ADHD in the United Kingdom: Regional and Socioeconomic Variations in Incidence Rates Amongst Children and Adolescents (2004-2013)

Abstract

Objective: To describe the incidence and distribution of ADHD within the United Kingdom, and to examine whether there was any association between ADHD incidence and socioeconomic deprivation. Method: The study used data from the Clinical Practice Research Datalink (CPRD). Patients diagnosed with ADHD before the age of 19 between January 1, 2004 and December 31, 2013 were stratified according to the region in which their general practice was based. Practice Index of Multiple Deprivation (IMD) score was used as a surrogate measure of patients’ deprivation status. Results: ADHD incidence was relatively stable between 2004 and 2013, but peaked in the last 2 years studied. Statistically significant (p ≤ .05) differences in incidence were observed between U.K. regions. In almost every year studied, incidence rates were highest among the most deprived patients and lowest among the least deprived patients. Conclusion: In the United Kingdom, ADHD may be associated with socioeconomic deprivation.

Keywords

ADHD incidence variation deprivation

Introduction

ADHD is a neurodevelopmental disorder characterized by three core symptoms: hyperactivity, impulsivity, and inattention (Bolea-Alamañac et al., 2014). In the United Kingdom, general practitioners (GPs) play a key role in the diagnosis, management, and treatment of the disorder. As gatekeepers to the U.K. health care system (Herrett et al., 2015; Murray et al., 2014), GPs will generally be the first port of call for individuals concerned that they or their child may have ADHD. After referring suspected cases to secondary care (such as pediatric or psychiatry services) for a confirmation of the diagnosis, GPs may prescribe medications and undertake monitoring measures as part of a shared care arrangement (National Institute for Health and Care Excellence [NICE], 2008a). Four medications are currently licensed in the United Kingdom for the treatment of ADHD (methylphenidate, dexamfetamine, lisdexamfetamine, and atomoxetine; Joint Formulary Committee, 2015), although pharmacological intervention may not be required in all cases (McCarthy et al., 2012).

McCarthy et al. (2012) and Holden et al. (2013) both detected increases in the incidence and prevalence of ADHD in the United Kingdom during the first decade of the 21st century. However, there is some evidence to suggest that the burden of the disorder is unevenly distributed. Rowlingson, Lawson, Taylor, and Diggle (2013) observed that primary care spending on methylphenidate varied significantly across England. In addition, a U.K. cohort study by Russell, Ford, Rosenberg, and Kelly (2014) found that ADHD was particularly prevalent among children living in circumstances of social and economic disadvantage.

The findings of Rowlingson et al. (2013) were based on national prescribing data from a single month in 2011. Similarly, the link between parentally reported ADHD and socioeconomic deprivation was based on a sample containing a relatively small number (n = 187) of affected children (Russell et al., 2014). The aim of this study was to establish whether the regional prescribing variations observed in the United Kingdom reflected regional variations in ADHD incidence, and to determine whether ADHD incidence showed any association with socioeconomic deprivation on a national scale. The study also sought to update the findings of earlier epidemiological studies, describing ADHD incidence rates among children and adolescents in the United Kingdom between the years 2004 and 2013.

Method

Data Source

A retrospective cohort study was performed using primary care consultation data from the Clinical Practice Research Datalink (CPRD). These data consist of information routinely recorded by general practitioners during their consultations with individual patients, including diagnoses made and medications prescribed. CPRD has been collating anonymized patient-level data from U.K. general practices since its inception (as the “General Practice Research Database”) in 1987 (Mansell, 2013). General practices that contribute data to CPRD are required to meet certain data quality requirements before they are declared “up to standard” for research purposes (Bhaskaran, Forbes, Douglas, Leon, & Smeeth, 2013). Only data from these “up to standard” practices were included in the study. The number of general practices sharing data with the CPRD has expanded steadily over time. At the time the study was conducted, CPRD held longitudinal, research-quality data for 684 U.K. general practices (Clinical Practice Research Datalink, 2015). This equates to around 9% of the United Kingdom’s general practices (Mansell, 2013) and records for approximately 13.5 million individuals (CPRD Knowledge Centre, personal communication, 2014), a large sample that is broadly representative of the U.K. population as a whole (Bushe, Wilson, Televantou, Belger, & Watson, 2015; Herrett et al., 2015; Holden et al., 2013; Thomas, Mitchell, & Batstra, 2014; West, Fleming, Tata, Card, & Crooks, 2014).

