Abstract
Objective:
To provide an overview of Joe Biederman’s contributions to child and adolescent psychiatry.
Method:
Nine colleagues described his contributions to: psychopharmacology, comorbidity and genetics, pediatric bipolar disorder, autism spectrum disorders, Tourette’s and tic disorders, clinical and neuro biomarkers for pediatric mood disorders, executive functioning, and adult ADHD.
Results:
Joe Biederman left us with many concrete indicators of his contributions to child and adolescent psychiatry. He set up the world’s first pediatric psychopharmacology clinic and clinical research program in child adolescent psychiatry. As a young faculty member he began a research program that led to many awards and eventual promotion to full professor at Harvard Medical School. He was for many years the most highly cited researcher in ADHD. He achieved this while maintaining a full clinical load and was widely respected for his clinical acumen.
Conclusion:
The world is a better place because Joe Biederman was here.
Introduction (By Stephen Faraone, PhD)
Joseph Biederman, a tireless clinician and persistent innovator in child and adolescent psychiatry, passed away on January 5, 2023. He left us an indelible legacy. His prolific and highly cited work in evidence-based diagnosis and treatment along with insights into the nature of mental disorders in youth has had profound implications for the lives of affected children and their caregivers around the world.
The tale of Joe Biederman is one steeped in resilience. Born in Prague, Czechoslovakia, Joe entered a world stained with the horrors of the Holocaust, which claimed most of his extended family. His parents survived, fled to Argentina, and began anew, a beacon of hope for their newborn son. Indeed, adversity was Joe’s cradle.
In Argentina, his parents set him on the path of healing and understanding by sending him to medical school in Buenos Aires. The journey led him to Jerusalem, where he studied psychoanalysis. Fortunately, he joined Dr. Robert Belmaker’s group which was researching groundbreaking psychopharmacologic methods. Witnessing the transformative impact of lithium on the lives of patients suffering from bipolar disorder was a turning point for Joe. The budding psychoanalyst metamorphosed into a psychopharmacologist.
Life took a tragic turn when Joe’s wife was diagnosed with cancer. The quest for her treatment brought the young Biederman family to Boston. Despite ultimately losing his wife and facing the daunting task of raising a 4-year-old son alone in a new land, Joe remained unyielding in the face of grief. He would rise from the ashes of adversity, stronger, and more determined. At the Massachusetts General Hospital, Joe set up the world’s first pediatric psychopharmacology clinic and a clinical research program in child and adolescent psychiatry. He would meet and marry Helen and together create a loving family with children and grandchildren that brought him a lifetime of joy.
Joe’s research started earning recognition as early as 1977, when he won the Research Prize in Medicine from the Hebrew University and Hadassah Medical School for his work on Cyclic AMP in the CSF of patients with schizophrenia. In 1978, he won the prestigious Bornstein Prize for Research in Psychiatry and Neurology for his research on the effects of L-Dopa and Propranolol on human CSF Cyclic Nucleotides. The 1980s saw him gaining more recognition with his fellowship in the American Academy of Child Psychiatry in 1986. The same year, he was made a Fellow of the American Psychiatric Association, further solidifying his standing in the field. In the 1990s, he began being consistently recognized as one of the best doctors in America, appearing in various listings of top physicians. He was awarded the American Psychiatric Association Blanche Ittelson Award for Excellence in Child Psychiatric Research. In 1995, he was inducted into the Hall of Fame of CHADD (the US advocacy group for ADHD) for his clinical work and research about ADHD.
After Joe rose to national prominence as a Full Professor at Harvard Medical School, adversity struck again when he was targeted by the anti-psychiatry movement fueled by articles in the New York Times containing lies and innuendo. These articles implied that he had manipulated his work for the benefit of pharmaceutical companies. A 4-year independent investigation found no ethical violations and nothing to question the integrity of his research or his clinical care. The investigation reported no evidence to support newspaper innuendos that he had violated federal regulations or had engaged in unethical behavior or scientific misconduct. Thus, he kept his Professorship at Harvard and his leadership positions at the Massachusetts General Hospital where he was awarded an Endowed Chair.
His passion for the field, his tenacity, and his willingness to challenge the norms have left a profound mark on the clinical research literature in psychiatry. Joe’s innovative work sparked many ground-breaking studies. These studies are summarized in Table 1 and described in more detail in this article. Joe’s high level of impact is well documented. His publications have been cited by colleagues 52,251 times, leading to an enviable lifetime h-index of 122. His citation metrics placed him in a select group: the top 0.01% of scientists across all fields of study (Ioannidis et al., 2019). In 2000, the Institute for Scientific Information recognized him as the fourth-highest producer of high-impact papers in psychiatry. His work continued to receive international recognition in the 2000s. He received the NARSAD Senior Investigator Award in 2002, and in 2005, the Institute for Scientific Information ranked him at #1 in terms of total citations to his papers published on ADHD in the prior decade. The same institute also placed him as the #2 highest producer of high-impact papers in psychiatry. In 2021 he received the World Federation of ADHD Gold Medal Award, which acknowledges his lifetime contributions to the field of ADHD. In 2023 he was awarded the Lifetime Achievement Award from the American Professional Society of ADHD and Related Disorder
Prof. Joseph Biederman’s Contributions to Child and Adolescent Psychiatry.
Why was Joe so productive and impactful? There are many reasons. He challenged the status quo whenever it did not fit with his clinical experience. His family-based follow-up studies dismantled the prevailing consensus that ADHD was only associated with conduct and oppositional defiant disorders. Instead, ADHD was associated with many disorders and that comorbidity was likely due, in part, to a shared genetics and underlying neurobiology. He raised these ideas in the 1990s, long before they were confirmed by contemporary genome wide association studies. His longitudinal studies of children and parallel studies of clinically referred adults debunked the belief that ADHD disappeared in adulthood, providing evidence of its continued clinical significance in many adults. He further challenged the status quo in his studies of bipolar disorder in youth. Before his work, the disorder was considered rare in youth. However, his research revealed the contrary and inspired a significant body of work that validated his findings about pediatric bipolar disorder, leading to new applications of existing treatments for these severely affected children.
Joe also realized that the problems of child psychiatry were too large for one person to address. So his productivity was multiplied by his mentorship and collegiality (Table 1), reflecting the sentiment of Socrates who said, “Education is the kindling of a flame, not the filling of a vessel.” The flames he kindled in his close colleagues at MGH and his more distant colleagues around the world will illuminate the dark areas of child psychiatry for decades to come.
Another reason for Joe’s productivity and impact was his insatiable curiosity and generative mind. As Albert Einstein famously said, “I have no special talents. I am only passionately curious.” This quote perfectly embodies Joe’s approach. If he found a weakness in his clinical practice or a phenomenon he did not understand, that weakness became a target for a research program. His ceaseless quest for knowledge propelled him to keep publishing papers even during his last days, a testament to his indomitable spirit. In fact, his colleagues are currently finalizing research papers that he had initiated but could not complete due to his untimely death. Throughout his illness, Joe remained an exemplary model for budding researchers, proving that the path of clinical research is one paved with resilience, innovation, and an unyielding commitment to improving lives.
Joe’s relentless pursuit of knowledge, his adaptability, and his unwavering commitment to improving lives created a legacy that will continue to shape the field of child psychiatry for generations to come. With this special issue of the Journal of Attention Disorders, we celebrate Joe’s legacy by reviewing his contributions to psychopharmacology, genetics and neurobiology, adult ADHD, executive function deficits, autism spectrum disorders, tic disorders, pediatric bipolar disorder, and pediatric depression. As you will see, Joe’s lifelong commitment to the field of child psychiatry has left a legacy that will continue to shape the field for generations to come.
Psychopharmacology (By Jeffrey H. Newcorn, MD)
Although the field of child psychopharmacology had several early champions—including prominent child psychiatrists such as Magda Campbell, John Werry, and Denny Cantwell—many would say that Joe is the father of modern child and adolescent psychopharmacology. His work in ADHD literally led the field for three decades. It is impossible to think of a medication that we use for ADHD today that was not studied or pioneered by Joe. Moreover, beyond that, he made major contributions to the pharmacotherapy of pediatric bipolar disorder, autistic spectrum disorder, and treatment of ADHD and comorbidity, in children and adults.
Joe’s early work in ADHD focused on the noradrenergic theory of the disorder. Building on studies of imipramine first done by Judy Rapoport in the 1970s (P. O. Quinn & Rapoport, 1975; Rapoport et al., 1974), he used the more noradrenergic desipramine (DMI) as the drug of choice. In a parallel group, double blind, placebo-controlled study of 62 children, many of whom had responded poorly to psychostimulant treatment, 68% of subjects were very much or much improved with DMI (Biederman et al., 1989). He further reported that the presence of comorbidity did not alter the response; nor was the medication preferentially better in youth with comorbidity (Biederman, Baldessarini et al., 1993). Joined now by Tom Spencer, who became a close, long-time colleague, he further studied the effects of DMI on youth with tic disorders (T. Spencer et al., 1993) reporting positive effects that were later confirmed in a more definitive double blind, placebo-controlled study (T. Spencer et al., 2002). Joe and Tom also found that DMI was associated with minimal growth suppression and could be considered an alternative for children who had stimulant-related growth suppression (T. Spencer et al., 1992). However, from early on, Joe advised caution in using this medication at doses above 3.5 mg/kg due to cardiovascular effects (Biederman et al., 1989). He considered using DMI blood levels to assess potential toxicity and reported, from a naturalistic study, that despite wide between-patient variability in serum DMI levels at the same dose, future within-subject DMI blood levels were highly predictable from knowledge of current levels, current dose, and the future dose (Biederman et al., 1997).
When concerns about cardiotoxicity of desipramine came to the fore, Joe and colleagues conducted population-based research on the topic (Biederman, Thisted, Greenhill, & Ryan, 1995). While they found only a slight, non-significant risk with treatment, they recommended extreme caution. Eventually, they decided that desipramine was not a viable option for large-scale use, and sought another medication to use in its place—one that would be especially useful for stimulant non-responders and individuals with ADHD + comorbidity, and with a better safety profile than DMI. Conversations with John Heiligenstein at Eli Lilly led them to a drug called tomoxetine (later re-named atomoxetine on guidance from FDA), which has a very similar chemical structure to fluoxetine but is highly noradrenergic and was all but ignored after failing in Phase 3 trials for depression. They conducted a Phase 1 trial in adults with ADHD and found substantially greater improvement in those treated with tomoxetine compared with placebo (T. Spencer et al., 1998). Based on these encouraging initial results, Eli Lilly launched a large-scale development program with atomoxetine, led by Joe and Tom, which resulted in FDA approval of this first nonstimulant in ADHD. Because this research (Michelson et al., 2001, 2002) found significant improvement in symptoms and functional status, for adults as well as children with ADHD, atomoxetine was the first medication of any kind to receive an FDA indication for treatment of ADHD in both adults and children (L. Adler et al., 2003; Faraone, Biederman, Spencer, Michelson, Adler, Reimherr, & Glatt, 2005; Faraone, Biederman, Spencer, Michelson, Adler, Reimherr, & Seidman, 2005; Michelson et al., 2003). Although atomoxetine was on average very effective, Joe’s team noted variability in response which was not explained by comorbidity (T. Spencer et al., 2006), and was likely accounted for by other factors. Better understanding the variable response profile of atomoxetine and other nonstimulant medications is central to developing a personalized treatment approach for ADHD—a very important topic in ADHD pharmacotherapy today. And, of course, this point was recognized very early on by Joe and his colleagues.
