The Feasibility of Diagnosing Psychiatric Disorders with Neuroimaging

 Introduction

In the past few years, the number of individuals seeking treatment for psychiatric disorders has increased significantly [1]. Mental illness statistics continue to rise year after year. In his 2010 book, Robert Whitaker reported that the number of mentally ill had tripled in the past two decades [2]. In 2019, 56.4% of individuals ages 18-25 received mental health treatment, compared to the 45.9% receiving treatment in 2008 [3]. Suicide is now the second leading cause of death in persons aged 18-34. As of December 2020, Hedegaard and colleagues reported that suicide rates have increased by 35% since 1999. Furthermore, the report stated that 90% of the people who died by suicide were confirmed to have shown symptoms of mental illness [4]. These statistics are concerning and bring about a number of questions, one being the effectiveness of prescription drugs. Just how effective are these treatments? Furthermore, what limits improvement within the fields of psychiatry and psychology?

One surprising limitation may be the Diagnostic and Statistical Manual of Mental Disorders (DSM-5). The DSM-5 classifies mental disorders using set symptom-based criteria and is the standard for clinical diagnoses. However, this manual does not come without fault and controversy. A growing number of researchers have cited concern about false positives that occur as a result of the Diagnostic Manual’s recently lowered diagnostic thresholds [5]. 

Neuroimaging, also known as brain scans, may be useful for improving diagnostic accuracy. Neuroimaging approaches involve assessing structural anatomy and functional activity. If health professionals can diagnose individuals based on brain abnormalities associated with psychiatric disorders, then there may be a lower chance of misdiagnosis and error. This article will explore neuroimaging literature to assess the feasibility of this approach. It will be organized by first considering current issues within the field of psychiatry and a review of neuroimaging methods before a discussion of potential strengths and limitations of the approach.

Current Issues in Psychiatry

Medication is often the first line of defense in treating individuals who are diagnosed with psychiatric and neurocognitive disorders, such as depression and Attention-Deficit/ Hyperactivity Disorder (ADHD) [8]. However, many psychiatrists, such as Daniel Cartal, have expressed concern in this overemphasis of medication and have questioned the accuracy of their success rates [6]. Throughout the years, the amount of psychotropic drugs prescribed has continued to increase. Between the years 2001 to 2010, there was a 22% increase in the prescription of psychotropic drugs. In a 2011 study, Pratt et al., found that antidepressant use had increased 400% in a decade [5]. In 2012, antidepressants were the second most commonly prescribed prescription drug [7]. However, few research articles have addressed more recent statistics of this increase in prescription drugs.  

Interestingly, this increase does not appear to have resulted in reduced prevalence of mental disorders [9]. One reason for this could be that antidepressants may not be as successful as consumers are led to believe. In 2008, clinical psychologist Irving Kirsch, PhD, led an analysis of all FDA trials for antidepressants and found that many antidepressants did not perform better than the placebo pill for mild and moderate depression [7]. However, some critics have cited issues with methodology in Kirsch’s analysis, so more research is needed in this area.

Another concern regards the overdiagnosis of psychiatric disorders, which may be exacerbated by the DSM-5. In mental health cases, overdiagnosis is said to occur, “when people do experience symptoms, but where definitions of disorders include potentially transitory or mild symptoms that reflect ordinary life experiences that, as such, are not amenable to improvement or management benefits through medical intervention” [10]. Put more simply, anyone experiencing mild symptoms of a psychiatric disorder can be diagnosed based on DSM criteria and prescribed medication. This may then lead to overtreatment since, in many cases, it is possible that an individual with mild symptoms may have been able to recover without medical intervention. This may expose individuals to drugs that cause them more harm than good [10]. For example, overdiagnosis of ADHD in children can result in overtreatment through major drugs such as Ritalin, Concerta, and Lisdexamfetamine. One study found that these medications can cause suicidal thoughts, weight loss, and can be toxic to liver cells [11].