Study Population and Study Period

The study population comprised patients diagnosed with ADHD before the age of 19, between January 1, 2004 and December 31, 2013. Data stored within CPRD are coded; terminology relating to patients’ clinical management is encoded using a standardized set of codes termed “Read codes” to promote consistency and uniformity (Chisholm, 1990). Individuals with a diagnosis of ADHD were identified by the presence of Read codes relating to the disorder in their CPRD record. To be eligible for inclusion as an incident case of ADHD, the earliest occurrence of a relevant code had to occur within the study window, and following at least 365 days continuous registration with their general practice. A list of Read codes denoting a diagnosis of ADHD and a further list of Read codes denoting drugs used in its treatment were compiled (both lists available at clinicalcodes.org, an online repository for clinical codes used in database research; Springate et al., 2014). The drugs selected encompassed all agents currently licensed for the treatment of ADHD in the United Kingdom—methylphenidate, dexamfetamine, lisdexamfetamine, and atomoxetine. All four drugs are licensed for use in patients between the ages of 6 and 18 years of age; atomoxetine is also approved for use in adults (Joint Formulary Committee, 2015). With the exception of dexamfetamine (which is also licensed for the treatment of narcolepsy), the drugs of interest examined by this study are solely licensed for the treatment of ADHD (Joint Formulary Committee, 2015).

Assessment of Geographical Location

Every practice contributing data to CPRD has a unique identifying number. Associated with this number is information about that practice’s geographical location within the United Kingdom. By looking at the practice identifier associated with a particular patient, their location within the United Kingdom can be discerned. CPRD subdivides the United Kingdom (a nation itself comprised of four “nations”—England, Scotland, Wales, and Northern Ireland) into 13 geographical regions. Scotland, Wales, and Northern Ireland comprise three of these regions; the remainder are regions situated within England (North West, North East, Yorkshire and the Humber, East Midlands, West Midlands, East of England, South West, South Central, London, and the South East Coast; West et al., 2014).

Assessment of Deprivation: “Practice-Level” Deprivation Score

England and Wales are divided up into approximately 35,000 defined geographical areas known as Lower Layer Super Output Areas (LSOA; Office of National Statistics, 2011). Generated for the purposes of statistical research, these areas each contain populations of between 1,000 and 3,000 people (Office of National Statistics, 2011). Similar geographic divisions are applied to Northern Ireland and Scotland, which are divided into smaller areas termed datazones (DZ; U.K. Data Service, 2014). Measurements relating to seven key indicators of socioeconomic deprivation are routinely compiled for each LSOA/DZ. These indicators examine household income, employment, health and disability, education and training, barriers to housing and services, crime, and the living environment (U.K. Data Service, 2014). An amalgamation of this information is used to calculate an Index of Multiple Deprivation (IMD) score for each LSOA/DZ, allowing each to be ranked in order of relative deprivation.

Every general practice contributing data to CPRD has an IMD score based on the LSOA/DZ in which it is situated. These scores are available to CPRD researchers, rounded to the nearest quintile. For the purposes of this study, “practice-level” IMD scores were used as a surrogate measure of patients’ deprivation status. This measure was deemed appropriate as patients would be expected to reside in the locality of their general practice, within a geographically defined catchment area (NHS Choices, 2014).

Assessment of Deprivation: “Patient-Level” Deprivation Score

For around 70% of English practices (covering just over 50% of all patients in CPRD), IMD scores can be provided for individual patients based on the LSOA in which their home address is situated (CPRD Knowledge Centre, personal communication, 2014). This direct measure of deprivation status was requested for the subset of ADHD patients for whom it was available. By comparing these individuals’ practice-level IMD score with their patient-level IMD score, it could be established whether practice-level deprivation scores provided an accurate reflection of patient-level deprivation scores.

Data Analysis

Incidence Calculation

The earliest occurrence of an ADHD-related Read code in each patient’s records was identified, and the calendar year in which this occurred was noted. The patient was then counted as a newly diagnosed incident case for that calendar year. The incidence denominator for each year comprised of person-time contributed by individuals who were considered “at risk” of developing ADHD in that year.

Incidence rates were expressed as cases per 10,000 person years at risk (PYAR) and presented with 95% confidence intervals (CI). Annual incidence rates were calculated and stratified according to patient gender, nation (England/Scotland/Wales/Northern Ireland), and practice-level deprivation (IMD) quintile. An overall incidence rate was calculated for the study period as a whole; this was stratified according to gender, nation, CPRD region (in the case of English patients), age group, and deprivation quintile. Multivariable Poisson modeling was used to determine incidence rate ratios (IRR) and accompanying 95% CIs and p values, adjusted for gender, nation, age group, and deprivation quintile. A regression was similarly conducted using only English patients; this was adjusted for gender, CPRD region, age group, and deprivation quintile. Statistical significance was set at p ≤ .05, and all statistical analyses were performed using Stata version 13 (StataCorp, College Station, Texas, United States).

Results

Overall and Annual Incidence Rates (the United Kingdom)

Over the 10-year study period, 10,284 new diagnoses of ADHD were recorded in under 19s in CPRD. The overall ADHD incidence rate for the study period was 11.67 cases per 10,000 PYAR (95% CI = [11.45, 11.90]). Incidence rates were at their lowest in 2008 (11.04 cases per 10,000 PYAR; 95% CI = [10.38, 11.75]) and highest in 2012 (12.56 cases per 10,000 PYAR; 95% CI = [11.84, 13.33]), as shown in Figure 1.

Figure 1.

U.K. incidence rate (2004-2013).