Atomoxetine was not the only nonstimulant medication that Joe had a major hand in developing. He was one of the leaders of the development program for guanfacine extended release (Biederman, Melmed, Patel, McBurnett, Donahue, & Lyne, 2008; Biederman, Melmed, Patel, McBurnett, Konow, Lyne, & Scherer, 2008; Sallee et al., 2007, 2009). He recognized that this medication could be an important alternative to stimulants in the care of youth with comorbid disruptive disorders and tic disorders, and especially those with autism spectrum disorder—given the high level of adverse effects with stimulants in the latter population. His seminal research with nonstimulants extended well beyond ADHD. Studies of second generation neuroleptics and mood stabilizers for disruptive behavior disorders (Biederman, Mick, Faraone, et al., 2006) and bipolar disorder (Biederman, Mick, Hammerness, Harpold, et al., 2005; Biederman, Mick, Wozniak, Aleardi, Spencer, & Faraone, 2005; Biederman, McDonnell, Wozniak, Spencer, et al., 2005; Biederman, Mick, Spencer, Doyle, Joshi, et al., 2007; Biederman, Hammerness, Doyle, Joshi, Aleardi, & Mick, 2008; Biederman, Mick, Faraone, et al., 2006) also led the field.
Joe’s work in ADHD psychopharmacology was by no means confined to the nonstimulants. He was a major force in leading the development of numerous methylphenidate (Biederman, Quinn, Weiss, et al., 2003; Biederman, Mick, Surman, Doyle, Hammerness, Harpold, Dunkel, et al., 2006; Biederman, Mick, Spencer, Surman, & Faraone, 2012; Biederman, Mick et al., 2011) and amphetamine (Biederman, Krishnan, Zhang, McGough, & Findling, 2007; McCracken et al., 2003; T. Spencer et al., 1999) stimulant formulations, and together with Tom Spencer, extended their use to adults with ADHD (T. Spencer et al., 1995, 2001, 2005; Weisler et al., 2006). But his interest in these medications went far beyond symptomatic improvement.
Joe was keenly aware of the impact of ADHD across multiple domains of function, and was equally interested to assess the impact of stimulant treatment on functional outcomes (Biederman, DiSalvo, Fried, et al., 2019; Boland et al., 2020). He showed that ADHD led to lower levels of educational and occupational attainment (Biederman, Petty, Fried, Kaiser, Dolan, Schoenfeld, Doyle, Seidman, & Faraone, 2008), and to lower earning potential at all levels of social class (Biederman, Faraone, Spencer, Mick, Monuteaux, & Aleardi, 2006; Biederman, Petty, Fried, Doyle, Mick, Aleardi, Monuteaux, Seidman, Spencer, Faneuil, et al., 2008). He conducted important studies on the beneficial effects of stimulant treatment of ADHD on driving (Biederman, Fried, Hammerness, Surman, Mehler, Petty, Faraone, Miller, Bourgeois, Meller, Godfrey, Baer, & Reimer, 2012; Biederman, Fried, Hammerness, Surman, Mehler, Petty, Faraone, Miller, Bourgeois, Meller, Godfrey, & Reimer, 2012) and the less than satisfactory impact of medication on executive function (Biederman, Seidman, Petty, Fried, Doyle, Cohen, Kenealy, & Faraone, 2008; Biederman, Mick et al., 2011). He also was very keenly attuned to issues in the clinical context and how they impacted treatment. To this end, he utilized the large data base available to him from the Partners Healthcare Data Registry to examine adherence to stimulant treatment, finding a major drop-off in those to whom medication was prescribed (Biederman, Fried, DiSalvo, Storch, et al., 2019; Biederman, Fried et al., 2020). To remedy this situation, he developed an app to help patients remember to take their medication and illustrated the important impact this had on treatment outcome (Biederman, Fried, DiSalvo, Woodworth, et al., 2019; Fried et al., 2020). Joe was always attentive to the potential for adverse effects and contributed much to our understanding of safety in all his clinical trials publications and especially contributed greatly to our understanding of the effects of stimulants on height and weight (Biederman, Faraone, Monuteaux, et al., 2003; Biederman et al., 2003; Biederman, Spencer, Monuteaux, & Faraone, 2010; Biederman et al., 2023; Faraone, Biederman, Monuteaux, & Spencer, 2005; Faraone, Biederman et al., 2008; T. J. Spencer et al., 1996; T. Spencer et al., 2006).
Understanding the impact of treatment on psychiatric comorbidity let to further clinically relevant contributions. Not only did Joe disseminate the concept of comorbidity in ADHD (Biederman, Newcorn, & Sprich, 1991), he showed that treating ADHD decreased the risk for developing comorbid disorders (Biederman, Monuteaux, Spencer, Wilens, & Faraone, 2009). At a time when many worried that stimulant treatment might increase risk for substance abuse, Joe published findings suggesting that stimulant treatment decreased risk (Biederman, Wilens, Mick, et al., 1999). As with much of his clinical acumen and research, he was way ahead of the curve with this notion—which was controversial at the time but later substantiated in large-scale population studies (Chang et al., 2012; P. D. Quinn et al., 2017). In a similar vein, he and his colleagues showed that stimulant treatment protected against the development of cigarette smoking, as well as alcohol and other substance use disorders (Wilens, Adamson et al., 2008).
Joe not only studied whether medications were safe and effective, he also studied the mechanisms through which they exerted their effects. To this end, he and his group joined forces with several colleagues to establish a program using neuroimaging to examine to the biological basis of ADHD and its treatment. Much of this work examined aspects of ADHD pathophysiology, finding volumetric, and functional differences in frontostriatal (Makris et al., 2007; Seidman, Valera et al., 2006), cerebellar (Makris et al., 2008; Valera et al., 2005), and limbic (Bush et al., 1999; Makris et al., 2010) regions and associated circuitry in youth and adults with ADHD. Other seminal imaging studies co-led with Tom Spencer demonstrated the key mechanism of methylphenidate, its binding to the presynaptic dopamine transporter (T. J. Spencer et al., 2005, 2007) and how the time course of this activity could alter the duration and clinical effects of stimulant treatment (T. J. Spencer et al., 2006). Other studies illustrated the functional impact of this activity (Bush et al., 2008). They also showed that modafinil (Madras et al., 2006) shared this same property with methylphenidate.
This brief review illustrates the breadth and impact of Joe’s work in child and adolescent psychopharmacology, extending across the lifespan. The scope of his work and the vision which drove it are reflected in the numerous publications cited here, and the even larger number of publications that could not be included in this brief overview. Though we mourn his loss, we celebrate his accomplishments and take solace in the fact that several prominent colleagues he mentored and collaborated closely with will continue this work, which has greatly improved the lives of many youth, worldwide.
Comorbidity and Genetics (By Stephen Faraone, PhD)
Joe Biederman’s seminal contributions to understanding psychiatric comorbidity in ADHD stem from his family studies of children with and without ADHD which were being published contemporaneously with his heavily cited review article on the topic (Biederman, Newcorn, & Sprich, 1991). Although family and twin studies had already been done by some groups and longitudinal studies by other groups, his team was the first to combine them in one design and to explore the familial association of ADHD with its comorbid conditions. They also completed one of the few adoption studies of ADHD, which further documented the genetic component to the disorder (Sprich et al., 2000).
His preliminary DSM-III study used the family study method to assess familial and psychosocial risk factors for ADD among the first-degree relatives of clinically referred children and adolescents with ADD. Comparisons were made with psychiatric and normal controls (Biederman et al., 1990). The results documented a wide range of comorbidity of ADD with anxiety, mood, and disruptive behavior disorders. Joe also set out to assess the familial association of ADHD with other disorders in more detail. Because he was not expert in genetic epidemiology, he sought the collaboration of Ming Tsuang, a prominent psychiatric geneticist at Harvard. I was junior faculty in Ming’s group and was fortunate that Ming asked me to take the lead in collaborating with Joe. That began a collaboration that lasted over 30 years.
Joe’s DSM-III-based family studies showed that the relatives of ADD children had a significantly higher risk for ADD, antisocial, anxiety, and mood disorders than relatives of psychiatric and typically developing controls. These findings suggested that DSM-III defined ADD might share genetic risk factors linked to these other mental health disorders. Innovative publication from his family study tested patterns of familial association between ADD and mood disorders: major depression, bipolar disorder, and dysthymia (Biederman, Faraone, Keenan, Steingard, & Tsuang, 1991; Biederman, Faraone, Keenan, & Tsuang, 1991).
Joe’s team found that relatives of both ADD proband subgroups (ADD + mood and ADD only) were at a significantly higher risk for both ADD and mood disorder than relatives of normal controls. They found no evidence for the cosegregation of ADD and mood disorder within families or for nonrandom mating, suggesting that ADD and mood disorders may share common familial vulnerabilities. Using a similar analytic design, Biederman, Faraone, Keenan, Steingard, & Tsuang (1991) evaluated the familial association between ADD and anxiety disorders. The results paralleled what was found for depression, showing that ADD and anxiety disorders segregated independently in families, suggesting that having a family history of ADD could increase the risk of developing anxiety disorders.
Continuing with this theme, his team focused on the familial association between ADD and disruptive behavior disorders (Faraone et al., 1991b). They found that the morbidity risk for ADD was highest among relatives of ADD probands who also had childhood conduct disorder or oppositional defiant disorder. Also, the risk for any antisocial disorder was highest among relatives of ADD probands who also had antisocial disorders. These results suggest that ADD with and without disruptive disorders may be distinct disorders but could also fall along a continuum of increasing familial etiological factors and severity of illness.
These family study data about ADD were based on families were ascertained via boys with ADD. In that era, only studying boys with ADD was common practice as girls with ADD were much rarer in clinical practice. In fact, Biederman’s group had a difficult time getting NIH funding for a study of girls due to a reviewer claiming that ADD was an essentially male disorder. Biederman’s group bucked that trend by completing the first family genetic study of girls with ADD (Faraone et al., 1991a). They found that, as they had reported for boys with ADD, the relatives of girls with ADD had higher risks for ADD, antisocial disorders, major depression, and anxiety disorders. The increased risk for ADD couldn’t be attributed to gender, generation, age, social class, or family intactness. These results suggest that ADD in girls is valid and shares a common biological substrate with boys.
