Temperament development in infancy: What we have learned about the origins of individual differences in the past 25 years

Abstract

The goal of this review article is to provide an overview of the last quarter century of infant temperament research. Beginning with a brief description of theoretical underpinnings, we focus on studies addressing early beginnings of reactivity and self-regulation. After the most widely accepted framework used to conceptualize temperament has been defined, the next section examines various methodologies for measuring temperament in the past 25 years. We then focus on work that aims to describe developmental changes and progression of temperament development. The following sections consider individual and contextual contributing factors, such as infant sex and cultural influences. Finally, we review infant temperament as a predictor of important child outcomes, focusing on behavior problems/symptoms. Overall, the past 25 years temperament research has increased our understanding of important developmental changes in different domains of reactivity and regulation, emphasizing biological underpinnings, such as underlying brain activity, as well as contributing factors (e.g., genetic/epigenetic contributions). The review ends with a discussion of remaining gaps in research and recommendations for future research, such as the need to harmonize datasets across laboratories to leverage latest quantitative methods resulting in reproducible science.

Defining infant temperament: Leading theoretical framework of the past 25 years

The psychobiological theory of temperament is most cited in current temperament research. This approach conceptualized temperament as constitutionally based individual differences in emotional, motor, and attentional reactivity, and self-regulation, demonstrating consistency across situations and relative stability over time (Rothbart et al., 2000; Rothbart, 2007). Reactivity refers to the latency, rise time, and duration of response to stimulation and self-regulation encompasses processes involved with modulating reactivity (Rothbart & Bates, 2006). Attentional functions (e.g., allocation/shifting) were conceptualized as foundational to emerging self-regulation (Gartstein et al., 2016). More broadly, Rothbart’s psychobiological model of temperament emphasizes the connection between temperament and biology in terms of underlying neurobehavioral systems and genetic contributions (including genetic and epigenetic effects) and has provided a foundation for a variety of studies focused biological underpinnings and correlates of temperament. Moreover, this theoretical framework offers a hierarchical definition of temperament in early childhood, including overarching domains of negative emotionality, surgency/positive affectivity, and regulatory capacity/effortful control, with related measurement tools.

Measurement approaches dominating infant temperament research in the last 25 years

Overall, parent-report surveys represent the most frequently used assessment method for children (Rothbart & Bates, 2006). The Infant Behavior Questionnaire-Revised (IBQ-R; Gartstein & Rothbart, 2003), designed for infants 3 to 12 months of age, and the Early Childhood Behavior Questionnaire (ECBQ; Putnam et al., 2006), for toddlers 18–36 months old, based on the psychobiological approach, have been widely used in the last 25 years. These instruments owe their widespread use in part to availability of shorter forms, more readily incorporated into otherwise demanding protocols (Putnam et al., 2014). For the IBQ-R, the original instrument includes 191 items which form 14 scales, shown to cluster into over-arching factors (Bosquet et al., 2016; Gartstein & Rothbart, 2003; Peterson et al., 2017). The short form includes 91 items, otherwise maintaining the scale and factor structure, whereas the very short form consists of 37 items and only yields three overarching factors: Negative Emotionality, Surgency/Positive Affectivity, and Regulatory Capacity/Orienting.

Laboratory tasks designed for earlier infancy tend to focus on reactivity to stimuli, measuring distress proneness or irritability, soothability, activity level and alertness, later incorporating ability to orient, interest/persistence, fear/avoidance, and joy/ pleasure, all becoming more prominent in the second half of the first year of life (Gartstein et al., 2016). The Laboratory Temperament Assessment Battery (Lab-TAB; Gagne et al., 2011; Planalp et al., 2017) offers a wide-ranging standardized protocol addressing multiple domains of reactivity and emerging regulation. Starting in toddlerhood, effortful control becomes the focus of several laboratory tasks. The ability to suppress a dominant response in favor of a sub-dominant yet a more adaptive one is often studied by using delay of gratification tasks (e.g., “snack delay” or “marshmallow test”; Conway, 2020; Kochanska et al., 2000; Luo & Pattanakul, 2020; Peake, 2017). Other laboratory observation tasks target behavioral inhibition (related to fear), such as via exposure to unfamiliar objects or people, or opportunities to engage in “risky” activities (Barker et al., 2014; Hendry et al., 2022).

