Skip to main content
PMC home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2023 Jul 12.
Published in final edited form as: Twin Res Hum Genet. 2023 Jan 12;25(6):226–233. doi: 10.1017/thg.2022.35

Shared Genetic Factors Contributing to the Overlap between Attention-Deficit/Hyperactivity Disorder Symptoms and Overweight/Obesity in Swedish Adolescent Girls and Boys

K N Javaras 1,2,3,*,†, M A Munn-Chernoff 3,*, E W Diemer 3,4,5,*, L M Thornton 3,*, C M Bulik 3,6,7,*, Z Yilmaz 3,8,*, P Lichtenstein 6,*, H Larsson 6,9,*, J H Baker 3,*
PMCID: PMC10191205  NIHMSID: NIHMS1895668  PMID: 36633106

Abstract

Attention-deficit/hyperactivity disorder (ADHD) and obesity are positively associated, with increasing evidence that they share genetic risk factors. Our aim was to examine whether these findings apply to both types of ADHD symptoms for female and male adolescents. We used data from 791 girl and 735 boy twins ages 16–17 years to examine sex-specific phenotypic correlations between the presence of ADHD symptoms and overweight/obese status. For correlations exceeding .20, we then fit bivariate twin models to estimate the genetic and environmental correlations between the presence of ADHD symptoms and overweight/obese status. ADHD symptoms and height/weight were parent- and self-reported, respectively. Phenotypic correlations were .30 (girls) and .08 (boys) for inattention and overweight/obese status and .23 (girls) and .14 (boys) for hyperactivity/impulsivity and overweight/obese status. In girls, both types of ADHD symptoms and overweight/obese status were highly heritable, with unique environmental effects comprising the remaining variance. Furthermore, shared genetic effects explained most of the phenotypic correlations, in girls. Results suggest that the positive association of both types of ADHD symptoms with obesity may be stronger in girls than boys. Further, in girls, these associations may stem primarily from shared genetic factors.

Keywords: Obesity, Attention-deficit/hyperactivity disorder, Inattention, Hyperactivity/Impulsivity, Genetics

Attention-deficit/hyperactivity disorder (ADHD)—characterized by pervasive and impairing inattention, hyperactivity, and impulsivity—affects 3.4% of children and adolescents worldwide (Polanczyk et al., 2015). Individuals with ADHD are also at greater risk for certain conditions, including obesity (Instanes et al., 2018), which has become increasingly prevalent among children and adolescents globally (Ng et al., 2014). Specifically, in meta-analyses, children and adolescents with ADHD (based on an actual or self-reported medical record diagnosis, a research diagnosis, or scores on validated ADHD rating scales) had 1.2 times the odds of obese status (Cortese et al., 2016) and 1.13 times the odds of overweight and/or obese status (Nigg et al., 2016), compared to children and adolescents without ADHD; both odds ratios (i.e., 1.2 and 1.13) were statistically significant. Notably, in the latter meta-analysis, when the association between ADHD and overweight/obese status was calculated separately for adolescents and for younger children, the association in adolescents was qualitatively larger than the association in children (Nigg et al., 2016). This finding converges with results from longitudinal studies showing that body mass index (BMI) trajectories for individuals with ADHD diverge upward from the trajectories of those without ADHD during the adolescent years (Porter et al., 2021; Schwartz et al., 2014). Although the meta-analytic associations (between ADHD and overweight/obese status) did not differ significantly by sex in analyses that combined studies conducted in all age groups (Cortese et al., 2016; Nigg et al., 2016), some individual studies suggest that, among youth, the association between ADHD and high BMI may be stronger in females (Aguirre Castaneda et al., 2016; Byrd et al., 2013; Kase et al., 2021; Kim et al., 2011; Nigg et al., 2016; van Egmond-Fröhlich et al., 2012). Further, meta-analytic results suggest that the positive association between ADHD and obesity pertains only to individuals not taking stimulant medications (Cortese et al., 2016). Finally, although the meta-analyses did not examine whether the association between ADHD and high BMI differed for different ADHD subtypes (i.e., inattention (IN), or hyperactivity/impulsivity (HI)), a recent study conducted across adolescence and young adulthood found a significant association between HI symptoms and BMI, but not IN symptoms and BMI (Kase et al., 2021).

In addition to evidence that ADHD and obesity co-occur within individuals, evidence also points to co-aggregation within families, with ADHD in a proband increasing the likelihood of high BMI in family members (Chen et al., 2017, 2019; Geuijen et al., 2019). The presence of co-aggregation suggests a role for familial influences, specifically genetic and common environmental factors. Regarding genetics, both ADHD and obesity are substantially influenced by genetic factors. Diagnosed ADHD is highly heritable (between .70 and .80) (Faraone et al., 2005; Faraone & Larsson, 2019). Further, ADHD symptoms are moderately to highly heritable in adolescence (Merwood et al., 2013), with higher estimates of heritability typically observed when the same person (e.g., a parent) rates symptoms for both twins (Brikell et al., 2015; Freitag et al., 2010), potentially because having a single rater reduces discordance in twins’ ratings due to differences in rater effects between informants (Brikell et al., 2015). Additionally, heritability is high for both subtypes of ADHD symptoms (i.e., IN and HI) in childhood and adolescence (Larsson et al., 2011), although the genetic influences on the two types of symptoms appear to overlap only partially (Freitag et al., 2010). BMI is also moderately to highly heritable in childhood and adolescence, with twin-based heritability estimates ranging from approximately .50 to .90 (Silventoinen et al., 2010). Estimates of BMI’s heritability tend to be higher in studies that use staff-measured (rather than self-reported) BMI, likely due to reduced measurement error for staff-measured BMI (Elks et al., 2012). BMI heritability estimates also tend to be higher in studies conducted in adolescents or in countries with high gross domestic product per capita (Elks et al., 2012; Min et al., 2013), possibly because of developmental and environmental differences, respectively, in the expression of adiposity-related genes (Min et al., 2013). Regarding common environment, childhood ADHD symptoms and obesity share some environmental risk factors (e.g., low parental socioeconomic status, maternal smoking during pregnancy), which may account for their co-occurrence (Donnchadha et al., 2020; Pauli-Pott et al., 2017) and some instances of co-aggregation. Notably, some of these risk factors (e.g., maternal smoking during pregnancy) may share overlapping genetic influences with ADHD and BMI (Obel et al., 2011; Skoglund et al., 2014).

