European Child & Adolescent Psychiatry

, Volume 21, Issue 1, pp 39–49 | Cite as

Examining the relationship between Attention-Deficit/Hyperactivity Disorder and overweight in children and adolescents

  • Michael Erhart
  • Beate Herpertz-Dahlmann
  • Nora Wille
  • Barbara Sawitzky-Rose
  • Heike Hölling
  • Ulrike Ravens-Sieberer
Original Contribution

Abstract

Although a higher prevalence of overweight/obesity was reported in clinical samples of patients with Attention-Deficit/Hyperactivity Disorder (ADHD), an association between overweight and ADHD has yet not been established in the general population in childhood. As both disorders are common and significantly affect psychosocial functioning, we investigated the prevalence of ADHD in overweight/obese youth and vice versa. In a cross-sectional nationally representative and community based survey 2,863 parents and their children aged 11–17 years rated symptoms on the Diagnostic and Statistical Manual of Mental Disorders-based German ADHD Rating scale. Weight and height were assessed by professionals. Body mass index was categorized according to national age and sex specific reference values. Overall, 4.2% of the respondents met criteria for ADHD. The prevalence of ADHD was significantly higher for overweight/obese (7%) than for normal weight (3.5%) and underweight (4.9%) children. In a logistic regression analysis controlling for age, gender, and socio-economic status, overweight/obese children were twice as likely to have an ADHD diagnosis (OR = 2.0). Vice versa, adjusting for all covariates, children with ADHD had an OR of 1.9 for overweight/obesity status. For all weight-status groups, children with ADHD more frequently reported eating problems as compared to their non-clinical counterparts. Overweight/obese respondents with ADHD displayed the highest level of health services utilization. A clinician should be aware of the significant risk for a child with ADHD to become overweight and for an overweight child to have ADHD. Longitudinal studies are needed to better understand the mechanisms underlying the association between ADHD and overweight/obesity.

Keywords

Children and adolescents Attention-deficit/hyperactivity disorder Overweight/obesity Eating problems Epidemiological study 

Introduction

Attention/Deficit Hyperactivity Disorder (ADHD) is one of the most common psychiatric conditions in childhood. It affects 3–10% of school-aged children [1], according to the classification criteria of the international classification of diseases (ICD-10) or Diagnostic and Statistical Manual of Mental Disorders (DSM-IV), respectively. ADHD is associated with a wide range of comorbidities, such as learning disorders, conduct issues, and mood and anxiety disorders.

In recent decades, childhood obesity has been identified as a rapidly increasing health problem in modern society. According to the National Health and Examination Survey (NHANES), 18.8% of 6–11-year-old and 17.4% of 12–19-year-old youth living in the USA are overweight [2]. According to the German Health Interview and Examination Survey for Children and Adolescents (KiGGS), 15% of German youth aged 3–17 years are overweight [3]. In treatment-seeking samples of overweight and obese children versus normal weight children, higher rates of psychiatric disorders have been reported repeatedly; however, epidemiological investigations have not always revealed increased rates of psychopathology in overweight/obese children [4]. However, comorbidity between ADHD and overweight/obesity has been explicitly studied only in recent years.

In recent reviews of the literature, all reviewed clinical studies reported a higher than average body weight in participants with ADHD [5, 6]. Similar findings were observed in large epidemiologic samples of the adult population, e.g. of the USA [7] and Germany [8].

However, there is a relative dearth of studies that investigate the association in question in larger epidemiological samples with children and adolescents [9]. Anderson et al. [10] found BMI z-scores were significantly higher in children and adolescents with disruptive disorders than in unaffected controls. Fuemmeler et al. [11] found self-reported ADHD symptoms in adolescence were associated with a 1.5-fold higher risk for early adulthood obesity and change in BMI from adolescence to early adulthood compared to those without ADHD symptoms. Lam and Yang [9] found 11–17-year-old patients with clinically high rates of ADHD symptoms showed a 1.4-fold higher risk for obesity compared to those with lower rates. However, a formal diagnosis of ADHD was not made, and information was obtained by the probands only.

Investigations based on clinical samples of obese subjects report higher prevalence levels of ADHD in obese subjects than what would be expected by chance [5]. However, studies with non-clinical samples have not confirmed this observation [12].

It was the aim of the present study to explore associations between ADHD and overweight/obesity in a large pediatric non-clinical sample. Comorbidity was assessed with respect to both the prevalence of ADHD in overweight and obese subjects and the prevalence of overweight/obesity in subjects with ADHD. ADHD problems were assessed according to ICD-10 and DSM-IV criteria. Problematic eating behavior patterns were studied as well.

Based on the review of the literature, we hypothesized that:
  1. 1.

    children and adolescents with overweight and obesity would be at higher risk for ADHD problems,

     
  2. 2.

    children and adolescents with ADHD problems would be at higher risk for overweight and obesity,

     
  3. 3.

    there would be differences in problematic eating behavior patterns between children and adolescents with and without ADHD problems.

     

Methods

Design and procedures

This study is part of the BELLA study mental health module of the KiGGS. The KiGGS study is a cross-sectional, nationally representative, and community-based survey that examined children and adolescents aged 0–17 years. The data were collected from May 2003 to May 2006 in 167 representative sample locations all over Germany. The objectives, procedures, design, and instruments of the KiGGS are described in detail elsewhere [13]. All participants were medically and physically examined and tested. Parents and children aged 11 years and older were asked to complete the KIGGS self-report questionnaire while visiting the KIGGS examination center. This questionnaire was comprised of several questions pertaining to demographic data, psychosocial well-being, bodily pain and life circumstances. In addition, the children and adolescents were asked to cooperate in the standardized BELLA telephone interview and to fill in the BELLA questionnaire approximately 3 weeks after their visit to the examination center. The questionnaire was sent to the participants shortly after the telephone interview had taken place. One parent also was interviewed with a computer-assisted telephone script and answered an additional questionnaire. The ethics committee of the Charité Universitätsmedizin Berlin approved the study. Written consent was obtained from all parents and from children aged 11 and older. The conceptualization, design, and procedure of the BELLA study and the KiGGS study are described elsewhere [14].