In the last 25 years, temperament researchers have increasingly relied on physiological measures, focusing on biological processes underlying temperament development. Interest in the Autonomic Nervous System has grown, particularly the complex interplay of the sympathetic (SNS) and parasympathetic (PNS) nervous systems, with the latter serving to modulate arousal, indexed by heart rate variability or vagal tone. High baseline vagal tone (commonly derived from respiratory sinus arrythmia; RSA), has been linked to better regulation and soothability in infants (Calkins & Fox, 2002) as well as greater readiness to approach strangers, regulation in frustrating situations, and lower levels of aggressions in toddlers (Calkins & Dedmon, 2000). The “vagal break”, regulation in the form of decreased vagal tone across experimental manipulations, was associated with greater attentional control and self-soothing behavior in infants (Calkins et al., 2007). More recent research has examined vagal tone/RSA in the context of parent/child interactions and bidirectional effects. For example, Porter and colleagues (2022) found that lower vagal tone was associated with higher levels of unilateral (i.e., mother-led) co-regulation in mother-infant dyads. A faster increase in infant vagal tone was associated with a shift towards symmetrical patterns of co-regulation, with infants making a greater contribution. Taken together, these findings show that higher levels of infant vagal tone may allow them to engage in more complex and effective means of emotional regulation (i.e. symmetrical patterns of co-regulation). In regard to SNS, children higher in “surgency”, exhibiting greater approach tendencies to novelty, sensation seeking and impulsivity, showed better emotion regulation skills when pre-ejection period (time between the beat of the heart and ejections of the blood into the aorta) baseline and reactivity scores were higher (Stifter et al., 2011).

Cortisol level markers represent another set of psychobiological measures utilized in the past quarter century of infant temperament studies. Cortisol can be described as the primary hormone resulting from the hypothalamic-pituitary-adrenocortical (HPA) axis activity, often referred to as the “stress hormone”. Cortisol levels are typically measured in saliva to study activation of HPA throughout the day and in response to environmental stimuli, often related to behavioral responses. Cortisol concentrations vary throughout the day following a rhythmic pattern (i.e., diurnal pattern) and are also reactive to external stressors (Lengua et al., 2024). Positive associations between shyness and higher levels of cortisol were found in several studies (e.g. Fortunato et al., 2008; Perez-Edgar et al., 2008; Watamura et al, 2003), while moderate levels were associated with approach behavior and greater cognitive and behavioral inhibitory control (Blair, Peters & Granger, 2004). Salivary alpha-amylase (sAA) was used to measure SNS activity related to temperament, with individual differences in sAA levels positively associated with infant approach behavior and positive affect in 24-month-olds (Fortunato et al., 2008).

In the last 25 years there has been considerable progress in the field of temperament and sleep research relying on actigraphy, wherein a device records infants’ physical activity from which information on infant sleep and awakenings is derived, providing an objective measure of sleep quality/quantity. Spruyt and colleagues (2008) gathered infant sleep data using a parental sleep diary and actigraphy demonstrating that “easy” temperament, characterized by greater approachability, rhythmicity, adaptability, and lower distractibility, was associated with greater amounts of sleep. Jian & Teti (2016) examined infant temperament moderation of links between maternal emotional availability at bedtime and infant sleep development. Surgency moderated the association between maternal emotional availability and infant sleep duration - surgent infants whose mothers were emotionally available at bedtime showed significantly longer night sleep duration.