Incidence by Gender and Age Group

After adjustment for nation, deprivation quintile, and age group, a large and statistically significant difference (p ≤ .001) in incidence rates was observed between males and females. Between 2004 and 2013, the overall incidence of ADHD among the male population at risk was 18.63 cases per 10,000 PYAR (95% CI = [18.24, 19.03]). The overall incidence rate in females was much lower (4.37 cases per 10,000 PYAR; 95% CI = [4.18, 4.57]). As shown in Figure 1, female incidence rates were relatively static from 2004 to 2010 but were notably higher in the last 3 years of the study period (peaking at 5.45 cases per 10,000 PYAR; 95% CI = [4.80, 6.21] in 2012).

Figure 2 shows the incidence of ADHD in males and females according to age group. In both males and females, ADHD was most commonly diagnosed at age 7 (1,057 new diagnoses in males, 238 new diagnoses in females). Thirty-five percent of all ADHD patients identified (n = 3,606) were diagnosed between the ages of 7 and 9.

Figure 2.

ADHD incidence by age of diagnosis.

Incidence by Nation (England, Scotland, Wales, and Northern Ireland)

As shown in Table 1, Northern Ireland’s overall incidence rate was the highest of the four U.K. nations (with 13.32 cases per 10,000 PYAR; 95% CI = [12.11, 14.66]). This was significantly higher than that of Scotland (p ≤ .001), England (p ≤ .001), and Wales (p = .015). Wales had the second highest incidence rate across the study period, significantly higher than that of England (p = .012) and Scotland (p = .010). Scotland’s overall incidence rate was the lowest of the four nations, although the difference between Scottish and English rates was not statistically significant (p = .359).

Table 1.

ADHD Incidence.

	ADHD cases	Person-years at risk (nearest whole year)	Incidence rate 2004-2013 per 10,000 person-years (95% CI)	Adjusted incidence rate ratio^a (95% CI)
The United Kingdom (total)	10,284	8,808,590	11.67 [11.45, 11.90]	N/A
Gender
Male	8,407	4,512,611	18.63 [18.24, 19.03]	1.00 (ref)
Female	1,877	4,295,979	4.37 [4.18, 4.57]	0.23 [0.22, 0.25]
Nation
England	7,984	6,851,691	11.65 [11.40, 11.91]	1.00 (ref)
Scotland	976	917,860	10.63 [9.99, 11.32]	0.97 [0.91, 1.04]
Wales	903	723,038	12.49 [11.70, 13.33]	1.09 [1.02, 1.17]
Northern Ireland	421	316,002	13.32 [12.11, 14.66]	1.26 [1.14, 1.39]
Deprivation quintile (practice-level)
1 (least deprived)	1,539	1,646,181	9.35 [8.89, 9.82]	1.00 (ref)
2	1,864	1,706,407	10.92 [10.45, 11.43]	1.16 [1.08, 1.24]
3	2,029	1,786,384	11.36 [10.87, 11.86]	1.19 [1.12, 1.27]
4	2,359	1,904,857	12.38 [11.89, 12.89]	1.30 [1.22, 1.39]
5 (most deprived)	2,493	1,764,762	14.13 [13.58, 14.69]	1.49 [1.39, 1.58]
Age group
1-3 years old	420	305,937	13.73 [12.48, 15.11]	1.00 (ref)
4-6 years old	1,984	847,226	23.42 [22.41, 24.47]	1.71 [1.54, 1.89]
7-9 years old	3,606	1,191,796	30.26 [29.29, 31.26]	2.20 [1.99, 2.44]
10-12 years old	2,222	1,390,343	15.98 [15.33, 16.66]	1.17 [1.06, 1.30]
13-15 years old	1,592	1,494,459	10.65 [10.14, 11.19]	0.78 [0.70, 0.87]
16-18 years old	460	3,578,831	1.29 [1.17, 1.41]	0.09 [0.08, 0.11]

Note. CI = confidence interval

Adjusted for gender, nation, deprivation quintile, age group.

In England, annual fluctuations in incidence rates broadly corresponded to those of the United Kingdom as a whole. Incidence rates were at their lowest in 2008 and highest in 2012 (peaking at 12.73 cases per 10,000 PYAR; 95% CI = [11.90, 13.62]). However, a decrease in incidence rates between 2007 (11.87 cases per 10,000 PYAR; 95% CI = [11.10, 12.70]) and 2008 (10.27 cases per 10,000 PYAR; 95% CI = [9.55, 11.05]) was observed in England but not observed in the U.K. data as a whole.

In Scotland, ADHD incidence was lowest in 2005 (7.60 cases per 10,000 PYAR; 95% CI = [6.00, 9.62]) and highest in 2013 (14.80 cases per 10,000 PYAR; 95% CI = [12.52, 17.48]). In contrast to England, 2008 saw a relatively high incidence of newly diagnosed ADHD in Scotland and peak annual incidence in Northern Ireland (15.46 cases per 10,000 PYAR; 95% CI = [11.62, 20.58]). In Wales, peak ADHD incidence was observed in 2007 (15.63 cases per 10,000 PYAR; 95% CI = [12.98, 18.83]).