Sample of the present study

For the BELLA mental health module, a sub-sample of subjects was selected from the KiGGS study’s national representative sample of 17,641 families. In the KiGGS study, the overall response rate was 66.6%. A random selection of 4,199 families with children aged 7–17 years were asked to participate in the BELLA-study on mental disorders. Of those eligible families, 70% gave informed consent to participate and 68% (1,389 girls and 1,474 boys) were examined. Due to missing parental information, the actual sample for this paper consisted of 2,414 respondents with complete information. It was examined if there are systematic differences in age, gender, socio-economic status, weight status, and prevalence of ADHD between the actual study sample and those KiGGS respondents asked to participate the BELLA sample.

Instruments

Attention deficit/hyperactivity symptoms were assessed with the German ADHD Rating scale (FBB-HKS/ADHS). The FBB-HKS includes 20 items describing the symptom criteria from the DSM-IV (Inattention: no close attention to details; has difficulty sustaining attention; does not seem to listen; fails to finish work; has difficulty organizing tasks and activities; avoids tasks that require mental effort; loses things necessary for tasks or activities; is easily distracted; is forgetful in daily activities. Hyperactivity: fidgets with hands or feet or squirms in seat; leaves seat in classroom; has difficulty playing quietly; runs about or climbs excessively; often sense of an extreme internal restlessness; permanently extremely restless; is “on the go” or acts as if driven by a motor. Impulsivity: blurts out answers; has difficulty awaiting turn; interrupts or intrudes on others; talks excessively). Additional items assess symptom onset, duration, pervasiveness, and functional impairment. Parents indicated the frequency of each statement or symptom on a 4-point scale ranging from never or rarely (0) to very often (3), with higher scores indicating greater ADHD-related behavior. The mean item score was calculated for every dimension. Patients were classified according to the DSM-IV criteria for ADHD, including subtypes of the disorder [predominantly inattentive, predominantly hyperactive-impulsive, combined e.g.: 6 of 9 symptoms of inattention rated as “often” or “very often”, and/or 6 of 9 symptoms of hyperactivity-impulsivity rated as “often” or “very often”; and another type of ADHD (refers to the DSM-IV code 314.9 “Attention-Deficit/Hyperactivity Disorder not otherwise specified”)]. [15]. The FBB-HKS/ADHS is part of the comprehensive Diagnostic System for Mental Disorders in Childhood and Adolescence (DISYPS) [16]. Responders were classified as either displaying the criteria for any ADHD subtype or not. The instrument allows us to assess those symptoms that are specified as the criteria for ADHD by the DSM-IV. Disorders that may mimic ADHD symptoms cannot be easily detected by the instrumentarium.

ADHD medication use was assessed using structured interviews with the parents. Trained physicians asked the parents about children’s current medications. All reported medications were classified according to the ICD-10’s list of indications for drugs. Consumption of any substance used for the treatment of ADHD resulted in a classification of currently on ADHD medication.

Parents were asked if their child had been ever diagnosed as suffering from ADHD by a physician or a psychologist.

Emotional and behavioral problems

The Strengths and Difficulties Questionnaire assesses positive and negative attributes with 25 items that focus on emotional symptoms, conduct problems, hyperactivity/inattention, peer relationship problems, and prosocial behavior. Each of the 25 items of the SDQ is scored on a 0–2-point scale with higher scores indicating higher symptom levels. The prosocial behavior scale was not used for the current analysis. Items of the four problem areas were summed up to generate a total difficulties score [17].

The SDQ impact supplement asks whether the respondent thinks that the young person has problems in at least one of the areas emotions, attention, behavior, or being able to get along with other people. If at least “minor” problems are reported, then the respondent is asked further questions about associated distress as well as social impairment in home life, friendships, classroom learning, and leisure activities [18]. The SDQ predictive algorithm uses the information gathered from the symptom and impact scores from parents and children to predict psychiatric disorders, including ADHD, conduct disorders or emotional disorders as unlikely, possible and probable [19].

Socio-economic status (SES) was assessed with the Winkler-Index, which classifies the families of the respondents into those with low, medium, and high SES. The index takes income, education, and parental employment into account [20].

Body weight and height were measured by trained staff in the examination center. BMI values (weight in kilograms divided by height in meters squared) were interpreted according to national age- and sex-specific reference values derived from 17 ‘historic’ studies including about 34,000 German children [21]. In accordance with scientific standards, we used the 90th and 97th percentiles to identify overweight and obesity, respectively [22].

Chronic somatic conditions of the children were assessed using structured interviews with the parents. Trained physicians asked the parent about their children’s current and past chronic health condition.

Eating behavior problems were assessed with two items taken from the SCOFF Questionnaire [23]: “Do you worry you have lost control over how much you eat?” and “Would you say that food dominates your life?” Both items could be answered with yes or no. These questions were answered by the children themselves.

Health system utilization was assessed by the number of visits to a physician, psychologist or psychotherapist in the past 12 months. This information was gathered from the parents of children aged 7–13 and from the children themselves who were at least 14 years old.

The Investigators were physicians and non-physician health professionals. Investigators were trained in handling the research tools like, e.g. structured interviews, telephone interviews, physical measures, etc. Within a pretest on 800 subjects investigators abilities were examined and re-trained if necessary. Detailed information are available elsewhere [13, 14].

Statistical analyses

Descriptive statistical analyses included calculation of frequencies, mean and standard deviations. It was examined if KiGGS respondents who refused to participate the BELLA module differed from those who participated. The propensity of refusal was modeled by a logistic regression formula including age, gender, socio-economic status (Winkler Index), BMI categories, and being diagnosed with ADHD. Cross-tabulations of ADHD and Obesity were tested for statistical significance using chi-square tests. Analyses were repeated for age groups and gender.

A hierarchical series of logistic regression models was conducted. ADHD status was regressed on BMI weight status (Model 1), controlling for age, gender, and SES (Model 2). To control for the potential effects of ADHD medication on weight status, the later analyses were repeated for all respondents who did not report ADHD medication use.

