Adolescents with ADHD
Adolescents with ADHD engage in more risk taking behaviour than adolescents without ADHD. Mind-wandering is an underlying process of ADHD and a potential underlying process of risk taking behaviour in ADHD. The aim of this study was to test if mind-wandering predicted risk taking behaviour above and beyond ADHD symptoms. 315 adolescents (age: M =17.2, range = 16 – 22) participated in this study, and filled in a questionnaire about mind-wandering, ADHD symptoms and risk taking behaviour. The results show that ADHD symptoms predict risk taking behaviour and a combination of ADHD symptoms and mind-wandering predict risk taking behaviour. However mind-wandering does not predict risk taking behaviour above and beyond ADHD symptoms. These findings suggest that mind-wandering is not the underlying process of risk taking behaviour in ADHD. In clinical practice there should be a focus on risk taking behaviour in ADHD. Further research should focus on the different characteristics of ADHD as a predictor for risk taking behaviour.
Attention deficit and hyperactivity disorder (ADHD) is a well-known mental disorder, and is a concern for the mental health (Kupfer, 2000). In the fifth edition of the Diagnostic and Statistical Manual of Mental Disorders (DSM-5) ADHD is presented as a neurobiological developmental disorder. The disorder is characterised by a persistent pattern of inattention, hyperactivity and impulsivity, which obstructs daily functioning or development to a significant extent (American Psychiatric Association, 2013). Besides those general ADHD symptoms, it is generally acknowledged that ADHD can be associated with other problems like risk taking behaviour (Mohr-Jensen & Steinhausen, 2016).
Different types of risk taking behaviour are more commonly seen among individuals with ADHD than individuals without ADHD. For example in daily life they engage in more risky sexual behaviour (Flory, Molina, Pelham, Gnagy, & Smith, 2010), substance abuse (Steve S. Lee, Humphreys, Flory, Liu, & Glass, 2011), gambling (Dai, Harrow, Song, Rucklidge, & Grace, 2016), and commit offenses more often than people without ADHD (Mohr-Jensen & Steinhausen, 2016). The beforementioned types of risk taking behaviour can have effect on different levels, both on a societal-level and an individual-level.
Effects on a societal-level and an individual-level caused by risk taking behaviour are mainly destructive. For instance individuals who engage in risk taking behaviour are more likely to be involved in a car accident (Jonah, 1986). Adding to this, the leading cause of deaths for adolescents is risk taking behaviour (Dahl, 2004). Both the car accidents and higher death rates result in higher healthcare costs and can have a great impact on the individual and their family (Harpin, 2005). An individual with ADHD has higher healthcare costs (Hakkaart-van Roijen et al., 2007), and gets expelled from school or dismissed from their job more often (Harpin, 2005). All of these effects can be caused by risk taking behaviour (Dahl, 2004). Unfortunately, the reason why individuals with ADHD engage in more risk taking behaviour than individuals without ADHD is not clear. Individuals with ADHD experience a chronic low arousal, which may be an explanation what might be associated with risk taking behaviour. When an individual experiences chronically low arousal commonly they have problems with controlling their attention as well (Matthies, Philipsen, & Svaldi, 2012). This can possibly be linked to an underlying mental process of ADHD.
Mind-wandering is associated with functional problems and is linked to psychopathologies like ADHD (Franklin et al., 2014). In addition, mind-wandering is an evident underlying mental process of ADHD (Mowlem et al., 2016). The thoughts of people in the general population are 24 to 50 percent of the time unrelated to their environment or current activities (Kane et al., 2016; Killingsworth & Gilbert, 2010). When these unrelated thoughts are deliberate, there is no problem. However, the problem arises when the mind-wandering is spontaneous (Smallwood & Schooler, 2015). Mowlem et al. (2016) studied the prediction of risk taking behaviour through mind-wandering. 108 participants filled in self-report questionnaires, concerning ADHD symptoms, mind-wandering and impairment. Risk taking behaviour was covered in a part of the impairment questionnaire. There is found that mind-wandering is an independent predictor for risk taking behaviour. Mind-wandering may be the underlying process which links risk taking behaviour and ADHD.