Measures of neural activity using electroencephalogram (EEG) have also become of considerable interest to infant temperament researchers, with differences in frontal left and right hemispheric activity in the Alpha band linked with lateralization of emotions/motivation. Relative right frontal activation is linked with withdrawal behaviors and emotions (e.g., behavioral inhibition, fear), and left hemisphere activation associated with approach and positive affect (Hane et al., 2008). Infants exhibiting relative left frontal activation expressed more positive affect and more readily approached an unfamiliar experimenter during a playful episode (Hane et al., 2008). Conversely, right frontal EEG asymmetry was associated with behavioral manifestations of fear (e.g., distress vocalizations, attempts to escape) during the presentation of unfamiliar masks and fear during stranger approach (Diaz & Bell, 2011).

Differences in temperament have been studied with functional (fMRI) and structural magnetic resonance imaging (fMRI), focusing on differences in neural activation and brain structure. For example, amygdala functioning was related to fear/reactivity to novelty, and the dopamine transmitter system linked with extraversion and surgency (Rothbart, 2007). In a seminal study, Schwartz et al. (2003) demonstrated differences in amygdala activity among adults who had been categorized as uninhibited vs. inhibited as toddlers. Those in the inhibited group showed greater functional MRI signal response within the amygdala to novel versus familiar faces, compared to those uninhibited as children. In a follow-up fMRI study (Schwartz et al., 2010) differences in thickness in the orbitofrontal cortex were also identified, wherein lateral differences in cortical thickness were linked with infant temperament variability. Greater left frontal cortical thickness was observed in adults categorized as low in reactivity in the first year of life, whereas adults previously categorized as highly reactive showed greater cortical thickness in the right frontal region. Leveraging fMRI, Ellis and colleagues (2021) found that brain regions involved in attention, supporting regulation, in infants are similar to those recruited in adulthood. That is, the basal ganglia and the anterior cingulate cortex were similarly activated during an attentional orienting task in the first year of life. Filippi et al. (2021) examined how infant amygdala functional connectivity related to observations of temperament in 4-months-old, reporting that greater amygdala-superior frontal gyrus and amygdala-cingulate connectivity was associated with lower positive affect in infants.

Developmental changes in temperament across early childhood

Although temperament has been described a set of relatively stable individual differences (Gartstein et al., 2016), research conducted in the past 25 years has demonstrated considerable changes, especially in infancy, coinciding with the generally rapid pace of development. Multiple studies examined temperament growth in samples of healthy/typically developing babies across the first year of life (Braungart-Rieker et al., 2010; Bridgett et al., 2013; Gartstein et al., 2010; Gartstein et al., 2018; Gartstein & Hancock, 2019; Groh et al., 2017). Gartstein et al. (2010) found increases in fear across infancy, especially rapid at the end of the first year of life, based on parent-report and laboratory observations. Braungart-Rieker et al. (2010) also reported increases in fear, as well as anger across infancy, in response to structured laboratory episodes. Across infancy, fear showed significant increases captured by a linear slope, along with some deceleration, reflected in a quadratic effect. A linear trajectory emerged in assessing growth for smiling/laughter measured via parent-report, with increases across the first year of life (Bridgett et al., 2013). Gartstein et al. (2018) found that a linear trajectory was optimal for parent-reported smiling/laughter and fear, both increasing after 6 months of age. Results of these growth-focused investigations demonstrate considerable changes for several domains of temperament across infancy. Importantly, studies linking these developmental trajectories with distal outcomes suggest the patterns of change are meaningful insofar as their parameters were predictive of emerging symptoms/behavior problems and parenting in early childhood (Bridgett et al., 2013; Gartstein et al., 2010).