Problematic eating behavior was cross-tabulated against all possible weight-status × ADHD combinations. Using a logistic regression formula, problematic eating behavior was regressed on ADHD, controlling for weight-status, age, gender, and SES.

All statistical analyses are based on weighted sample data to represent the age-, gender-, regional- and citizenship-structure of the German population (reference data from 12.31.2004). The number of cases reported in the tables and text refer to this weighted data and thus might deviate from the number of cases reported in the earlier description of the sample.

The actual sample size of more than 2,400 responders allowed for the detection of associations between weight status and ADHD at a small magnitude (w-effect size measure of 0.1 [24]) at an alpha level of P = 0.05 and power of 0.99.

All analyses were conducted with SPSS software, version 15. The power calculation was performed with GPower.

Results

Overall, the data of 2,414 children and adolescents with complete information on ADHD status could be analyzed. The propensity of refusal was modeled by a logistic regression formula including age, gender, socio-economic status (Winkler Index), BMI categories, and being diagnosed as ADHD. Those who refused to participate were more likely to be older (OR = 1.10 for a 1 year difference), and to be living in a family with lower- (OR = 1.99) or medium (OR = 1.19) socio-economic status. Refusal to participate was in no way associated with BMI status or ADHD diagnosis. A total of 101 responders (4.2%) could be classified as ADHD cases based on the aforementioned criteria. Regarding subtypes, 46 (1.9%) of the ADHD cases were predominantly inattentive, 9 (0.4%) were predominantly hyperactive-impulsive, 15 (0.6%) were combined and 31 (1.3%) were “another type of ADHD”.

Table 1 shows the socio-demographic and socio-economic characteristics of the sample stratified for responders with and without ADHD. Responders with ADHD were more likely than those who did not meet criteria to be younger, male, non-immigrants and to display emotional problems. There was a statistical tendency for responders with ADHD to be more likely to live in a family with low SES. About 4% of the ADHD respondents were classified as extremely underweight compared to 2.2% of the non-ADHD respondents. Overweight was observed in 19.8% of the children with ADHD; an additional 8.9% of ADHD children displayed obesity. Of the respondents without ADHD, only 10.0% were overweight, and an additional 6.7% were obese. According to national German norms [21], the mean BMI z score of the 101 respondents with ADHD was 0.51.
Table 1

Socio-demographic and socio-economic characteristics of participants with and without ADHD

 

No ADHD (N = 2,313)

ADHDa (N = 101)

P

Mean age (SD)

12.68 (3.21)

11.98 (3.19)

0.033 (t test)

7–10 years

34.3%

44.6%

0.065 (χ2 test)

11–13 years

25.9%

17.8%

 

14–17 years

39.8%

37.6%

 

Females

50.2%

27.7%

0.001 (χ2 test)

With immigration status

9.8%

1.0%

0.001 (χ2 test)

Low socio-economic statusb

22.0%

31.7%

0.059 (χ2 test)

Medium socio-economic status

47.8%

44.6%

 

High socio-economic status

30.3%

23.8%

 

Emotional problems (SDQ)c Unlikely

94.3%

74.9%

0.001 (χ2 test)

Possible

3.8%

19.1%

 

Probable

1.9%

6.1%

 

Extreme underweight (0–3rd percentile)d

2.2%

4.0%

0.010 (χ2 test)

Underweight (3–10th percentile)d

5.8%

5.0%

 

Normal weight (10–90th percentile)d

75.3%

62.4%

 

Overweight (90–97th percentile)d

10.0%

19.8%

 

Obese (97–100th percentile)d

6.7%

8.9%

 

aADHD classification according to the FBB/HKS (any subtype)

bSocio-economic status according to the Winkler Index

cStrengths and Difficulties questionnaire predictive algorithm to predict any emotional psychiatric disorder

dWeight status classification of the body mass index according to the Kromeyer–Hauschild criteria

Table 2 shows the presence of ADHD in 7.4% of the extreme underweight responders, 8.0% of the responders with overweight and 5.5% of the obese responders. For underweight or normal weight responders, lower prevalence rates of 3.6 and 3.5%, respectively, were observed.
Table 2

Percentage of ADHD caseness for different BMI statuses

Girls and boys 7–17 years

Extreme underweight (0–3rd percentile)

Underweight (3–10th percentile)

Normal weight (0–90th percentile)

Overweight (90–97th percentile)

Obese (97–100th percentile)

χ2(P)

No ADHD (n)

50 (92.6%)

134 (96.4%)

1,735 (96.5%)

230 (92.0%)

154 (94.5%)

13.347

ADHD (n)

4 (7.4%)

5 (3.6%)

63 (3.5%)

20 (8.0%)

9 (5.5%)

(0.010)

  

Under- and extreme underweight (0–10th percentile)

Normal weight (10–90th percentile)

Overweight and obese (90–100th percentile)

  

Girls 7–17 years

 No ADHD (n)

 

93 (100.0%)

872 (98.0%)

189 (95.0%)

 

7.443

 ADHD (n)

 

0 (0.0%)

18 (2.0%)

10 (5.0%)

 

(0.012)

Boys 7–17 years

 No ADHD (n)

 

91 (91.0%)

863 (95.0%)

195 (90.7%)

 

8.789

 ADHD (n)

 

9 (9.0%)

45 (5.0%)

20 (9.3%)

 

(0.024)

7–10 years

 No ADHD (n)

 

60 (96.8%)

608 (95.0%)

121 (91.7%)

 

3.000

 ADHD (n)

 

2 (3.2%)

32 (5.0%)

11 (8.3%)

 

(0.223)

11–13 years

 No ADHD (n)

 

49 (92.5%)

419 (97.0%)

127 (98.4%)

 

4.540

 ADHD (n)

 

4 (7.5%)

13 (3.0%)

2 (1.6%)

 

(0.103)

14–17 years

 No ADHD (n)

 

75 (96.2%)

707 (97.5%)

136 (89.5%)

 

21.830

 ADHD (n)

 

3 (3.8%)

18 (2.5%)

16 (10.5%)

 

(<0.001)

ADHD classification according to the FBB/HKS (any subtype)

Weight status classification of the body mass index according to the Kromeyer–Hauschild criteria

Analyses were stratified by age and gender groups. Due to the smaller sample sizes with this classification, the BMI categories underweight and extreme underweight were collapsed. The overweight and obese responders were also analyzed as one category. In girls, 5.0% of the overweight/obese children displayed ADHD, compared to 2.0% of the normal weight children. In boys, 9.3% of the overweight/obese children displayed ADHD, compared to 5.0% of the normal weight responders. The analyses stratified for age groups showed no statistical significant associations between weight status and ADHD for 7–10 and 11–13 year olds. In the 14–17-year-old group, however, 10.5% of the overweight/obese children showed ADHD, compared to 3.8% of the underweight/extreme underweight children and 2.5% of the normal weight responders.