Bivariate behavioral genetic models can be used to elucidate the extent to which genetic and environmental influences on ADHD and obesity overlap. To our knowledge, two studies have conducted this type of analysis (Chen et al., 2019; Do et al., 2019). Do and colleagues (2019) fitted behavioral genetic models to data on retrospectively-reported childhood ADHD symptoms and adolescent through young adult BMI, from twins and siblings in a nationally-representative longitudinal study conducted in the United States. In females, the best-fitting model included shared genetic influences, but not shared unique environmental influences, between overall ADHD symptoms and BMI; in this model, genetic correlations (i.e., correlations between the genetic influences on ADHD symptoms and BMI) ranged from .20 to .24, depending on the BMI timepoint. In contrast, in males, the best-fitting model included shared unique environmental influences, but not shared genetic influences, between ADHD symptoms and BMI; in this model, environmental correlations (i.e., correlations between the environmental influences on ADHD symptoms and BMI) ranged from .16 to .19, depending on the BMI timepoint. Do and colleagues (2019) also explored whether these results held when models were fitted to IN and HI symptoms separately. For IN, results were similar to those for overall ADHD symptoms, with genetic influences playing a (qualitatively) greater role in the overlap between symptoms and BMI in females: genetic correlations ranged from .10 in adolescence to .25 in adulthood for females, but from 0 in adolescence to .07 in adulthood for males. In contrast, there was no evidence of sex differences in the genetic overlap between HI symptoms and BMI, with genetic correlations ranging from .08 in adolescence to .25 in adulthood for both females and males. In another study, Chen and colleagues (2019) fitted behavioral genetic models to data on ADHD and (clinical) obesity for full and half siblings, using data from the Swedish national registers. Results suggest that overlapping genetic influences contribute to the overlap between ADHD and obesity, with a (significant) genetic correlation of .30 in both sexes combined. The correlations for common and unique environmental influences on ADHD and obesity were .19 and −.03, respectively, and not significant. Taken together, these studies provide evidence that the genetic influences on ADHD and obesity overlap, at least in females. However, the retrospective-reporting of ADHD symptoms in Do et al. (2019) and the clinical ascertainment of ADHD and obesity cases in Chen et al. (2019) are well-acknowledged sources of bias that may impact results, underscoring the importance of independent replications that address these methodological limitations. Further, given evidence that the use of stimulant medications attenuates the association between ADHD and obesity (Cortese et al., 2016; Nigg et al., 2016), it is important to investigate genetic and environmental influences on this association in a sample naïve to prescription stimulants.

Thus, we conducted bivariate behavioral genetic analyses to examine the relative contributions of shared genetic and environmental influences on ADHD symptoms and overweight/obese status, using data from a sample of Swedish adolescent twins. Importantly, ADHD symptoms were reported close to the time of their occurrence, information on BMI was available for the whole sample (i.e., rather than only for those who received a clinical diagnosis of obesity), and the sample was largely naïve to prescription stimulant use given the extremely low prevalence of prescription stimulants in Sweden at the relevant time (Ekman & Gustafsson, 2000). Given evidence that the strength of the association between ADHD symptoms and BMI or obesity may differ by symptom type (Kase et al., 2021) and sex (Aguirre Castaneda et al., 2016; Byrd et al., 2013; Kase et al., 2021; Kim et al., 2011; Nigg et al., 2016; van Egmond-Fröhlich et al., 2012), as well as prior research suggesting that the degree of genetic overlap between ADHD symptoms and BMI may differ for females and males depending on ADHD symptom type (Do et al., 2019), we conducted separate analyses for IN and HI symptoms, and analyses were stratified by sex.

Materials and Methods

Participants

Our sample included participants from the Swedish Twin Study of CHild and Adolescent Development (TCHAD), which includes twins born between May 1985 and December 1986 in Sweden (Lichtenstein et al., 2007). Twins and their parents were identified via the Swedish Medical Birth Register (Cnattingius et al., 1990) and invited to participate by completing parent and self-report questionnaires at four time points: Wave 1 (ages 8–9), Wave 2 (ages 13–14), Wave 3 (ages 16–17), and Wave 4 (ages 18–19).

For this study, we only used Wave 3 data (ages 16–17) because of evidence that the association between ADHD and obesity strengthens with age, emerging only in late adolescence in some samples (Nigg et al., 2016). Although participants were slightly older at Wave 4, there were fewer participants available for analysis, a significant limitation given the sample sizes required for behavioral genetic modelling, especially for categorical phenotypes. At Wave 3, 74% of the twins’ parents and 82% of the twins responded to the questionnaire (Lichtenstein et al., 2007).

Our sample included 791 girls and 735 boys with known zygosity and at least some relevant variables (i.e., ADHD symptoms and BMI) observed. These participants comprised 389 complete and 13 incomplete female twins pairs and 357 complete and 21 incomplete male twin pairs (see Table 1a for the number of complete and incomplete pairs by sex and zygosity). Opposite sex dizygotic twin pairs were not included because analyses were stratified by sex (see below).

Table 1a:

Number of Complete and Incomplete Twin Pairs, by Sex and Zygosity

Girls Boys
Monozygotic Dizygotic Monozygotic Dizygotic
Complete pairs, number 225 164 208 149
Incomplete pairs, number 3 10 13 8
Open in a new tab

Consent was obtained from all twins and their parents who participated. Ethical approval for each wave was provided by the Ethics Committee at Karolinska Institutet in Stockholm, Sweden, and the current study was approved by the University of North Carolina Institutional Review Board.

Measures

Zygosity.

Zygosity was determined by an algorithm derived from discriminant analyses of responses to zygosity questionnaires for 106 same-sex twin pairs with known zygosity (i.e., determined based on 16 DNA polymorphisms) (Lichtenstein et al., 2007). In instances where responses yielded discordant classifications (N=86 in the full TCHAD sample), zygosity was set to unknown, and the twin pair was not included in analyses.

ADHD.

Parents completed a checklist assessing the presence or absence of 14 ADHD symptoms in their child over at least the past six months. The checklist was developed based on the ADHD criteria from the Diagnostic and Statistical Manual of Psychiatric Disorders, 4th edition (American Psychiatric Association, 1994), and details regarding its development and validation have been described elsewhere (Larsson et al., 2006, 2011). We focused on the presence of IN and HI symptoms, defined as having at least one symptom (Li et al., 2019) on the IN subscale (6-items) or HI subscale (8-items), respectively.