To build upon these preliminary data, Joe completed two NIMH funded studies using the then new definition of ADHD in the DSM-IV. This comprised longitudinal studies of boys and girls with DSM-IV defined ADHD, normal controls, and their first-degree relatives (Biederman et al., 1992). These studies have sometimes been misunderstood in by others as only being relevant to ADHD patients referred to a tertiary care psychopharmacology clinic. That view was wrong for two reasons. First, the subset that came from that clinic included many children with new onsets of ADHD, not tertiary referrals. Second, half of the sample were recruited from pediatric clinics. Joe’s team found no difference between the pediatric and psychiatric referrals in age at onset of ADHD symptoms or the prevalence of mood, disruptive behavior, anxiety, and substance use disorders in the patients or their families. Significant impairments of intellectual, academic, interpersonal, and family functioning did not differ between the pediatric and psychiatric referral sources.
In Joe’s longitudinal studies, all first-degree family members of these ADHD and comparison youth were assessed comprehensively and the patients were followed into young adulthood. They found that ADHD probands were more likely to have conduct, mood, and anxiety disorders compared to controls. Relatives of ADHD probands had a higher risk for ADHD, antisocial disorders, major depressive disorder, substance dependence, and anxiety disorders (Biederman et al., 1992). They suggested that ADHD and major depressive disorders might share familial vulnerabilities, and that ADHD plus conduct disorder could be a distinct subtype.
Although Joe recognized the importance of genetic risk factors for ADHD, he also sought to document environmental risks (Biederman, Milberger, Faraone, Kiely, Guite, & Mick, 1995). His group found a positive association between Rutter’s adversity indicators and the risk for ADHD and its associated psychiatric, cognitive, and psychosocial impairments. Their work supported the significance of adverse family-environment variables as risk factors for both ADHD and the comorbidities and impairments associated with ADHD.
In the same year, the team showed that psychiatric comorbidity, familiality, and learning disabilities influenced the severity and pattern of neuropsychological function in ADHD. Their results showed that neuropsychological performance in ADHD was especially affected by familial status and the presence of learning disabilities. Their work also suggested that alterations of cerebral dominance and frontal networks might be involved in ADHD (Seidman, Benedict et al., 1995; Seidman, Biederman et al., 1995).
Joe’s group devoted much time to investigating the comorbidity between ADHD and major depression with the goal of ruling out confounds that might account for the comorbidity (Biederman, Faraone, Mick, Moore, & Lelon, 1996; Biederman, Ball, Monuteaux, Mick, Spencer, & McCreary et al., 2008; Braaten et al., 2001). He further validated the diagnoses of depression in his ADHD sample by showing they could not be accounted for by demoralization secondary to having ADHD (Biederman, Mick, & Faraone, 1998). When looking at the families of ADHD and control youth, they found a familial link between ADHD and depression, with ADHD families with antisocial disorders showing the greatest risk for depression. In the absence of antisocial disorders, ADHD also imparted a familial risk for depression (Faraone & Biederman, 1997, 1998). His group also showed that the comorbidity and familial association between ADHD and depression could not be accounted for by symptoms shared by the disorders (Milberger et al., 1995).
Because Joe’s writings about bipolar disorder in youth with ADHD had generated much controversy, he decided to use family-study data to test his ideas (Biederman, Faraone, Wozniak, & Monuteaux, 2000; Biederman, Faraone, Petty, et al., 2013). He reasoned that if the bipolar disorder he had diagnosed in youth was valid, it ought to run in families. This work found that ADHD and bipolar disorder ran together in families. Of note, the parents of youth with ADHD were at increased risk for bipolar disorder. That was a key validation finding because although methods of diagnosis for bipolar disorder in youth were controversial, the diagnosis In adults was not. Joe’s findings about the familial transmission of bipolar disorder were further validated by a meta-analysis by his team on that topic (Faraone et al., 2012). He also showed that the comorbidity and familial association between ADHD and bipolar disorder could not be accounted for by symptoms shared by the disorders (Milberger et al., 1995).
Realizing that his sample was primarily Caucasian, Joe sought to determine if his results generalized to the African-American population. He first reviewed the extant literature (V. J. Samuel et al., 1997) and used that information to design a culturally sensitive study. He first documented high rates of psychiatric comorbidity among African-American patients with ADHD (V. J. Samuel et al., 1998). He also found significantly higher risks for ADHD and other disorders like oppositional defiant disorder, antisocial personality disorder, major depression, generalized anxiety, and substance use disorders among the relatives of ADHD probands (V. Samuel et al., 1999).
His DSM-IV longitudinal family study of girls documented that girls with ADHD had prototypical core symptoms of the disorder, high levels of comorbid psychopathology, and dysfunction in multiple domains (Biederman, Faraone, Mick, Williamson, et al., 1999). Moreover, the transmission of ADHD and comorbid disorders generalized to families of girls with ADHD. Neither proband gender nor subtype influenced the familial transmission of ADHD (Faraone, Biederman, Mick, et al., 2000). The co-occurrence of ADHD and comorbid psychopathology in families was similar for families ascertained through boy and girl probands (Faraone, Biederman, Mick, Doyle, et al., 2001). Their work suggested more similarities than differences between the sexes in symptoms of ADHD, the nature of psychiatric comorbidity, neuropsychology, social impairments and the familial transmission of ADHD and comorbid disorders (Doyle et al., 2001; Doyle, Biederman et al., 2005; Faraone, Biederman, Mick, Doyle, et al., 2001; Greene et al., 2001; Seidman et al., 1997; Seidman, Biederman et al., 2006). They also found that patterns of familial co-transmission of ADHD, bipolar disorder and conduct disorder were similar for boys and girls (Faraone, Biederman, & Monuteaux, 2000, 2001). Psychosocial adversity and low social class, maternal psychopathology, and family conflict in particular increased the risk for ADHD and associated morbidity independently of gender and other risk factors, but gender modified the risk for adverse cognitive and interpersonal outcomes (Biederman et al., 2002).
With the dawn of molecular genetics, Joe was a guiding light as I secured NIH grants to conduct candidate gene studies (Lasky-Su et al., 2006, 2008; Lasky-Su, Banaschewski et al., 2007; Lasky-Su, Faraone et al., 2007; Lasky-Su et al., 2007), meta analyses of these studies (Faraone, Doyle, Mick, & Biederman, 2001; Faraone et al., 2014), and linkage studies (Doyle et al., 2008, 2010; Faraone, Doyle et al., 2008). These approaches were not fruitful so the team abandoned them for the method of genomewide association (Doyle et al., 2008; Lasky-Su et al., 2010; Mick et al., 2010, 2011; Neale et al., 2010) which eventually led, via a huge international consortium to the discovery of the first genomewide significant loci for ADHD (Demontis et al., 2019). I was continually amazed by Joe’s ability to learn about new technologies, find the right team to execute a project and inject his clinical wisdom into the research protoco.
In summary, Joe’s family-genetic studies of ADHD taught us much about psychiatric comorbidity and the familial transmission of ADHD and comorbid disorders. It is a testament to the quality of his work that his main findings about the comorbidity of ADHD with depression and bipolar disorder, albeit once controversial, have been replicated many times by other groups. Strikingly, the genetic inferences he drew from this work in the 1990s have been confirmed by genomewide association studies two decades later (Demontis et al., 2019, 2021, 2023; van Hulzen et al., 2017).
Pediatric Bipolar Disorder (By Janet Wozniak, MD)
In 1992, Joe noted a curious finding in his comparison sample of outpatient child psychiatry prepubertal patients coming for clinical care: a significant minority of parents were endorsing the symptoms on the mania module, suggesting a Bipolar I diagnosis. He said, “Either these are erroneous endorsements of the symptoms, and we need to re-train our trained raters on this module, or we have an important finding to report.” Fresh out of training, I was already biased against the notion that young children could have mania, so I assumed the former, but, in 1995, after a close examination which dispelled all our skepticism, we published a manuscript describing in detail 43 youngsters, each experiencing a full array of manic symptoms (Wozniak, Biederman, Kiely, et al., 1995). This publication spurred an emerging paradigm shift in the diagnosis of mood disordered children and launched my deeply gratifying 30+ year clinical and research career with Joe.
Because the idea that bipolar disorder could onset in the earliest years of life was novel, we proceeded in logical research steps, building a converging evidence base supporting the validity of pediatric bipolar disorder (Biederman, Kwon, Wozniak, Mick, et al., 2004; Biederman, Petty, Wilens, Spencer, et al., 2009; Biederman, Petty et al., 2011; Doyle, Wilens et al., 2005; Henin et al., 2007, 2009; Wozniak et al., 2003). Because most of these youngsters also had a full array of severely impairing ADHD symptoms, we examined this comorbidity first, asking whether the children had ADHD (without mania at all), mania (without ADHD at all), or both conditions. Our clinical research suggested that both conditions were validly diagnosed in these highly impaired children and adolescents (Biederman, Faraone, Mick, Wozniak, et al., 1996; Faraone, Biederman, Wozniak, et al., 1997; Wilens, Biederman, Wozniak, et al., 2003). Many investigations were descriptive (Biederman, Faraone, Wozniak, et al., 2005), comparing children with bipolar disorder + ADHD with children with uncomplicated ADHD (Biederman, Russell, Soriano, et al., 1998). Compared with those with ADHD only, youth with both bipolar disorder and ADHD had higher rates of comorbidity (including severely impairing depression, psychosis, conduct disorder, and anxiety disorders) and significantly worse functioning including hospitalization, even by the age of 8 years. The story of pediatric bipolar disorder unexpectedly became a story of the role of irritability as a symptom. Our manuscripts relay the qualitatively and quantitatively extreme form of irritability in pediatric mania (Mick et al., 2005; Serra et al., 2017; Wozniak et al., 2005; Wozniak & Biederman, 1997). This manic-level irritability is distinct from the irritability noted in youth with ADHD and other disorders and compellingly described by parents as the most impairing part of the clinical picture. Clinicians and parents of these severely emotionally dysregulated children welcomed the studies, finding hope in a diagnosis with attendant treatments for a severely impairing clinical picture.
Our family-genetic studies found that children with bipolar disorder versus ADHD had a higher rate of bipolar disorder in first-degree relatives, including adult relatives for whom the diagnosis of bipolar disorder was not controversial (Biederman, Faraone, Petty, et al., 2013; Biederman, Chan et al., 2018; Faraone et al., 2012; Wozniak, Biederman, Mundy, et al., 1995; Wozniak et al., 2010; Wozniak, Faraone, Martelon, et al., 2012; Wozniak et al., 2013). Family studies are valuable “external validators” of illness, separate from the diagnosis made in the child, and provide validating evidence that these children have a bona fide form of bipolar disorder. Not only are these studies of scientific interest, but they provided important data in the heated debate as to whether children could “really” suffer from mania.