Examinations of the relationships between ADHD subtypes and BMI status were carried out but were limited in reliability due to the small number of cases per subtype. Overweight/obesity was observed for 39% of predominantly inattentive ADHD cases, 22% of predominantly hyperactive-impulsive cases, 13% of combined cases, and 23% of other ADHD cases. There was a statistical tendency (P = 0.060) for a higher percentage of overweight/obesity (39.1%) in the predominantly inattentive type, compared to any other subtype (20.0%).

A hierarchical series of logistic regression analysis was conducted next to predict ADHD presence from weight status. In the base model, ADHD status was regressed on the 3-category weight-status variable. The reference group was normal weight (10–90th percentile). The two focal groups were underweight/extreme underweight (0–10th percentile) and overweight/obese (90–100th percentile). Table 3 shows that overweight/obese responders demonstrated a 2.1 times higher risk (OR) for ADHD compared to the normal weight group.
Table 3

Predicting ADHD with BMI status, controlling for socio-demographic and socio-economic characteristics

 

Covariates

Exp (B)

95% CI limits

Under

Over

Model 1 (N = 2,414)

BMI 10–90th percentile

1.000

  

0–10th percentile

1.316

0.640

2.710

90–100th percentile

2.096

1.333

3.297

Constant

0.036

  

Model 2 (N = 2,414)

BMI 10–90th percentile

1.000

  

0–10th percentile

1.385

0.669

2.869

90–100th percentile

1.959

1.233

3.114

Age

0.934

0.877

0.995

Male

1.000

  

Female

0.381

0.244

0.596

High SES

1.000

  

Medium SES

1.712

0.980

2.990

Low SES

1.290

0.773

2.155

Constant

0.250

  

Model 2 (cases without ADHD medication) (N = 2,376)

BMI 10–90th percentile

1.000

  

0–10th percentile

0.720

0.256

2.022

90–100th percentile

1.663

1.008

2.744

Age

0.904

0.844

0.968

Male

1.000

  

Female

0.373

0.230

0.606

High SES

1.000

  

Medium SES

2.884

1.468

5.664

Low SES

2.274

1.208

4.282

Constant

0.228

  

ADHD classification according to the FBB/HKS (any subtype)

Weight status classification of the body mass index according to the Kromeyer–Hauschild criteria

SES Socio-economic status according to the Winkler Index

In the second model, age, gender, and SES were entered as additional covariates. After adjusting for these covariates, overweight/obese responders still showed a 2.0-fold higher risk for ADHD than their normal weight peers. Higher age and female gender was associated with lower risk for ADHD; the OR for every year of age was 0.9, and the OR for being female was 0.4. Compared to high SES responders, those with medium or low SES had a 1.7 and 1.3-fold higher chance (OR) of ADHD, respectively. In addition, it was tested for potential confounding effects of other somatic disorder. The presence of asthma, bronchitis, atopic dermatitis, cardiovascular disorder, anemia, epilepsia, thyroid disorder, diabetes, scoliosis, migraine and other were included as dummy-coded additional predictors. The OR for 90–100th percentile BMI slightly decreased from 2,096 to 1,733 but was still statistically significant. To control for potential confounding by other mental problems that frequently co-occur with ADHD the SDQ scores emotional-problems, conduct-problems and peer-problems (all categorized as normal–borderline–abnormal) were included as additional predictors. However, the ORs for 90–100th percentile BMI displayed only marginal changed: 2,096 versus 2.024.

The analyses without the other somatic and mental health problems were repeated but this time overweight versus normal weight was used as the outcome, and ADHD served as the predictor. After adjusting for the other covariates, children with ADHD showed a 1.9-fold higher risk for overweight/obesity.

To control for potential effects of ADHD medication on weight status, the latter analyses were repeated without the 38 responders currently taking ADHD medication. Again, overweight/obese responders showed higher risk for ADHD. However, the OR of 1.7 was slightly lower than it had been with medicated children in the sample. The ORs for the other covariates were similar to those obtained in the previous analysis. Higher age (OR = 0.9) and female gender (OR = 0.4) were associated with lower risks for ADHD. A medium (OR = 2.9) or low (OR = 2.3) SES was associated with a higher risk of ADHD; these ORs were larger than those obtained from the analyses with medicated children included. The analyses were repeated with weight status (normal vs. overweight) as the outcome and ADHD as the predictor. After adjusting for the other covariates, children with ADHD were 1.7 times more likely to be overweight/obese than those without the disorder. It was not possible to control for other psychiatric and somatic medication that are known to affect weight. If children with ADHD had a higher chance for the consumption of such medicaments then this could also contribute to their higher risk for obesity.