Overweight/Obese Status.

Height and weight were provided via twin self-report. We used age- and sex-specific means and standard deviations (SD) from a sample of Swedish children born in 1981 to remove values 6 standard deviations above and below the mean for height and log(weight) (Werner & Bodin, 2006). We then used the World Health Organization 2007 package for R to calculate age- and sex-adjusted BMI z-scores (de Onis et al., 2007). In the present analyses, we used overweight/obese status, which was defined as being at or above +1 standard deviation for the age- and sex-adjusted BMI z-scores.

Analyses

Descriptive statistics and phenotypic correlations were calculated using R Statistical Software (version 4.0.2). Phenotypic tetrachoric correlations were calculated using the corCI() function from the psych package. Bivariate twin analyses were performed only if the correlation between the presence of ADHD symptoms and overweight/obese status exceeded .20, similar to previously used approaches (Baker et al., 2018; Koren et al., 2014). Although bivariate twin models can be fitted to variables with small phenotypic correlations, there is limited value in decomposing correlations that do not statistically differ from zero into genetic and environmental components.

Bivariate twin analyses were conducted in Mx (Neale et al., 2006; Neale & Cardon, 1992). Bivariate twin analyses were performed between each ADHD variable and overweight/obese status (provided their phenotypic correlation exceeded .20), stratified by sex. Of note, we did not formally test for sex differences in bivariate twin model parameters, since prior work with this sample has demonstrated that we have insufficient power to do so (Baker et al., 2017). Cholesky decomposition was used to determine the relative magnitude of additive genetic (a), common environmental (c), and unique environmental (e) effects on the liability to the presence of ADHD symptoms, overweight/obese status, and their covariance. Additive genetic effects represent the accumulation of genetic factors influencing a phenotype, whereas common environmental effects are those factors that make members of a family similar for a given phenotype, and unique environmental effects are those factors that make members of a family dissimilar for a given phenotype. Measurement error is also included in unique environmental effects. With respect to their covariance, we estimated the genetic (rg), common environmental (rc), and unique environmental (re) correlations between the two phenotypes. The genetic correlation is the correlation between the genetic factors responsible for the presence of ADHD symptoms and the genetic factors responsible for overweight/obese status. If rg is estimated at unity, it means that the same genetic factors contribute to the development of ADHD symptoms and overweight/obesity.

Full ACE models were fitted first. We then tested whether any parameter estimates (i.e., A, C, or E) could be set to zero and, thus, removed from the model. These submodels were tested against the full model by using minus two log-likelihood (−2LL) and standard chi-square (χ2) difference tests to determine model fit (Neale et al., 2006). In addition, Akaike’s Information Criterion (AIC) (Akaike, 1987) was used for model comparisons. A lower AIC value indicates a better and more parsimonious fitting model.

Results

Summary statistics for the analysis sample are presented in Table 1b.

Table 1b:

Analysis Sample Summary Statistics, by Sexa,b

Girls
(n = 791)		 Boys
(n = 735)
Age, years 16.7 (0.5) 16.7 (0.5)
ADHD symptoms
 At least one IN symptom 33.4% 35.3%
 At least one HI symptom 21.1% 15.9%
Overweight/obese status 8.1% 11.7%
Open in a new tab
a

Abbreviations: ADHD = attention-deficit/hyperactivity disorder; IN = inattention; HI = hyperactivity/impulsivity.

b

Statistics presented are mean (standard deviation) for continuous variables, and percentage for categorical variables.

Phenotypic tetrachoric correlations between ADHD variables and overweight/obese status were larger than .20 in girls (IN: .30 (95% confidence interval [CI] = .12, .46); HI: .23 (.03, .40)), but not in boys (IN: .08 (−.09, .24); HI: .14 (−.07, .31)). Thus, the following results pertain to girls only.

Model-fit indices are presented in Table 2, with parameter estimates for the full and best-fitting models shown in Table 3.

Table 2:

Bivariate Cholesky Decomposition Model-Fitting Results for Girlsa,b

Inattention & Overweight/Obesity (Girls)
Model Model Fit Comparative Fit
Number Model −2LL df AIC Δχ2 Δdf p-value
1 ACE-ACE
rg, rc, re 		1138.148 1415 −1691.852 --- --- ---
2 AE-AE
rg, re 		1142.201 1418 −1693.799 4.053 3 .256
3 CE-CE
rc, re 		1154.947 1418 −1681.053 16.799 3 .001
4 E-E
re 		1272.188 1421 −1569.812 134.040 6 <.0001
Hyperactivity/Impulsivity & Overweight/Obesity (Girls)
Model Model Fit Comparative Fit
Number Model −2LL df AIC Δχ2 Δdf p-value
1 ACE-ACE
rg, rc, re 		987.380 1418 −1848.620 --- --- ---
2 AE-AE
rg, re 		987.689 1421 −1854.311 .309 3 .958
3 CE-CE
rc, re 		1006.304 1421 −1835.696 18.924 3 <.0001
4 E-E
re 		1115.017 1424 −1732.983 127.637 6 <.0001
Open in a new tab
a

A = additive genetic effects; C = common environmental effects; E = unique environmental effects; rg = correlation between presence of ADHD symptoms and overweight/obesity that is due to additive genetic influences; rc = correlation between presence of ADHD symptoms and overweight/obesity that is due to common environmental influences; re = correlation between presence of ADHD symptoms and overweight/obesity that is due to unique environmental influences; −2LL = −2 x log likelihood; AIC = Akaike’s Information Criterion; Δχ2 = chi-square difference test statistic.

b

Statistics for the best-fitting model appear in bold.