In a series of manuscripts, we systematically explored all the comorbidities present in youth diagnosed with bipolar disorder (Wilens, Biederman, Forkner, et al., 2003): conduct disorder (Biederman, Mick, Prince, et al., 1999; Biederman, Faraone, Wozniak, & Monuteaux, 2000; Biederman, Mick, Faraone, & Wozniak, 2004; Biederman, Fitzgerald et al., 2018; J. Biederman, Mick, Wozniak et al., 2003; Biederman, Mick, Faraone, et al., 2006; T. J. Spencer et al., 2001; Wozniak et al., 2001, 2019), substance use disorders (T. Spencer et al., 1999; Wilens et al., 2004, 2016), anxiety disorders (Biederman, Wozniak, Martelon, et al., 2013; Harpold et al., 2005; Joshi, Mick, Wozniak, Geller, et al., 2010; Joshi, Wozniak, Petty, Vivas, et al., 2010; Wozniak et al., 2002), and autism spectrum disorder (Joshi, Biederman, Petty, Goldin, et al., 2013; Joshi et al., 2018; Joshi, Biederman et al., 2012; Wozniak et al., 1997). We explored the differences between unipolar and bipolar depression (Biederman et al., 2014; Wozniak et al., 2004). Comorbidity with mania emerged as a meaningful clinical and scientific marker as we identified clinically relevant subtypes of these co-occurring disorders. Joe ardently maintained that the bidirectional overlap of conduct disorder and mania provided the important possibility to divert some conduct disordered youth from a criminal course by quelling the impulsivity, aggression, and poor judgment of comorbid mania.
An important external validator emerged in the Child Behavior Checklist (CBCL). As a parent completed checklist of symptoms, the CBCL is free from clinician biases. Subscales on the CBCL (Attention, Aggression and Anxiety/Depression) repeatedly demonstrated association with pediatric bipolar disorder, suggesting that these subscales were a useful screening instrument for this severely impairing disorder (Biederman, Wozniak, Kiely, Ablon, Faraone, Mick, et al., 1995; Biederman, Petty, Monuteaux, Evans, et al., 2009; Biederman, Petty, Day, Goldin, Spencer, Faraone, et al., 2012; Biederman, Martelon, Faraone, et al., 2013; Biederman et al., 2022; Uchida et al., 2014; Yule et al., 2019). Treatment (Biederman, Mick, Bostic, Prince, et al., 1998; Biederman, Mick, Prince, et al., 1999; Liu et al., 2011; Potter et al., 2009).
Our group performed a series of clinical trials for the treatment of pediatric bipolar disorder, examining the effectiveness of second-generation antipsychotics (olanzapine, risperidone, aripiprazole, quetiapine, ziprasidone, and paliperidone; Biederman, McDonnell, Wozniak, Spencer, et al., 2005; Biederman, Mick, Hammerness, Harpold, et al., 2005; Biederman, Mick, Wozniak, Aleardi, Spencer, & Faraone, 2005; Biederman, Mick, Faraone, et al., 2006; Biederman, Mick, Spencer, Dougherty, Aleardi, & Wozniak, 2007; Biederman, Mick, Spencer, Doyle, Joshi, et al., 2007; Frazier et al., 1999; Joshi, Petty, Wozniak, Faraone, et al., 2013; Joshi, Petty et al., 2012; Tohen et al., 2007; Wozniak et al., 2009) as well as anticonvulsants (valproic acid, carbamazepine, and lamotrigine; Biederman, Joshi, Mick, Doyle, et al., 2010; Joshi, Wozniak, Mick, Doyle, et al., 2010; Wozniak, Faraone, Martelon, et al., 2012; Wozniak, Hammerness et al., 2012) and complementary and alternative treatments (Wozniak et al., 2007, 2015; Wozniak, DiSalvo, Farrell, Vaudreuil, Uchida, Ceranoglu, et al., 2022; Wozniak, Farrell, DiSalvo, et al., 2022). These open label trials, some of which included pre-school aged children, provided an early evidence base for addressing symptoms of mania in these severely impaired children. Many of the medications which demonstrated effectiveness in open label trials eventually gained FDA approval for the treatment of pediatric bipolar disorder and Joe’s team showed that the results of the open-label studies were useful predictors of the safety and efficacy later reported in placebo controlled trials (Biederman, Petty, Woodworth, Lomedico, O’Connor, Wozniak, & Faraone, 2012).
Longitudinal studies are difficult to do, but because “what happens in the future” remains a key question asked by parents of the children we treat, we tackled this important clinical question as well (Biederman, Mick, Faraone, Van Patten, Burback, & Wozniak, 2004; Biederman, Petty, Byrne, Wong, et al., 2009; Wozniak et al., 2011, 2018; Wozniak, DiSalvo, Farrell, Joshi, Uchida, Faraone, et al., 2022). Our results, which aligned with the few other research groups studying this topic, demonstrated a waxing/waning but persistent and chronic course in most of the children and adolescents studied. Examining functioning over time, in addition to presence of symptoms or subthreshold syndromes, was key to our longitudinal studies as an important indicator of persistence. In fact, subthreshold bipolar is highly impairing in its own right, and studies of children who do not meet the full DSM criteria for mania was the next logical step in the Biederman panoply of pediatric bipolar disorder studies (Vaudreuil et al., 2019; Wozniak et al., 2017).
After years of clinical research, the idea that the age of onset of bipolar disorder cleaves a distinct pathophysiologic subtype of bipolar disorder has become apparent (Biederman, Mick, Faraone, Spencer, et al., 2000; Faraone, Biederman, Mennin, et al., 1997; Mick et al., 2003), and studies of adults with early versus late onset bipolar disorder as well as studies of neuroimaging and genetic biomarkers provided long-awaited biomarker validation of the pediatric onset form of bipolar disorder (Martelon et al., 2012; Mick et al., 2008, 2009; Moore et al., 2006, 2007; Wozniak, Gönenç, Biederman, et al., 2012).
Joe’s clinical insight that children could have mania, and his perseverance in the study of pediatric onset bipolar disorder despite opposition and doubt, sparked a new research field and a new standard in the diagnosis of bipolar disorder in children. This new standard of diagnosis and treatment has improved the lives of the youth and families whose lives had been upended by this severe disorder.
Autism Spectrum Disorders (By Gagan Joshi, MD)
Although Joe was most prolific in the area of ADHD, the field of autism also benefited from his scientific approach. Joe recognized the presence and impact of autistic traits in patients not diagnosed with an autism spectrum disorder (ASD). His inclusion of an autism assessment module in the diagnostic interview schedule filled a crucial gap in the battery of assessments for psychopathologies (Joshi et al., 2011). Crucially, he dropped the then DSM requirement that ADHD and autism could not be diagnosed in the same person. Consequently, his work enabled him to document the prevalence of autism spectrum disorders in youth with ADHD and the prevalence of ADHD and other disorders in large cohorts of youth and adults with ASDs (Joshi, Petty, Wozniak, Henin, et al., 2010; Joshi, Biederman, Petty, Goldin, et al., 2013; Joshi et al., 2014).
Joe and his team challenged the prevailing notion that psychopathologies in autism were merely consequences rather than co-occurring disorders. To investigate the relationship between comorbid disorders he investigated their prevalence rates and clinical correlates in the context of comorbidity. They reported, among some patients with ASDs, the typical clinical presentation and correlates of ADHD and bipolar disorder. Contrary to the prevailing notion, had higher rates of non-ASD psychopathology than expected (Joshi, Biederman, Petty, Goldin, et al., 2013; Joshi, Faraone et al., 2017). Additionally, Joe’s team documented the presence of an ADHD-associated neural profile in individuals with autism and ADHD, dispelling reservations about the typical expression of ADHD in intellectually capable populations with ASD. Furthermore, his work established the importance of identifying psychiatric comorbidity in autism by documenting expected responses to anti-ADHD and anti-manic treatments in autism populations with those comorbidities (Joshi, DiSalvo, Wozniak, et al., 2020; Joshi, Biederman et al., 2012).
Joe was a strong advocate for early diagnosis. He taught us that proper treatment and early intervention hinge upon timely identification. He dedicated himself to investigating ways to facilitate early screening and rapid diagnosis of autism in psychiatrically referred populations. Through his research he identified a symptom profile in the Child Behavior Checklist (CBCL) that strongly indicated a child’s struggles with autism, providing clinicians with a valuable screening tool for both psychopathologies and autism (Biederman, Petty, Fried, Wozniak, et al., 2010). Furthermore, he identified a profile of neuropsychological impairments associated with autism in commonly used neuropsychological tests, which has the potential to serve as an additional screener for autism (Fried et al., 2016).
As a clinician-researcher, Joe understood the challenges faced by clinicians and sought answers through extensive data analysis. He directed efforts to study the impact of significant autistic traits in individuals who did not meet the diagnostic threshold for autism according to the prevailing DSM criteria. Together, his team documented the associated morbidity and functional impairments related to these traits, emphasizing the need for early intervention rather than expecting children to outgrow them (Biederman et al., 2015; Kotte et al., 2013; Uchida et al., 2013). A subsequent publication on the persistence of these disabling traits as children transitioned into adulthood, further underscored the importance of early intervention (Joshi, DiSalvo, Faraone, et al., 2020).
Joe’s multidimensional approach to understanding ASDs extended to using cutting-edge neuroimaging techniques. Through extensive neuroimaging projects, his team identified abnormalities in neurochemical and neural connectivity associated with autism, contributing to a better understanding of the condition (Anteraper et al., 2019, 2020; Arnold Anteraper et al., 2019; Joshi, Biederman, Wozniak, Goldin, et al., 2013; Joshi, Arnold Anteraper et al., 2017). Notably, the magnetic resonance spectroscopy (MRS) investigation that shed light on abnormal brain glutamate activity in intellectually capable individuals with autism also offered a potential target for treatment interventions.
Recognized as an early adopter of technology in psychiatric research, Joe collaborated with the Massachusetts Institute of Technology (MIT) to develop laboratory driving and workplace simulation environments. These simulations allowed for the investigation of driving and workplace deficits in adults with autism. His studies revealed subtle but significantly impairing correlates of autism that adversely affected driving and workplace functioning (Fried et al., 2013; Reimer et al., 2013).
Tourettes and Tic Disorders (By Barbara Coffey, MD)
The literature on persistent tics and Tourette’s disorder in the early 20th century was dominated by the psychoanalytic concept that tics were a manifestation of conflict, or worse, a sign of a neurodegenerative disorder. Treatment was psychodynamic, lengthy, and often not successful. It was not until the 1970s with the work of Drs. Arthur and Elaine Shapiro, Jerry Young, and Todd Feinberg, that the scientific era began. The Drs. Shapiro published the first clinical trial in Tourette syndrome in 1968 using haloperidol, introducing the novel concept that this was a neurobiological disorder (Shapiro & Shapiro, 1968). This work greatly influenced Joe’s transition from psychoanalysis to psychopharmacology.
Scientific work on Tourette’s and the comorbid psychiatric disorders has exploded in the past two or three decades. With advances in neuroimaging, and enhancement in understanding of genomics and clinical phenomenology, there is a much better appreciation for the biopsychosocial nature of the disorder. Investigation and treatment of persistent tics and Tourette’s disorder has brought neurologists and psychiatrists together; neurologists have tended to focus on the tics, and psychiatrists/psychologists on the many comorbid psychiatric disorders and their impact on patients’ families, educational, and occupational lives.