Responders were classified according to their ADHD status and their weight status (underweight/extreme underweight, normal weight, overweight/obese). Self-reported eating behavior problems were then examined in the resulting six combinations (ADHD × weight status). Table 4 shows that for each weight group a higher proportion of responders with ADHD report “being worried about loss of control over eating”. Using logistic regression, “being worried about loss of control over eating” was regressed on ADHD (yes/no), controlling for weight status, age, gender, and SES. The OR for ADHD was 2.3.
Table 4

ADHD/weight status configuration and eating behavior problems (11–17-year-olds only)

 

N

χ2

P

 

Yes (%)

No (%)

Lost control over how much I eata

 BMI 0–10th percentile

  No ADHD

125

6.4

93.6

116.693

<0.001

  ADHD

7

14.3

85.7

  

 10–90th percentile

  No ADHD

1,113

18.5

81.5

  

  ADHD

31

22.6

77.4

  

 90–100th percentile

  No ADHD

261

44.4

55.6

  

  ADHD

18

61.1

38.9

  

Food dominates lifea

 BMI 0–10th percentile

  No ADHD

125

23.2

76.8

37.509

<0.001

  ADHD

7

14.3

85.7

  

 10–90th percentile

  No ADHD

1,110

21.5

78.5

  

  ADHD

30

36.7

63.3

  

 90–100th percentile

  No ADHD

260

36.5

63.5

  

  ADHD

18

55.6

44.4

  

ADHD classification according to the FBB/HKS (any subtype)

Weight status classification of the body mass index according to the Kromeyer–Hauschild criteria [21]

aItems taken from the SCOFF questionnaire

That “food dominates life” was claimed by 23.2% of underweight/extreme underweight respondents without ADHD and 14.3% of the underweight/extreme underweight respondents with ADHD. However, in all other weight status groups, those with ADHD reported to a greater extent than controls that food dominates their life (see Table 4).

Using logistic regression, “food dominates life” was regressed on ADHD (yes/no), controlling for weight status, age, gender, and SES. The OR for ADHD was 2.1 and statistically significant.

Regarding health services utilization, it was shown that responders with overweight/obesity more frequently visited a physician for somatic reasons than their normal weight or underweight/extremely underweight peers. This ranking could be observed for responders regardless of ADHD status. Overweight/obese children most frequently visited a psychologist, psychotherapist, or psychiatrist. Results are shown in Table 5.
Table 5

ADHD/weight status and health services utilization in the past 12 months (7–17 year olds)

 

ANOVA

N

Mean

SD

F

P

Number of visits to physician (without psychologists/psychotherapists)

 No ADHD

  BMI 0–10th percentile

178

3.46

3.45

4.583

<0.001

  10–90th percentile

1,683

3.97

3.76

  

  90–100th percentile

372

4.55

5.00

  

 ADHD

     

  BMI 0–10th percentile

9

4.27

4.32

  

  10–90th percentile

62

5.38

4.79

  

  90–100th percentile

27

5.94

5.03

  

Number of visits to psychiatrist/psychologists/psychotherapists

 No ADHD

  BMI 0–10th percentile

184

0.25

1.84

2.537

0.012

  10–90th percentile

1,705

0.27

2.30

  

  90–100th percentile

380

0.38

3.49

  

 ADHD

  BMI 0–10th percentile

9

1.47

3.17

  

  10–90th percentile

62

0.73

2.67

  

  90–100th percentile

28

1.69

3.58

  

ADHD classification according to the FBB/HKS (any subtype)

Weight status classification of the body mass index according to the Kromeyer–Hauschild criteria

To test the hypothesis that—impulsivity could foster the development of eating patterns that lead to an increased risk for obesity—we examined the relationship between impulsivity symptom severity in ADHD cases and BMI status (controlled for age and gender in a multiple linear regression). Impulsivity symptoms were associated with higher BMI values—the standardized regression coefficient was β = 0.16 and tendentially statistically significant (P = 0.066) within the 101 ADHD cases. The regression analyses were repeated (a) with severity of inattention and (b) hyperactivity instead of impulsivity. The standardized regression coefficients for inattention (β = 0.06) and hyperactivity (β = 0.11) were smaller and statistically not significant (P = 0.509 and 0.215).

In addition to the analyses presented, we also examined if children and adolescents with ADHD spent more time watching TV/videos, playing computer/video games, or browsing the internet (analyses not reported). After adjusting P values for multiple testing, the only statistically significant result occurred for self-reported time spent playing computer/video games. About 32.8% of the adolescents aged 11–17 with ADHD reported playing computer/video games for 1–2 h and over 2 h a day, compared to 16.0% of their peers without ADHD. After controlling for socio-economic status, however, this difference was no longer statistically significant.

Discussion

To our knowledge, this is the first epidemiological nationwide study using a screening instrument based on formal classification criteria of DSM-IV that demonstrates that overweight and obese children are at a higher risk for ADHD than their normal weight counterparts. Likewise, children with ADHD proved to be at higher risk for overweight than children without ADHD. In extremely underweight children, we also found a tendency towards a higher risk for ADHD. Yet, this tendency was less pronounced and not statistically significant.

The results are based on a population-based sample with a prevalence rate of ADHD (4.2%) that corresponds to other epidemiological studies. A recent review of 102 studies calculated a worldwide prevalence estimate of 5.3% [1]. The examined prevalence of the subtypes (1.9% inattentive, 0.4% hyperactive-impulsive, 0.6% combined, 1.3 any other type) should not be interpreted further because the actual number of cases are very small and our applied screening instrument does not provide a formal ADHD diagnosis.

Our finding of an association between ADHD and overweight/obesity is in line with the results from clinical studies [5, 25, 26]. However, findings in population-based studies have been contradictory. While in one study a diagnosis of ADHD was not associated with overweight or obesity [8], results suggesting a positive connection were obtained by other studies [9, 27]. Interestingly, we found stronger effects than in previous studies. After adjusting for socio-demographic and socio-economic covariates, we found that children and adolescents with ADHD had a 1.9-fold increase in risk for overweight/obesity. Lam and Yang [9] and Waring and Lapane [27] reported 1.4- and 1.5-fold increases, respectively, in the risk for obesity in children with ADHD. The differences might be attributable to the way ADHD problems were measured. We assessed ADHD problems with a standardized DSM-IV based instrument [15]. Waring and Lapane [27] asked the parents “has a doctor or health/professional ever told you that your child has … ADHD…?”. Lam and Yang [9] used a structured interview with questions based on ADHD-related questionnaires. However, they only asked the children themselves. Yet not all children with ADHD might be aware of their disorder. Both of the latter studies found effects only for the obese respondents, while our study found effects for both overweight and obese children.