Table 3:

Full and Best-Fitting Model Parameter Estimates (and 95% Confidence Intervals) for Girlsa,b

Inattention & Overweight/Obesity (Girls)
Model Inattention Obesity Correlations
Number Model a2 c2 e2 a2 c2 e2 r g r c r e
1 ACE-ACE
rg, rc, re 		.48
(.09, .83)		 .30
(.19, .64)		 .22
(.12, .37)		 .45
(.04, .92)		 .45
(.00, .82)		 .10
(.03, .26)		 −.42
(−1.00, .28)		 1.00
(.23, 1.00)		 .61
(.08, .90)
2 AE-AE
rg, re 		.80
(.66, .89) 		--- .20
(.11, .34) 		.91
(.77, .97) 		--- .09
(.03, .23)
.22
(−.02, .45) 		--- .57
(−.02, .94)
Hyperactivity/Impulsivity & Overweight/Obesity (Girls)
Model Hyperactivity/Impulsivity Obesity Correlations
Number Model a2 c2 e2 a2 c2 e2 r g r c r e
1 ACE-ACE
rg, rc, re 		.86
(.46, .94)		 .00
(.00, .36)		 .14
(.06, .27)		 .68
(.06, .98)		 .23
(.00, .80)		 .10
(.03, .24)		 .27
(−.40, 1.00)		 −.95
(−1.00, 1.00)		 .31
(−.48, .90)
2 AE-AE
rg, re 		.86
(.73, .94) 		--- .14
(.06, .27) 		.91
(.77, .97) 		--- .09
(.03, .23) 		.23
(−.02, .46) 		--- .31
(−.48, .90)
Open in a new tab
a

A = additive genetic effects; C = common environmental effects; E = unique environmental effects; rg = correlation between presence of ADHD symptoms and overweight/obesity that is due to additive genetic influences; rc = correlation between presence of ADHD symptoms and overweight/obesity that is due to common environmental influences; re = correlation between presence of ADHD symptoms and overweight/obesity that is due to unique environmental influences.

b

Statistics for the best-fitting model appear in bold. 95% confidence intervals are presented in parentheses.

For IN and overweight/obese status, the model that only includes additive genetic and unique environmental effects provided the best fit to the data. This model (Model 2) estimated the heritability of IN and overweight/obesity to be .80 (.66, .89) and .91 (.77, .97), respectively. The remaining variance was attributable to unique environmental effects. The correlation due to additive genetic effects was .22 (−.02, .45), and the correlation due to unique environmental effects was .57 (−.02, .94). Based on these estimates, approximately 5% (=.222) and 32% (=.572) of the genetic and unique environmental effects, respectively, between IN and overweight/obesity were overlapping. Model 2’s parameter estimates can also be used to calculate the proportion of the phenotypic correlation (between IN and overweight/obesity) that is due to additive genetic effects (71%) and unique environmental effects (29%) (Munn et al., 2010).

Similarly, for HI and overweight/obese status, the best-fitting model (Model 2) did not include common environmental variances or covariance. The heritability of HI was .86 (.73, .94), and unique environmental effects were estimated at .14 (.06, .27). Heritability estimates for overweight/obesity were again .91 (.77, .97), with the remainder due to unique environmental effects. The genetic and unique environmental correlations were .23 (−.02, .46) and .31 (−.48, .90), respectively, indicating that 5% of the genetic effects and 10% of the unique environmental effects between HI and overweight/obesity were overlapping. Using Model 2’s parameter estimates to calculate the proportion of the phenotypic correlation (between HI and overweight/obesity) that is due to additive genetic effects and unique environmental effects yielded values of 85% and 15%, respectively.

Discussion

We examined the phenotypic overlap of IN and HI symptoms with overweight/obesity in a sample of adolescents, stratified by sex. Further, in instances where the overlap was significant, we also performed bivariate twin analyses examining the relative importance of genetic and environmental contributions to the overlap.

We found a statistically significant phenotypic association between the presence of both types of ADHD symptoms and overweight/obese status in girls, but not in boys. Although meta-analyses combining age groups have not found significant sex differences in the association between ADHD and overweight and/or obesity (Cortese et al., 2016; Nigg et al., 2016), some studies conducted in youth suggest that the association between ADHD and high BMI may be stronger or more robust in girls than in boys (Aguirre Castaneda et al., 2016; Kase et al., 2021; Kim et al., 2011; Nigg et al., 2016; van Egmond-Fröhlich et al., 2012). Given that reduced self-regulation has been described as central to the deficits observed in ADHD (Shiels & Hawk, 2010), findings of a stronger ADHD – BMI association in female youth converge with research suggesting that a positive association between reduced self-regulation and adiposity may emerge earlier in girls (Javaras et al., 2020). Several explanations have been proposed for why this may be the case, including the lower prevalence of prescription stimulant use in girls, the lower prevalence of the hyperactive/impulsive subtype (vs. the inattentive subtype) in girls with ADHD, the greater frequency of obesity-related behaviors (e.g., screen time) in girls with ADHD, and the greater prevalence of comorbid pathology (e.g., depression, eating disorders) in girls with ADHD (Aguirre Castaneda et al., 2016; Byrd et al., 2013; Javaras et al., 2020; Kim et al., 2011; van Egmond-Fröhlich et al., 2012). Although our study was not designed to shed light on the validity of these explanations, it does suggest that sex differences in prescription stimulant use are unlikely to explain the qualitatively stronger female ADHD – BMI association in our sample, since prescription stimulant use in our sample was likely to be rare given its extremely low prevalence in Sweden during the relevant time frame (e.g., in 1998, 0.08% of children age 5–18 years in Sweden were prescribed stimulants) (Ekman & Gustafsson, 2000). Further, because we examined the association of high BMI with IN and HI symptoms separately, our findings suggest that sex differences in the different symptom types are unlikely to explain the stronger female ADHD – BMI association.

In girls, both IN and HI symptoms were highly heritable, as expected given that prior research in the current sample has found high heritability for IN and HI symptoms (Larsson et al., 2011). The broader literature suggests that ADHD symptoms are generally moderately to highly heritable (Merwood et al., 2013), with higher estimates typically found when the same person (e.g., a parent) rates symptoms for both twins (Brikell et al., 2015; Freitag et al., 2010), as was the case in the current study. In the current study, overweight/obese status was also highly heritable in girls and at the high end of prior twin-based estimates (Silventoinen et al., 2010). Although our study used self- (i.e., twin-) reported BMI, which tends to decrease estimates of heritability, heritability estimates tend to be higher for adolescents and for countries, such as Sweden, with high gross domestic product per capita (Elks et al., 2012; Min et al., 2013).

For both types of ADHD symptoms and for overweight/obese status, model fitting results suggest that non-genetic influences include unique environmental factors, but not common environmental factors. These findings are in keeping with prior twin analyses of BMI symptoms, which suggest that common environmental influences on BMI are present in early adolescence, but wane by late adolescence (Lajunen et al., 2009). With regard to ADHD symptoms, most twin analyses of ADHD have not found evidence for common environmental influences (Polderman et al., 2007).