Individual factors contributing to reactivity and regulation in early childhood

The psychobiological model, as well as other theories of temperament (Gartstein et al., 2016), have emphasized biological underpinnings of individual differences, traditionally interpreted as genetic contributions. Planalp and Goldsmith (2020) found that temperament similarity was almost universally higher for monozygotic twins than dizygotic twins. The observed genetic effect was strongest for infants with withdrawn/inhibited profiles, in line with previous research demonstrating more significant genetic contributions for negative facets of temperament compared to positive ones (e.g., Clifford et al., 2015; Planalp et al., 2017). Focusing on genetic mechanisms, Zwir and colleagues (2020) found 51 sets of single nucleotide polymorphisms (SNPs) that identified 736 genes explaining 48% of the variability in temperament in a sample of 2149 healthy Finnish children. More recently, epigenetic influences on temperament development have been described (Gartstein & Skinner, 2018). Research with infants has focused on candidate genes, most frequently NR3C1, the glucocorticoid receptor (GR) gene that codes for a protein involved in modulating the hypothalamic-pituitary-adrenal (HPA) axis, as well as SLC6A4, the serotonin transporter gene encoding a protein responsible for transporting the neurotransmitter serotonin back into the neurons. For instance, SLC6A4 methylation was linked with infant soothability in the context of in utero Selective Serotonin Reuptake Inhibitor (SSRI) exposure (Gartstein et al., 2016). Higher neonatal methylation was associated with increased soothability from 3 to 6 months among infants of depressed mothers prenatally exposed to SSRIs, interpreted to suggest that prenatal SSRI exposure may confer a developmental benefit. With respect to NR3C1, Ostlund and colleagues (2016) reported that exploratory sex-specific analyses revealed an association between prenatal stress and increased methylation of NR3C1 exon 1F for girls, but not boys, in infancy. Increased methylation was also significantly associated with greater fearfulness for girls, suggesting an experience-dependent pathway to fearfulness for girls via epigenetic modification of the glucocorticoid receptor gene. Fuemmeler et al. (2016) examined DNA methylation status of differentially methylated regions (DMRs) for nine imprinted genes, wherein expression is determined by its parental origin – only one of the two copies (the one inherited from the mother or the father) is active. Higher methylation levels at the intergenic MEG3-IG regions were associated with greater surgency whereas PEG3 methylation was positively associated with negative affectivity. It should also be noted that prenatal effects, for example those resulting from maternal depression, trauma and stress exposure, are thought to be transmitted to the offspring via epigenetic mechanisms, contributing to temperament (Gartstein & Skinner, 2018).

Infant sex is another individual attribute consistently associated with temperament and implicated in temperament differences. Fewer differences between boys and girls are found in the first year of life, becoming more evident in toddlerhood (Else-Quest et al., 2006). In infancy, boys were described as higher in activity level and approach behaviors (Gartstein & Rothbart, 2003; Gagne et al., 2013), whereas girls exhibited greater novel object approach hesitation, assessed in multi-cultural studies, measured via laboratory observations and parent report (Cosentino-Rocha et al., 2014; Gartstein et al., 2016; Gartstein & Rothbart, 2003). With recent advances in quantitative methods, research questions about differences between boys and girls have been approached using machine learning techniques. Specifically, algorithmic modeling techniques were used to discern the extent to which the 14 IBQ-R subscale scores accurately classified participating children as boys and girls, as well as into age groups, using a large representative sample of infants (N = 4438; Gartstein et al., 2022). Infant sex-based classification was superior in the oldest age group (>48 weeks of age), suggesting temperament differences between boys and girls are accentuated with development.

Contextual factors influential in temperament development

Parent-child interactions and parental characteristics are chief among contextual factors relevant to temperament development, especially in early childhood, as these represent a critical aspect of the infant’s social milieu (Lengua et al., 2024). Parent–child interactions were shown as critical contributors to fear/reactivity to novelty (Buss & Kiel, 2013; Fox et al., 2001), parental sensitivity/responsiveness in particular. Infants with more sensitive mothers were slower to increase in fear between four and sixteen months of age (Braungart-Rieker et al., 2010). Maternal sensitivity also served a protective function relative to fear and right frontal EEG activation early in infancy (e.g., Gartstein et al., 2018; Hardin et al., 2021). Reciprocity was associated with slower increases in fearfulness across the first year of life, and tempo of interactions with lower levels of fear at four months. Reciprocity and positive emotional tone/warmth of mother–infant interactions predicted positive affectivity in four-month-old infants, and tempo was linked with more rapid increases in positive affectivity across infancy (Gartstein et al., 2018). Hane and Fox (2006) examined high quality maternal caregiving, operationalized as a sum of acceptance, sensitivity, availability, and appropriateness of pace in feeding ratings, with intrusiveness score subtracted. These authors reported associations with greater left frontal alpha activation (associated with approach tendencies, positive affect, and more effective regulation), less fearfulness, more joint attention, and less negative affect. Low quality caregiving was associated with the opposite pattern of results.