Stress, excitement, and anxiety have long been described as major precipitants of tic exacerbations in the typically fluctuating course of Tourette’s disorder. But careful dissection of stress and anxiety (as either trait or state) had not taken place despite enhanced understanding of the clinical phenomenology of Tourette’s disorder in the scientific era. Fortunately, Joe’s pediatric psychopharmacology unit at MGH was a setting in which each referred child or adolescent was evaluated comprehensively with a semi-structured diagnostic interview; thus, a detailed, nuanced clinical picture was painted for each new patient that addressed all disorders and problems, not only the reason for referral. This approach allowed Joe’s team to lead the field in demonstrating the extent and impact of the psychiatric co-occurring conditions in Tourette’s disorder, previously conceptualized as a “neurological disorder” (B. J. Coffey et al., 2000).
Anxiety and mood disorders began to stand out as highly prevalent, even when OCD or ADHD were not present. We found that whether a child was initially referred through a specialized Tourette’s clinic or through a general child and adolescent psychopharmacology program, mood and anxiety disorders were pervasive and often associated with significant overall functional impairment. In addition, we established that major depression and bipolar disorder were better predictors of overall illness impairment, manifest by hospitalization or impaired Global Assessment of Functioning scores, than tic severity in children and adolescents with Tourette’s disorder (B. J. Coffey et al., 2000).
Joe’s team was the first to comprehensively characterize the role and impact of anxiety disorders in youth with Tourette’s, independent of OCD or other psychiatric disorders (B. J. Coffey et al., 2000). Importantly, our finding that separation anxiety disorder predicted high tic severity among our Tourette’s patients was recently replicated as a sub-type of anxiety disorder unique to youth with Tourette’s disorder (Vermilion et al., 2021). We confirmed in a prospective cohort design that while tics tend to persist through childhood into later adolescence, functional impairment generally remits, information that is usually a relief for parents with a newly diagnosed child with Tourette’s disorder.
In another notable study of the course of tic disorders in the context of a diagnosis of ADHD, we showed that while ADHD tends to persist, tics tend to remit over time, and the courses are essentially independent of one another (T. Spencer et al., 1998). This finding has been highly important to parents, who are often most concerned about their child’s tics, which in the long run are usually less problematic than the child’s symptoms and impairments due to ADHD.
In summary, this work from Joe’s team set the foundation for many and varied research spin offs. Among those that followed were two longitudinal, prospective studies of Pediatric Autoimmune Neuropsychiatric Disorders Associated with Streptococcus (PANDAS), which found no evidence that children with PANDAS were more likely to have strep or to experience tic exacerbations precipitated by streptococcal infection than children without PANDAS (Kurlan et al., 2008; Leckman et al., 2011). Another iteration was exploration of the role of inflammation in Tourette’s disorder, showing that youth with tics were more likely than healthy controls to have elevations of inflammatory cytokines (B. Coffey et al., 2015). And another line of work which followed Joe’s extensive clinical trial experience and mentorship focused on controlled trials of novel agents in treatment of Tourette’s disorder, ranging from omega-3 fish oil to pramipexole to most recently, deutetrabenazine (B. Coffey et al., 2021). Joe’s legacy as a brilliant clinician, investigator, and mentor lives on in this work.
Clinical and Neural Biomarkers for Pediatric Mood Disorders (By Mai Uchida, MD)
One of Joe’s major clinical concerns was the fact that most youth with psychopathology started treatment many years after the onset of illness. Thus, he sought methods to improve the early identification of psychopathology risk, especially for mood disorders. Studies indicated that the prevalence of major depression, dysthymia, and bipolar spectrum disorders steadily increase in childhood and peak in adolescence (Biederman, Martelon, Faraone, et al., 2013; Uchida et al., 2014; Uchida, Fitzgerald et al., 2018). It had been estimated that, by adolescence, up to 25% of youth in the U.S. will experience a major depressive disorder (Yule et al., 2019) and up to 4% will develop bipolar disorder (Uchida, Fitzgerald et al., 2018). These studies also indicated that pediatric-onset mood disorders are associated with very high levels of distress and disability. Joe taught us that early identification of the risk for mood disorders in children can lead to the clinician becoming more vigilant in monitoring for specific signs of distress and have important implications for the initiation of treatment and the reduction of long-term morbidity. Here, we discuss our studies on the clinical and neurobiological correlates to the development of pediatric mood disorders.
Joe’s team first focused on clinical risk factors using the Childhood Behavior Checklist (CBCL). Their efforts led to two major findings: first, subsyndromal scores on the CBCL Anxiety/Depression subscale (defined as scores between 0.5 and 2 SDs above the mean, or 55 and 70) were a risk factor for the development of depression; second, the combined positive scores (>2 SD above the mean per scale, or >210) for three subscales (Anxiety/Depression, Aggressive Behavior, and Attention Problems) was a risk factor for development of bipolar 1 disorder (Biederman et al., 2022; Uchida et al., 2021).
In a prospective study of children, participants were stratified by risk according to the presence or absence of familial mood disorders and the presence or absence of subsyndromal scores on the CBCL Anxiety/Depression subscale. Participants were followed over 10 years for the development of mood disorders (MDD or BP-1 disorder). Children in the high-risk group (positive for both familial histories of mood disorders and subsyndromal scores on the CBCL Anxiety/Depression subscale) had significantly higher rates of mood and anxiety disorders than other groups (Uchida, Fitzgerald et al., 2018). Children with either a positive familial history or subsyndromal CBCL Anxiety/Depression scores had an intermediate risk for developing mood and anxiety disorders, and children in the low-risk group (neither family history nor subsyndromal scores) had the lowest risk. When analyzed for the development of BP-1 disorder, there were no significant differences between groups. These findings suggested that a CBCL Anxiety/Depression subsyndromal score would be a useful clinical predictor for unipolar depression but not bipolar disorder.
Identifying risk factors for bipolar disorder is of particular clinical importance in children. MDD and bipolar disorder can both present initially with depressive symptoms, making differentiation between the two diagnoses one of the most difficult challenges for clinicians. Manic switches from an initial presentation of depression are more common in children than adults and many first-line treatments for depression can worsen the course of bipolar disorder. Thus, Joe emphasized the importance of identifying risk factors for developing bipolar disorder separate from major depression. One such risk factor that had been studied is a positive score on three subscales of the CBCL: Anxiety/Depression, Attention Problems, and Aggressive Behavior. Joe’s group referred to this score as the severe dysregulation profile. One study that compared children with BP-1 disorder and a positive severe dysregulation profile, children with BPD-1 without a positive severe dysregulation profile, and matched controls found that the severe dysregulation profile identified a severe subgroup of youth with BP-1 disorder. This group had significantly different patterns of psychiatric comorbidities, psychosocial dysfunction, cognitive deficits, and risk for a first-degree relative with BP-1 disorder (Biederman, Martelon, Faraone, et al., 2013). When evaluated as a screening tool, a conditional probability analysis found that the CBCL severe dysregulation profile was significantly more likely to be positive in children with BP-1 disorder than those with ADHD or controls. The positive predictive values of this score were 99% and 92% when compared with controls or the ADHD group, respectively (Uchida et al., 2014). These findings were replicated in another study, which found that a score on the CBCL severe dysregulation profile of >195 correctly identified 86% of participants with BP-1 disorder, with 80% sensitivity, 87% specificity, 61% positive predictive value, and 95% negative predictive value (Yule et al., 2019).
Building on the clinical correlates of pediatric mood disorders, Joe’s team sought to identify neurobiological correlates through imaging studies. Initial work focused on correlating scores on the CBCL Anxiety/Depression subscale and emotional dysregulation profile with white matter microstructure using results from diffusion tensor imaging (DTI) studies. Analyses found a negative correlation between CBCL Anxiety/Depression scores and fractional anisotropy (FA) in the anterior cingulum-colossal (CC-CG) region, with stronger negative association localized to the right anterior region (aCC + aCG; Hung et al., 2020). Additionally, a negative correlation was identified between CBCL emotional dysregulation scores and FA in both the anterior and posterior subdivisions of the CC-CG bundles. Furthermore, a positive correlation was identified between CBCL emotional dysregulation scores and the radial diffusivity (RD) located in both the anterior and posterior subdivisions of the CC-CG bundles (Uchida et al., 2021). Importantly, negative correlations were observed between FA in the posterior cingulum and posterior cerebral cortex and CBCL emotional dysregulation scores but not between these regions and CBCL anxiety/depression scores. Taken together, these studies suggest two distinct models of DTI structural connectivity that correlate with unipolar versus bipolar depression development in children. First, the stronger association between decreased structural connectivity in the anterior cingulate region and CBCL anxiety/depression scores suggest that this could be a neural marker for the risk for developing unipolar depression. Second, the negative correlation between FA in both the posterior cingulum and posterior cerebral cortex and CBCL emotional dysregulation scores suggest this could represent a neural marker for the risk for developing bipolar disorder.
Joe’s team’s imaging studies also sought to identify biomarkers correlated to familial risk for developing depression. Two studies were conducted that compared fMRI data between unaffected children of parents with a history of depression (at-risk group) and controls with no parental history of depression. They found multiple atypical areas of resting state functional connectivity in the at-risk group, specifically: hypoconnectivity within the cognitive control network, hypoconnectivity between the dorsolateral prefrontal cortex and subgenual anterior cingulate cortex, and hyperconnectivity between the right amygdala and right inferior frontal gyrus (a key area for top-down emotion modulation; Chai et al., 2016). This data was used to classify at-risk and control cases with greater accuracy than clinical rating scales. In a similar study, fMRI data was obtained while at-risk and control groups were shown images of positive, neutral, and negative facial expressions. Compared with the control group, the at-risk group showed hyperactivation in the amygdala and multiple cortical regions in response to fearful expressions and hypoactivation in the anterior cingulate cortex and supramarginal gyrus in response to positive expressions (Chai et al., 2015). These findings are consistent with behavioral evidence that children at risk for developing depression direct more attention to negative stimuli than positive stimuli.
Additional fMRI studies evaluated intrinsic functional brain architecture that correlates with participants’ ability to regulate emotion. In one recent study, participants were asked to attend to neutral images, attend to negative images, and reappraise negative image such that they viewed them more positively. Results showed that reappraisal success correlated with greater activation in the medial and lateral prefrontal cortex, as well as lesser resting state functional connectivity between the right amygdala and medial prefrontal and posterior cingulate cortices (Uchida et al., 2015). Reappraisal success also correlated with less trait anxiety and more positive daily emotions. These findings suggest that poor reappraisal abilities are linked to specific neural markers which could be a risk factor for depression. In other words, these markers possibly identify individuals who are not able to find positive ways to view negative stimuli or move on from negative stimuli (those with a so-called “glass half-empty” perspective).