In girls, most of the overlap between IN (or HI) and overweight/obese status was due to genetic effects, with the remainder due to unique environmental effects. The finding that genetic effects account for more of the overlap between ADHD symptoms and BMI in females is consistent with results from a prior behavioral genetics study (Do et al., 2019). The estimated genetic correlations in the current study (.22 for IN and overweight/obese status, and .23 for HI and overweight/obese status) are somewhat higher than analogous estimates (.10 and .08, respectively) from the prior study (Do et al., 2019), which may be attributed to differences in study methodology, such as differences in the reporting of ADHD symptoms in the current and prior study. Regardless of these numeric differences, results from the current study and prior behavioral genetic research (Chen et al., 2019; Do et al., 2019) converge in suggesting a modest genetic correlation between ADHD and high BMI, at least in females. Further, genetic correlations based on GWAS data were .22 for ADHD and childhood obesity (Demontis et al., 2019) and .32 for ADHD and BMI (Mota et al., 2020), the former being especially consistent with estimated genetic correlations in the current study. Collectively, these results suggest that IN/HI symptoms and overweight/obesity may have overlapping genetic risk factors, at least in females. Although the identity of the overlapping genetic variants remains to be fully elucidated, recent research implicates dopaminergic genes – particularly the Dopamine-DARPP32 Feedback in cAMP Signaling pathway – in the genetic association between ADHD and obesity (Mota et al., 2020), consistent with earlier proposals (Liu et al., 2008).

The current study should be interpreted in light of several limitations. First, due to the low levels of ADHD pathology in TCHAD, participants were considered to have ADHD symptoms if they reported at least one ADHD symptom. Therefore, most participants with ADHD symptoms would not receive a diagnosis of ADHD, and these findings may not generalize to clinical populations. Second, as a result of the low prevalence of overweight and obesity in the Swedish population during the study period, we were unable to examine associations of ADHD symptoms with overweight and obese status separately. The relation between ADHD pathology and weight status is potentially complex, and further research is needed to determine whether genetic correlations with ADHD symptoms differ for overweight and obese status. Third, overweight/obese status was determined based on self-reported height and weight, which may have resulted in underestimates, especially in female adolescents (Bowring et al., 2012; Sherry et al., 2007). Although differential measurement error by sex could account for sex differences in the association between ADHD symptoms and overweight/obese status, a recent meta-analysis found that restricting analyses to studies using directly measured height and weight had minimal effect on estimates of the ADHD-obesity association (Cortese et al., 2016). Fourth, confidence intervals for genetic and environmental correlations between ADHD symptoms and overweight/obese status were wide and included zero. However, as noted above, estimates of the genetic correlations were generally very similar to those found in other studies, increasing confidence in these results. Fifth, AIC and the standard chi-square distribution for log-likelihood tests can lead to selection of an overly parsimonious model for some model comparisons (Javaras et al., 2009; Sullivan & Eaves, 2002). However, in the current study, use of a more lenient threshold (Javaras et al., 2009) would not have changed the conclusions regarding the best-fitting model. Sixth, as is standard in twin analyses, our study assumes random mating, equal environments, and additive genetic effects. Other designs are needed to investigate the potential role of epigenetics and gene by environment interactions in the co-occurrence between ADHD and obesity, as well as the validity of the random mating and equal environment assumptions with regards to ADHD symptoms and obesity, although existing evidence suggests that the latter assumption is a reasonable one for ADHD (Cronk et al., 2002).

These limitations aside, the present study provides additional evidence that the overlap between ADHD symptoms and overweight/obese status is stronger in female youth compared to male youth, and that the source of this overlap is primarily genetic in girls. These findings have both clinical and research implications. Clinically, obesity-related prevention efforts may need to begin earlier for girls with ADHD. In terms of research, future molecular genetic research focused on identifying additional genetic variants shared by both ADHD and obesity may benefit from stratifying analyses by sex. Finally, future research should focus on examining genetic and environmental overlap between ADHD and obesity in samples that are larger and from countries other than the U.S. and Sweden. Further, given that the association between ADHD and obesity may not become substantial in males until later in development, future investigations should also focus on older ages.

Financial Support:

KNJ’s work on this project was supported by: the National Institute of Mental Health (grant number T32-MH076694); the National Institute of Diabetes and Digestive and Kidney Diseases of the National Institutes of Health (grant number K23-DK120517); the Martini Family Foundation; and the McLean Hospital Women’s Mental Health Innovation Fund. MAMC’s work on this project was supported by: the National Institute on Alcohol Abuse and Alcoholism (grant number K01-AA025113). ZY’s work on this project was supported by: the National Institute of Mental Health (grant numbers T32-MH076694, K01-MH109782). CMB is supported by: the National Institute of Mental Health (grant numbers R01-MH120170, R01-MH124871, R01-MH119084, R01-MH118278); the Brain and Behavior Research Foundation Distinguished Investigator Grant; the Swedish Research Council (Vetenskapsrådet, award: 538–2013-8864); and the Lundbeck Foundation (grant number R276–2018-4581). HL acknowledges funding / financial support from: the European Union’s Horizon 2020 research and innovation programme (under grant agreement No 965381); the Swedish Research Council (grant number 2018–02599); and the Swedish Brain Foundation (grant number FO2021–0115).

The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health.

Footnotes

Conflicts of Interest:

KNJ, MAM-C, EWD, LMT, ZY, PL, and JHB declare the absence of any financial, professional, or personal relationships with the potential to bias the work. CMB reports the following relationships: Shire (grant recipient, Scientific Advisory Board member); Lundbeckfonden (grant recipient); Pearson (author, royalty recipient); Equip Health Inc. (Clinical Advisory Board). HL reports receiving the following (all outside the submitted work): grants from Shire Pharmaceuticals; personal fees from and serving as a speaker for Medice, Shire/Takeda Pharmaceuticals, and Evolan Pharma AB; and sponsorship for a conference on attention-deficit/hyperactivity disorder from Shire/Takeda Pharmaceuticals and Evolan Pharma AB.

Ethical Standards:

The authors assert that all procedures contributing to this work comply with the ethical standards of the relevant national and institutional committees on human experimentation and with the Helsinki Declaration of 1975.