Gartstein (2020) examined infant EEG asymmetry in the context of the Still Face Procedure, reporting that higher levels of intensity/stimulation in mother-child interactions were associated with greater left frontal activation for infants high in surgency/approach. This interaction was interpreted as evidence for goodness-of-fit, wherein more surgent/approach-oriented infants demonstrated a more adaptive asymmetry pattern when their mothers’ play was characterized by greater stimulation. Diaz and colleagues (2019) conducted a study examining infant EEG asymmetry, toddler negative affect, and maternal parenting behaviors. An interaction between infant frontal asymmetry and maternal behaviors was noted, such that sensitive maternal behaviors were associated to lower levels of toddler negative emotionality only for infants displaying relative left frontal activation. Maternal intrusiveness was related to higher levels of toddler negative affect only when they displayed right frontal asymmetry as infants.

Interactions between the quality parent-child exchanges and temperament in infancy have also been studied with respect to more distal outcomes (e.g. childhood obesity). In a review of parent-child interaction, self-regulation, and obesity prevention, Anderson and Keim (2016) reported that low maternal sensitivity in mother-child interactions at 6 months old was a significant predictor of higher weight status and increased risk for obesity through middle childhood (age 12). Importantly, difficult child temperament (i.e., distress proneness) exacerbated the effects of low maternal sensitivity increasing obesity risk.

Maternal depression and to a lesser extent anxiety have also been associated with infant temperament, with these effects largely understood as indirect, conferred by mother-infant interactions (Ierardi et al., 2019), although the latter is primarily an interpretation rather than a test of mediational pathways. Higher severity of maternal depressive symptoms predicted steeper increases in fearful reactivity (Gartstein et al., 2010). Gartstein and Hancock (2019) reported that anxiety symptoms were associated with decreases in approach and perceptual sensitivity (i.e., attention to subtle stimuli in the environment) scores, as hypothesized, whereas maternal depression predicted increased approach tendencies during the same mid-infancy period. Thus, maternal symptoms of anxiety and depression played different roles with respect to development of surgent tendencies and the authors suggested that infants of mothers reporting higher levels of depression may engage in more approach behaviors in an attempt to compensate for mothers’ withdrawal, consistent with earlier findings (e.g., Pelaez, Field, Pickens, & Hart, 2008). Greater maternal effortful control and extraversion were associated with more initial infant smiling and laughter, and were interpreted as playing a protective role, whereas parenting stress predicted decreases in positive affectivity (Bridgett et al., 2013).

Temperament researchers have become increasingly interested in the effect of maternal depression on infant EEG asymmetry, associated with temperamental approach/positive affectivity and avoidance/negative emotionality. Wen and colleagues (2017) linked maternal sensitivity and postnatal depression to relative right frontal EEG activation in infants classified as belonging in a “high mother time spent” group (defined as spending more than 50% of the time with their mother). Lower levels of maternal sensitivity and greater maternal depression severity predicted relative right frontal dominance at 6 months. Wen et al. (2017) also found that greater relative right frontal activation at 6 months was associated with higher infant negative emotionality at 12 months. Goodman et al. (2019) conducted a study with mother-infant dyads, recruiting mothers with a documented history of depression and measuring infant frontal EEG asymmetry across five contexts (e.g. peekaboo, play, feeding). Mothers’ prenatal depressive symptoms, rather than postnatal or concurrent, were associated with infants’ consistent right frontal activation across contexts, controlling for infants’ observed affect. Thus, maternal depression during pregnancy was conducive to a trait-like presentation of right frontal dominance, potentially increasing offspring risk for developing their own symptoms later.