With the above-reviewed research, Joe’s team took key initial steps toward developing predictive models for pediatric mood disorders, but work remains in identifying, validating, optimizing, and disseminating such models. In recent years, the team began to use machine learning to improve upon statistical models. One study identified the first quantitative model for predicting the development of bipolar disorder in children. In that study, participants were assessed at baseline and followed over 10 years. The Balanced Random Forest algorithm was used to identify variables associated with the development of bipolar disorder. In the 10% of participants who developed bipolar disorder, multiple risk factors were identified, including: CBCL Externalizing and Internalizing Behaviors, CBCL total t-score, problematic school function (indexed through the CBCL School Competence scale), CBCL Anxiety/Depression scale, and CBCL Aggression scale. This model predicted subsequent bipolar disorder development with 75% sensitivity and 76% specificity (Uchida et al., 2022). Another study showed that polygenic risk scores for ADHD, MDD, BPD, disruptive behaviors disorders and aggression could modestly predict mood disorders in a clinically referred sample (Barnett et al., 2022). Models such as these, once validated in other samples, could offer great clinical utility and could even be incorporated into baseline assessments via integration into electronic medical record systems and public health records. More work remains in this area, but machine learning algorithms have exciting implications for early identification of pediatric mood disorders.
In summary, Joe’s team contributed much to the predictive modeling of pediatric mood disorders with new findings about clinical and neurobiological risk factors. Identified predictors include familial risk, clinical rating scales (specifically, CBCL anxiety/depression subsyndromal scores for depression and the CBCL severe dysregulation profile for bipolar disorder), and neural markers in white matter microstructure, functional resting state connectivity, and polygenic risk scores. As this area continues to expand into new fields such as machine learning, new predictive models should seek to build on existing findings and explore new modalities. This work could lead to earlier identification, improved treatment courses, and better outcomes for pediatric patients with mood disorders.
Executive Functioning and ADHD (By Timothy Wilens, MD)
Longstanding interest exists regarding the overlap and role of Executive Functioning (EF) in ADHD. While many definitions exist, EF is commonly described as a set of mental operations that include response inhibition, working memory, set shifting, and interference control that are utilized to plan and execute a goal (Willcutt et al., 2005). Clinicians assess deficits in EF through neuropsychological tests like the Digit Span Memory Test, the Stroop test, Wisconsin Card Sort, and the Go/No-Go task. Similarly, EF deficits can be identified through behavioral observation using clinical checklists like the Behavioral Inventory of Executive Functioning (BRIEF), Barkley Scale of EF, and the Conners 3-P assessment. Not only has EF been shown to be relatively stable over time—children with deficits in EF grow up to be adults with deficits in EF—but also, EF deficits have been associated with impaired school performance, behavioral issues, and deficits in social functioning (Abreu-Mendoza et al., 2018; Romero-López et al., 2020; Ros & Graziano, 2018). For over three decades, Joe and colleagues have been studying the overlap of EF in both youth and adults with ADHD (Biederman, Petty, Fried, Fontanella, Doyle, Seidman, & Faraone, 2006; Biederman et al., 2017).
Among the first to highlight the important overlap of ADHD and deficits of EF, sometimes referred to as “cognitive comorbidity” of ADHD, Joe and colleagues found that 33% of children between the ages of 6 and 17 years with ADHD manifest neuropsychologically defined deficits in EF (Biederman, Monuteaux, Doyle, et al., 2004). These deficits, determined using a binary impairment indicator, were operationalized as a score greater than 1.5 standard deviations from the mean of controls’ scores on two or more of the eight neuropsychological measures examined (Biederman, Petty, Fried, Doyle, Spencer, et al., 2007). In these youth, they identified higher-than-expected rates of grade repetition, learning disabilities, and lower IQ score in comparison to youth with ADHD and no apparent EF deficits (J. Biederman et al., 2004). Of note, no gender differences were found regarding the frequency of EF deficits in these youth. Interestingly, when following a cohort of male youth over 7 years, the Biederman team found that 69% of individuals with neuropsychologically-defined EF deficits at baseline continued to experience said deficits at follow-up, while an additional 25% of individuals without EF deficits at baseline later developed EF deficits (Biederman, Petty, Fried, Doyle, Spencer, et al., 2007). There was specificity in the longer-term outcomes associated with deficits in EF in ADHD. For example, follow-up of both girls and boys with ADHD through adolescence did not show that EF predicted cigarette smoking or substance use disorders. In contrast, cigarette smoking and substance use disorders appeared to increase the rates of EF deficits in ADHD and controls (Wilens et al., 2011).
Not surprisingly, this group reported similar rates of comorbidity in adults. Using neuropsychological measures, they found that 31% of adults with ADHD had clinically relevant EF deficits (Biederman, Petty, Fried, Fontanella, Doyle, Seidman, & Faraone, 2006). Reflecting results found in youth, in adults with ADHD who manifested EF deficits compared to those without EF deficits, higher rates of grade repetition and a significant average decrease in IQ score were reported. They also reported that adults with ADHD with versus without EF deficits had significantly lower levels of overall socioeconomic status, education, and occupation. Likewise, EF deficits were associated with self-regulation issues faced by adults with ADHD (Seidman et al., 1998). Deficits in emotional self-regulation (DESR), often considered an “executive function,” and ADHD were further examined in a 2011 study in which a pattern of inheritance of ADHD with DESR was found—suggesting that DESR is a specific familial subtype of ADHD. (C. B. Surman et al., 2011). In a subsequent study, DESR severity was found not to correlate with test-based executive function measures suggesting a “bottom-up” neuropsychological etiology (C. B. H. Surman et al., 2015).
Better understanding deficits in EF is not just of academic interest. It is a tremendous clinical concern. In fact, one of the major reasons Joe created his EF research program was to figure out how neuropsychological testing should be used in his clinic. He also believed that understanding the relationship between EF and ADHD in youth and adults would highlight potential targets of treatment. Clinicians have long known that substantial problems with organization—often a proxy symptom of deficits in EF—were not adequately addressed with pharmacological treatment and often required additional, but less well described, psychosocial treatments. Providing quantitative evidence to support this observation, in one of the first systematic meta-analyses of studies examining pharmacological response in adolescents and adults with ADHD with and without EF deficits, Joe’s research showed that the impact of methylphenidate (MPH) on the core symptoms of ADHD was more robust than on the neuropsychological constituents of EF (Biederman, Mick et al., 2011).
Joe’s team worked to understand the relationship between test-based and behavior-based executive function measures, and their relationship to daily role function. They found that these two domains of measurement predict overlapping but not identical aspects of functional impairment (Biederman, Petty, Fried, Black, Faneuil, Doyle, Seidman, & Faraone, 2008). They also piloted various instruments to elucidate the presence of EF deficits: the Cambridge Automated Neuropsychological Tests and Battery (CANTAB), a computerized neuropsychological assessment for evaluating EF in clinical care and trials (Fried et al., 2015, 2021) and the Behavioral Rating Inventory of Executive Functioning (BRIEF), a behavioral assessment for evaluating EF challenges in clinical care in adults. Eventually the lab’s work led to collaborative expansion of Adult ADHD assessment to include EF deficits not recognized in the core DSM-5 criteria (Silverstein et al., 2019). Biederman’s team also piloted the sensitivity to intervention effects of the CANTAB and BRIEF by using these measures in clinical trials that directly targeted EF in patients with ADHD (Biederman, Mick et al., 2011; C. B. Surman et al., 2013).
In his writings and lectures, Joe underscored the need to extend studies of treatment of ADHD to include EF, encouraging clinicians and researchers to identify and monitor both ADHD and EF outcomes. Multiple studies from the Biederman lab showed impact on EF in ADHD when using nicotinic analogs and anti-Alzheimer’s medications (Biederman et al., 2017; Wilens, Biederman, Millstein, et al., 1999; Wilens, Vitulano et al., 2008). The work of the Biederman lab highlights the historical and present-day view that deficits in EF need to be identified and interventions employed to enhance short- and longer-term outcomes of individuals with these concurrent problems. Through a multifaceted effort, Joe and his team contributed much to our current understanding of pragmatic approaches to identifying and supporting executive function challenges.
Adult ADHD (By Craig Surman, MD)
In the past three decades, it has become clear that ADHD produces significant morbidity and can be treated in adults as well as in children. Joe’s lab was a prominent contributor to the scientific evidence on ADHD in adulthood. He and his collaborators answered several key questions through carefully designed and conducted studies. These include whether ADHD is a valid diagnosis in adults, what criteria allow its identification, whether it has unique neurobiological and neuropsychological features, and what treatments alleviate ADHD-related suffering in adulthood.
Joe and this team began systematically characterizing the longitudinal experience of children with and without ADHD in the 1980s. The formal longitudinal studies conducted by the group spanned up to 16 years, following girls (Biederman, Petty, Monuteaux, Fried, et al., 2010) and boys (Biederman, Petty, Woodworth, Lomedico, Hyder, & Faraone, 2012) with ADHD. In agreement with other longitudinal studies, this work affirmed that children with ADHD usually grow up to be adults with ADHD (Uchida, Spencer et al., 2018). Differentiating syndromic and symptomatic remission (Biederman, Mick, & Faraone, 2000) was central to understanding the continuity of ADHD’s impact into adulthood, because the natural history of ADHD involves a morphing of both the symptom patterns over time, with a decline in hyperactivity and impulsivity traits and persistence of inattentive traits and shifting environmental demands. A meta-analysis subsequently showed that their findings were consistent with a broader body of literature (Faraone, Biederman, & Mick, 2006). Their meta-analysis also showed why some prior work had erroneously concluded that ADHD disappeared in adulthood. That work had not included as persistent cases those youth who, although not meeting full DSM criteria, continued to have impairing symptoms of ADHD into adulthood.
The lab’s research strongly influenced the evolution of current diagnostic criteria for adult ADHD. In the late 1980s to early 1990s, these criteria were in flux between different versions of the Diagnostic and Statistical Manual. The DSM III-R (American Psychiatric Association, 1987) required impulsive-hyperactive traits to be present for the diagnosis, while the DSM IV (American Psychiatric Association, 1994) asserted that inattentive symptoms were sufficient for diagnosis of an ADHD presentation. By comparing referred and non-referred adults, the lab showed that ADHD under the DSM IV criteria met heritability and morbidity criteria for a clinical syndrome in adults. Further, this inattentive manifestation of ADHD was clearly associated with functional impairment and morbidity (Biederman, Baldessarini et al., 1993). Biederman and colleagues also explored the real-world impact of ADHD in adults, finding significant impairment across multiple measures of life function (Biederman, Faraone, Spencer, Mick, Monuteaux, & Aleardi, 2006; C. B. Surman et al., 2009).
