Stephen V. Faraone, PhDA working group of the International League Against Epilepsy (ILAE), consisting of twenty experts spanning the globe (U.S., U.K., France, Germany, Japan, India, South Africa, Kenya, Brazil), recently published a “consensus paper” summarizing and evaluating what is currently known about comorbid epilepsy with ADHD, and best practices.

ADHD is two to five times more prevalent among children with epilepsy. The authors suggest that ADHD is underdiagnosed in children with epilepsy because its symptoms are often attributed either to epilepsy itself, or to the effects of antiepileptic drugs (AEDs).

The working group did a systematic search of the English-language research literature. It then reached consensus on practice recommendations, graded on the strength of the evidence.

Three recommendations were graded A, indicating they are well-established by evidence:

  • Children with epilepsy with comorbid intellectual and developmental disabilities are at increased risk of ADHD.
  • There is no increased risk of ADHD in boys with epilepsy compared to girls with epilepsy.
  • The anticonvulsant valproate can exacerbate attentional issues in children with childhood absence epilepsy (absence seizures look like staring spells during which the child is not aware or responsive). Moreover, a single high-quality population-based study indicates that valproate use during pregnancy is associated with inattentiveness and hyperactivity in offspring.

Four more were graded B, meaning they are probably useful/predictive:

  • Poor seizure control is associated with increased risk of ADHD.
  • Data support the ability of the Strengths and Difficulties Questionnaire (SDQ) to predict ADHD diagnosis in children with epilepsy: “Borderline or abnormal SDQ total scores are highly correlated with the presence of a validated psychiatric diagnosis (93.6%), of which ADHD is the most common (31.7%).” The SDQ can therefore be useful as a screening tool.
  • Evidence supports the efficacy of methylphenidate in children with epilepsy and comorbid ADHD.
  • Methylphenidate is tolerated in children with epilepsy.

At the C level of being possibly useful, there is limited evidence that supports that atomoxetine is tolerated in children with ADHD and epilepsy, and that the combined use of drugs for ADHD and epilepsy (polytherapy) is more likely to be associated with behavioral problems than monotherapy. In the latter instance, “Studies are needed to elucidate whether the polytherapy itself has resulted in the behavioral problems, or the combination of polytherapy and the underlying brain problem reflects difficult‐to‐control epilepsy, which, in turn, has resulted in the prescription of polytherapy.”

All other recommendations were graded U (for Unproven), “Data inadequate or conflicting; treatment, test or predictor unproven.” These included three where the evidence is ambiguous or insufficient:

  • Evidence is conflicted for the impact of early seizure onset on the development of ADHD in children with epilepsy.
  • Tolerability for amphetamine in children with epilepsy is not defined.
  • Limited evidence exists for the efficacy of atomoxetine and amphetamines in children with epilepsy and ADHD.

There were also nine U-graded recommendations based solely on expert opinion. Most notable among these:

  • Screening children with epilepsy for ADHD beginning at age 6.
  • Reevaluation of attention function after any change in antiepileptic drug.
  • Screening should not be done within 48 hours following a seizure.
  • ADHD should be distinguished from childhood absence epilepsy based on history and an EEG with hyperventilation.
  • Multidisciplinary involvement in transition and adult ADHD clinics is essential as many patients experience challenges with housing, employment, relationships, and psychosocial well‐being.

REFERENCES
Stéphane Auvin et al., “Systematic review of the screening, diagnosis, and management of ADHD in children with epilepsy. Consensus paper of the Task Force on Comorbidities of the ILAE Pediatric Commission,” Epilepsia (2018), doi: 10.1111/epi.14549. [Epub ahead of print].

 

Stephen V. Faraone, PhDRoughly one in thirty adult women have ADHD. Research results indicate that psychostimulants (methylphenidate and amphetamines) offer the most effective course of treatment in most instances. But during pregnancy, such treatment also exposes the fetus to these drugs.

Several studies have set out to determine whether such exposure is harmful. The largest compared 5,571 infants exposed to amphetamines and 2,072 exposed to methylphenidate with unexposed infants. It found no increased risks for adverse outcomes due to amphetamine or methylphenidate exposures.