Thus, it should not be surprising that some mental health professionals also raise concerns about the DSM-5. Both psychologists and psychiatrists initially rely on the DSM for a clinical diagnosis, but revisions of the latest DSM volume have not come without criticism. One issue that may be driving overdiagnosis could be the lowering of diagnostic thresholds in DSM-5 revisions. These lower thresholds cause many psychologists to believe that it may “inappropriately classify more behaviors as disorders” [12]. Put more simply, psychologists feel that some commonplace human behaviors have been added as symptom-criteria used to classify disorders. For example, nearly everyone has experienced some form of grief, anxiety, and distractibility for a few weeks at some point, but that does not mean they have a psychiatric disorder [5].

Financial conflicts of interest disclosed by DSM advisors may also further complicate the issue of overdiagnosis. For example, 78% of DSM-IV ADHD advisors “disclosed links to drug companies as a potential financial conflict of interest” [11]. In other words, health professionals advising changes in the DSM-IV had been previously (or continued to be) financially associated with drug companies. It is possible that this conflict of interest impacts decision making in diagnostic criteria as it may relate to prescription treatment.

Yet, despite concerns over the manual, 94% of psychologists still use it when making a clinical diagnosis [12]. With the aforementioned limitations, there is a clear reason to explore other treatment and diagnosis options. One potential tool is neuroimaging, with the hope of identifying psychiatric disorders by a biological marker. 

Types of Neuroimaging

Neuroimaging comes in many varieties and forms. Structural anatomical neuroimaging shows the anatomical matter of the brain. Computed tomography (CT) scans create images by consolidating several two-dimensional images into a single three-dimensional image. Comparatively, magnetic resonance imaging (MRI) uses magnetic technology that creates an image based on the displacement of atoms.  

While CT scans are generally used to diagnose neurological disorders and other diseases  (such as cancerous tumors), MRI may be beneficial in allowing researchers to measure the volume of brain structures that might be implicated in psychiatric disorders [13, 14]. 

For the purposes of this article, we will focus on functional neuroimaging, which displays brain activity over time. Two common functional modalities are positron emission tomography (PET) and functional magnetic resonance imaging (fMRI).

PET scans are administered by injecting an individual with a radioactive isotope tracer that is tracked by gamma rays to generate an image of the brain. Isotopes can be specified to study certain biological mechanisms implicated in a neurological disease. For example, scientists have developed a tracer that binds to the beta-amyloid protein indicated in Alzheimers’ Disease [13]. PET scans can also measure metabolic blood flow activity in the brain, also known as regional cerebral blood flow (rCBF). By tracking rCBF, researchers can identify active and inactive brain structures during rest and other cognitive tasks. Some psychiatriatric disorders may be associated with increased, or decreased activity in specific brain areas. Thus, PET scans can help researchers identify potential functional differences associated with a disorder [13].

 Similar to PET scans, fMRI creates images by tracking rCBF associated with action potentials via movement of hemoglobin and oxygen. It allows researchers to image functional activity of brain regions over time. Rather than measuring activity via a radioactive isotope injection, fMRI measures the ratio of weakly magnetic deoxygenated hemoglobin to oxygenated hemoglobin (which has no magnetic properties) [13]. The imaging device then generates a functional image of neural activity based on this ratio. fMRI may be preferable in some cases, as it does not involve radiation, and can thus be used for vulnerable populations such as children and pregnant women. This decreased exposure to radiation also means that individuals can be tested multiple times with fMRI. Finally, fMRI has higher quality spatial resolution, meaning that it produces clearer images than those generated by PET during tasks in the scanner [13].

Given their ability to assess functional brain differences that may be associated with  psychiatric disorders, PET and fMRI are often explored for purposes of diagnoses.

Neuroimaging & Psychiatric Disorders

Psychiatric disorders have long been classified on categorical diagnoses based on presenting symptoms [12]. However, with the introduction of neuroimaging technology, scientists have been able to learn about brain abnormalities and  differences that may correlate with psychiatric disorders. This section will focus on the neuroscientific perspective of major depressive disorder and bipolar disorder. However, it is notable that psychiatric disorders are not solely caused by brain abnormalities. Rather, they are multifactorial and are likely caused by an interaction of biological, psychological, and sociocultural factors such as genetics, learning, and environment.

PET Scans and Major Depressive Disorder

Major depressive disorder (MDD) is a distressing and impairing disorder. It is characterized by feelings of hopelessness, pessimism about future events, difficulty concentrating, loss of pleasure and motivation, and much more.  Research has thus sought to identify brain changes associated with MDD in order to identify possible neurochemical and pathological causes that in turn, might help develop treatments.