Incidence by CPRD Region (England)

Within each English region, annual incidence rates fluctuated between the years 2004 and 2013 without any consistent pattern. However, the South East Coast region had both the highest number of ADHD diagnoses during the study period (n = 1,461, 18.3% of the England’s total cases) and the highest overall incidence rate of ADHD in under 19s (see Table 2). This was significantly higher (p ≤ .001) than that of the Yorkshire and the Humber region, which had the lowest incidence rate of England’s 10 CPRD regions.

Table 2.

ADHD Incidence by English Region.

	ADHD cases	Person-years at risk (nearest whole year)	Incidence rate 2004-2013 per 10,000 person-years (95% CI)	Adjusted incidence rate ratio^a (95% CI)
England (total)	7,984	6,851,691	11.65 [11.40, 11.91]	N/A
English region
North West	1,078	1,043,531	10.33 [9.73, 10.97]	1.00 (ref)
North East	168	155,931	10.77 [9.26, 12.53]	0.98 [0.83, 1.15]
Yorkshire and the Humber	209	258,378	8.09 [7.06, 9.26]	0.73 [0.63, 0.85]
East Midlands	335	281,963	11.88 [10.67, 13.22]	1.14 [1.00, 1.28]
West Midlands	791	769,010	10.29 [9.59, 11.02]	0.96 [0.87, 1.05]
East of England	1,174	795,373	14.76 [13.94, 15.63]	1.53 [1.40, 1.66]
South West	673	683,147	9.85 [9.13, 10.62]	0.99 [0.90, 1.09]
South Central	1,175	1,001,399	11.73 [11.08, 12.42]	1.30 [1.19, 1.42]
London	920	942,061	9.77 [9.15, 10.42]	0.92 [0.84, 1.01]
South East Coast	1,461	920,899	15.86 [15.07, 16.70]	1.67 [1.54, 1.81]

Note. CI = confidence interval.

Adjusted for gender, deprivation quintile, age group.

Incidence by Deprivation (IMD) Quintile

When stratified according to deprivation quintile, the U.K.’s diagnostic data suggested a significant link between deprivation and ADHD incidence. In almost every year studied, incidence rates were highest in the most deprived patients and lowest in the least deprived patients (see Figure 3). Underlying this U.K. trend were England’s diagnostic data. Patients belonging to practices in the most deprived areas of England (IMD quintile 5) had the highest incidence of ADHD overall (13.84 cases per 10,000 PYAR; 95% CI = [13.23, 14.47]). This was significantly higher (p ≤ .001) than the incidence rates for Quintiles 1, 2, 3, and 4. At the opposite end of the deprivation scale, patients belonging to the least deprived quintile (1) had a significantly lower incidence (p ≤ .001) of diagnosed ADHD than patients in any other quintile (9.24 cases per 10,000 PYAR; 95% CI = [8.72, 9.80]). Patient-level deprivation data were accessible for 80.5% of English ADHD patients (n = 6,424). In 4,476 of these patients, their patient-level IMD quintile was either the same as their practice-level quintile or higher. That is to say, in 69.7% of instances, patients were either as deprived as their practice-level IMD suggested, or more deprived.

Figure 3.

Annual incidence rate (2004-2013) by deprivation quintile.

In the other three nations of the United Kingdom, evidence for an association between deprivation and ADHD was somewhat weaker (see Figure 4). In Scotland (as in England), patients belonging to practices in the most deprived areas (IMD quintile 5) had the highest incidence of diagnosed ADHD; rates were significantly higher (p ≤ .001) than those in the less deprived quintiles (Quintiles 1, 2, 3, and 4). In Wales, patients in the most deprived quintile had the highest incidence of ADHD, significantly higher (p ≤ .001) than that observed in the least deprived quintile. In Northern Ireland, there was no clear association between ADHD and deprivation.

Figure 4.

ADHD incidence by deprivation quintile.

Discussion

This study found that there were statistically significant differences in ADHD incidence rates between the United Kingdom’s constituent nations, and between individual regions within England. The finding of significant geographical differences within the United Kingdom is probably unsurprising. In the United States, significant differences in diagnostic and treatment rates for ADHD have been observed between states, and between different communities within the same state (Fulton et al., 2009; McDonald & Jalbert, 2013). Furthermore, Rowlingson et al. (2013) had observed regional variations in methylphenidate spending in England that had suggested such variations might be present. That study identified a notable area in the South East of England where medical practices’ methylphenidate spending was 4 times the national average; this study found that the CPRD’s South East Coast region had the highest ADHD incidence rate of all CPRD regions.

It may be the case that these differences in diagnostic rates are explicable by national and regional differences in diagnostic and management procedures. All areas of the United Kingdom would be expected to take the 2008 NICE guidance on the diagnosis and management of ADHD as a primary resource (NICE, 2008b). However, it is possible that a child diagnosed with ADHD in one part of the United Kingdom may not have had the disorder recognized and diagnosed had they lived in another part of the country. All four constituent nations of the United Kingdom have distinct budgets for health care and must prioritize spending according to national needs and priorities (National Audit Office, 2012). Similarly, different regions within each nation have their own allocated budgets which must be used to provide a well-rounded health service to the local populace. It has been acknowledged that different areas of the United Kingdom provide inconsistent levels of service provision for ADHD (NICE, 2013), potentially resulting in different levels of case recognition.