Data Availability:

The data for this study cannot be made publicly available because of ethical limitations on data sharing through the European Union’s General Data Protection Regulation (GDPR). Application from researchers to access data from the Swedish Twin Registry is via the register website: https://ki.se/en/research/swedish-twin-registry-for-researchers.

References

  1. Aguirre Castaneda RL, Kumar S, Voigt RG, Leibson CL, Barbaresi WJ, Weaver AL, . . . , & Katusic SK (2016). Childhood attention-deficit/hyperactivity disorder, sex, and obesity. Mayo Clinic Proceedings, 91(3), 352–361. 10.1016/j.mayocp.2015.09.017 [DOI] [PMC free article] [PubMed] [Google Scholar]
  2. Akaike H (1987). Factor analysis and AIC. Psychometrika, 52, 317–332. 10.1007/BF02294359 [DOI] [Google Scholar]
  3. American Psychiatric Association. (1994). Diagnostic and Statistical Manual of Mental Disorders: DSM-IV (4th ed.). American Psychiatric Press. [Google Scholar]
  4. Baker JH, Brosof LC, Munn-Chernoff MA, Lichtenstein P, Larsson H, Maes HH, & Kendler KS (2018). Associations between alcohol involvement and drive for thinness and body dissatisfaction in adolescent twins: A bivariate twin study. Alcoholism: Clinical and Experimental Research, 42(11), 2214–2223. 10.1111/acer.13868 [DOI] [PMC free article] [PubMed] [Google Scholar]
  5. Baker JH, Munn-Chernoff MA, Lichtenstein P, Larsson H, Maes H, & Kendler KS (2017). Shared familial risk between bulimic symptoms and alcohol involvement during adolescence. Journal of Abnormal Psychology, 126(5), 506–518. 10.1037/abn0000268 [DOI] [PMC free article] [PubMed] [Google Scholar]
  6. Bowring AL, Peeters A, Freak-Poli R, Lim MSC, Gouillou M, & Hellard M (2012). Measuring the accuracy of self-reported height and weight in a community-based sample of young people. BMC Medical Research Methodology, 12(1), 175. 10.1186/1471-2288-12-175 [DOI] [PMC free article] [PubMed] [Google Scholar]
  7. Brikell I, Kuja-Halkola R, & Larsson H (2015). Heritability of attention-deficit hyperactivity disorder in adults. American Journal of Medical Genetics Part B: Neuropsychiatric Genetics, 168(6), 406–413. 10.1002/ajmg.b.32335 [DOI] [PubMed] [Google Scholar]
  8. Byrd HCM, Curtin C, & Anderson SE (2013). Attention-deficit/hyperactivity disorder and obesity in US males and females, age 8–15 years: National Health and Nutrition Examination Survey 2001–2004. Pediatric Obesity, 8(6), 445–453. [DOI] [PMC free article] [PubMed] [Google Scholar]
  9. Chen Q, Hartman CA, Kuja-Halkola R, Faraone SV, Almqvist C, & Larsson H (2019). Attention-deficit/hyperactivity disorder and clinically diagnosed obesity in adolescence and young adulthood: A register-based study in Sweden. Psychological Medicine, 49(11), 1841–1849. 10.1017/S0033291718002532 [DOI] [PMC free article] [PubMed] [Google Scholar]
  10. Chen Q, Kuja-Halkola R, Sjölander A, Serlachius E, Cortese S, Faraone SV, . . . , & Larsson H (2017). Shared familial risk factors between attention-deficit/hyperactivity disorder and overweight/obesity—A population-based familial coaggregation study in Sweden. Journal of Child Psychology and Psychiatry, 58(6), 711–718. 10.1111/jcpp.12686 [DOI] [PubMed] [Google Scholar]
  11. Cnattingius S, Ericson A, Gunnarskog J, & Källén B (1990). A quality study of a medical birth registry. Scandinavian Journal of Social Medicine, 18(2), 143–148. 10.1177/140349489001800209 [DOI] [PubMed] [Google Scholar]
  12. Cortese S, Moreira-Maia CR, St. Fleur D, Morcillo-Peñalver C, Rohde LA, & Faraone SV (2016). Association between ADHD and obesity: A systematic review and meta-analysis. American Journal of Psychiatry, 173(1), 34–43. 10.1176/appi.ajp.2015.15020266 [DOI] [PubMed] [Google Scholar]
  13. Cronk NJ, Slutske WS, Madden PAF, Bucholz KK, Reich W, & Heath AC (2002). Emotional and behavioral problems among female twins: An evaluation of the equal environments assumption. Journal of the American Academy of Child & Adolescent Psychiatry, 41(7), 829–837. 10.1097/00004583-200207000-00016 [DOI] [PubMed] [Google Scholar]
  14. Demontis D, Walters RK, Martin J, Mattheisen M, Als TD, Agerbo E, . . . , & Neale BM (2019). Discovery of the first genome-wide significant risk loci for attention deficit/hyperactivity disorder. Nature Genetics, 51(1), 63–75. 10.1038/s41588-018-0269-7 [DOI] [PMC free article] [PubMed] [Google Scholar]
  15. Do EK, Haberstick BC, Williams RB, Lessem JM, Smolen A, Siegler IC, & Fuemmeler BF (2019). The role of genetic and environmental influences on the association between childhood ADHD symptoms and BMI. International Journal of Obesity, 43(1), 33–42. 10.1038/s41366-018-0236-5 [DOI] [PMC free article] [PubMed] [Google Scholar]
  16. Donnchadha SÓ, Bramham J, & Greene C (2020). Rethinking the association between overweight/obesity and ADHD in children: A longitudinal and psychosocial perspective. Irish Journal of Psychological Medicine, 1–14. 10.1017/ipm.2019.61 [DOI] [PubMed] [Google Scholar]
  17. Ekman JT, & Gustafsson PA (2000). Stimulants in AD/HD, a controversial treatment only in Sweden? European Child & Adolescent Psychiatry, 9(4), 312–313. 10.1007/s007870070036 [DOI] [PubMed] [Google Scholar]
  18. Elks CE, den Hoed M, Zhao JH, Sharp SJ, Wareham NJ, Loos RJF, & Ong KK (2012). Variability in the heritability of body mass index: A systematic review and meta-regression. Frontiers in Endocrinology, 3. 10.3389/fendo.2012.00029 [DOI] [PMC free article] [PubMed] [Google Scholar]