Numerous PET scans have found that individuals with MDD have decreased activity in their subgenual anterior cingulate at rest. The subgenual anterior cingulate, located in the frontal lobe, is associated with executive functions such as self-regulation and selective attention. Thus, functional differences in the subgenual anterior cingulate appear to be associated with MDD [14].

Another PET scan study focused on dopamine, a neurotransmitter that is implicated in mood, reward, addiction, and motivation [16]. One major symptom of MDD is lack of motivation and pleasure, hence why dopamine is a neurotransmitter of interest for researchers. Like all neurotransmitters, it must be regulated and maintain equilibrium. This is the job of the dopamine transporter (DAT), DAT terminates dopamine related activity through reuptake (reabsorption) after the neurotransmitter has triggered the appropriate signalling. Dopaminergic neurons can be found in the striatum and ventral tegmental area, which are areas of the brain associated with motivation and reward activation. One recent PET scan study found that individuals with MDD may have lower amounts of dopamine transporter availability in these brain areas [17]; low levels of DAT cause decreased dopamine availability. In turn, this decreased availability may be associated with deficits in motivation, suggesting that decreased dopamine activity may be one factor that contributes to MDD.
While these studies do not show instances in which PET scans diagnosed psychiatric disorders, they show the important contributions that PET scans can make by identifying possible neural abnormalities associated with a disorder.

fMRI and Bipolar Disorder

Neurobiological markers for psychiatric disorders are often identified by comparing neuroimaging scans between clinical groups and “healthy controls” (HC), a case-controlled approach [18]. 

Many studies use this approach in the context of neutral facial processing. For example, individuals are shown an image of a neutral face and asked to classify the type of mood that the face exhibits. Researchers then analyze brain activity in specific brain areas to test differences between the clinical sample and HC. One area of interest is the dorsal prefrontal cortex (DPFC), which is associated with emotional processing, among other functions. Therefore, this part of the brain is quite active as individuals classify the moods of neutral faces.

One study found that patients with Bipolar Disorder (BD) -- a mental disorder characterized by severe shifts in mood -- were more likely to classify neutral faces as exhibiting negative moods [18]. Upon further analysis of their scans during this task, experts have found decreased DPFC activity in individuals with BD. This study demonstrated that decreased DPFC activity during certain cognitive tasks may be associated with BD [18]. Thus, showing us that neuroimaging may have the potential to help researchers and health professionals recognize potential biomarkers (measurable indicators of biological and psychological state). By recognizing disorder-specific neural differences, researchers may then be able to develop improved therapies, medications, and other treatment techniques in the future. Furthermore, it may lead to increased understanding of this disorder.

Limitations

While these studies show promising evidence that neuroimaging can detect neural differences in psychiatric groups, they may not be enough to conclude that neuroimaging can be used to diagnose psychiatric disorders.

Individual differences may affect neuroimaging. For example, even the brains of healthy controls can show anatomical and functional differences. However, these dissimilarities do not necessarily indicate behavioral issues [19]. Furthermore, differences in age, sex, medication, and the phase of illness may further act as confounding variables [15].

Moreover, due to methodology issues, such as small sample and effect sizes, neuroimaging studies often have low generalizability. This means that what is common for a small number of individuals with a specific psychiatric disorder may not be common for most individuals with the disorder. Thus, it cannot be assumed that the same functional abnormalities are present in all individuals with the psychiatric disorder, unless multiple replication and longitudinal studies are done in the future [19]. Unreported comorbid psychiatric disorders may further confound the methodological process. For example, an individual with bipolar disorder may also experience unreported depression and obsessive-compulsive disorder. In a case like this, if the researcher is unaware of the other existing disorders, they may come to believe that bipolar disorder is associated with specific neural differences, thus jeopardizing the study’s validity.