It was not until DSM 5 (American Psychiatric Association, 2013) that specific criterion emerged for ADHD in adults. These were criteria were directly influenced by Biederman and colleagues’ work. The adult manifestation of ADHD was codified in DSM 5 as needing fewer symptoms than pediatric ADHD, and an age of onset by age 12 years was established as sufficient to identify the syndrome. These elements of our current diagnostic criteria were consistent with the lab’s findings of the longitudinal course of ADHD symptom burden (Biederman, Petty, Monuteaux, Fried, et al., 2010; Biederman, Petty, Woodworth, Lomedico, Hyder, & Faraone, 2012; Uchida, Spencer et al., 2018) and demonstration that recall of ADHD symptoms by age 12 years differentiated adults who otherwise met the same criteria and syndrome correlates (Chandra et al., 2021; Kessler et al., 2006). They were also motivated by the lab’s work studying subthreshold and late onset forms of ADHD in adulthood (Faraone, Biederman, Doyle, Murray, et al., 2006; Faraone, Biederman, Spencer, Mick, et al., 2006; Faraone et al., 2007, 2009).
We now appreciate that there is a high rate of comorbidity between mental health conditions identified via DSM nosology in adulthood. The research group showed this to be true across both men and women with ADHD (Biederman et al., 1994). Isolating the presence of ADHD in the context of comorbidity, as well as symptom overlap between ADHD and other conditions, was a central challenge that required rigorous methods. Specifically, the research group operationalized rater-trained structured clinical interviewing covering DSM-categorized conditions. This allowed identification of both current and past major mental health syndromes in the volunteers that took part in studies. Systematic identification of comorbidity also allowed isolation of ADHD-specific treatment effects during clinical trials because those with comorbid conditions could be excluded (C. Surman et al., 2010), or separately, how comorbidity impacted response to treatment could be evaluated (Biederman, Mick, Spencer, Surman, & Faraone, 2012).
Systematic identification of comorbidity patterns within research subjects was also used to evaluate the risk that the presence of ADHD confers for other conditions. This led to an understanding that with ADHD came a higher risk for more severe and earlier onset substance use disorders as children transition into adulthood (Biederman, Wilens, Mick, Milberger, Spencer, & Faraone, 1995; Wilens et al., 1997). Similarly, ADHD was also determined to convey risk for bulimia nervosa among women (C. B. Surman et al., 2006). A series of explorations of the connection between ADHD and bipolar disorder led to a range of important findings, including that bipolar disorder and ADHD may represent a unique phenotype (Biederman, Faraone, Keenan, & Tsuang, 1991). Adult ADHD was also found to confer unique risk for disordered patterns of sleep, independent of presence of other comorbid conditions (C. B. Surman et al., 2009).
Earlier descriptions of how ADHD manifests in adults had suggested that affective behavior patterns such as temper and anger could be a core element of the adult diagnosis. Through a series of studies, the laboratory showed that the dimensional capacity of emotional self-regulation was often deficient in ADHD adults (C. B. Surman et al., 2013), and its co-occurrence with ADHD may reflect a separate phenotype (C. B. Surman et al., 2011). In a series of studies, his team documented high rates of deficient emotion self-regulation among adults with ADHD that increased the disability associated with the disorder(Biederman, DiSalvo et al., 2020; C. B. Surman et al., 2013; C. B. H. Surman et al., 2015). They also showed that emotional dysregulation was transmitted in families and could not be accounted for by other DSM disorders or non-familial environmental factors
The journey to characterize ADHD in adults also required development of new tools to capture syndrome specific symptom and functional impairments. These tools were developed through inter-academic center collaboration, and included a systematic clinical interview, the Adult ADHD Investigator Symptom Rating Scale (AISRS; Biederman, Mick, Spencer, Surman, Hammerness, Doyle, Dougherty, et al., 2006), as well as a self-report scale, the Adult ADHD Self-Report Scale (ASRS; L. A. Adler et al., 2006). A screener developed from this latter instrument (Kessler, Adler, Barkley, Biederman, et al., 2005) was central to determination of the prevalence of ADHD in the United States during the National Comorbidity Replication Survey (Kessler, Adler, Ames, Demler, et al., 2005), which identified that 4.4% of adults aged 18 to 55 years had ADHD. Both scales remain in regular use as part of scientific evaluations of ADHD and its treatment. Laboratory measures of ADHD-specific functional impairment were also validated for two critical areas of threat to adults with ADHD: workplace performance (Fried et al., 2012) and driving performance (C. B. H. Surman et al., 2017). With the emergence of technology for remote care, Joe piloted how a SMS messaging system could improve adult ADHD patient engagement in treatment (Biederman, Fried, DiSalvo, Woodworth, et al., 2019; Fried et al., 2019).
Pilot work by the research program also helped establish that ADHD in adults was heritable and manifested as a genetically influenced neurobiological condition. This was supported by clear familial patterns of ADHD among the children of adult probands (Biederman, Faraone, Mick, Spencer, Wilens, Kiely, Guite, Ablon, et al., 1995). Using state-of-the-art technology, the research program also explored the neurobiological basis of ADHD. This line of research identified differences in brain structure (Makris et al., 2007; Semrud-Clikeman et al., 1994) and function (Bush et al., 1999; Mattfeld et al., 2016) in adults with ADHD compared to those without ADHD, contributing to our current understanding that implicates differences in a range of executive and attention circuits. Imaging studies were also used to identify differences in ADHD individuals with and without other conditions such as bipolar disorder (Biederman, Makris, Valera, Monuteaux, Goldstein, Buka, Boriel, Bandyopadhyay, Kennedy, et al., 2008) and to identify dopamine system substrate involved in stimulant treatment effects of ADHD (T. J. Spencer et al., 2005). A quest for neuropsychological correlates of ADHD revealed that a range of executive function deficits were more common in both ADHD children and adults, but that behaviorally defined ADHD by the DSM criteria existed without such deficits (Seidman et al., 1998).
The research group also conducted many investigations of potential treatments for adult ADHD, building at first on agents that worked for pediatric ADHD, but expanding to pilot new therapies, some of which are now used across the lifespan. This work included positive pilot studies of desipramine (Wilens et al., 1996), methylphenidate (T. Spencer et al., 1995), atomoxetine (T. Spencer et al., 1998), and memantine (C. B. Surman et al., 2013). Just as importantly, other agents were explored and found to not have promise, such as galantamine (Biederman, Mick, Faraone, Hammerness, Surman, Harpold, Dougherty, Aleardi, & Spencer, 2006). Behaviorally focused cognitive training was also found to relieve ADHD burden (Safren et al., 2010).
Through systematic research, Biederman and colleagues helped build the modern foundation we rely on for the assessment and treatment of self-regulation problems that occur in adults. Driving each of the myriad investigations were core questions whose answers now pragmatically guide how we approach and relieve ADHD suffering across the lifespan.
Summary (By Thomas Spencer, MD)
The authors have admirably carried out the daunting task of reviewing and narrating the immense contributions of Joe Biederman’s research career. An essential part of Joe’s genius was to assemble, mentor, and work closely with such amazing colleagues. He worked with each team to define the important goals and outline a plan of attack. Importantly, this included the methodological rigor of epidemiologic and statistical approaches provided by Steve Faraone and an army of statisticians, clinician researchers and research assistants. During Joe’s extended peak period of productivity, he had over 100 people on staff, published 50 to 70 scientific papers per year, submitted dozens of grants, mentored dozens of students and colleagues, presented at meetings of the major national and international organizations and carried over 500 active patients. He did not command from a distance. He wrote side by side with each of us and wrote each manuscript, presentation, and grant, line by line. His rich and precise command of language and logic was remarkable even though English was his third language.
As evident in this review, Joe orchestrated a vast array of interconnected activities that extended from exploratory pilot studies to hypothesis testing in fully funded, methodologically sophisticated, controlled, multiyear projects. He simultaneously worked on multiple projects in different stages across topics and disorders, with complementary methodologies including characterization, diagnostic issues, longitudinal outcome, family functioning, neuropsychology, genetics, neurochemistry, neuroimaging, and treatment. And although he was best known for his work on ADHD, he made substantial contributions to the literature on mood and anxiety disorders. While challenging, Joe’s multi-prong approach facilitated an unparalleled understanding of the interconnectedness of each component. Instead of being restricted to the tunnel vision of a single specialty clinic or a single specialty research focus, he was able to approach the totality of each person’s experience and the complexity of each condition.
There is a tension between lumpers and splitters in understanding psychopathology. Of course, there is some truth in each approach. In child psychiatry, before Joe, there was a tendency to always lump. Each individual’s condition had to be explained by a single cause, most often thought to be entirely a reaction to environmental factors. In part this was due to an aversion to giving children any psychiatric diagnoses. Instead of diagnoses being seen as the key to precise understanding (diagnosis: “to know thoroughly”), psychiatric diagnoses were often considered erroneous and stigmatizing. In part this came from a notion of childhood as being uniformly idyllic. For instance, it was thought that children did not possess a sufficient maturity of their mental apparatus to manifest clinical depression. This is at odds to an increasing appreciation of the prevalence of serious pathology in childhood. Landmark epidemiologic research has established that 50% of all mental health disorders begin on or before age 14 years (Kessler, Berglund, Demler, et al., 2005). The Youth Risk Behavior Surveillance System (YRBSS) has documented high rates of suicidality every year since 1991. Yet each new report is usually greeted with surprise and incredulity. The YRBSS is a nationwide set of surveys conducted by the CDC in students grades 9 through 12 (Kolbe et al., 1993). In the latest YRBSS survey(Center for Disease Control and Prevention, 2021), a sample of 17,232 youth, 30% of girls and 13% of boys had seriously considered attempting suicide in the previous 12 months. About 23.6% of girls and 11.6% of boys reported making a suicide plan. About 13.3% of girls and 6.6% of boys reported an attempted suicide. Childhood is a difficult time for many and requires clearheaded guidance.
I met Joe in 1985 working together on pediatric psychiatric inpatient units, including those of the Massachusetts Department of Mental Health. It was clear that each hospitalized child had many years of escalating difficulties before needing these most dramatic interventions. Joe recruited me to work in an outpatient setting and on research projects to define and detect problems at an earlier stage to prevent such catastrophic deterioration. I was dismayed to find that even children referred as outpatients already experienced years of escalating difficulties. Delays in referral were often due to public misconceptions from attacks on Child Psychiatry. In a sample of 424 consecutive referrals to the MGH outpatient clinic, structured diagnostic evaluations revealed that 40% met criteria for clinical depression. Even though the average age at referral was 11 years old, those with depression had already experienced 3.2 years of clinical depression, 28% of their young lives. Further, 96% of the depressed children had other psychiatric disorders usually starting even earlier. Seventy-five percent also had ADHD, onsetting at an average age of 3.5 years and 40% also had anxiety disorders, onsetting at 5 years old (Biederman, Faraone, Mick, & Lelon, 1995).
Many accepted clinical maxims were challenged by Joe. Examples include the DSM-IV-TR proscription against diagnosing ADHD in the presence of Pervasive Developmental Disorder. Attentional symptoms were thought to be always part of the condition and could not be diagnosed (or treated) separately. Moreover, mood problems in individuals with ADHD were considered to be expressions of demoralization or a variant expression of the ADHD itself, so called “bad” ADHD. While sometimes this is true, as detailed in this article, Joe, his colleagues, and others have been able to disentangle mood disorders and ADHD with careful, structured observation over time, family studies, large community samples, imaging, and genetics.
