number of statistical tests and are thus liable to false positives (Ebstein et al. 2002). Moreover, the number of participants with particular genotype combinations and the predicted phenotype becomes increasingly small (Van Gestel & Van Broeckhoven 2003). In our study, the risk of capitalization on chance was limited because we included only two genes, and the smallest subgroup for the combination of polymorphisms of the two genes was = 20. Furthermore, we enhanced the power to find replicable outcomes by careful observational assessment of parenting, and we included assessment of the (risk) environment (Caspi et al. 2003; Rutter 2006; Rutter et al. 2006) with a validated standard questionnaire. Better measurement of the environment is, however, crucial (Luan et al. 2001; Rutter 2006; Wong et al. 2003), and in future studies our self-report assessment of daily hassles may be replaced by observational or even psychophysiological measures. Finally, the families in the current study had an externalizing toddler (75th percentile or above on the CBCL). Therefore, our findings may only apply to parents who perceive their children as difficult and non-compliant and who already at an early stage have difficulty managing their children. It is important to note, however, that the families in our study were two-parent families from predominantly middle-class background and without psychiatric disorders. The level of their daily hassles was similar to a non-select community sample of parents with 4 year olds. Nevertheless, replication in unselected samples is needed.

To our knowledge, the current study is the first to find a molecular genetic basis of observed human parenting, although parenting has been the subject of numerous studies in the past decades, resulting in the multivolume Handbook of Parenting (Bornstein 2002) and several specialized journals (e.g. Parenting). Because of the intriguing findings on the molecular genetic basis of parenting in studies on rodents (e.g. Meaney 2007; Meaney & Szyf 2005) and on non-human primates (e.g. Suomi 1999), it is time to address similar issues in the parenting of human offspring. Sensitive parenting is a complex activity in which affective information must be perceived and processed and a course of action chosen from a variety of competing activities. In the somewhat demanding and mildly stressful conditions of daily life, the ability to remain focused on the child’s signals and needs seems crucial for sensitive parenting. We showed that dopamine-related genetic differences may play a role in sensitive parenting. Molecular genetic studies promise to elucidate differential susceptibility of parents to environmental stressors in taking sensitive care of their children.


This study is part of the research project SCRIPT, conducted at the Centre for Child and Family Studies, Leiden University, the Netherlands. The study is supported by grant 2200.0097 from ZonMw (Netherlands Organization for Health Research and Development) to M.H. van IJ. and Femmie Juffer. Support from the Netherlands Organization for Scientific Research NWO (VENI, VIDI and SPINOZA Prize) to J.M., M.J.B.-K. and M.H. van IJ. respectively, is also gratefully acknowledged. We thank Drs Jantien van Zeijl, Mirjam Stolk, Lenneke Alink, Femke Pijlman, Femmie Juffer and Hans Koot for their contributions to the various parts of the study and Base-Clear for the genotyping. Last but not least, we thank the parents and children who participated in our study as well as the students who assisted in various phases of the SCRIPT project.