Of interest to women’s mental health, most prevalence studies of attention deficit hyperactivity disorder (ADHD) have reported a significant excess of boys over girls, with ratios varying from 18:1 to 4:1 (1). The actual prevalence depends on whether the classification system used is categorical or dimensional. It also depends on the choice of informant, the age and gender composition of the population studied, and the instrument used to measure impairment. Widely divergent estimates of prevalence, from 4% to 17% of the population, have been reported (estimates are lower, 0.8% to 1.7%, for hyperkinetic disorder, the term used by the International Classification of Diseases, 10th edition, ICD-10) (1). All questionnaire estimates show an excess of ADHD in boys. There is less consistency with respect to the influence of age and comorbidity. In general, interview studies suggest a decline in prevalence with age. The Buitelaar review found high rates of comorbidity of ADHD and oppositional defiant disorder (25%), conduct disorder (15%), anxiety (25%), and depression (25%) (1). Studies of externalizing and internalizing comorbidity patterns in children aged 6 to 12 years old (with ADHD combined and ADHD inattentive subtypes) indicate significant comorbidity between ADHD combined subtype and oppositional defiant disorder or conduct disorder (2).

Possible explanations for the male excess are rater bias, referral bias, threshold differences, and behavioral or learning differences between the sexes. Faraone and colleagues (3) have found that boys have an increased risk over girls for ADHD diagnosed according to criteria of the Diagnostic and Statistical Manual of Mental Disorders, 3rd edition (DSM-III) only in families that exhibit antisocial disorders. These investigators postulate that gender differences might be important in that they provide clues to the genetic heterogeneity of ADHD.

In terms of etiology, Rhee and colleagues (4,5) have compared the constitutional variability model with the polygenic multiple threshold model. The latter assumes a continuum of liability with a higher threshold for girls, while the constitutional variability model implies unique male susceptibility. Evidence for male susceptibility is buttressed by the fact that the birth of boys is associated with relatively more labor and delivery problems and boys show greater immaturity at birth. Boys exhibit greater degrees of overactivity and inattention. In this way of thinking, girls with ADHD represent a brain-damaged group. The polygenic model postulates that those of the sex that is less frequently affected (females in the case of ADHD) should have more affected relatives than those of the more frequently affected sex. Indeed, Rhee and colleagues have shown that cotwins and cosiblings of girls with ADHD have a higher number of ADHD symptoms on average (mean = 5.54, sd = 4.35) than the cotwins and cosiblings of boys with ADHD (mean = 4.24, SD = 4.0), a finding supportive of the polygenic multiple threshold model (4).

In their review of gender differences in prevalence and in the association of ADHD and hyperkinetic disorder (HKD), Heptinstall and Taylor (6) point to a number of inconsistencies and contradictions in the literature, namely that population samples and clinic-referred samples differ in their characteristics. A study on the discriminant validity (e.g., relationship to criteria external to the defining symptoms) of DSM-IV ADHD in a nationally representative Australian sample of 3,597 children, aged 6 to 17 years, found a prevalence of 7.5% (6.9% if only those with impairment are included) for ADHD (7). The inattentive subtype was found to be more common than the combined and hyperactive-impulsive subtypes. This study, which included parent-rated impairment measures, concluded that ADHD was more prevalent among males in all three subtypes, with the male to female ratio for combined type being approximately twice that of hyperactive-impulsive and inattentive type (7).

In their meta-analysis of gender differences in ADHD (8), Gaub and Carlson found that among population samples not referred to clinics girls with ADHD displayed lower levels of inattention, internalizing behavior, and peer aggression than boys with ADHD, but clinic-referred samples showed similar gender levels of both impairment and comorbidity. According to observations of classroom behavior of 403 boys and 99 girls with ADHD, boys engaged in more rule-breaking and externalizing impulsive behaviors (disruptive behavior disorders, or DBD) than did girls (9). It was also found that children with ADHD and DBD manifested more interference with classroom routine behaviors than did children with ADHD and anxiety, although comorbid anxiety did not always inhibit the rate of disruptive behavior.

A recent meta-analysis by Gershon has pointed out that identification of ADHD in females is difficult, as fewer females than males are evaluated in specialized clinics (10). While epidemiologic samples estimate gender differences at 3:1, clinical samples range closer to 9:1. This supports the Australian epidemiologic findings: male:female ratio of 1.7:1 for hyperactive-impulsive types and 4.6:1 for combined types, respectively, somewhat lower and somewhat higher than in previous community-based studies. Combined subtypes had higher scores on the anxious/depressed scale of the Achenbach Child Behavior Checklist (CBCL), and on all three externalizing scales of the CBCL. Combined type symptoms were rated as causing greater disruption to family activities than inattentive and hyperactive-impulsive symptoms. All children with combined type symptoms met criteria for impairment, compared with 93% of the inattentive and 86% of the hyperactive-impulsive types (7).

Comparing results with those of Gaub and Carlson, Gershon (10) examined potential moderators of effect size estimates of the male:female ratio. These include publication status, referral source, rater effects, assessment of IQ, age of subjects, and diagnostic system used. He found that ADHD females manifested significantly fewer externalizing problems, but significantly more internalizing problems than ADHD males. Girls performed worse on full scale and verbal IQ. Teachers rated girls as less inattentive and having fewer externalizing problems than boys. Clinically referred samples, as was expected, tended to manifest more severe symptoms than community samples. This study also found a possible gender bias in rating scales, all tending to score boys higher than girls. Large gender differences persist, however, even when this is taken into account.

Sex differences in prevalence are not explained by heritability differences between girls and boys. Rietveld and colleagues (11) studied a large community sample of twins 3 to 12 years old for overactive behavior and attention problems. They used a cross-sectional twin design controlling for developmental, gender, and rater contrast contributions. They found that heritability ranged from 68% to 76% across age groups. In general, patterns of additive genetic, dominance, and unique environmental effects were similar in boys and girls, with a rater contrast effect found at 3 years of age.