More macro environmental factors, such exposure to adversity and chaotic home environments, were also shown as influential in temperament development. Wiseman et al. (2021) utilized data from the Avon Longitudinal Study of Parents and Children (ALSPAC) to study links between adverse childhood experiences (ACEs), infant temperament, and post-traumatic stress disorder (PTSD) in adulthood. Temperamental traits related to more intense reactivity/”difficult” temperament and higher activity level were associated with an increased probability of ACEs (e.g., trauma exposure). Trait intensity moderated the relationship between ACES and the development of PTSD, such that individuals higher in intensity were more likely to display PTSD symptoms given traumatic exposure(s). Maternal ACEs indirectly predicted poor self-regulation in early infancy via prenatal internalizing symptoms, with anxiety/depression serving as mediators (Mattera et al., 2022). Early household chaos was also related to subsequently lower orienting/regulatory abilities, especially for girls (Petrenko et al., 2019). Vernon-Feagans et al. (2012) reported associations between household chaos (instability/disorganization) and infant distress proneness.

Culture and related effects

Cross-cultural variability in early temperament development has been consistently reported over the past 25 years (e.g. Chen, 2018; Gartstein et al., 2006; Gartstein et al., 2005; Kirchoff et al., 2019). Culture encompasses a set of attitudes, values, goals, and practices of a group that provide a shared approach to family life, and shared experiences of a cultural group have a profound influence on individual development (Bornstein, 2013). East-West comparisons have been frequently reported in the literature, for example, Chinese and South Korean toddlers were found to be more inhibited compared to Australian and Italian toddlers in a study by Rubin et al. (2006) comparing five countries (including Canada). Individualism/Collectivism distinctions have played a major role in interpretations of East-West temperament differences, wherein eastern cultures are thought to embrace more collectivistic values, defined as emphasizing group over individual needs, whereas western cultures are generally viewed as more individualistic, prioritizing personal freedom and interests. Individualistic cultural orientation has often been associated with more approach/positive affectivity related traits and/or lower levels of negative emotionality, whereas collectivism has been linked with more avoidant tendencies and negative emotionality as well as less prominent approach/positive affectivity (Desmarais et al., 2021; Putnam & Gartstein, 2017).

More recent temperament studies compared a larger number of countries. A meta-analysis of temperament data from 18 countries (Putnam & Gartstein, 2017) indicated a consistent pattern for the three overarching factors of the ECBQ, with higher levels of negative emotionality and lower levels of surgency and effortful control in East Asian compared to Northern European cultures. Joint Effort Toddler Temperament Consortium (JETTC, Gartstein & Putnam, 2018) compared toddler temperament from 14 different countries around the world, confirming documented differences between Eastern and Western cultures, with higher levels of negative emotionality in toddlers from Asian countries (China, Korea, Turkey), low levels in the US and Western European toddlers (Belgium, Finland, Netherlands, Italy), and mid-levels of negative emotionality for Eastern European countries (Russia and Romania). Finnish and Belgium samples were rated highest on surgency, and significantly different from Korean, Turkish and Chinese toddlers, with Russian and Romanian toddler scores in the mid-range.

Global Temperament Project (GTP) by Putnam and colleagues (2024) leveraged data from 59 countries, providing critical insights into cross-cultural variability of early childhood temperament. Globally, higher levels of surgency were characteristic of cultures that place an emphasis on short-term, as opposed to long-term, goals. Lower levels of negative emotionality were more common in wealthy, individualistic countries, primarily in Northern and Western Europe. Conversely, higher levels of negative emotionality were found in Southern Asia and South America. Cross-cultural differences in negative emotionality were substantially greater compared to surgency or regulatory capacity, indicating this domain of temperament may be more strongly influenced by cultural factors. Putnam and colleagues proposed that social acceptability of negative emotions may influence how parents respond to such displays from their children, either by rewarding or discouraging them. Societal views of negative emotions can also impact how parents perceive the frequency, duration, and intensity of child negative emotions, potentially impacting parent-report.

