The prevalence of psychiatric comorbidities in adult ADHD compared with non-ADHD populations: A systematic literature review

1ADHD(attention-deficit/hyperactivity disorder) is a common psychiatric disorder presenting persistent inattention and/or with hyperactivity/impulsivity, which is associated with considerable problems in personal, social, and occupational areas. While ADHD is well studied in children, it is recently being studied in adults as well. According to a previous meta-analysis, 65% of children who were diagnosed with ADHD have persistent ADHD symptoms in adulthood. In addition, the prevalence of ADHD in adults is known to reach 2.5%, which is moderate compared to its prevalence in children, which is about 5%.

2Although comorbid psychiatric disorders are common in both adults and children, the comorbidity rate is higher in adults; as many as 80% of adults with ADHD are reported to have at least one comorbid psychiatric disorder. In clinical adult ADHD samples, substance use disorder (SUD), mood disorder, anxiety disorder, and antisocial personality disorder (ASPD) are the most common comorbid disorders, and these mental disorders can adversely affect patient prognosis. Furthermore, research revealed that comorbid psychiatric disorders cause considerable functional impairment in individuals with ADHD and place a great burden on society.

3For this reason, several cross-sectional studies have been conducted on various populations including clinical and general settings over 30 years to evaluate the prevalence of comorbid psychiatric conditions in adults with ADHD. However, the prevalence of comorbid psychiatric disorders varied according to characteristics of the subjects, including country, race, gender, and other socioeconomic characteristics as well as the screening or diagnostic tools applied. Moreover, since ADHD has been recognized in adults, diagnostic tools for adult ADHD and its comorbid disorders have changed over time, and the interest in clinical diagnoses and optimal treatments in adults with ADHD has also increased. These factors might have contributed to the divergent prevalence rate of ADHD and comorbid disorders in adults.

4However, to the best of our knowledge, despite the high prevalence of psychiatric disorders documented in previous studies in the adult ADHD subjects and their importance in the clinical field, no systematic literature review has specifically compared the prevalence of comorbid psychiatric disorders between adults with and without ADHD. Considering the high prevalence of adult ADHD and its impact on quality of life, a perspective on frequent comorbid psychiatric disorders would be helpful for individuals with ADHD and clinicians. Thus, the aim of our study was to ascertain the difference in the prevalence rates of comorbid psychiatric disorders between adults with and without ADHD including both clinical and general populations.

5The methodology of the present review followed the Preferred Reporting Items for Systematic Reviews and Meta-Analysis (PRISMA). Following PRISMA guideline, we conducted research based on PICO (Population, Intervention, Comparison and Outcome). The target population was adults with diagnosed ADHD. We compared the prevalence of comorbid psychiatric disorders between ADHD and non-ADHD patients. We searched electronic libraries of PubMed, EMBASE, PsycINFO, PsycNET, and Google Scholar for publications regarding the epidemiology and prevalence rate of comorbidities of adult ADHD published from 1 January 1990 to 1 August 2022. The initial search was conducted by two authors (WSC, YSW) using the following terms: Prevalence AND (ADHD OR ADD OR Attention Deficit) AND Adult AND (comorbidity OR comorbid) in titles or abstracts. Each database was updated as appropriate when preparing this submission for publication. Full electronic search strategies are provided in the S1 Text.