Another study studied 3,331 infants exposed to amphetamines, 1,515 exposed to methylphenidate, and 453 to atomoxetine. Comparing these infants to unexposed infants, it found a slightly increased risk of preeclampsia, with an adjusted risk ratio of 1.29 (95% CI 1.11-1.49), but no statistically significant effect for placental abruption, small gestational age, and preterm birth. When assessing the two stimulants, amphetamine and methylphenidate, together, it found a small increased risk of preterm birth, with an adjusted risk ratio of 1.3 (95% CI 1.10-1.55). There was no statistically significant effect for preeclampsia, placental abruption, or small gestational age. Atomoxetine use was free of any indication of increased risk.

Another study involving 1,591 infants exposed to ADHD medication (mostly methylphenidate) during pregnancy, reported increased risks associated with exposure. The adjusted odds ratio for admission to a neonatal intensive care unit was 1.5 (95% CI 1.3-1.7), and for central nervous system disorders was 1.9 (95% CI 1.1-3.1). There was no increased risk for congenital malformations or perinatal death.

Six studies focused on methylphenidate exposure. Two, with a combined total of 402 exposed infants, found no increased risk for malformations. Another, with 208 exposed infants, found a slightly greater risk of cardiovascular malformations, but it was not statistically significant. A fourth, with 186 exposed infants, found no increased risk of malformations, but did find a higher rate of miscarriage, with an adjusted hazard ratio of 1.98 (95% CI 1.23-3.20). A fifth, with 480 exposed infants, also found a higher rate of miscarriage, with an odds ratio of 2.07 (95% CI 1.51-2.84). But although the sixth, with 382 exposed infants, likewise found an increased risk of miscarriage (adjusted relative risk 1.55 with 95% CI 1.03-2.06), it also found an identical risk for women with ADHD who were not on medication during their pregnancies (adjusted relative risk 1.56 with 95% CI 1.11-2.20). That finding suggests that all women with ADHD have a higher risk of miscarriage, and that methylphenidate exposure is not the causal factor.

Summing up, while some studies have shown increased adverse effects among infants exposed to maternal ADHD medications, most have not. There are indications that higher rates of miscarriage are associated with maternal ADHD rather than fetal exposure to psychostimulant medications. One study did find a small increased risk of central nervous system disorders and admission to a neonatal intensive care unit. But, again, we do not know whether that was due to exposure to psychostimulant medication, or associated with maternal ADHD.

If there is a risk, it appears to be a small one. The question then becomes how to balance that as yet uncertain risk against the disadvantage of discontinuing effective psychostimulant medication. As the authors of this review conclude:
It [ADHD] is associated with significant psychiatric comorbidities for women, including depression, anxiety, substance use disorders, driving safety impairment, and occupational impairment. The gold standard treatment includes behavioral therapy and stimulant medication, namely methylphenidate and amphetamine derivatives. Psychostimulant use during pregnancy continues to increase and has been associated with a small increased relative risk of a range of obstetric concerns. However, the absolute increases in risks are small, and many of the best studies to date are confounded by other medication use and medical comorbidities. Thus, women with moderate-to-severe ADHD should not necessarily be counseled to suspend their ADHD treatment based on these findings.

They advise that when functional impairment from ADHD is moderate to severe, the benefits of stimulant medications may outweigh the small known and unknown risks of medication exposure, and that “If a decision is made to take ADHD medication, women should be informed of the known risks and benefits of the medication use in pregnancy, and take the lowest therapeutic dose possible.”

REFERENCES
Allison S. Baker, Marlene P. Freeman, “Management of Attention Deficit Hyperactivity Disorder During Pregnancy,” Obstetrics and Gynecology Clinics of North America, vol. 45, issue 3 (2018), 495-509.

Stephen V. Faraone, PhDA team of Spanish researchers has published a systematic review of 16 studies with a total of 728 participants exploring the effects of physical exercise on children and adolescents with ADHD. Fourteen studies were judged to be of high quality, and two of medium quality.

Seven studies looked at the acute effects of exercise on eight to twelve-year-old youths with ADHD. Acute means that the effects were measured immediately after periods of exercise lasting up to 30 minutes. Five studies used treadmills and two used stationary bicycles, for periods of five to 30 minutes. Three studies “showed a significant increase in the speed reaction and precision of response after an intervention of 20–30 min, but at moderate intensity (50–75%).” Another study, however, found no improvement in mathematical problem solving after 25 minutes using a stationary bicycle at low (40–50%) or moderate intensity (65–75%). The three others found improvements in executive functioning, planning, and organization in children after 20- to 30-minute exercise sessions.