  19. Faraone SV, & Larsson H (2019). Genetics of attention deficit hyperactivity disorder. Molecular Psychiatry, 24(4), 562–575. 10.1038/s41380-018-0070-0 [DOI] [PMC free article] [PubMed] [Google Scholar]
  20. Faraone SV, Perlis RH, Doyle AE, Smoller JW, Goralnick JJ, Holmgren MA, & Sklar P (2005). Molecular genetics of attention-deficit/hyperactivity disorder. Biological Psychiatry, 57(11), 1313–1323. 10.1016/j.biopsych.2004.11.024 [DOI] [PubMed] [Google Scholar]
  21. Freitag CM, Rohde LA, Lempp T, & Romanos M (2010). Phenotypic and measurement influences on heritability estimates in childhood ADHD. European Child & Adolescent Psychiatry, 19(3), 311–323. 10.1007/s00787-010-0097-5 [DOI] [PubMed] [Google Scholar]
  22. Geuijen PM, Buitelaar JK, Fliers EA, Maras A, Schweren LJS, Oosterlaan J, . . . , & Rommelse NN (2019). Overweight in family members of probands with ADHD. European Child & Adolescent Psychiatry, 28(12), 1659–1669. 10.1007/s00787-019-01331-7 [DOI] [PMC free article] [PubMed] [Google Scholar]
  23. Instanes JT, Klungsøyr K, Halmøy A, Fasmer OB, & Haavik J (2018). Adult ADHD and comorbid somatic disease: A systematic literature review. Journal of Attention Disorders, 22(3), 203–228. 10.1177/1087054716669589 [DOI] [PMC free article] [PubMed] [Google Scholar]
  24. Javaras KN, Armstrong JM, Klein MH, Essex MJ, Davidson RJ, & Goldsmith HH (2020). Sex differences in the relationship between childhood self-regulation and adolescent adiposity. Obesity, 28(9), 1761–1769. 10.1002/oby.22926 [DOI] [PMC free article] [PubMed] [Google Scholar]
  25. Javaras KN, Hudson JI, & Laird NM (2009). Fitting ACE structural equation models to case-control family data. Genetic Epidemiology, 34(3), 238–245. 10.1002/gepi.20454 [DOI] [PMC free article] [PubMed] [Google Scholar]
  26. Kase BE, Rommelse N, Chen Q, Li L, Andersson A, Du Rietz E, . . . , & Hartman CA (2021). Longitudinal associations between symptoms of ADHD and BMI from late childhood to early adulthood. Pediatrics, 147(6), e2020036657. 10.1542/peds.2020-036657 [DOI] [PubMed] [Google Scholar]
  27. Kim J, Mutyala B, Agiovlasitis S, & Fernhall B (2011). Health behaviors and obesity among US children with attention deficit hyperactivity disorder by gender and medication use. Preventive Medicine, 52(3), 218–222. 10.1016/j.ypmed.2011.01.003 [DOI] [PubMed] [Google Scholar]
  28. Koren R, Munn-Chernoff MA, Duncan AE, Bucholz KK, Madden PAF, Heath AC, & Agrawal A (2014). Is the relationship between binge eating episodes and personality attributable to genetic factors? Twin Research and Human Genetics, 17(2), 65–71. 10.1017/thg.2013.92 [DOI] [PMC free article] [PubMed] [Google Scholar]
  29. Lajunen H-R, Kaprio J, Keski-Rahkonen A, Rose RJ, Pulkkinen L, Rissanen A, & Silventoinen K (2009). Genetic and environmental effects on body mass index during adolescence: A prospective study among Finnish twins. International Journal of Obesity, 33(5), 559–567. 10.1038/ijo.2009.51 [DOI] [PMC free article] [PubMed] [Google Scholar]
  30. Larsson H, Dilshad R, Lichtenstein P, & Barker ED (2011). Developmental trajectories of DSM-IV symptoms of attention-deficit/hyperactivity disorder: Genetic effects, family risk and associated psychopathology. Journal of Child Psychology and Psychiatry, 52(9), 954–963. 10.1111/j.1469-7610.2011.02379.x [DOI] [PubMed] [Google Scholar]
  31. Larsson H, Lichtenstein P, & Larsson J-O (2006). Genetic contributions to the development of ADHD subtypes from childhood to adolescence. Journal of the American Academy of Child & Adolescent Psychiatry, 45(8), 973–981. 10.1097/01.chi.0000222787.57100.d8 [DOI] [PubMed] [Google Scholar]
  32. Li T, Mota NR, Galesloot TE, Bralten J, Buitelaar JK, IntHout J, . . . , & Franke B (2019). ADHD symptoms in the adult general population are associated with factors linked to ADHD in adult patients. European Neuropsychopharmacology, 29(10), 1117–1126. 10.1016/j.euroneuro.2019.07.136 [DOI] [PubMed] [Google Scholar]
  33. Lichtenstein P, Tuvblad C, Larsson H, & Carlström E (2007). The Swedish Twin study of CHild and Adolescent Development: The TCHAD-study. Twin Research and Human Genetics, 10(1), 67–73. 10.1375/twin.10.1.67 [DOI] [PubMed] [Google Scholar]
  34. Liu L-L, Li B-M, Yang J, & Wang Y-W (2008). Does dopaminergic reward system contribute to explaining comorbidity obesity and ADHD? Medical Hypotheses, 70(6), 1118–1120. 10.1016/j.mehy.2007.10.012 [DOI] [PubMed] [Google Scholar]
  35. Merwood A, Greven CU, Price TS, Rijsdijk F, Kuntsi J, McLoughlin G, . . . , & Asherson PJ (2013). Different heritabilities but shared etiological influences for parent, teacher and self-ratings of ADHD symptoms: An adolescent twin study. Psychological Medicine, 43(9), 1973–1984. 10.1017/S0033291712002978 [DOI] [PMC free article] [PubMed] [Google Scholar]
  36. Min J, Chiu DT, & Wang Y (2013). Variation in the heritability of body mass index based on diverse twin studies: A systematic review. Obesity Reviews, 14(11), 871–882. 10.1111/obr.12065 [DOI] [PMC free article] [PubMed] [Google Scholar]