6First, articles obtained from the initial search were de-duplicated by EndNote 20. Then, inclusion/exclusion screening was performed by lead authors (WSC, YSW) based on exclusion criteria of non-relevant articles (e.g., did not focus on adult patients or did not include psychiatric comorbidity data), non-English articles, full text not available, abstract-only papers, and articles that were not peer-reviewed. We included all types of research except a systematic review or meta-analysis defined by title and method. The initial inclusion/exclusion review was based on titles and abstracts; if the relevance of the article was unclear, a full-text review was performed to determine the eligibility of each study. After this initial process, the full texts of all included articles were retrieved to evaluate our detailed eligibility criteria. Articles were included in the study if they 1) used samples of adult populations aged 18 years or older, 2) defined clear ADHD and non-ADHD groups by clinically diagnoses or using any diagnostic criteria (e.g., DSM (Diagnostic and Statistical Manual of Mental Disorders)) or tools for screening/diagnosing ADHD in adults (e.g., ASRS (Adult ADHD Self-report Scale)), 3) defined the prevalence rate of comorbid psychiatric disorders using any diagnostic tools for each psychiatric disorder (e.g. SCID (Structured Clinical Interview for DSM-IV)), and 4) directly compared ADHD and non- ADHD groups using statistical analysis. Any discrepancy between the two lead authors during study selection was resolved through discussion, and other authors were consulted if necessary. The inclusion consistency between the two authors was 94.1% (32/34).

7Microsoft Excel was used to develop a data extraction spreadsheet, and all included full-text articles were reviewed by both researchers (WSC, YSW), who also conducted the initial data search and study selection process. The extracted data were reviewed for consistency, and any queries that arose were resolved by discussion among the researchers. The lead author decided whether to include/exclude data by reviewing the specific articles.

8Because there are various methods for diagnosing ADHD and psychiatric disorders in adults, we extracted the following variables from the articles ultimately included: 1) data describing the study characteristics, such as year of publication, country, or study design; 2) data describing the target population, like sample size, age range or mean age/SD, or gender composition; 3) diagnostic tools for adult ADHD and comorbid psychiatric disorders, whether clinical diagnosis was performed, and the diagnostic criteria for ADHD/psychiatric comorbidity; 4) study results including the prevalence rate of ADHD in the target population, prevalence rate of each psychiatric comorbidity in each ADHD and non-ADHD group, and any statistically significant comparable variables including odd ratios(ORs) with 95% confidence intervals or chi-square (χ2) test variables.

9Based on several studies targeting nation-wide psychiatric comorbidities, assuming that the prevalence of comorbid psychiatric disorders is higher in groups of psychiatric patients, we decided to divide the study populations into general population group studies and clinical group studies. A clinical group study was defined as one in which the study population included patients who had previously been clinically diagnosed with any psychiatric disorder or had visited/been admitted either voluntarily or involuntarily to a hospital for treatment. A general group study was defined as that in which the whole study population was not diagnosed with any psychiatric disorders before the start of each study.

10In addition, considering the specificity that the prevalence of ADHD among incarcerated people was five to 10 times higher than that of the general population, and that the prevalence of comorbid psychiatric disorders among inmates was higher than that of the general population, we classified data of incarcerated patients separately from other population groups. The incarcerated group study was defined as a study that only included incarcerated participants.

11In total, 1768 articles were identified by the search method described above, and 335 duplicates were removed. After the duplicates were excluded, an additional 1121 articles were excluded by screening titles and abstracts. The remaining 292 articles were read in full and included in the analysis if they met the inclusion criteria of our study. Based on our study criteria, 260 articles were excluded for reasons noted in Fig 1. Thus, 32 studies comparing the prevalence rates of comorbid psychiatric disorders between ADHD and non-ADHD adult subjects were selected for systematic review.

12Of the 32 studies comparing the prevalence of comorbid psychiatric disorder between subjects with and without adult ADHD, according to our classification criteria, 11 studies involved general populations, 18 studies included psychiatric populations, and three studies focused on incarcerated populations. One of the three studies dealing with incarcerated populations involved only female inmates, and the other one study involved only male inmates.

13In this review, 12 diagnostic tools including clinical diagnostic criteria like DSM or ICD(International Classification of Disease), were used to evaluate adult ADHD. In addition, five diagnostic tools were mainly used for comorbid psychiatric disorders. The most used evaluation tool for adult ADHD was Adult ADHD Self-Report Scale (ASRS), which was used in 12 studies. Five studies used the ASRS alone to evaluate ADHD in adults, and the rest of the studies used more than one tool to evaluate ADHD. The next most frequently used evaluation tools were Conner’s Adult ADHD Diagnostic Interview for DSM-IV(CAADID) and the Wender Utah Rating Scale (WURS).