Infant temperament as a predictor of behavior problems: Contributions of biology and culture

Researchers’ interest in temperament development has increased over the past quarter century in part because early reactivity and regulation are predictive of critical outcomes, behavior problems/symptoms chief among these. Infant and toddler temperament were shown to predict emerging internalizing and externalizing symptoms (Gartstein et al., 2012). Specifically, higher negative emotionality and lower effortful control were linked to both externalizing and internalizing symptoms. Externalizing behaviors were associated with all aspects of effortful control in toddlerhood, with particularly strong effects observed for inhibitory control, both concurrently and longitudinally. Higher surgency was associated with greater externalizing behavior problems, whereas low surgency increased the likelihood of internalizing symptoms. Trait-by-trait moderation involving negative emotionality and regulatory capacity/effortful control was also reported, wherein greater reactivity and poorer regulation resulted in more severe behavioral/emotional problems.

With respect to fine-grained aspect of temperament, perhaps the most consistent findings have been reported for links between fear, behavioral inhibition, and associated right frontal alpha activation, with anxiety and internalizing symptoms/disorders. Infant fear and behavioral inhibition have been consistently demonstrated as risk factors for anxiety, with those who show stable patterns more likely to develop anxiety disorders later (Clauss & Urbano Blackford, 2012). Toddlers reacting with high fear to low threat situations (e.g., display of unfamiliar masks), described as dysregulated fear, were at greater risk for social anxiety (Buss et al., 2013). Fewer studies focused on internalizing symptoms/anxiety and frontal EEG asymmetry in early childhood, yet connections have been demonstrated (Smith & Bell, 2010). Stability in frontal EEG asymmetry (10 and 24 month) was linked to maternal rating of behavior problems. Infants with stable right frontal EEG asymmetry were rated higher in internalizing behaviors.

Externalizing/disruptive behavior problems have also been linked with behavioral and neurophysiological temperament measures. Degnan et al. (2011) conducted Latent Profile Analysis (LPA) relying on behavioral measures of exuberance collected between and 36 months, when frontal EEG asymmetry was assessed. LPA revealed a high, stable exuberance group, and membership in this profile was associated with greater externalizing/disruptive behavior and lower social competence later. These effects were particularly strong for children who displayed left frontal EEG asymmetry – the neurophysiological signature of temperamental approach. Links between left frontal activation in infancy and emerging externalizing symptoms were also demonstrated by Smith and Bell (2010).

Whereas fearfulness and fear sensitivity have been often considered maladaptive, contributing to risk for later anxiety and depression (Fox et al., 2021; Sandstrom et al., 2020), Grossman (2023) proposed an alternative evolutionary-developmental framework – the fearful ape hypothesis. According to this framework, fearfulness and sensitivity to fear may be adaptive in the context of strong interdependence and cooperative care, as fearfulness may facilitate care-based responding from members of the community, not just mothers. Grossman (2023) noted the human fear paradox, wherein fearfulness represents a risk factor for internalizing disorders, yet has also been linked to greater sensitivity and accuracy in detecting fear in others, as well as more cooperative behaviors (Grossman, 2018; Hrdy & Burkart, 2020; Marsh, 2015; Warneken, 2015). This paradox may be due to the research focus on Western, educated, industrialized, rich, and democratic (WEIRD) samples of convenience. WEIRD societies place an emphasis on independence, which impacts socialization goals, educational and clinical practices (Chiao & Blizinsky, 2010; Tsai, 2017). Therefore, what traits are considered “adaptive” and “healthy” are largely governed by cultural values and norms. Grossman (2023) acknowledged clinical significance of extreme fearfulness in the etiology of internalizing disorders in WEIRD samples (Henderson et al., 2015; Sandstrom et al., 2020), encouraging the field to take evolutionary, developmental, and cultural factors into account when researching fear, anxiety and depression.