Mind-wandering can cause difficulties for someone to focus on a relevant task or event, which can make it more difficult to ensure good performance. Because someone’s thoughts are constantly ‘on the go’, it can be difficult to think through and plan activities. When this happens, possible risks may be overlooked. Furthermore, it is possible that people who engage in a lot of mind-wandering due to chronically low arousal seek distraction to reduce this problem. Risk taking behaviour can be seen as highly salient activities, which results in an increase of arousal. When the arousal is experienced, the excessive mind-wandering will reduce, what leads to a sense of relief (Mowlem et al., 2016). Although Mowlem et al. (2016) shows mind-wandering can predict risk taking behaviour, there needs to be more evidence to show in what extent mind-wandering can predict risk taking behaviour.
In this study, we investigate if mind-wandering can predict risk taking behaviour above and beyond ADHD symptoms. Most research about this topic is focuses on adults (Franklin et al., 2014; Jonkman, Markus, Franklin, & Van Dalfsen, 2017; Mowlem et al., 2016; Seli, Smallwood, Cheyne, & Smilek, 2015). This research however focuses on adolescents, since adolescents engage in more risky decision making than adults (Defoe, Dubas, Figner, & Van Aken, 2015). All the participants fill in self-report questionnaires about ADHD symptoms, mind-wandering, and risk taking behaviour. Firstly, we check whether ADHD symptoms predict risk taking behaviour. Secondly, we check if mind-wandering can predict risk taking behaviour above and beyond ADHD symptoms.
It is expected that mind-wandering can predict risk taking behaviour above and beyond ADHD symptoms. Thereby, we expect a positive prediction of risk taking behaviour through ADHD symptoms. Furthermore, we expect mind-wandering to predict risk taking behaviour above and beyond ADHD symptoms.
315 adolescents participated in this study (age: M =17.2, SD = 1.3, range = 16 – 22, 104 male, 205 female, 1 genderneutral and 6 did not want to report). 49 participants indicated being diagnosed with AD(H)D and 19 participants use AD(H)D medication. The participants were recruited through their public high school and secondary vocational education, which in Dutch is called Middelbaar Beroepsonderwijs. The participants did not get any reward for participating. All the participants signed an informed consent form and were allowed to stop with the questionnaire at any moment.
The ADHD self-report questionnaire (Kooij & Buitelaar, 1997) was used to identify ADHD symptoms. Participants were among other things asked if they felt restless and if they were easily distracted. The questionnaire included 23 items, as five symptoms were reformulated in two items. All the items were scored on a four point Likert scale (0 never or rarely to 3 very often) (cf. Kooij, Boonstra, Swinkels, Bekker, de Noord, & Buitelaar, 2008). The sum of the points were used for the analysis and the maximum score was 69.
The risk taking behaviour self-report questionnaire was used to identify risk taking behaviour. Participants were asked to indicate how often they participated in risk taking behaviour (for example drinking alcohol, having unsafe sex). The questionnaire included 28 items on a five point Likert scale (0 never to 4 weekly). The points were added up for the analysis and the maximum score was 112. The risk taking behaviour self-report questionnaire was based on existing questionnaires (Figner & Weber, 2011; Pat-Horenczyk et al., 2007; Pollak, Poni, Gershy, & Aran, 2017).
The Mind Excessively Wandering Scale (MEWS) (Mowlem et al., 2016) was used to identify mind-wandering. Participants were for example asked if they had difficulty controlling their thoughts and if they had difficulty switching of their thoughts. The MEWS included 12 items on a four point Likert scale (0 never or rarely to 3 almost always or constantly). The points were added up for the analysis and the maximum score was 36. This questionnaire was translated from English to Dutch.