An intuitive plausible explanation for the relationship between overweight and ADHD has been proposed: impulsivity and poor behavioral regulation as well as the delay aversion associated with ADHD could foster the development of eating patterns that lead to an increased risk for obesity [28]. Children with ADHD might also tend to use food as a means of gratification more than their peers without ADHD [27, 29]. Our additional results are consistent with this assumption: We found that children with ADHD more often endorsed “lost control over eating” and “food dominates life” than their peers without ADHD. This difference was apparent in underweight, normal weight and overweight children. The study of Pagoto et al. [7] found binge eating disorder, but not depression, partially mediates the associations between ADHD and both overweight and obesity. However, their study was on adults with childhood episodes of ADHD assessed retrospectively. In another recent study, we demonstrated that children with ADHD were more delay aversive when confronted with an edible reward compared to other kinds of reward. In addition, children with ADHD ate more snacks than normal or overweight controls. Especially children with ADHD and obesity were prone for immoderate and uncontrolled eating [30]. In other studies, children with ADHD also showed marked deficits on several tasks of delay aversion. Children with ADHD had more difficulties waiting for healthy food when, e.g. fast food was immediately available [31, 32]. Lack of impulse control and impulsivity had been generally shown to play a major role in obesity, binge eating and “loss of control over eating” [33, 34]. Inattention, as another core symptom of ADHD, might limit awareness of hunger and satiety [32, 34]. In a recent study, we found age-based BMI to be best predicted by maternal BMI and behavioral ratings of inattentiveness and impulsivity [30].

Patterns of physical activity could play a complex role in the association between ADHD and overweight. On the one hand, it is intuitively plausible that children with ADHD would have higher physical activity levels than their healthy peers. On the other hand, children and adolescents with ADHD may spend more time playing computer games and watching TV or videos [27].

Another explanation is that of a genetic linkage between ADHD and obesity. A study of people with seasonal affective disorder found that the presence of the 7-repeat allele (7R) of the dopamine 4 receptor gene was associated with higher levels of childhood ADHD symptoms as well as with maximum lifetime BMI. In contrast, the absence of the 7R did not result in higher ADHD scores or maximum lifetime BMI [35]. It has been argued that addiction and food intake are regulated—at least partly—by the same dopamine brain reward mechanism. Some cases of obesity might thus be the consequence of a food addiction, which could be more prevalent in individuals with ADHD [36]. Davis et al. [37] found three DRD3 genotypes associated with significantly elevated ADHD scores. It has been argued that the DRD3 gene is associated with brain regions that play a role in the reward processes characteristic of addictive behavior.

Contrary to other studies [27], we found that the association between overweight and ADHD was less pronounced in children who were not taking ADHD medication. Weight loss and reduced appetite are known side effects of current stimulant medications used to treat ADHD [38, 39]. Thus, children on ADHD medication could be expected to be at a lower risk for obesity than their unmedicated peers with ADHD. However, ADHD medication might also correspond to cases with sever forms of ADHD and especially higher levels of impulsivity. Controlling for ADHD medication could therefore eliminate these children with a greater risk for obesity from the pool of children with ADHD.

Our study has some limitations. There is a considerable attrition rate from the sample of eligible subjects contacted for the KiGGS study to those who were actually examined for the current paper. However, our analyses showed no sign of a systematic deviation about the frequencies of BMI status groups and ADHD caseness. Systematic differences in the age-gender structure are controlled for. Our study is focused on the association between ADHD and overweight/obesity. It is less likely that the results of such analyses are largely affected from non-representativeness of the sample: there is no strong theory suggesting systematic differences in the strength of the ADHD–overweight association between those who participated and those who did not. Another matter of concern is the fact that ADHD was assed using a screening instrument. We acknowledge that it thus is not possible to estimate the true prevalence of ADHD. Yet even in the absence of a formal ADHD diagnosis, our results provide some important indications on the association between ADHD and overweight/adipositas though the exact magnitude of this relationship cannot be validly estimated. Because of the cross-sectional design, we could not determine any chronological relationships between ADHD and obesity. However, such sequelae would be important to know to better understand the possible mechanisms underlying both disorders. In reality, the relationship between ADHD and overweight/obesity might be more complex than the explanations outlined above. Although we analyzed a large sample of the general population, due to the prevalence rates of ADHD and overweight, there was only a small sample of children with ADHD and overweight under study. Thus, the results might be less precise, especially those for the ADHD subtypes. Although current ADHD medication was thoroughly assessed in our study, we had insufficient information on medication history and its potential effects on weight status.

The strengths of our study are large and mostly representative sample from the general population of children and adolescents in Germany—except for a known deviation in the socio-economic status. The calculated prevalence rate of ADHD in this study corresponds well to other worldwide rates. Weight and height were assessed by medical staff, resulting in precise BMI measurement. In contrast to similar studies, ADHD was assessed using a standardized, validated, strictly DSM-IV based instrument with good diagnostic properties.

Conclusions

Our results have some important clinical implications. A clinician should know that overweight/obese children have approximately a twofold higher “risk” (OR) for ADHD than normal weight children. Because of the relationship between ADHD and overweight/obesity, symptoms of ADHD should be actively sought in children presenting with obesity. Clinicians treating obesity should be aware that undiagnosed ADHD might be a potential factor in treatment failure due to a lack of motivation and compliance. Thus, it would be necessary to adjust such programs to the specific needs and characteristics of these children. Children with a diagnosis of ADHD should be monitored for their weight status. Attention should also be given to their eating behavior, especially to loss of control over eating. Diagnosis and treatment of ADHD might lead to an improvement in BMI [40].

Notes

Acknowledgments

The authors would like to thank all coworkers who carried out the field work of the KiGGS and the BELLA study. The authors especially thank all children, adolescents and their parents who participated in the study. The authors Dr. Ravens-Sieberer and Dr. Erhart confirm that they have full access to all the data in the study, and takes responsibility for the integrity of the data and the accuracy of the data analysis.