The human fear paradox represents one example of how biological dispositions and culturally shaped environments co-act to influence temperament development. These tandem effects are likely bidirectional, as biological predispositions shape cultural values and vice versa. That is, aggregate characteristics of the population, in part a function of genetic composition, influence cultural priorities in socialization/child-rearing, which in turn influence temperament development. As genetic effects are subject to natural selection, culturally influenced environmental pressures enhance the survival and reproductive fitness of those possessing genes that give rise to behavioral phenotypes (i.e., temperament) advantageous under certain conditions (Putnam & Gartstein, 2017). Genetic effects, foundational to individual differences in reactivity/regulation, can be expected to shape cultural parameters, for example, with respect to differential consideration of threats/rewards related to most critical societal concerns (Chiao & Blizinsky, 2010; Way & Lieberman, 2010).

Remaining gaps in research and future directions

Over the past 25 years temperament research has been primarily driven by the psychobiological model of temperament (Gartstein et al., 2016) and has focused on demonstrating developmental trajectories, individual and contextual contributing factors, links between temperament and subsequent behavior problems/symptoms, as well as biological markers of reactivity and regulation. While significant advances have been made in all these areas, important research questions remain. For example, parallel changes in behavioral and biological indicators of temperament across infancy and early childhood have not been examined and should be considered in the future because shifts occurring in tandem (e.g., increases in fearfulness and relative right frontal alpha activation) may be most consequential for the development of behavior problems/symptoms. A great deal more can be learned about prenatal influences transmitted via epigenic mechanisms relying on genome-wide approaches. That is, recent molecular genetic evidence has cast doubt on the candidate gene approach (e.g., Border et al., 2019). Complex genetic interactions with genes critical to fetal neurodevelopment playing a role in the origins of psychopathology have been implicated - not only do multiple genes have synergistic effects as networks, but a single gene can have multifaceted effects on psychological symptoms or disorders. Pleiotropic (i.e., associated with more than one disorder) loci within genes that show heightened expression in the brain beginning in utero play prominent roles in neurodevelopmental processes and have also been implicated in disorders such as major depression (Cross-Disorder Group of the Psychiatric Genomics Consortium, 2019), thus should be considered with respect to temperament.

Mediational models incorporating parental characteristics (e.g., depression/anxiety symptoms and disorders) and different aspects of parent-infant interactions should also be examined in infancy/toddlerhood to complement existing studies with older children (e.g. Chung, Lanier, & Wong, 2022; Van Dijk et al., 2020). In addition, the study of temperament should continue to leverage advanced quantitative techniques, such as machine learning, to provide more robust answers to lingering research questions. Because these techniques rely on large samples, typically larger than what can be collected in a single laboratory, multi-site studies and broader efforts to harmonize datasets across laboratories are required. There is also a need for cross-cultural, longitudinal, multi-method studies that integrate biological, behavioral, and environmental measures to track temperament development globally, discerning universal and culture-specific effect. Lastly, much of early childhood temperament research has utilized a risk-focused approach with respect to predicting later psychopathology. Strength-based/resilience-oriented work is well underway with older children and adolescents (Zimmerman, 2013), yet in early childhood have been largely limited to research addressing the protective role of effortful control/self-regulation (Eisernberg et al., 2004). Future directions of research should involve approaching temperament predictors of developmental psychopathology from a strengths-based framework (see Kim-Cohen, 2007 and Masten et al., 2021) to provide a more holistic view of its origins.

Highlights.

• Temperament research in the last 25 years has relied on the psychobiological theory

• Parent-report, observations and physiological indices/biomarkers served as measures

• Studies addressed temperament development, contributing factors, and outcomes

• Emphasis on biological underpinning and correlates yielded important findings

• Future directions include additional areas of study and data harmonization efforts