  37. Mota NR, Poelmans G, Klein M, Torrico B, Fernàndez-Castillo N, Cormand B, . . . , & Arias Vásquez A (2020). Cross-disorder genetic analyses implicate dopaminergic signaling as a biological link between Attention-Deficit/Hyperactivity Disorder and obesity measures. Neuropsychopharmacology, 45(7), 1188–1195. 10.1038/s41386-019-0592-4 [DOI] [PMC free article] [PubMed] [Google Scholar]
  38. Munn MA, Stallings MC, Hyun Rhee S, Sobik LE, Corley RP, Rhea SA, & Hewitt JK (2010). Bivariate analysis of disordered eating characteristics in adolescence and young adulthood. International Journal of Eating Disorders, 43(8), 751–761. 10.1002/eat.20854 [DOI] [PMC free article] [PubMed] [Google Scholar]
  39. Neale MC, Boxer SM, Xie G, & Maes HH (2006). Mx: Statistical Modeling (7th ed.). Department of Psychiatry. [Google Scholar]
  40. Neale MC, & Cardon LR (1992). Methodology for genetic studies of twins and families. Kluwer Academic. [Google Scholar]
  41. Ng M, Fleming T, Robinson M, Thomson B, Graetz N, Margono C, . . . , & Gakidou E (2014). Global, regional, and national prevalence of overweight and obesity in children and adults during 1980–2013: A systematic analysis for the Global Burden of Disease Study 2013. The Lancet, 384(9945), 766–781. 10.1016/S0140-6736(14)60460-8 [DOI] [PMC free article] [PubMed] [Google Scholar]
  42. Nigg JT, Johnstone JM, Musser ED, Long HG, Willoughby MT, & Shannon J (2016). Attention-deficit/hyperactivity disorder (ADHD) and being overweight/obesity: New data and meta-analysis. Clinical Psychology Review, 43, 67–79. 10.1016/j.cpr.2015.11.005 [DOI] [PMC free article] [PubMed] [Google Scholar]
  43. Obel C, Olsen J, Henriksen TB, Rodriguez A, Jarvelin M-R, Moilanen I, . . . , & Gissler M (2011). Is maternal smoking during pregnancy a risk factor for Hyperkinetic disorder?—Findings from a sibling design. International Journal of Epidemiology, 40(2), 338–345. 10.1093/ije/dyq185 [DOI] [PubMed] [Google Scholar]
  44. Pauli-Pott U, Reinhardt A, Bagus E, Wollenberg B, Schroer A, Heinzel-Gutenbrunner M, & Becker K (2017). Psychosocial risk factors underlie the link between attention deficit hyperactivity symptoms and overweight at school entry. European Child & Adolescent Psychiatry, 26(1), 67–73. 10.1007/s00787-016-0870-1 [DOI] [PubMed] [Google Scholar]
  45. Polanczyk GV, Salum GA, Sugaya LS, Caye A, & Rohde LA (2015). A meta-analysis of the worldwide prevalence of mental disorders in children and adolescents. Journal of Child Psychology and Psychiatry, 56(3), 345–365. 10.1111/jcpp.12381 [DOI] [PubMed] [Google Scholar]
  46. Polderman TJC, Derks EM, Hudziak JJ, Verhulst FC, Posthuma D, & Boomsma DI (2007). Across the continuum of attention skills: A twin study of the SWAN ADHD rating scale. Journal of Child Psychology and Psychiatry, 48(11), 1080–1087. 10.1111/j.1469-7610.2007.01783.x [DOI] [PubMed] [Google Scholar]
  47. Porter PA, Henry LN, Halkett A, & Hinshaw SP (2021). Body mass indices of girls with and without ADHD: Developmental trajectories from childhood to adulthood. Journal of Clinical Child & Adolescent Psychology, 1–13. 10.1080/15374416.2020.1852942 [DOI] [PMC free article] [PubMed] [Google Scholar]
  48. Schwartz BS, Bailey-Davis L, Bandeen-Roche K, Pollak J, Hirsch AG, Nau C, . . . , & Glass TA (2014). Attention deficit disorder, stimulant use, and childhood body mass index trajectory. Pediatrics, 133(4), 668–676. 10.1542/peds.2013-3427 [DOI] [PMC free article] [PubMed] [Google Scholar]
  49. Sherry B, Jefferds ME, & Grummer-Strawn LM (2007). Accuracy of adolescent self-report of height and weight in assessing overweight status: A literature review. Archives of Pediatrics & Adolescent Medicine, 161(12), 1154–1161. 10.1001/archpedi.161.12.1154 [DOI] [PubMed] [Google Scholar]
  50. Shiels K, & Hawk LW (2010). Self-regulation in ADHD: The role of error processing. Clinical Psychology Review, 30(8), 951–961. 10.1016/j.cpr.2010.06.010 [DOI] [PMC free article] [PubMed] [Google Scholar]
  51. Silventoinen K, Rokholm B, Kaprio J, & Sørensen TIA (2010). The genetic and environmental influences on childhood obesity: A systematic review of twin and adoption studies. International Journal of Obesity, 34(1), 29–40. 10.1038/ijo.2009.177 [DOI] [PubMed] [Google Scholar]
  52. Skoglund C, Chen Q, D′Onofrio BM, Lichtenstein P, & Larsson H (2014). Familial confounding of the association between maternal smoking during pregnancy and ADHD in offspring. Journal of Child Psychology and Psychiatry, 55(1), 61–68. 10.1111/jcpp.12124 [DOI] [PMC free article] [PubMed] [Google Scholar]
  53. Sullivan PF, & Eaves LJ (2002). Evaluation of analyses of univariate discrete twin data. Behavior Genetics, 32(3), 221–227. 10.1023/a:1016025229858 [DOI] [PubMed] [Google Scholar]
  54. van Egmond-Fröhlich AWA, Widhalm K, & de Zwaan M (2012). Association of symptoms of attention-deficit/hyperactivity disorder with childhood overweight adjusted for confounding parental variables. International Journal of Obesity, 36(7), 963–968. 10.1038/ijo.2012.78 [DOI] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Data Availability Statement

The data for this study cannot be made publicly available because of ethical limitations on data sharing through the European Union’s General Data Protection Regulation (GDPR). Application from researchers to access data from the Swedish Twin Registry is via the register website: https://ki.se/en/research/swedish-twin-registry-for-researchers.

RESOURCES