Conflict of interest

The BELLA study module of the German Health Interview and Examination Survey for Children and adolescents (KiGGS) was financed by a Grant of the German Research Foundation. The KiGGS survey was financed and carried out on the request of the German Ministry of Health. The writing and submission of the manuscript was not contingent on the approval or censorship of the German Research Foundation or the German Ministry of Health. There was no direct financial support for the preparation of the manuscript. Dr. Herpertz-Dahlmann is a consultant to Eli Lilly and has received industry funding from AstraZeneca, Eli Lilly, Novartis and Janssen Cilag. All other authors declare no conflict of interest.

References

  1. 1.
    Polanczyk G, de Lima MS, Horta BL, Biedermann J, Rohde LA (2007) The world wide prevalence of ADHD: A systematic review and metaregression analysis. Am J Psychiatry 164:942–948PubMedCrossRefGoogle Scholar
  2. 2.
    Ogden CL, Carroll MD, Curtin LR, McDowell MA, Tabak CJ, Flegal KM (2006) Prevalence of overweight and obesity in the United States, 1999–2004. JAMA 295:1549–1555PubMedCrossRefGoogle Scholar
  3. 3.
    Kurth B-M, Schaffrath-Rosario A (2007) The prevalence of overweight and obese children and adolescents living in Germany. Results from the German Health Interview and Examination Survey for Children and Adolescents (KiGGS). Bundesgesundheitsblatt Gesundheitsforschung Gesundheitsschutz 50:736–743PubMedCrossRefGoogle Scholar
  4. 4.
    Hebebrand J, Herpertz-Dahlmann B (2009) Psychological und psychiatric aspects of pediatric obesity. Child Adolesc Psychiatr Clin N Am 18:49–65PubMedCrossRefGoogle Scholar
  5. 5.
    Cortese S, Angriman M, Maffeis C, Isnard P, Konofal E, Lecendreux M, Purper-Ouakil D, Vincenzi B, Dalla Bernardina B, Mouren M-C (2008) Attention-Deficit/Hyperactivity Disorder (ADHD) and obesity: a systematic review of the literature. Crit Rev Food Sci Nutr 48:524–537PubMedCrossRefGoogle Scholar
  6. 6.
    Cortese S, Morcillo Peñalver C (2010) Comorbidity between ADHD and obesity: exploring shared mechanisms and clinical implications. Postgrad Med 122:88–96PubMedCrossRefGoogle Scholar
  7. 7.
    Pagoto SL, Curtin C, Lemon SC, Bandini LG, Schneider KL, Bodenlos JS, Ma Y (2009) Association between adult attention deficit/hyperactivity disorder and obesity in the US population. Obesity 17:539–544PubMedCrossRefGoogle Scholar
  8. 8.
    de Zwaan M, Gruss B, Müller A, Philipsen A, Graap H, Martin A, Glaesmer H, Hilbert A (2011) Association between obesity and adult Attention-Deficit/Hyperactivity Disorder in a German community-based sample. Obes Facts 4:204–211PubMedCrossRefGoogle Scholar
  9. 9.
    Lam LT, Yang L (2007) Overweight/obesity and attention deficit and hyperactivity disorder tendency among adolescents in China. Int J Obes Relat Metab Dis 31:584–590Google Scholar
  10. 10.
    Anderson SE, Cohen P, Naumova EN, Must A (2006) Relationship of childhood behavior disorders to weight gain from childhood into adulthood. Ambul Pediatr 6:297–301PubMedCrossRefGoogle Scholar
  11. 11.
    Fuemmeler BF, Østbye T, Yang C, McClernon FJ, Kollins SH (2011) Association between Attention-Deficit/Hyperactivity Disorder symptoms and obesity and hypertension in early adulthood: a population-based study. Int J Obes 35(6):852–862CrossRefGoogle Scholar
  12. 12.
    Mustillo S, Worthman C, Erkanli A, Keeler G, Angold A, Costello EJ (2003) Obesity and psychiatric disorder: developmental trajectories. Pediatrics 111:851–859PubMedCrossRefGoogle Scholar
  13. 13.
    Kurth B-M, Kamtsiuris P, Hoelling H, Schlaud M, Dölle R, Ellert U, Kahl H, Knopf H, Lange M, Mensink GBM, Neuhauser H, Schaffrath-Rosario A, Scheidt-Nave C, Schenk L, Schlack R, Stolzenberg H, Thamm M, Thierfelder W, Wolf U (2008) The challenge of comprehensively mapping children’s health in a nation-wide health survey: design of the German KiGGS-Study. BMC Public Health 8:196PubMedCrossRefGoogle Scholar
  14. 14.
    Ravens-Sieberer U, Kurth B-M (2008) The mental health module (BELLA study) within the German Health Interview and Examination Survey of Children and Adolescents (KiGGS): study design and methods. Eur Child Adolesc Psychiatry 17:10–21PubMedCrossRefGoogle Scholar
  15. 15.
    Döpfner M, Lehmkuhl G (2000) Diagnostik-System für Psychische Störungen im Kindes- und Jugendalter nach ICD-10 und DSM-IV (DISYPS-KJ), 2nd edn. Huber, BernGoogle Scholar
  16. 16.
    Döpfner M, Steinhausen H-C, Coghill D, Dalsgaard S, Poole L, Ralston SJ, Rothenberger A (2006) Cross-cultural reliability and validity of ADHD assessed by the ADHD Rating Scale in a pan-European study. Eur Child Adolesc Psychiatry 15:S46–S55CrossRefGoogle Scholar
  17. 17.
    Goodman R (1997) The Strengths and Difficulties Questionnaire: A research note. J Child Psychol Psychiatry 38:581–586PubMedCrossRefGoogle Scholar
  18. 18.
    Goodman R (1999) The extended version of the Strengths and Difficulties Questionnaire as a guide to child psychiatric caseness and consequent burden. J Child Psychol Psychiatry 40:791–799PubMedCrossRefGoogle Scholar
  19. 19.
    Goodman R, Renfrew D, Mullick M (2000) Predicting type of psychiatric disorder from Strengths and Difficulties Questionnaire (SDQ) scores in child mental health clinics in London and Dhaka. Eur Child Adolesc Psychiatry 9:129–134PubMedCrossRefGoogle Scholar