14Nineteen studies provided data comparing the prevalence of mood disorders (including depressive disorders and bipolar disorders) between ADHD patients and non-ADHD individuals. In the general populations, the prevalence of any depressive disorder in the non-ADHD group was estimated at 1.2% to 12.5%, compared to 8.6% to 55% in the ADHD group. In clinical populations, the prevalence of any depressive disorder in the non-ADHD group was estimated at 5.8% to 39.6%, compared to 15.4% to 39.7% in the ADHD group. In the general population, the prevalence of any bipolar disorder in the non-ADHD group was estimated at 0.2% to 3.6% compared to 4.48% to 35.3% in the ADHD group. In clinical populations, the prevalence of any bipolar disorder in the non-ADHD group was estimated at 2.0% to 19.5%, compared to 7.4% to 80.0.% in the ADHD group. There were no differences reported in the prevalence of mood disorders between ADHD and non-ADHD groups in the incarcerated population studies. Detailed information from each study is summarized in Table 1.

15Sixteen studies provided data comparing the prevalence of anxiety disorders including obsessive-compulsive disorder, somatoform disorders and trauma/stress-related disorders between ADHD patients and non-ADHD individuals. In general population, the prevalence of any anxiety disorders in the non-ADHD group was estimated at 0.5% to 9.5% compared to 4.3% to 47.1% in ADHD group. In clinical populations, the prevalence of any anxiety disorders in the non-ADHD group was estimated at 5.4% to 40% compared to 3.9% to 84% in the ADHD group. Only one study of incarcerated populations showed a difference in the prevalence of social phobia between non-ADHD and ADHD individuals. Detailed information from each study is summarized in Table 2.

16Twenty-two studies provided data comparing the prevalence of substance use disorders (including addiction to alcohol, opioids, stimulants, cannabis, anxiolytics, and nicotine) and gambling disorders between ADHD and non-ADHD individuals. In general populations, the prevalence of any substance use disorder in the non-ADHD group was estimated at 0% to 16.6% compared to 2.3% to 41.2% in the ADHD group. In clinical populations, the prevalence of any substance use disorder in the non-ADHD group was estimated to be 2.0% to 72.2% compared to 10.0% to 82.9% in the ADHD group. Two studies compared the prevalence of gambling disorder between ADHD and non-ADHD patients, and there was one study for each general/psychiatric population group, showing a statistically significant difference in prevalence. Two studies of incarcerated populations showed differences in the prevalence of benzodiazepine use disorder and drug dependence. Detailed information from each study is summarized in Table 3.

17Fourteen studies provided data comparing the prevalence of personality disorders (including borderline personality disorder and antisocial personality disorder) between ADHD and non-ADHD individuals. In general populations, the prevalence of any personality disorders in the non-ADHD group was estimated at 0% to 3.9% compared to 0.31% to 33.8% in the ADHD group. In clinical populations, the prevalence of any personality disorder in the non-ADHD group was estimated at 6.6% to 34.4% compared to 21.9% to 65.95% in the ADHD group. Two studies of incarcerated populations showed differences in the prevalence of borderline personality disorder and antisocial personality disorder. The prevalence of antisocial personality disorder was higher in the ADHD group in both studies, and the prevalence of borderline personality disorder was higher in one study. Detailed information from each study is summarized in Table 4.

18In our systematic review, we included 32 studies conducted over 16 years dealing with the prevalence of psychiatric comorbidities between adults with and without ADHD. To our knowledge, this is the first systematic review comparing the prevalence of comprehensive comorbid psychiatric disorders between adults with and without ADHD and including both general and clinical populations. Articles published from 2006 to 2022 were included in this review. This might because the interest in adult ADHD had increased based on the results of longitudinal follow-up studies on children and adolescents with ADHD.