Nine studies examined longer-term effects, following regular exercise over a period of many weeks. One reported that twenty consecutive weekly yoga sessions improved attention. Another found that moderate to vigorous physical activity (MVPA) led to improved behavior beginning in the third week, and improved motor, emotional and attentional control, by the end of five weeks. A third study reported that eight weeks of starting the school day with 30 minutes of physical activity led to improvement in Connors ADHD scores, oppositional scores, and response inhibition. Another study found that twelve weeks of aerobic activity led to declines in bad mood and inattention. Yet another reported that thrice-weekly 45-minute sessions of MVPA over a ten-week period improved not only muscle strength and motor skills, but also attention, response inhibition, and information processing.

Two seventy-minute table tennis per week over a twelve-week period improved executive functioning and planning in addition to locomotor and object-control skills.

Two studies found a significant increase in brain activity. One involved two hour-long sessions of rowing per week for eight weeks, the other three 90-minute land-based sessions per week for six weeks. Both studies measured higher activation of the right frontal and right temporal lobes in children, and lower theta/alpha ratios in male adolescents.

All 16 studies found positive effects on cognition. Five of the nine longer-term studies found positive effects on behavior. No study found any negative effects. The authors of the review concluded that physical activity “improves executive functions, increases attention, contributes to greater planning capacity and processing speed and working memory, improves the behavior of students with ADHD in the learning context, and consequently improves academic performance.” Although the data are limited by lack of appropriate controls, they suggest that, in addition to the well-known positive effects of physical activity, one may expect to see improvements in ADHD symptoms and associated features, especially for periods of sustained exercise.

REFERENCES
Sara Suarez-Manzano, Alberto Ruiz-Ariza, Manuel De La Torre-Cruz, Emilio J. Martínez-López, “Acute and chronic effect of physical activity on cognition and behaviour in young people with ADHD: A systematic review of intervention studies,” Research in Developmental Disabilities, vol. 77, 12-23 (2018).

Stephen V. Faraone, PhDA study conducted at Auburn University in Alabama recruited 54 college students to address this question. All had previously been diagnosed with ADHD. All lived independently, and all were taking a prescribed ADHD medication. Students with severe comorbid psychiatric conditions were excluded. Three students dropped out, leaving a final sample size of 51.

Each student completed a total of four half-hour assessments, scheduled at monthly intervals. At each first assessment, researchers counted the participant’s ADHD medication pills and transferred them to an electronic monitoring bottle – a bottle with a microchip sensor in the cap that automatically tracks the date and time of every opening. This enabled them to compare students’ subjective estimates at subsequent assessments with the objective evidence from pill counts and from the data output from the electronic monitoring bottles.

Overall, students reported missing about one in four (25 percent) of their prescribed doses. But the objective measures showed they were in fact skipping closer to half their doses. According to pill counts they were missing 40 percent of their doses, and according to the electronic monitoring bottles, 43 percent. The odds of obtaining such a result due to chance with a sample of size were less than one in a hundred (p < 0.01).

In other words, college students with ADHD significantly overestimate their adherence rates to their medications. The authors concluded, “without additional strategies in place, expecting adolescents and young adults with ADHD to remember a daily task that requires no more than a few seconds to accomplish, such as medication taking, is unrealistic.” They suggest using smartphone reminder applications (“apps”) and text messaging services.

The authors caution that this was the first such study, and that it had a small sample size. Moreover, the study was not randomized. Students responded to advertisements posted on campus, and thus self-selected.

Pending the outcome of larger studies with randomization, the authors suggest that wherever possible, prescribing physicians adopt objective measures of medication adherence, as an aid to ensuring greater efficacy of treatment.

REFERENCES
Megan R. Schaefer, Scott T. Wagoner, Margaret E. Young, Alana Resmini Rawlinson, Jan Kavookjian, Steven K. Shapiro, Wendy N. Gray, “Subjective Versus Objective Measures of Medication Adherence in Adolescents/Young Adults with Attention-Deficit Hyperactivity Disorder,” Journal of Developmental & Behavioral Pediatrics, Published online July 11, 2018, DOI: 10.1097/DBP.0000000000000602.