Alternatively, the regional and national differences in diagnostic rates may reflect genuine differences in ADHD incidence across the United Kingdom. That is to say, populations in some parts of the United Kingdom may have a higher proportion of individuals with some genetic susceptibility to ADHD and/or higher exposure to environmental risk factors that promote its onset. One environmental factor suggested to play a role in the etiology of ADHD is sunlight. In 2013, Arns, van der Heijden, Arnold, and Kenemans reported an inverse association between regional solar intensity and ADHD prevalence across 49 U.S. states, and across several countries. That finding has been contested elsewhere (Hoffmann et al., 2014), and this study’s findings did not appear to suggest an association between ADHD and solar intensity in the United Kingdom. The South East Coast of England had the highest incidence of ADHD of all English regions, despite its southerly latitude and its relatively high solar intensity (Met Office, 2014). In addition, Scotland had the lowest ADHD incidence of all four U.K. nations despite being the most northerly and having the lowest solar intensity overall (Met Office, 2014). This does not rule out a relationship between ADHD and sunlight but does suggest that in a country the size of the United Kingdom, in the United Kingdom’s position geographically, regional differences in ADHD incidence do not appear to be influenced by regional differences in solar irradiation.

Exposure to socioeconomic deprivation is another purported risk factor for ADHD, and this study did observe a clear association between ADHD and socioeconomic deprivation. In England, Scotland, and Wales, ADHD incidence rates were highest among patients belonging to practices in the most deprived areas (IMD quintile 5). In all three nations, and in half of England’s 10 CPRD regions, incidence rates among individuals in Quintile 5 (the most deprived quintile) were significantly higher (p ≤ .05) than those of individuals in Quintile 1 (the least deprived quintile). These findings lend support to the theory that an individual’s likelihood of being diagnosed with ADHD may be increased by exposure to socioeconomic deprivation. This observation is in line with the findings of several studies from around the world (Döpfner, Breuer, Wille, Erhart, & Ravens-Sieberer, 2008; Froehlich et al., 2007; Hjern, Weitoft, & Lindblad, 2010; Nomura et al., 2012) and from the United Kingdom specifically (Green, McGinnity, Meltzer, Ford, & Goodman, 2005; Russell et al., 2014). It is beyond the capabilities of this study to identify an underlying reason for this apparent link between deprivation and the development of ADHD. The measure of deprivation used by this study (the Index of Multiple Deprivation 2010) assesses several distinct aspects of socioeconomic deprivation, rather than just one specific characteristic or risk factor that could then be investigated further. However, by identifying that local deprivation and ADHD often coexist, this study highlights the need for adequate service provision in deprived areas of the United Kingdom.

In line with current consensus, this study found that ADHD incidence rates were significantly higher in males than in females in every year studied and across the study period as a whole. The overall incidence rate observed among males was approximately 4.3 times that of females. This gender imbalance is not exceptional when compared with other studies in the literature. Epidemiological studies have typically found ADHD to be 2 to 4 times more common in males than in females (NICE, 2008b). Although the association between ADHD and gender is long-standing, in recent years its underlying reasons have come under increased scrutiny. Boys are more likely to have ADHD characterized by impulsivity and hyperactivity, whereas inattentive symptoms tend to predominate in girls (Kooij et al., 2010). It has therefore been hypothesized that ADHD in males is simply more visible and attention-grabbing than it is in females, leading to higher rates of recognition and diagnosis.

In both males and females, ADHD diagnosis was most commonly observed during patients’ primary school years (especially between the ages 7 and 9). This is broadly in line with the findings of earlier U.K. studies. One study observed peak incidence rates in 6- to 17-year-olds; the mean age of diagnosis among this group was 9.8 years (SD = 2.8; Holden et al., 2013). Another reported peak incidence rates in 6- to 12-year-olds (McCarthy et al., 2012). Given that the underlying causes of ADHD are yet to be fully understood, it is possible that the disorder commonly develops or first manifests itself around this time in patients’ lives. However, it may be the case that existing ADHD is particularly likely to be recognized as a problem during a child’s early years of formal schooling. Although a child may have exhibited tendencies toward hyperactivity, impulsivity, and inattention at an earlier stage, its disruptive impacts on early schooling could be the catalyst for seeking a medical assessment and subsequent diagnosis.

Across the United Kingdom as a whole, ADHD incidence peaked in 2012 and was higher in the last 2 years of the study than it had been in any of the preceding 8 years. This study’s results showed some agreement with those presented by Holden et al. (2013), which examined the period 1998 to 2010. Both studies found that incidence fell between 2004 and 2005, before rising in 2006 and 2007 and then falling in 2008. However, the continued decline in incidence rates observed by the earlier study in 2009 and 2010 was not observed by this study. Concordance with the findings of McCarthy et al. (2012) was somewhat mixed. That study (covering the years 2003-2008) observed peak incidence rates in 2006 and a slight decline in the following year; this study observed an increase in incidence rates between 2006 and 2007. Although comparing incidence rates across the three U.K. studies is an interesting exercise, drawing firm conclusions from these comparisons is not possible. Each study used slightly different case definitions, focused on different study populations (treated and untreated vs. treated only, all ages vs. under 19s) and used somewhat different sampling populations. What is clear is that ADHD diagnostic rates among children and adolescents in the United Kingdom have not been on a continual upward trend over the last decade, or even during the last 5 years. It is unclear whether the increases observed in the last 2 years of this particular study are outliers, or will be sustained in the coming years.