19The studies included in this review most commonly used ASRS as a diagnostic tool for adult ADHD (12 of 32), followed by CADDID and WURS. As a self-report scale, due to its simplicity and cost-effectiveness, ASRS can be preferred by ADHD investigators. However, the variability in the evaluation tools for adult ADHD is thought to be due to the lack of established diagnostic criteria. Many psychiatric disorders are diagnosed based on observation of an assessor and complaint of a patient. The lack of standardized diagnostic tools can lead to over-diagnosis or under-diagnosis of psychiatric disorders and can cause inconsistency in diagnosis, which complicates comparison studies. Heterogeneity of diagnostic tools observed in this study not only prevented meta-analysis, but also might affect the variability in the reported prevalence of adult ADHD.

20In addition, the prevalence of ADHD in adults varied from 1.1% to 8.8% in general population samples and from 5.12% to 35.75% in psychiatric population samples. A previous systematic review of ADHD prevalence in and adult psychiatric population shows a similar range, from 6.9% to 38.75%. In this previous study, the authors assumed that this variation might be due to the diversity of diagnostic methods and the inclusion and exclusion criteria in the studies. Similarly, in our study, the aforementioned variability of diagnostic methodologies for ADHD might have affected this various range of prevalence. In a general population, the estimated mean prevalence rate of ADHD in adults was 2.8% in a previous study. Except in two studies included in our review, targeting special populations of army soldiers and twins of Sweden, which had higher than estimated prevalence, the range of ADHD prevalence in the general population was 1.1% to 2.9% in our study, similar to that previously observed.

21The most frequent comorbid psychiatric disorder in the ADHD group was SUD, its prevalence ranging from 2.3% and 41.2% in the general population and between 10.0% and 82.9% in the clinical population; two of three studies showed significant prevalence difference between ADHD and non-ADHD subjects. This finding correlates with a previous meta-analysis that reported that almost one out of every four adolescent and adult patients with SUD presents with ADHD, which supports SUD as one of the most frequent comorbid psychiatric conditions in adult ADHD. There are some theoretical opinions of shared key characteristics and pathophysiology between ADHD and SUD, like dopaminergic dysregulation of motivational and reward systems, or reduced frontal function including executive functions and response inhibition. In addition, considering that childhood ADHD is a prominent risk factor for substance misuse and development of SUD due to the most frequent comorbidities in childhood ADHD, like conduct disorder or oppositional defiant disorder, untreated and preserved ADHD in adults might have influenced the cross-sectional difference of prevalence rate between ADHD and non- ADHD patients.

22Mood disorders, including depressive disorders and bipolar disorders, were also frequently observed comorbid psychiatric disorders in ADHD subjects compared to non-ADHD subjects. The estimated prevalence of depressive disorders in the ADHD group ranged from 8.6% to 55% in the general population and 15.4% to 39.7% in the clinical population. Also, the prevalence of bipolar disorder in the ADHD group was estimated at 4.45% to 35.3% in the general population and 7.4% to 80.8% in the clinical population. For depression, previous studies have also shown a higher prevalence of depressive disorders in young adults with ADHD compared to non-ADHD subjects as well as higher risk of suicidal behavior. This can be explained by a previous cross-sectional study showing the association between ADHD symptoms and depressive symptoms in young adults as identified by low hedonic responsivity. In addition, according to biologic aspects of depression and ADHD, the two disorders might share similar pathophysiologic regions of the brain including decreased activity in the prefrontal, amygdala, and hippocampus regions. Furthermore, 10 studies showed a higher prevalence of bipolar disorder, including current hypomania diagnosed by SCID-II, in ADHD subjects than in non-ADHD subjects. Considering that the worldwide prevalence rate of bipolar disorder is estimated as 1–3%, which was similar to that of the non-ADHD general population in our study, the prevalence of bipolar disorder in the ADHD group was greater than 3% in all 10 studies. This finding correlated with previous studies reporting reciprocal high comorbidity rates between ADHD and bipolar disorder, which suggests possible shared genetic effects or diagnostic overlap between the two disorders.