As in the rest of medicine, decisions to treat and choice of treatment modality in psychiatry depend on a clear understanding of the risks and benefits of deploying each potential treatment. This is only possible if we understand the natural course of the disorder with and without each treatment modality. As outlined in this article, Joe and teams’ longitudinal studies followed probands with ADHD, anxiety, and mood disorders with matched controls for decades. Treatment was naturalistic.
Exposure to ADHD treatments can worsen several conditions such as tics, anxiety, mood, and SUD in some individuals and should be used with care and close monitoring. However, Joe, his colleagues and other research groups have documented how carefully monitored, early treatment with medication, on average, reduces the risk of the subsequent development of additional disorders and associated dysfunction. A series of original studies and meta-analyses revealed a robust protective effect of ADHD medication treatment on mood disorders, suicidality, criminality, substance use disorders, accidents and injuries, traumatic brain injuries, motor vehicle crashes, and educational outcomes (Boland et al., 2020). Similarly, the meta-analyses demonstrated a protective effect of medication treatment on academic outcomes, accidents and injuries, and mood disorders. The authors concluded that these findings suggest that ADHD medication treatments are associated with decreases in the risks for a wide range of ADHD-associated functional outcomes supporting efforts aimed at early diagnosis and treatment of individuals with ADHD.
What was the source of Joe’s phenomenal creativity? He had a constant flow of new ideas and was never known to suffer writers’ block. Joe was fearless. He had a sign in his office with an Einstein quote “If we knew what we were doing, it wouldn’t be called research.” Joe was innovative, disruptive and generative but never expected his hypotheses to be accepted unless and until they were replicated by independent research. He was strongly moved by his patients’ conditions and personally available 24/7 to patients, supervisees, and colleagues. Several of us set up our clinics on the same day so we could consult with each other in real time. We were supposed to be experts at our craft, but, like all practitioners we each had patients who didn’t respond or were just confusing. We valued the opportunity for hallway consults. Clinical care is extremely difficult to do well and Child Psychiatry was under constant attack. Because we had a tight nexus of trusted colleagues, we could be open about our doubts and misgivings. The mutual support enabled us to be self-critical in order to try new approaches and innovate. In the clinic Joe would take careful note of characteristics of each apparent disorder, the time course, the family history of related disorders and response to school interventions, pharmacologic treatment, psychological treatment, and psychosocial intervention. From this data he generated hypotheses that he tested first in case series and open studies. Findings led to rigorous testing with refined hypotheses with controlled studies and blind assessments as detailed in this article. As noted by Dr. Faraone’s section, observed comorbidity was not due to an artifact of psychiatric referral as it was confirmed in independently ascertained psychiatrically and pediatrically referred samples as well as community based epidemiologic samples both nationally and internationally. Psychiatric comorbidity was already well recognized in adult research. Kessler et al.’s (2006, 2009) landmark adult and adolescent studies addressed the incidence, onsets and relationship of psychiatric disorders. Their findings on the bi-directionality of comorbidity in individuals with ADHD closely matched Joe’s earlier findings in samples pediatrically and psychiatrically ascertained child samples as well as adult samples.
When someone like Joe is so remarkably talented, accomplished and productive it is tempting to mythologize and imagine that it came easily. It did not. He was gifted but worked very hard. To do his best work he needed input from a large and diverse group of colleagues. He enjoyed writing alongside others and having someone else to bounce ideas off of as he worked. All the authors on this article are extremely grateful to have been able to work so closely with Joe. The mission of academic medicine is to move the field forward. I was extremely fortunate to be part of a remarkable team that achieved great accomplishments in research, publications, and teaching. All in the service of providing and improving compassionate psychiatric care.
Footnotes
Acknowledgements
The authors thank Helen Charlupski for her edits and review of the manuscript.
Declaration of Conflicting Interests
The author(s) declared the following potential conflicts of interest with respect to the research, authorship, and/or publication of this article: Stephen V. Faraone: Over the past 3 years, Dr. Faraone received income, potential income, travel expenses, and/or research support from Aardvark, Aardwolf, AIMH, Akili, Arbor, Atentiv, Axsome, Genomind, Ironshore, Johnson & Johnson/Kenvue, Kanjo, KemPharm/Corium, Medice, Noven, Ondosis, Otsuka, Rhodes, Shire/Takeda, Sky Therapeutics, Supernus, Tris, and Vallon. With his institution, he has US patent US20130217707 A1 for the use of sodium-hydrogen exchange inhibitors in the treatment of ADHD. He also receives royalties from books published by Guilford Press: Straight Talk about Your Child’s Mental Health; Oxford University Press: Schizophrenia: The Facts; and Elsevier: ADHD: Non-Pharmacologic Interventions. In addition, he is the program director of 
. Dr. Faraone is supported by the European Union’s Horizon 2020 research and innovation program under grant agreement No 965381; NIMH grants U01AR076092-01A1, 1R21MH1264940, R01MH116037; 1R01NS128535-01; Oregon Health and Science University, Otsuka Pharmaceuticals, Noven Pharmaceuticals Incorporated, and Supernus Pharmaceutical Company. Jeffrey Newcorn over the past 3 years, has received financial compensation as a consultant, advisory board member, and lecturer from Shire Pharmaceuticals (the study sponsor and manufacturer of the study drug Vyvanse (Lisdexamfetamine)). Dr. Newcorn also receives financial compensation from other companies which either develop or assess medicines used for the treatment of ADHD. Janet Wozniak, MD receives research support from PCORI, Demarest Lloyd, Jr. Foundation, and the Baszucki Brain Research Fund. In the past, Dr. Wozniak has received research support, consultation fees or speaker’s fees from Eli Lilly, Janssen, Johnson and Johnson, McNeil, Merck/Schering-Plough, the National Institute of Mental Health (NIMH) of the National Institutes of Health (NIH), Pfizer, and Shire. She is the author of the book, “Is Your Child Bipolar” published May 2008, Bantam Books. Her spouse receives royalties from UpToDate; consultation fees from Indorsia, Cozen O’Connor, Noctrix, FoxRothschild, Sterne Kessler Goldstein & Fox, Teladoc Health, Inc., and The International Law Firm of Winston & Strawn LLP; and research support from Merck, NeuroMetrix, American Regent, NIH, NIMH, the RLS Foundation, and the Baszucki Brain Research Fund. In the past, he has received honoraria, royalties, research support, consultation fees or speaker’s fees from: Emalex, Disc Medicine, Avadel, HALEO, OrbiMed, CVS, Otsuka, Cambridge University Press, Advance Medical, Arbor Pharmaceuticals, Axon Labs, Boehringer-Ingelheim, Cantor Colburn, Covance, Cephalon, Eli Lilly, FlexPharma, GlaxoSmithKline, Impax, Jazz Pharmaceuticals, King, Luitpold, Novartis, Neurogen, Novadel Pharma, Pfizer, Sanofi-Aventis, Sepracor, Sunovion, Takeda, UCB (Schwarz) Pharma, Wyeth, Xenoport, and Zeo. Gagan Joshi is supported by the National Institute of Mental Health (NIMH) of the National Institutes of Health (NIH) under Award Number K23MH100450. In 2023, he receives research support from the Demarest Lloyd, Jr. Foundation as a primary investigator (PI) for investigator-initiated studies. Additionally, he receives research support from Genentech as a site PI for multi-site trials. Dr. Joshi is also a consultant for EuMentis Theraputics. In the past 3 years, he has received speaker’s honorariums from the American Academy of Child and Adolescent Psychiatry, American Physician Institute, Asian College of Neuro-psychopharmacology, Hackensack Meridian Health, University of Colorado-Colorado Springs, and Kennedy Krieger Institute; he received research support from F. Hoffmann-La Roche Ltd. s a site PI for multi-site trial. Through Mass General Brigham Innovation, Dr. Joshi receives royalties from a licensed method for treating autism spectrum disorder. Barbara Coffey receives income, research support, or honoraria payments from: American Academy of Child and Adolescent Psychiatry, Emalex, Florida Department of Health, Galen Mental Health, Harvard Medical School /Psychiatry Academy, New Venture Fund, NIMH/NINDS, Partners Healthcare, Skyland Trail, Tourette Association of America, Zynerba. Mai Uchida has received honorarium from American Physicians Institute and Mochida Pharmaceuticals and royalties from the book “Ask The Geniuses About the Future” (Magazine House Publishing) and has consulted to Moderna and Guidepoint. Timothy Wilens receives or has received grant support from NIH(NIDA). Dr. Timothy Wilens has co/edited books: Straight Talk About Psychiatric Medications for Kids (Guilford Press), ADHD in Adults and Children (Cambridge University Press), Massachusetts General Hospital Comprehensive Clinical Psychiatry (Elsevier), Massachusetts General Hospital Psychopharmacology and Neurotherapeutics (Elsevier), and Update on Pharmacotherapy of ADHD (Elsevier Press). Dr. Wilens has a licensing agreement with Ironshore (BSFQ Questionnaire) and 3D Therapy. Dr. Wilens is Chief, Division of Child and Adolescent Psychiatry and (Co) Director of the Center for Addiction Medicine at Massachusetts General Hospital. He serves as a clinical consultant to U.S. Minor/Major League Baseball, Gavin Foundation and Bay Cove Human Services. Craig Surman has received, in his lifetime, consulting fees from Eisai, Ironshore, Kaylon, Mcneil, NLS Pharma, Neurocentria, Nutricia, Otsuka, Pfizer, Adlon/Purdue, Rhodes, Shire, Somaxon, Sunovion, Supernus, Takeda, and Teva. He has also received payments for lectures for Alcobra, Arbor, McNeil, Janssen, Janssen-Ortho, Novartis, Shire, and Reed/MGH Academy (funded by multiple companies) as well as GME CME (funded by multiple companies). Royalties have been given to Craig B. H. Surman from Berkeley/Penguin for ““FASTMINDS” How to Thrive if You have ADHD (or think you might)” and from Humana/Springer for “ADHD in Adults: A Practical Guide to Evaluation and Management.” Additionally, Dr. Surman has conducted clinical research at Massachusetts General Hospital supported by Abbot, Cephalon, Hilda and Preston Davis Foundation, Eli Lilly, Magceutics/Neurocentria, Jazz, Johnson & Johnson/McNeil, Lundbeck, Merck, Nordic Naturals, Shire, and Takeda. Thomas J. Spencer receives royalties from Cambridge University Press for book publication and from MGH Corporate Licensing for ADHD rating scales, holds a US patent (#14/027,676) for a nonstimulant treatment for ADHD (no license fees), and has a patent pending (#61/233,686) for a method to prevent stimulant abuse (no license fees).
Funding
The author(s) received no financial support for the research, authorship, and/or publication of this article.