Stephen V. Faraone, PhDA Norwegian team based at the University of Bergen recently performed a population study using the country’s detailed national health registries. With records from more than two and a half million Norwegians, the team examined what, if any, associations could be found between ADHD and nine autoimmune diseases: ankylosing spondylitis, Crohn’s disease, iridocyclitis, multiple sclerosis, psoriasis, rheumatoid arthritis, systemic lupus erythematosus, type 1 diabetes, and ulcerative colitis.

After adjusting for age and maternal education, the team found no association between ADHD and five of the nine autoimmune disorders: type 1 diabetes, rheumatoid arthritis, iridocyclitis, systemic lupus erythematosus, and multiple sclerosis. In the case of ankylosing spondylitis, it found no association with males with ADHD, but a negative association with females. Females with ADHD were less likely to have ankylosing spondylitis. The adjusted odds ratio (aOR) was 0.56 (95% CI 0.32-0.96).

Positive associations were found for only three autoimmune diseases. The strongest was for psoriasis, with adjusted odds ratios of 1.6 (95% CI 1.5-1.7) for females and 1.3 (95% CI 1.2-1.4) for males. When further adjusted for education, smoking, and body mass index (BMI), however, the adjusted odds ratio for females with ADHD dropped to 1.3 (95% CI 1.0-1.6).

The second strongest association was with Crohn’s disease. But here it was only among women. The odds ratio in this case was 1.4 (95% CI 1.2-1.8). Males with ADHD were actually less likely to have Crohn’s disease, with an odds ratio of 0.71 (95% CI 0.54-0.92).

Finally, females with ADHD were slightly more likely to have ulcerative colitis, with a barely significant odds ratio of 1.3 (95% CI 1.1-1.5), while no such association was found for males with ADHD, whose odds ratio was a statistically nonsignificant 0.9.

Given the large sample size of over two and a half million, this is no underpowered study. It found no association between ADHD and the generic category of autoimmune disorders. Furthermore, it is a stretch to argue that there are any clear and clinically meaningful links between ADHD and any of the specific disorders that were analyzed in this study. The small and often opposite effect sizes may simply reflect limitations with the data (presumed autoimmune disorders were identified based on drugs prescribed), or to other unidentified confounding factors.

REFERENCES
Tor‐Arne Hegvik, Johanne Telnes Instanes, Jan Haavik, Kari Klungsøyr, Anders Engeland, “Associations between attention‐deficit/hyperactivity disorder and autoimmune diseases are modified by sex: a population‐based cross‐sectional study,” European Child & Adolescent Psychiatry, vol. 27 (2018), 663-675.

Stephen V. Faraone, PhDAn international team of researchers has carefully examined the best current evidence and found strong evidence for an association between asthma and ADHD by combining a meta-analysis of prior data with a new analysis of the Swedish population.

The meta-analysis identified 46 datasets with a total of more than 3.3 million persons. It computed an unadjusted odds ratio (OR) of 1.7, which indicates that ADHD patients have about twice the risk of developing asthma compared with people without ADHD. Limiting the meta-analysis to studies that adjusted for confounding factors, 30 datasets with more than a third of a million participants still led to an adjusted odds ratio of 1.5 (95% CI 1.4 – 1.7). The likelihood of obtaining this result by chance in such a large sample would be less than one in ten thousand.

When the team further checked this result against the results for the Swedish population ofmore than one and a half million persons, the odds ratio was an almost identical 1.6. Adjusting for confounding factors reduced it to 1.5 (95% CI 1.41 – 1.48). That means the findings are very robust: asthma and ADHD are associated, with an odds ratio of 1.5, after adjusting for confounding factors.

What does this small but statistically very reliable association between asthma and ADHD mean? For researchers, it suggests that the two disorders may have common risk factors and that the search for these shared risk factors might lead to improved treatments. These risk factors might also be shared with two other somatic conditions for which ADHD patients are at increased risk: obesity and eczema. It is possible that common inflammatory processes account for this overlap among disorders. Clinicians should be aware that children with asthma have an increased risk for ADHD, although given the small association, systematic screening may not be warranted. But given that ADHD might interfere with asthma medication compliance, the disorder should be considered among noncompliant youth, especially those who show other evidence of inattention, poor memory or disorganization.

