Environmental Factors as the Cause of ADHD

Attention-deficit/hyperactivity disorder (ADHD) is a neurodevelopmental disorder that affects millions of children and adolescents worldwide, with a prevalence of approximately 5% in the general population (Schwarz, 2016). The symptomatology of ADHD includes an array of persistent problems, such as difficulties with attention, hyperactivity and impulsive behaviour (Willoughby, 2003). The aetiology of ADHD is complex and under continuous investigation, with the growing consensus in the current body of research indicating a substantial interplay between environmental and genetic influences (Pingault et al., 2015). Research into the possible environmental contributors to the manifestation of ADHD has investigated various factors, including maternal smoking during pregnancy (He, Chen, Zhu, Hua, & Ke, 2017), socioeconomic disadvantage (Russell, Ford, Williams, & Russell, 2016), lead exposure (Hong et al., 2014) and premature birth (Sucksdorff et al., 2015). Research exploring the relationship between the prenatal environment and ADHD has also found associations between higher levels of prenatal testosterone (determined by measurement of 2D:4D ratio) and ADHD symptoms in boys (Eichler et al., 2018; Myers, van’t Westeinde, Kuja-Halkola, Tammimies, & Bölte, 2018). In this essay, the findings of the existing literature exploring ADHD and its environmental influences, as well as those of the literature exploring the genetic underpinnings of the disorder, will be assessed and a conclusion about the current state of the knowledge of the aetiology of ADHD and its future direction will be drawn.

Despite the fact that there has been much research into the environmental influences on ADHD and other neurodevelopmental disorders, researchers have highlighted the problems with confounding variables in both environmental and gene-by-environment interaction studies (Keller, 2014), as well as the inconsistency of their findings (Polańska, Jurewicz, & Hanke, 2012). The review by Keller (2014) demonstrates that insufficient control of covariates and confounders is prevalent in this body of literature and in the context of the research into the environmental influences on ADHD, this is significant, as many of the previously mentioned environmental correlates of ADHD have shown associations with each other in past studies. Examples of this include associations between heavy smoking and low socioeconomic status (Hiscock, Bauld, Amos, Fidler, & Munafò, 2012), maternal smoking and premature birth (Dahlin, Gunnerbeck, Wikström, Cnattingius, & Edstedt Bonamy, 2016), low socioeconomic status and environmental exposure to lead (Benfer, 2017) and the association between prenatal testosterone levels and maternal smoking during gestation (Rizwan, Manning, & Brabin, 2007).

Furthermore, the influence of geographical location on the occurrence of ADHD has also been investigated (Saez, Barceló, Farrerons, & López-Casasnovas, 2018). In a population-based retrospective cohort study involving 116 children diagnosed with ADHD and 5077 non-ADHD controls, Saez et al. (2018) found that the risk of ADHD occurrence was positively correlated with proximity to agricultural areas/residential streets, as well as motorways and industrial estates, with ADHD risk increasing for those living less than 100 and 300 metres away from agricultural areas and motorways/industrial estates respectively. The authors considered agricultural areas in particular as a proxy for pesticide exposure, and a growing body of literature has also found associations between exposure to pesticides and ADHD symptoms (Richardson et al., 2015; Roberts, Dawley, & Reigart, 2018). Although it is important to note that a proxy is influenced by other confounding variables, research involving the direct measurement of pesticide levels such as those present in blood or urine has also found similar associations (Richardson et al., 2015; Yu et al., 2016). Richardson et al. (2015) determined the influence of pesticides directly by measuring the level of pyrethroid (a class of pesticide) metabolites in the urine samples of their participants, and found that children with detectable levels of pyrethroid metabolites in their urine were more than twice as likely to be diagnosed with ADHD than controls.

Moreover, with the aim of investigating the relationship between the class of pesticides known as organophosphates and ADHD occurrence, Yu et al. (2016), in a case-control study involving 97 children and teenagers who were doctor-diagnosed with ADHD and 110 controls, found a dose-response relationship between urinary concentrations of the organophosphate pesticide DMP and ADHD symptoms, with higher concentrations of DMP being associated with a two to three-fold increase in risk of ADHD diagnosis. Although the findings of Yu et al. (2016) were in agreement with the findings of Richardson et al. (2015), particularly with regards to the increase in ADHD occurrence associated with higher urinary pesticide, there are a number of methodological differences between the two studies and the most significant is the greater control of a wider range of potential confounders by Yu et al. (2016), such as blood lead concentrations, maternal smoking and parental education level as well as genetic covariates including familial history of neuropathology and genes related to the dopaminergic system. The type of pesticide investigated was also different in each of the two studies, as Yu et al. (2016) investigated the effects of an organophosphate pesticide whereas Richardson et al. (2015) investigated the effects of the pyrethroid class of pesticide as a variable. Both Richardson et al. (2015) and Yu et al. (2016) suggest that further studies investigating the effect of pesticide exposure on ADHD occurrence should involve a longitudinal design as these are required to establish causality, and an increased effort should be made in these studies to control for both environmental and genetic confounders. This suggestion can also be extended to other studies investigating environmental correlates of ADHD, particularly those involving other chemical influences such as lead exposure, as the absence of a longitudinal design involving controls of other correlating factors as well as a control group for comparison has the effect of ensuring that any relationship remains correlational rather than causal.