Strength and Weaknesses

The study’s main strength was its data source. CPRD is one of the largest primary care databases in the world (Thomas et al., 2013). It provided a large sample of real-life patient data, allowing several thousand real-world ADHD patients to be identified, their characteristics scrutinized, and statistically significant observations to be made. Its population is drawn from all four nations of the United Kingdom, and has been evaluated as being representative of the United Kingdom’s general population (Bushe et al., 2015; Herrett et al., 2015; Holden et al., 2013; Thomas et al., 2013; West et al., 2014). As such, this study’s findings should be generalizable to the U.K. population as a whole (Thomas et al., 2013). Furthermore, the validity of medical diagnoses in CPRD has been confirmed by several studies and for several different conditions (Herrett et al., 2015; Herrett, Thomas, Schoonen, Smeeth, & Hall, 2010), though not for ADHD specifically.

The reliable identification of valid ADHD cases represented the study’s biggest challenge. Within CPRD, prescriptions for drugs are not directly linked with their indication for use. Therefore, a patient with Read code(s) referring to ADHD in their records, plus documented prescriptions for a licensed ADHD medication may not conclusively represent a diagnosed, pharmaceutically treated ADHD patient (however suggestive this combination may be). As stated in NICE’s guidance on the management of ADHD, diagnosed ADHD may not require pharmacological intervention in all cases (NICE, 2008a). Identifying these diagnosed, non-pharmaceutically treated patients posed further potential problems. First, ADHD patients who received no pharmaceutical treatment from their GPs may have received pharmaceutical treatment elsewhere (for example, as a hospital outpatient). This would not necessarily have been detectable using CPRD data. Second, the presence of a single relevant Read code in a patient’s CPRD record may not conclusively denote a diagnosed, untreated ADHD patient. Holden et al. (2013) determined that, in untreated patients, the presence of two ADHD-related diagnostic codes was required to denote a diagnosis of ADHD. They hypothesized that, in cases of suspected ADHD, GPs would document a provisional diagnosis of ADHD in a patient’s records before referring them for specialist assessment. If specialist assessment confirmed a diagnosis of ADHD, this would be confirmed by the presence of a second ADHD-related Read code in patients’ records. This “two code” hypothesis assumed that prescribers acted uniformly in their diagnostic and documentary practices, and the authors conceded that it may have led to “fully diagnosed” ADHD patients being overlooked. It is unclear whether this or the earlier study’s approach to identifying untreated patients was best. However, a sensitivity analysis revealed that the majority of untreated patients defined by this study (82.4%) had at least two ADHD-related diagnostic codes in their CPRD record.

Estimating patients’ deprivation status using the IMD details of their general practice made use of readily available data, but posed the risk of ecological fallacy. It was recognized that patients would not always reside in the same LSOA/DZ as their general practice; in some cases, they may live in areas with a radically different level of socioeconomic deprivation. In these patients, their practice-level IMD would not give an accurate impression of their exposure to deprivation. Despite this, in the sample of English patients for whom data were available, practice-level IMD and patient-level IMD showed relatively close correspondence.

Conclusion

Statistically significant differences in ADHD incidence were observed between the United Kingdom’s four constituent nations, and between England’s 10 CPRD regions. In addition, ADHD incidence showed a positive association with socioeconomic deprivation. Taking the United Kingdom as a whole, annual ADHD incidence rates remained relatively stable between 2004 and 2013, but were highest in the last 2 years studied.

Footnotes

Authors’ Note

This study is based on data from the Clinical Practice Research Datalink obtained under license from the U.K. Medicines and Healthcare products Regulatory Agency (MHRA). However, the interpretation and conclusions contained in this article are those of the authors alone. The study protocol was approved by the independent scientific advisory committee (ISAC) for CPRD research (Reference Number: 15_036R).

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This article was produced as part of a PhD program funded by the University of Manchester.

Author Biographies

Adrian J. Hire, MPharm, is a PhD student based within Manchester Pharmacy School at the University of Manchester.

Darren M. Ashcroft, PhD, is a professor of pharmacoepidemiology at the University of Manchester and director of the Centre for Pharmacoepidemiology and Drug Safety at Manchester Pharmacy School.

David A. Springate, PhD, is a research fellow in the Institute of Population Health at Manchester University. He uses electronic health records to answer a broad range of questions in primary health care.

Douglas T. Steinke, PhD, is a senior lecturer in pharmacoepidemiology at Manchester Pharmacy School, University of Manchester. His research interest is in medicine’s use in chronic diseases, health services, and drug utilization research.