Lastly, the adolescents were asked to fill in some demographical questions: age, sex diagnosis and medication use.
During school hours the participants filled in the questionnaires, instead of their regular curriculum. This study is part of a larger research, therefore the participants filled in four additional questionnaires. In average it took the participants 15 to 25 minutes to fill in the questionnaire completely. The participants were asked to fill in the full questionnaire without leaving any questions unanswered. When the participants had a laptop or a computer at their disposal the questionnaires were filled in online using the website Qualtrics. The remaining participants filled in the questionnaire in a paper booklet. During the session the participants were neither allowed to talk to each other nor use their cell phone.
To analyse the data a multiple regression blockwise entry was used. ADHD symptoms was the independent variable and risk taking behaviour was the dependent variable in the first block. The second block consists of risk taking behaviour as the dependent variable and mind-wandering and ADHD symptoms as the independent variable.
The following assumptions will be checked: (1) the independent variable is continuous; (2) the sample is random; (3) the independent variable has variance; (4) multicollinearity; (5) correlations outside the model; (6) homoscedasticity; (7) uncorrelated residues; (8) normally distributed; (9) independent participants; (10) linearity.
Possible outliers will be detected using the MAD method (Leys, Ley, Klein, Bernard, & Licata, 2013), using the moderately conservative version of 2.5. When there are outliers detected the analysis will be run with and without outliers and both will be reported. When less than 25% of the items of a questionnaire is missing the data will be included, the total score will be imputed by the mean score of the filled-in items of that participant on that questionnaire. When the participant filled in less than 25% of the questionnaire, the data on that questionnaire of that participant will be discarded.
One participant did not fill in the ADHD and risk taking behaviour questionnaire and two participants did not fill in the risk taking behaviour questionnaire. These three participants are not included in the data analysis. The Median Absolute Deviation (MAD) method was used to detect outliers (Leys et al., 2013). Because there were no outliers according to the MAD method no participants needed to be excluded in the further analysis. In Table 1 the mean, median, standard deviation (SD) and range of the ADHD, mind-wandering and risk taking behaviour are presented.
A block wise multiple regression analysis was used to test if mind-wandering predicts risk taking behaviour above and beyond ADHD symptoms. The assumptions multicollinearity and homoscedasticity were violated. Because the mind-wandering and ADHD questionnaire have a strong linear relationship, which is shown in Table 2, the assumption of multicollinearity is violated. Usually one of the independent variables will be excluded in the analysis, but because the main question is based on the prediction of both the variables they are both included in the analysis. Because we are not interested in the significance testing and only in the estimation of the parameters a non-parametric alternative is not used, even though the homoscedasticity assumption is violated (Field, 2013).
The first regression is carried out to test whether ADHD symptoms significantly predict risk taking behaviour. The regression indicates that ADHD symptoms are a significant predictor of risk taking behaviour, F(1,310) = 79.87, p < .001, with an R2 of .205. In other words, ADHD symptoms can predict risk taking behaviour in adolescents with 20.5 percent. The second regression is carried out to test whether mind-wandering adds significantly to the prediction of ADHD symptoms and risk taking behaviour. The regression indicates that the combination of ADHD symptoms and mind-wandering is a significant predictor of risk taking behaviour, F(2,309) = 40.53, p < .001, with an R2 of .208. In other words, ADHD symptoms and mind-wandering can predict risk taking behaviour in adolescents with 20.8 percent. However when the model is further investigated it becomes clear mind-wandering is not significant in this prediction, ? = -.08, p = .29. The prediction of risk taking behaviour through ADHD symptoms is still significant.
These results are partially in line with the expectations. It was expected that ADHD symptoms predict risk taking behaviour. But it was also expected that mind-wandering predicted risk taking behaviour above and beyond ADHD symptoms, which is not in line with the results.
Descriptives ADHD, mind-wandering and risk taking behaviour questionnaires.