REFERENCES
Samuele Cortese, Shihua Sun, Junhua Zhang, Esha Sharma, Zheng Chang, Ralf Kuja-Halkola, Catarina Almqvist, Henrik Larsson, Stephen V Faraone, “Association between attention deficit hyperactivity disorder and asthma: a systematic review and meta-analysis and a Swedish population-based study,” Lancet Psychiatry, Published online July 24, 2018.
https://www.thelancet.com/journals/lanpsy/article/PIIS2215-0366(18)30224-4/fulltext

This two-year study examined the effect of digital media use on ADHD symptoms in over 2500 adolescents. An earlier meta-analysis found that traditional media use (TV and video console games) was modestly associated with ADHD-like behaviors (Nikkelen et al 2014). The current study extends the examination to a large sample, with modern digital media delivery of high-intensity stimuli, including mobile platforms. The authors used the Current Symptom Self-Report Scale (Barkley R 1998) to establish ADHD symptoms at baseline and at six-month assessments over a 24 month period. None of the subjects reported having ADHD at study entry. Subjects were considered to be ADHD symptom positive (the primary binary outcome) is they had greater than or equal to six inattentive and/or hyperactive-impulsive symptoms rated on this frequency-based scale (0-3).

Modern digital media use was surveyed on a frequency basis for 14 media activities (including checking social media sites, texting, browsing, downloading or streaming music, posting pictures, online chatting, playing games, online shopping, and video chatting). The most common media activity was high-frequency checking of social media. Of note, high-frequency engagement in each of the digital media activities was significantly, but moderately, associated with having ADHD symptoms at each six-month follow-up (OR 1.10), even after adjusting for covariates.

High-frequency media use at baseline seemed to be associated with development of ADHD symptoms. Among the 495 students who reported no high-frequency media use at baseline, 4.6% met ADHD symptom criteria at follow-up. Among 114 students scoring 7 for high-frequency media use at baseline 9.5% met the symptoms criteria. For the 51 students with a score of 14 for high-frequency media use at baseline, the rate was 10.5% (both comparisons were statistically significant).

This study is important in that it notes that an association between high-frequency digital media use (in current platforms and modalities) may be associated with the development of ADHD-like symptoms. A significant limitation of the study, as noted by the authors, is that ADHD-like symptoms do not establish a diagnosis of ADHD and do not assess impairment; therefore, these results must be interpreted with some caution. It does highlight that even with the current level of understanding it might be prudent for clinicians to recommend limiting high-frequency media use for adolescent patients.

REFERENCES
Barkley RA. Attention-Deficit Hyperactivity Disorder: A Clinical Workbook. 2nd ed. New York, NY: Guilford Press; 1998.

Nikkelen SW, Valkenburg PM, Huizinga M, Bushman BJ. Media use and ADHD-related behaviors in children and adolescents: a meta-analysis. Dev Psychol. 2014;50(9):2228-2241. doi:10.1037/a0037318

Ra CK, Junhan Cho J, Stone MD, De La Cerda J, Goldenson NI, Moroney E, Tung I, Lee SS, Leventhal AM. Association of Digital Media Use With Subsequent Symptoms of Attention-Deficit/Hyperactivity Disorder Among Adolescents JAMA. 2018;320(3):255-263. doi:10.1001/jama.2018.8931

This article provides a review of the cardiovascular effects of ADHD medications including potential effects on blood pressure, heart rate and risk of cardiovascular events (myocardial infarction, sudden death and stroke).

The article notes that meta-analyses have generally found that the effects of stimulant medications and atomoxetine were generally similar on systolic blood pressure (1-3 mm Hg) and heart rate (2-5 beats/minute); these were felt to be of limited clinical significance, except for patients with elevated blood pressures or heart rate antecedent to starting these ADHD therapies. However, as these are average changes, changes in individual patients may vary and important to monitor. Additionally, the meta-analysis and observational data available also do not find significantly higher risks for MI or stroke in patients receiving stimulant medications. These findings are complicated by the use of clinical trial data in the meta-analysis which specifically limit is the enrollment of patients with higher risks of pre-existing cardiovascular illnesses and the observational data were of relatively short treatment exposures.

This article is important for clinicians because it reviews the cardiovascular safety profiles of current ADHD medications and also recommends monitoring of blood pressure and pulse at baseline and during treatment. Furthermore, the authors recommend baseline screening patients for significant cardiovascular histories via family history of cardiovascular disease and sudden death.