References

Arns

van der Heijden

K. B.

Arnold

L. E.

Kenemans

J. L.

(2013). Geographic variation in the prevalence of attention-deficit/hyperactivity disorder: The sunny perspective. Biological Psychiatry, 74, 585-590. doi:10.1016/j.biopsych.2013.02.010

Bhaskaran

Forbes

H. J.

Douglas

Leon

D. A.

Smeeth

(2013). Representativeness and optimal use of body mass index (BMI) in the UK Clinical Practice Research Datalink (CPRD). BMJ Open, 3(9), e003389. doi:10.1136/bmjopen-2013-003389

Bolea-Alamañac

Nutt

D. J.

Adamou

Asherson

Bazire

Coghill

. . .Young

S. J.

(2014). Evidence-based guidelines for the pharmacological management of attention deficit hyperactivity disorder: Update on recommendations from the British Association for Psychopharmacology. Journal of Psychopharmacology, 28, 179-203. doi:10.1177/0269881113519509

Bushe

Wilson

Televantou

Belger

Watson

(2015). Understanding the treatment of attention deficit hyperactivity disorder in newly diagnosed adult patients in general practice: A UK database study. Pragmatic and Observational Research, 6, 1-12. doi:10.2147/POR.S74161

Chisholm

(1990). The read clinical classification. British Medical Journal, 300(6732), 1092. Retrieved from http://www.pubmedcentral.nih.gov/articlerender.fcgi?artid=1662793&tool=pmcentrez&rendertype=abstract

Clinical Practice Research Datalink. (2015). CPRD linked data. Retrieved from http://www.cprd.com/dataAccess/linkeddata.asp

Döpfner

Breuer

Wille

Erhart

Ravens-Sieberer

(2008). How often do children meet ICD-10/DSM-IV criteria of attention deficit-/hyperactivity disorder and hyperkinetic disorder? Parent-based prevalence rates in a national sample—Results of the BELLA study. European Child & Adolescent Psychiatry, 17(Suppl. 1), 59-70. doi:10.1007/s00787-008-1007-y

Froehlich

T. E.

Lanphear

B. P.

Epstein

J. N.

Barbaresi

W. J.

Katusic

S. K.

Kahn

R. S.

(2007). Prevalence, recognition, and treatment of attention-deficit/hyperactivity disorder in a national sample of US children. Archives of Pediatrics & Adolescent Medicine, 161, 857-864. doi:10.1001/archpedi.161.9.857

Fulton

B. D.

Scheffler

R. M.

Hinshaw

S. P.

Levine

Stone

Brown

T. T.

Modrek

(2009). National variation of ADHD diagnostic prevalence and medication use: Health care providers and education policies. Psychiatric Services, 60, 1075-1083.

10.

Green

McGinnity

Á.

Meltzer

Ford

Goodman

(2005). Mental health of children and young people in Great Britain, 2004. Retrieved from http://www.esds.ac.uk/doc/5269/mrdoc/pdf/5269technicalreport.pdf

11.

Herrett

Gallagher

A. M.

Bhaskaran

Forbes

Mathur

van Staa

Smeeth

(2015). Data resource profile: Clinical Practice Research Datalink (CPRD). International Journal of Epidemiology, 44(3), 827-836. doi:10.1093/ije/dyv098

12.

Herrett

Thomas

S. L.

Schoonen

W. M.

Smeeth

Hall

A. J.

(2010). Validation and validity of diagnoses in the General Practice Research Database: A systematic review. British Journal of Clinical Pharmacology, 69, 4-14. doi:10.1111/j.1365-2125.2009.03537.x

13.

Hjern

Weitoft

G. R.

Lindblad

(2010). Social adversity predicts ADHD-medication in school children—A national cohort study. Acta Paediatrica, 99, 920-924. doi:10.1111/j.1651-2227.2009.01638.x

14.

Hoffmann

M. S.

Polanczyk

G. V.

Kieling

Dos Santos

I. P.

Willcutt

E. G.

Rohde

L. A.

Salum

G. A.

(2014). Attention-deficit/hyperactivity disorder and solar irradiance: A cloudy perspective. Biological Psychiatry, 76(8), e19-e20. doi:10.1016/j.biopsych.2013.07.044

15.

Holden

S. E.

Jenkins-Jones

Poole

C. D.

Morgan

C. L.

Coghill

Currie

C. J.

(2013). The prevalence and incidence, resource use and financial costs of treating people with attention deficit/hyperactivity disorder (ADHD) in the United Kingdom (1998 to 2010). Child and Adolescent Psychiatry and Mental Health, 7, Article 34. doi:10.1186/1753-2000-7-34

16.

Joint Formulary Committee. (2015). British national formulary. Available from http://www.medicinescomplete.com

17.

Kooij

S. J. J.

Bejerot

Blackwell

Caci

Casas-Brugué

Carpentier

P. J.

. . .Asherson

(2010). European consensus statement on diagnosis and treatment of adult ADHD: The European Network Adult ADHD. BMC Psychiatry, 10, Article 67.

18.