Mean Median Standard deviation Range
ADHD 19.707 18 11.142 0 – 59
Mind-wandering 14.098 14 8.8082 0 – 36
Risk taking behaviour 18.897 16.5 12.623 0 – 73
Correlations ADHD, mind-wandering and risk taking behaviour questionnaires.
ADHD Mind-wandering Risk taking behaviour
Pearson correlation ADHD 1.00 .756 .453
Mind-wandering .756 1.00 .307
Risk taking behaviour .453 .307 1.00
Significance ADHD . .001 .001
Mind-wandering .001 . .001
Risk taking behaviour .001 .001 .
In this study we have investigated the prediction of risk taking behaviour. The main question was if mind-wandering predicts risk taking behaviour above and beyond ADHD symptoms. This has been investigated through self-report questionnaires in adolescents. This study did not detect any evidence that mind-wandering predicts risk taking behaviour above and beyond ADHD symptoms. However the results do show that ADHD symptoms predict risk taking behaviour.
These results are not in line with the data obtained in the study of Mowlem et al. (2016). They suggested that mind-wandering could be used to predict risk taking behaviour instead of ADHD. However they based the degree of risk taking behaviour on an section of an impairment questionnaire. Which may have resulted in a limited presentation of risk taking behaviour. Risk taking behaviour is possibly better operationalized in this study than in the study of Mowlem et al. (2016). Mind-wandering may predict those limited parts of risk taking behaviour above and beyond ADHD symptoms. However, when all the aspects of risk taking behaviour are taken into account, mind-wandering does not predict risk taking behaviour above and beyond ADHD symptoms. The different operationalisation of risk taking behaviour may explain the divergent results.
All the questionnaires used in this study are self-report questionnaires, which can be a
limitation. Adolescents have a tendency to report more or less than the reality, due to selective memory, attribution or exaggeration. The deviant answers can vary per question (Brener, Billy, & Grady, 2003). In other words an adolescent may report more problems than he actually experiences on the first question, less problems on the second question and he may report quite accurately on the third question. Which can result in an inaccurate representation of the ADHD symptoms, degree of mind-wandering or degree of risk taking behaviour and. Which could result in an inaccurate conclusion. A second limitation in this study is that the mind-wandering and the ADHD questionnaires have a strong linear relationship. It is possible that they both measure roughly the same problems, but the ADHD questionnaire is more widespread. When the ADHD questionnaire is more widespread it may contain more aspects which are important for the prediction of risk taking behaviour than the mind-wandering questionnaire. Therefore, mind-wandering may miss important aspects for the prediction of risk taking behaviour.
Furthermore, besides the role of mind-wandering as an underlying process it overlaps with regular ADHD symptoms. The main characteristics of ADHD are inattention, hyperactivity and impulsivity. Mind-wandering is strongly related to the inattention part of ADHD. So maybe it is not the inattention part of ADHD which can predict mind-wandering but one of the other characteristics. In the study of Lee & Hinshaw (2006) they found that inattention symptoms could not predict risk taking behaviour. In contrast to hyperactivity and impulsivity symptoms which could predict risk taking behaviour. An explorative analysis was run to check if this is the same case on our data. EXPLORATIEVE ANALYSE!!
There are several implications for both clinical practise as further research on base of the results of the current study. Regarding clinical practice, when an individual is diagnosed with ADHD, clinician should always keep risk taking behaviour in mind as a potential threat. First concerning further research, the operationalisation could be objectified. For example, instead of a self-report questionnaire of risk taking behaviour, the participants could keep score of certain behaviour for a period of time. Furthermore, research could focus on the different characteristics of ADHD in combination with risk taking behaviour. When it is more clear which characteristics , inattentiveness, hyperactivity or impulsivity can predict risk taking behaviour best, prevention of this behaviour can be focussed.
Overall this study shows an evident link between ADHD and risk taking behaviour. However there is no evidence for mind-wandering as an underlying process for this problem. In clinical practice it is important to directly assess risk taking behaviour when ADHD is diagnosed.