Furthermore, the study by Saez et al. (2018) also found an association between proximity to motorways and ADHD occurrence and motorways can be seen as a proxy for air pollution, given the greater concentrations of major pollutants such as nitrogen dioxide found closer to expressways (Beckerman et al., 2008). Associations have been found between ADHD symptoms and nitrogen dioxide in particular, and this remains even after controlling for other environmental confounders such as socioeconomic status, number of smokers in home and prematurity (Morales et al., 2009). Markevych et al. (2018) diagnosed 2044 children with ADHD during an observational period that also involved calculating the particulate matter (PM), nitrogen dioxide and vegetation levels of 186 postal code areas.

During analysis, associations were found between increases in both nitrogen dioxide and particulate matter levels and ADHD occurrence, supporting the findings of Morales et al. (2009). However, it should be noted that in the study by Markevych et al. (2018), few possible confounders excluding proximity to psychiatrists were controlled for and in the context of the validity of their results, this is problematic. As nitrogen dioxide and particulate matter levels tend to increase in areas with greater urbanisation and traffic density (Huang et al., 2013), the inclusion of these variables without controlling for other environmental confounders associated with greater urbanisation, such as lead exposure (Schur & John, 2014), as well as other variables related to air pollution and those of a sociological nature, make it difficult to ascertain the extent of the association between nitrogen dioxide/particulate matter and ADHD, and lack of control of these confounders makes a causal relationship harder to identify.

Although there has been a concentrated effort to investigate the influence of environmental factors on ADHD and many possible correlates have been assessed, there is also a substantial genetic basis for ADHD as the disorder has been shown to run in families, with heritability estimates ranging between 50-80% (Faraone & Larsson, 2018) and adoption studies have shown that adopted children are more similar to their biological parents than to their adopted parents with regards to ADHD behaviour (Langley, 2018; Sprich, Biederman, Crawford, Mundy, & Faraone, 2000). In a review of the literature investigating the genetic basis of ADHD, Langley (2018) outlined some of the issues regarding the existing adoption studies of ADHD in particular, noting the prevalence of small sample sizes in these studies, which brings into question their representativeness to the general population despite the usefulness and consistency of their findings. Furthermore, investigating the genetic basis of ADHD requires disentangling the genetic factors from possible environmental correlates and the most frequently employed method for achieving this is the use of twin studies (Wood & Neale, 2010). Due to the fact that monozygotic (MZ) twins share 100% of their genes and dizygotic (DZ) twins share 50% of their genes respectively, twin studies are often designed with the assumption that all twins have an equally similar rearing environment and that any difference on a particular variable can then be attributed to genetic factors (Felson, 2014).

Twin studies have found that the genetic influences on ADHD act on a continuum and that heritability estimates are similar within normal variables of behaviour associated with ADHD as well as at extremes high enough to be diagnosed with the disorder (Larsson, Anckarsater, Råstam, Chang, & Lichtenstein, 2012; Levy, Hay, McStephen, Wood, & Waldman, 1997). In a study involving 8,500 twin pairs, Larsson et al. (2012) found a strong genetic association between both extreme and subthreshold clinical assessments of ADHD symptomatology, with both assessments showing very similar heritability rates of .62 for the extreme criteria and .60 for the subthreshold criteria respectively. Support for a genetic continuum was also found in a study of 2,143 twin pairs by Grevin et al. (2016), however their results also differed from the study by Larsson et al. (2012) as a significant shared environmental influence was found for low-extreme ADHD traits compared with high-extreme ADHD traits, and the high-ADHD traits showed significantly higher heritability, with a lower environmental influence.

Another method employed in the investigation of the genetic aetiology of ADHD is genome-wide association studies (GWAS), although these have yielded mixed results (Langley, 2018). Meta-analyses of earlier GWAS involving ADHD have typically found no significant associations (Neale et al., 2010; Zavats et al., 2015) and this calls into question the high heritability found in twin studies (50-80%), with some researchers having remarked that this heritability might be influenced by the faith in the equal-environment assumption typically upheld in twin studies (Felson, 2014), although simultaneously acknowledging that this may be offset by the general robustness and consistency in the findings of twin studies compared with other investigations into the genetic roots of ADHD. In the largest internationally collaborative GWAS to date, involving over 20,000 ADHD cases as well as 35,000 controls, Demontis et al. (2017) found 12 significant genome-wide associations which are the first significant GWAS results involving ADHD and they also give support to the dominant hypothesis of the previous twin studies, which is that ADHD as disorder is an extreme expression of a number of continuous heritable traits and behaviours (Grevin et al., 2016; Larsson et al., 2012). Additionally, due to the lack of significant results in earlier genome-wide studies of ADHD, these findings can be seen as representing a positive step in this particular field, although it can also be seen as being too early to understand their developments fully and future studies will need to be undertaken to investigate the relationship between each of the 12 independent gene loci identified and ADHD symptoms separately if a causal relationship is to be established.

In conclusion, based on the existing literature, it cannot be stated that ADHD is solely a consequence of environmental factors. Furthermore, it is clear that despite the extensive research investigating the aetiology of ADHD, the definitive causes of this disorder remain elusive and in order to elucidate the aetiology of ADHD further, future studies with a longitudinal design and involving the careful control of confounders are a necessity. However, as these studies are costly and can be difficult to implement, this effort can be also guided by further literature reviews of studies investigating both the environmental and genetic factors, which will help with refining a future methodology and avoiding making the same errors as previous studies, which in turn will increase the validity of any future findings and help the research effort move forward.