Mansell

(2013). HRA approvals clear CPRD for broad data linkage. Retrieved from http://www.pharmatimes.com/article/13-10-01/HRA_approvals_clear_CPRD_for_broad_data_linkage.aspx

19.

McCarthy

Wilton

Murray

M. L.

Hodgkins

Asherson

Wong

I. C. K.

(2012). The epidemiology of pharmacologically treated attention deficit hyperactivity disorder (ADHD) in children, adolescents and adults in UK primary care. BMC Pediatrics, 12, Article 78. doi:10.1186/1471-2431-12-78

20.

McDonald

D. C.

Jalbert

S. K.

(2013). Geographic variation and disparity in stimulant treatment of adults and children in the United States in 2008. Psychiatric Services, 64, 1079-1086. doi:10.1176/appi.ps.004442012

21.

Met Office. (2014). Solar energy. Retrieved from http://www.metoffice.gov.uk/renewables/solar

22.

Murray

M. L.

Hsia

Glaser

Simonoff

Murphy

D. G. M.

Asherson

P. J.

. . .Wong

I. C. K.

(2014). Pharmacological treatments prescribed to people with autism spectrum disorder (ASD) in primary health care. Psychopharmacology, 231, 1011-1021. doi:10.1007/s00213-013-3140-7

23.

National Audit Office. (2012). Healthcare across the UK: A comparison of the NHS in England, Scotland, Wales and Northern Ireland. Retrieved from http://www.nao.org.uk/wp-content/uploads/2012/06/1213192es.pdf

24.

National Institute for Health and Care Excellence. (2008a). Attention deficit hyperactivity disorder: Diagnosis and management of ADHD in children, young people and adults. Retrieved from http://publications.nice.org.uk/attention-deficit-hyperactivity-disorder-cg72/guidance

25.

National Institute for Health and Care Excellence. (2008b). Attention deficit hyperactivity disorder (ADHD): Full guideline. Retrieved from http://www.nice.org.uk/guidance/cg72/evidence/adhd-full-guideline-241963165National

26.

Institute for Health and Care Excellence. (2013). NICE support for commissioning for attention deficit hyperactivity disorder. Retrieved from http://www.nice.org.uk/guidance/qs39/resources/qs39-attention-deficit-hyperactivity-disorder-nice-support-for-commissioning2

27.

NHS Choices. (2014). Tips on how to choose a GP—Patient choice—NHS choices. Retrieved from http://www.nhs.uk/choiceintheNHS/Yourchoices/GPchoice/Pages/ChoosingaGP.aspx

28.

Nomura

Marks

D. J.

Grossman

Yoon

Loudon

Stone

Halperin

J. M.

(2012). Exposure to gestational diabetes mellitus and low socioeconomic status: Effects on neurocognitive development and risk of attention-deficit/hyperactivity disorder in offspring. Archives of Pediatrics & Adolescent Medicine, 166, 337-343. doi:10.1001/archpediatrics.2011.784

29.

Office of National Statistics. (2011). Super Output Area (SOA). Retrieved from http://www.ons.gov.uk/ons/guide-method/geography/beginner-s-guide/census/super-output-areas—soas-/index.html

30.

Rowlingson

Lawson

Taylor

Diggle

P. J.

(2013). Mapping English GP prescribing data: A tool for monitoring health-service inequalities. BMJ Open, 3(1), e001363. doi:10.1136/bmjopen-2012-001363

31.

Russell

Ford

Rosenberg

Kelly

(2014). The association of attention deficit hyperactivity disorder with socioeconomic disadvantage: Alternative explanations and evidence. Journal of Child Psychology and Psychiatry, and Allied Disciplines, 55, 436-445. doi:10.1111/jcpp.12170

32.

Springate

D. A.

Kontopantelis

Ashcroft

D. M.

Olier

Parisi

Chamapiwa

Reeves

(2014). ClinicalCodes: An online clinical codes repository to improve the validity and reproducibility of research using electronic medical records. PLoS ONE, 9(6), e99825. doi:10.1371/journal.pone.0099825

33.

Thomas

K. H.

Martin

R. M.

Davies

N. M.

Metcalfe

Windmeijer

Gunnell

(2013). Smoking cessation treatment and risk of depression, suicide, and self harm in the Clinical Practice Research Datalink: Prospective cohort study. British Medical Journal, 347, Article f5704. doi:10.1136/bmj.f5704

34.

Thomas

Mitchell

G. K.

Batstra

(2014). Attention-deficit/hyperactivity disorder: Are we helping or harming? British Medical Journal, 347, Article f6172. doi:10.1136/bmj.f6172

35.

UK Data Service. (2014). UK Data Service Census—Deprivation data. Retrieved from http://census.ukdataservice.ac.uk/get-data/related/deprivation.aspx

36.

West

Fleming

K. M.

Tata

L. J.

Card

T. R.

Crooks

C. J.

(2014). Incidence and prevalence of celiac disease and dermatitis herpetiformis in the UK over two decades: Population-based study. The American Journal of Gastroenterology, 109, 757-768. doi:10.1038/ajg.2014.55