Prenatal testosterone exposure is associated with delay of gratification and attention problems/overactive behavior in 3-year-old boys

Highlights

• Girls are better able to delay gratification than boys.

• Boys with higher prenatal testosterone levels are less able to delay gratification.

• Boys with higher prenatal testosterone levels show more attentions problems/overactive behavior.

• Boys with more attention problems/overactive behavior are less able to delay gratification.

Abstract

Sex differences in self-control become apparent during preschool years. Girls are better able to delay their gratification and show less attention problems and overactive behavior than boys. In this context, organizational effects of gonadal steroids affecting the neural circuitry underlying self-control could be responsible for these early sex differences. In the present study testosterone levels measured in amniotic fluid (via ultra performance liquid chromatography and tandem mass spectrometry) were used to examine the role of organizational sex hormones on self-control. One hundred and twenty-two 40-month-old children participated in a delay of gratification task (DoG task) and their parents reported on their attention problems and overactive behavior. Girls waited significantly longer for their preferred reward than boys, and significantly more girls than boys waited the maximum period of time, providing evidence for sex differences in delay of gratification. Boys that were rated as suffering from more attention problems and overactive behavior waited significantly shorter in the DoG task. Amniotic testosterone measures were reliable in boys only. Most importantly, boys who waited shorter in the DoG task and boys who were reported to suffer from more attention problems and overactive behavior had higher prenatal testosterone levels. These findings extend our knowledge concerning organizational effects of testosterone on the brain circuitry underlying self-control in boys, and are of relevance for understanding how sex differences in behavioral disorders are connected with a lack of self-control.

Introduction

Delaying an immediate smaller reward for the sake of a later, larger one is incredibly difficult at all ages. Interestingly, the ability to show self-control by resisting predominant impulses for the pursuit of long-term goals already evolves early in life. Self-control in preschoolers is typically investigated with the delay of gratification (DoG) paradigm (Mischel and Ebbesen, 1970), which gives children the choice between an immediate smaller or less preferred reward (e.g. one marshmallow) and a later, larger or more preferred reward (e.g. two marshmallows). The relationship between delay time and parents’ as well as caregivers’ ratings of self-control corroborate the validity of the DoG paradigm as a measure of self-control (Duckworth et al., 2013).

A meta-analysis consisting of 33 studies confirmed a small, but significant, female advantage in DoG, which was substantially larger when waiting time was measured continuously instead of a dichotomous measurement of choice behavior (Silverman, 2003). In line with the female advantage in the DoG paradigm, 2135 parents reported better self-control in their 3- to 6-year-old daughters in comparison with their sons on a questionnaire (Tao et al., 2014).

These sex differences fit with the finding that substantially more boys than girls are diagnosed with attention-deficit/hyperactivity disorder (ADHD; Ramtekkar et al., 2010). Impulsive behavior (Duckworth and Steinberg, 2015) is one of the main DSM-5 (APA, 2013) diagnostic features of ADHD and is reflected by reward-delay impulsivity, defined as the choice of an immediate small reward over a delayed large reward (Patros et al., 2016). Consequently, children with ADHD differ from healthy children in their ability to delay gratification (for a meta-analysis see Patros et al., 2016). Additionally, parents’ estimates of attention deficits and hyperactive behavior on questionnaires correlate with the ability to delay gratification in samples of healthy children (Duckworth et al., 2013; Paloyelis et al., 2009). Sex differences in self-control also help to explain the gender gap in school grades. While girls achieve higher grades in the major subjects, they differ from boys in measures of self-control not IQ (Duckworth and Seligman, 2006; Duckworth et al., 2015; Matthews et al., 2009; Weis et al., 2013).

Many sex differences are affected by sex hormones. Testosterone especially, has organizational and activational effects. There is little and conflicting evidence concerning activational effects of testosterone on self-control in adults (Doi et al., 2015; Ortner et al., 2013; Peper et al., 2013). Since there are little variations in circulating sex hormones before puberty, prenatal (for a review see Cohen-Bendahan et al., 2005) or early postnatal testosterone exposure (for a review see Hines et al., 2016) having organizational effects on the brain is likely to play a role in the emergence of the early evident sexual dimorphism in self-control in childhood. Nevertheless, it is indisputable that self-control is affected by the social environment to a great extent. For example, parents serve as a model in the development of self-control in their children and influence their children’s self-control via caregiving behavior (for an overview see Bridgett et al., 2015). It also cannot be ruled out that boys and girls face different expectations regarding their self-control abilities (Bembenutty, 2007; Funder et al., 1983).

Supporting the hypothesis that self-control is influenced by prenatal testosterone, male rats normally make more impulsive choices than female rats in a delay-based impulsive choice task. However, these behavioral sex differences disappear when female rats are neonatally injected with testosterone (Bayless et al., 2013). In humans, there is only indirect evidence for prenatal testosterone effects on self-control from studies using the second-to-fourth-digit-ratio (2D:4D). 2D:4D is larger in females than in males early in life and is used as a marker for prenatal testosterone exposure (for a review see Breedlove, 2010). A study with preschoolers found that children who were less able to delay gratification had lower 2D:4D ratios (and supposedly higher prental testosterone levels; Da Silva et al., 2014). However, as this study did not apply separate analyses for boys and girls, the significant results might be due to the fact that both measures (2D:4D and delay of gratification) show sex differences. Another study with adults reported that women with smaller digit ratios on their right hands prefer smaller, more immediate over larger, delayed amounts of money (Lucas and Koff, 2010). Importantly, the specific relationship of 2D:4D with prenatal sex hormone levels is still under debate (Richards, 2017). Only one research group found the assumed negative relationship between 2D:4D and testosterone from amniotic fluid (a more direct measure) in girls (Ventura et al., 2013). Another report on a correlation between 2D:4D and the ratio between testosterone and estradiol from amniocentesis (Lutchmaya et al., 2004) did not separate the sexes for the correlational analyses. Clearly, evidence for an effect of prenatal steroids on self-control in children from a more direct measure of prenatal sex hormones, namely testosterone levels measured in amniotic fluid, is missing.

Amniocentesis is an invasive method of sampling amniotic fluid, normally conducted during the second trimester of pregnancy and is used only in cases of increased risk for genetic and chromosomal anomalies, like increased maternal age. Sex differences in testosterone from amniotic fluid emerge between week 8 and 24 of pregnancy peaking around week 17 (Judd et al., 1976). This corresponds quite well with the timing of amniocentesis and resembles the testosterone peak in serum (Finegan et al., 1989; although the exact relationship between amniotic and serum testosterone is still not completely understood - for a review see Constantinescu and Hines, 2012). Other sex dimorphic behaviors in early childhood like sex-typed play behavior and empathic behavior have already been found to be associated with prenatal testosterone levels from amniotic fluid (for a review see Auyeung et al., 2013).

The present study explores, for the first time, the effects of prenatal testosterone from amniotic fluid on self-control in early childhood. Self-control behavior was measured in 40-month-old children using a modified version of the DoG task (Mischel and Ebbesen, 1970). Attention problems/overactive behavior were obtained from the parents’ report in a questionnaire. Based on previous research, girls were expected to wait longer for their preferred reward than boys and boys to be reported to show more attention problems/overactive behavior than girls. According to the results from an animal study (Bayless et al., 2013) and 2D:4D studies in humans (Da Silva et al., 2014; Lucas and Koff, 2010), we expected children with higher amniotic testosterone levels to be less able to delay gratification in the DoG task and to show more attention problems/ overactive behavior. Furthermore, we expected children showing more attention problems/overactive behavior to be less able to delay gratification in the DoG task.