23In anxiety disorder, almost two of three studies showed a higher prevalence in the ADHD group than the non-ADHD group. The prevalence rate in the ADHD group was estimated to range from 4.3% to 47.1% in the general population and from 3.9% to 84% in the clinical population. Only one study of a clinical population dealing with psychiatric outpatients in the US showed a higher prevalence of adjustment disorder in the non-ADHD group. These findings correlate with previous studies that revealed a high prevalence of anxiety in the adult ADHD population. ADHD seems to show different characteristics from anxiety, namely fearlessness and impulsivity. Therefore, various theories have been suggested to explain this phenomenon using developmental or biologic aspects in children and adolescents. Similarly, in adults, as far as we know, the two disorders have been related to several common neuroanatomical regions like the dorsolateral prefrontal cortex or the anterior cingulate cortex, which are critically involved with the executive function control network. In addition, considering a previous study about increased risk of accidents in ADHD over the lifespan, traumatic events might have influenced the higher prevalence of anxiety disorders. From a developmental viewpoint, as in depressive disorders, this frequently higher prevalence of anxiety disorder might represent the social and relational difficulties induced by ADHD.

24The estimated prevalence of personality disorders in the ADHD group was ranged from 0.31% to 33.8% in the general population and from 21.9% to 65.95% in the clinical population. Previous studies have reported that personality disorders, mostly cluster B or C personality disorders, are present in almost 50% of adults with ADHD. The association between ADHD and personality disorders might be mediated by the symptomatic dimensions of ADHD such as emotional dysregulation and oppositional symptoms. In our review, most studies showed a higher prevalence of cluster B personality disorders in ADHD than in non-ADHD groups. Specifically, in the clinical population, more than 20% of adult ADHD subjects were estimated to have comorbid cluster B personality disorder including borderline personality disorder and antisocial personality disorder. Additionally, most clinical population studies included patients diagnosed with substance use disorders, which correlates with previous observational studies of young male adults with ADHD that revealed associations of antisocial personality disorder with ADHD.

25There are several limiting factors in this review. As mentioned previously, there is significant heterogeneity across studies diagnosing both ADHD and comorbid psychiatric disorders. This prohibited meta-analysis. Furthermore, except for two international studies, the included studies were conducted in high-income regions like Europe or North America. In a previous epidemiologic study investigating cross-national ADHD prevalence in adults, the prevalence differed by country income, with higher rates observed in higher-income countries. This difference in prevalence among countries affects the degree of interest in the disease, which might be why the included studies were mainly conducted in Europe and North America. In addition, we did not differentiate patients according to ADHD or comorbid psychiatric disorder treatment status, which might also have affected the prevalence of comorbid psychiatric disorders. Of the included studies, most explored the prevalence cross-sectionally, so we could not infer a correlation or antecedent relationship between ADHD and comorbidities. Only limited estimates of the associations between ADHD and comorbidities can be provided by our review at the study level. Additionally, we did not assess the risk of bias in each study. To include as many studies as possible to reflect broad and various studies of different countries, populations, and comorbid psychiatric disorders, we omitted the risk of bias assessment. Also, there was no pre-registration for our systematic literature review. These points are limitations to this study.

26In conclusion, our findings indicate a higher prevalence of comorbid psychiatric disorders in ADHD subjects compared to non-ADHD subjects, whether they were previously diagnosed with other psychiatric disorders or not. Furthermore, our results suggest a complex association between the multiple comorbidities of ADHD. Given that ADHD is often unrecognized and under-diagnosed in adults, screening for ADHD might be beneficial for patients presenting multiple psychopathologies including substance abuse, mood disorders, and anxiety disorders. In the future, research on standardization of ADHD diagnosis in adults and its comorbid psychopathologies will be required to distinguish adult ADHD from comorbid psychiatric disorders, and to enable comparison among study conditions. This standardization will aid in differential diagnosis and allow provision of earlier treatment in adult ADHD. In addition, research on the neurobiological and developmental bases of ADHD and its comorbid psychiatric disorders should continue to improve the understanding of the connectivity and associations between various comorbid psychiatric disorders and ADHD in adults.