REFERENCES
https://www.healio.com/psychiatry/journals/psycann/2018-7-48-7/%7B426ecc52-e3d9-4f38-afc6-34cbf88548c7%7D/review-of-cardiovascular-effects-of-adhd-medications#divReadThis

Lenard Adler, MD ADHD in AdultsThis article reviews the phenomenology of emotional dysregulation in adult ADHD. The article discusses whether symptoms of emotional dysregulation (ED) are co-traveling symptoms that travel with symptoms of adult ADHD or whether they are part of the core symptoms of the condition.

Symptoms of ED include rapidly shifting affect, changeable mood, mood lability, impulsivity and emotional overactivity. Barkley et al. (7 from article) have posited that ED symptoms are part of a set of executive function deficits which are critical to the core of ADHD symptomtology (8).

In contrast, Wender (13) and co-workers have posited that ED symptoms should be part of the diagnostic criteria of adult ADHD and are included in the Utah criteria they defined and the Wender-Reimherr Adult Attention-Deficit Disorder Scale (WRAADS). Work from Adler and Kessler and co-workers (21) found that ED symptoms tracked separately from symptoms of inattention, hyper-activity impulsivity and executive function and were more likely to load on the Combined presentation of ADHD and be present in sub-threshold cases. The article also reviews a number of scales which have been used to assess ED, including the WRAADS, the expanded Adult ADHD Clinician Diagnostic Scale (ACDS) v.1.2, expanded Adult ADHD Self-Report Scale (ASRS) v1.1 Symptom Checklists and the Brown Adult ADD Scale (BAADS).

This article is important to clinicians because: 1) it defines and highlights the importance of recognizing symptoms of ED, 2) describes assessment methods and 3) notes the lower efficacy of standard adult ADHD pharmacotherapies of stimulants and atomoxetine than on core ADHD symptoms and 4) highlights the potential utility of cognitive behavioral therapy in treating ED symptoms in adults with ADHD.

REFERENCES
https://www.healio.com/psychiatry/journals/psycann/2018-7-48-7/%7Bd9674afc-698a-4cd9-9a24-81e3e4d4d944%7D/emotional-dysregulation-in-adult-adhd

Stephen V. Faraone, PhDAn international team of researchers recently published a meta-analysis of randomized controlled trials examining the efficacy of meditation-based therapies. Thirteen randomized controlled clinical trials (RCTs) were included: seven, with 270 participants, focused on children and adolescents; the other six, with 339 participants, on adults. Because only one of the RCTs was appropriately blinded, the results discussed below, although promising, must be considered preliminary.

Among children and adolescents, meta-analysis revealed a significant, medium effect size (SMD = -0.44, 95% CI -0.69 to -0.19) on ADHD symptoms for meditation therapy versus no treatment. There was virtually no heterogeneity among studies and no sign of publication bias. Improvements in inattention and hyperactivity/impulsivity had similar effect sizes. Neuropsychological measures of inhibition and attention indicated small-to-medium effect sizes, but failed to achieve statistically significance, perhaps due to the small numbers of trials and participants (159 and 179, respectively).

For adults, the significant effect size on ADHD symptoms was medium-to-large (SMD = -.66, 95% CI -1.21 to -0.11). Once again, there was little sign of publication bias. But in this case, there was great heterogeneity among the studies. Improvements in inattention and hyperactivity/impulsivity were again comparable, although they fell just short of statistical significance for the latter. Neuropsychological measures of the efficacy of medication therapy produced statistically significant medium effect sizes for inhibition (SMD = -0.54) and working memory (SMD = – 0.42), with virtually no heterogeneity or sign of publication bias.

Although these results are promising, the authors of the meta-analysis concluded, “Despite statistically significant effects on ADHD combined core symptoms, due to paucity of RCTs, heterogeneity across studies and lack of studies at low risk of bias, there is insufficient methodologically sound evidence to support meditation-based therapies for ADHD.”

REFERENCES
Junhua Zhang, Amparo Díaz-Román, Samuele Cortese, “Meditation-based therapies for attention-deficit/hyperactivity disorder in children, adolescents and adults: a systematic review and meta-analysis,” Evidence-Based Mental Health, Published Online First: 10 July 2018. doi:10.1136/ebmental-2018-300015 (2018).