  20. 20.
    Winkler J, Stolzenberg H (1999) Social status scaling in the German National Health Interview and Examination Survey. Gesundheitswesen 61:S178–S183PubMedGoogle Scholar
  21. 21.
    Kromeyer-Hauschild K, Wabitsch M, Kunze D (2001) Percentiles for the body mass index for children aged 0–18 years. Monatszeitschrift Kinderheilkunde 149:807–818CrossRefGoogle Scholar
  22. 22.
    Herpertz-Dahlmann B (2009) Adolescent eating disorders: definitions, symptomatology, epidemiology and comorbidity. Child Adolesc Psychiatr Clin N Am 18:31–47PubMedCrossRefGoogle Scholar
  23. 23.
    Morgan JF, Reid F, Lacey H (1999) The SCOFF questionnaire: assessment of a new screening tool for eating disorders. Br Med J 319:1467–1468CrossRefGoogle Scholar
  24. 24.
    Cohen J (1988) Statistical power analysis for the behavioural sciences. Erlbaum, New YorkGoogle Scholar
  25. 25.
    Altfas JR (2002) Prevalence of attention deficit/hyperactivity disorder among adults in obesity treatment. BMC Psychiatr 13:9CrossRefGoogle Scholar
  26. 26.
    Holtkamp K, Konrad K, Müller B, Heussen N, Herpertz S, Herpertz-Dahlmann B, Hebebrand J (2004) Overweight and obesity in children with Attention-Deficit/Hyperactivity Disorder. Int J Obes Relat Metab Dis 28:685–689CrossRefGoogle Scholar
  27. 27.
    Waring ME, Lapane KL (2008) Overweight in children and adolescents in relation to Attention-Deficit/Hyperactivity Disoorder: results from a national sample. Pediatrics 122:e1–e6PubMedCrossRefGoogle Scholar
  28. 28.
    Davis C (2010) Attention-Deficit/Hyperactivity Disorder: associations with overeating and obesity. Curr Psychiatr Rep 12(5):389–395CrossRefGoogle Scholar
  29. 29.
    Agranat-Meged A, Ditcher C, Stein M, Leibenson L, Stein M, Galili-Weisstub E (2005) Childhood obesity and attention deficit/hyperactivity disorder: a newly described co-morbidity in obese hospitalised children. Int J Eat Dis 37:357–359CrossRefGoogle Scholar
  30. 30.
    Wilhelm C, Konrad K, Willmes K, Holtkamp K, Marx I, Herpertz-Dahlmann B (2011) Eating behaviour and impulsivity in overweight and obese children with and without attention deficit/hyperactivity disorder (submitted)Google Scholar
  31. 31.
    Bitsakou P, Psychogiou L, Thompson M, Sonuga-Barke EJS (2009) Delay Aversion in Attention Deficit/Hyperactivity Disorder: an empirical investigation of the broader phenotype. Neuropsychologia 47:446–456PubMedCrossRefGoogle Scholar
  32. 32.
    Davis C, Levitan RD, Smith M, Tweed S, Curtis C (2006) Associations among overeating, overweight, and attention deficit/ hyperactivity disorder: a structural equation modeling approach. Eat Behav 7:266–274PubMedCrossRefGoogle Scholar
  33. 33.
    Braet C, Claus L, Verbeken S, Van Vlierberghe L (2007) Impulsivity in overweight children. Eur Child Adolesc Psychiatry 16:473–483PubMedCrossRefGoogle Scholar
  34. 34.
    Nederkoorn C, Jansen E, Mulkens S, Jansen A (2007) Impulsivity predicts treatment outcome in obese children. Behav Res Ther 45:1071–1075PubMedCrossRefGoogle Scholar
  35. 35.
    Levitan RD, Masellis M, Lam RW, Muglia P, Basile VS, Jain U, Kaplan AS, Tharmalingam S, Kennedy SH, Kennedy JL (2004) Childhood inattention and dysphoria and adult obesity associated with dopamine D4 receptor gene in overeating woman with seasonal affective disorder. Neuropsychopharmacology 29:179–196PubMedCrossRefGoogle Scholar
  36. 36.
    Cassin SE, van Ranson KM (2007) Is binge eating experienced as an addiction? Appetite 49:687–690PubMedCrossRefGoogle Scholar
  37. 37.
    Davis C, Patte K, Levitan RD, Carter J, Kaplan AS, Zai C, Reis C, Curtis C, Kennedy JL (2009) A psycho-genetic study of the association between the symptoms of binge eating disorder and those of attention deficit (hyperactivity) disorder. J Psychiatr Res 43:687–696PubMedCrossRefGoogle Scholar
  38. 38.
    Poulton A, Cowell CT (2003) Slowing of growth in height and weight on stimulants: a characteristic pattern. J Paediatr Child Health 39:180–185PubMedCrossRefGoogle Scholar
  39. 39.
    Barkley RA, McMurray MB, Edelbrock CS, Robbins K (1990) Side effects of methylphenidate in children with attention deficit hyperactivity disorder: a systemic, placebo-controlled evaluation. Pediatrics 86:184–192PubMedGoogle Scholar
  40. 40.
    Levy LD, Fleming JP, Klar D (2009) Treatment of refractory obesity in several obese adults following management of newly diagnosed attention deficit hyperactivity disorder. Int J Obes 33:326–334CrossRefGoogle Scholar

Copyright information

© Springer-Verlag 2011

Authors and Affiliations

  • Michael Erhart
    • 1
    • 3
  • Beate Herpertz-Dahlmann
    • 2
  • Nora Wille
    • 1
  • Barbara Sawitzky-Rose
    • 1
  • Heike Hölling
    • 4
  • Ulrike Ravens-Sieberer
    • 5
  1. 1.University Medical Center Hamburg-EppendorfHamburgGermany
  2. 2.University Hospital AachenAachenGermany
  3. 3.Central Research Institute of Ambulatory Health Care in GermanyBerlinGermany
  4. 4.Robert Koch-InstituteBerlinGermany
  5. 5.Research Unit Child Public Health, Department of Child and Adolescent Psychiatry, Psychotherapy, and Psychosomatics University Medical Center Hamburg-EppendorfHamburgGermany

Personalised recommendations