In most neuroimaging cases, findings are correlational, as specific abnormalities may seem to correlate with certain psychiatric disorders. Thus, these types of neuroimaging studies should not be used to make causal statements. Neuroimaging studies are limited in their validity of potential diagnostic use, thus, it may be inappropriate to suggest that functional or anatomical abnormalities cause psychopathology. For example, it would be inappropriate to say that Patient X has MDD because neuroimaging showed anatomical and functional abnormalities in the anterior cingulate. Rather, it may be more appropriate to say that Patient X’s abnormalities in the anterior cingulate may have some correlation with MDD.

Finally, several clinical disorders may show similar brain abnormalities. For example, it is common to find decreased blood flow in the frontal lobe in more than one disorder. Patients with schizophrenia, Parkinson’s disease, Alzheimer's disease, and depression may all show the same pattern of decreased blood flow in the same brain area [19]. In order to narrow down which disorder is occurring, the healthcare provider must still consult the DSM for a diagnosis. This further exhibits why neuroimaging alone cannot be used for diagnostic purposes.

Neuroimaging may be beneficial in the study of psychiatric disorders, specifically in terms of decreasing the chance of false positives, in increasing understanding of disorders, and recognizing potential biomarkers. However due to the array of limitations involving low generalizability, low validity, methodological issues, and comorbidities, it is likely not currently a feasible method of clinical diagnosis.

Conclusion

Some popular psychiatrists, such as Dr. Daniel Amen, have suggested that neuroimaging tools should be used in clinical practice for diagnosis and treatment [20]. Nevertheless, neuroimaging has made important contributions to the understanding of psychiatric disorders and the brain in general. These contributions may indirectly aid in the future development of more sophisticated prescription drugs that will in turn, help treat such disorders. Perhaps, with time, neuroimaging technology will become advanced enough to diagnose and treat these disorders. But until analysis methods, generalizability, and other methodological limitations are solved, traditional methods remain the best option. However, as research continues on the use of neuroimaging research, there is promise to improve our understanding of psychiatric disorders in the future.

References

1. Jorm, A. (2018). Improving the impact of treatments and services. Australian & New Zealand Journal of Psychiatry, 52(10), 915–916. https://doi.org/10.1177/0004867418802031

2. Whitaker, R. (2010). Anatomy of an epidemic: Magic bullets, psychiatric drugs, and the astonishing rise of mental illness in America. Crown Publishers/Random House.

3. Substance Abuse and Mental Health Services Administration. (2020). Key substance use and mental health indicators in the United States: Results from the 2019 National Survey on Drug Use and Health (HHS Publication No. PEP20-07-01-001, NSDUH Series H-55). Rockville, MD: Center for Behavioral Health Statistics and Quality, Substance Abuse and Mental Health Services Administration. Retrieved from https://www.samhsa.gov/data

4. Hedegaard H., Curtin S. C., Warner, M. (2020). Increase in suicide mortality in the United States, 1999–2018. National Center for Health Statistics Data Brief, 362.

5. Wakefield J. C. (2016). Diagnostic Issues and Controversies in DSM-5: Return of the False Positives Problem. Annual review of clinical psychology, 12, 105–132. https://doi.org/10.1146/annurev-clinpsy-032814-112800

6. Flaskerud, J. (2011). Revelations on the State of Psychiatry Today. Issues in Mental Health Nursing, 32(4), 266-268. https://doi.org/10.3109/01612840.2010.519816

7. Smith, L. B. (2012). Inappropriate prescribing. Monitor on Psychology, 43(6), 36. https://www.apa.org/monitor/2012/06/prescribing

8. Partridge, B., Lucke, J., Hall, W. (2014). Over-diagnosed and over-treated: a survey of Australian public attitudes towards the acceptability of drug treatment for depression and ADHD. BMC Psychiatry, 14(74). https://doi.org/10.1186/1471-244X-14-74

9. Ormel, J., Kessler, Ronald C.B; Schoevers, Roberta. (2019). Depression: More Treatment but no Drop in Prevalence. Current Opinion in Psychiatry, 32(4), 348-354. https://doi.org/10.1097/YCO.0000000000000505

10. Thombs, B. D., Turner, K. A., Shrier., I. (2019). Defining and Evaluating Overdiagnosis in Mental Health: A Meta-Research Review. Psychotherapy and Psychosomatics, 88(4), 193-202. https://doi.org/10.1159/000501647

11.  Thomas, R., Mitchell, G. K., Batstra, L. (2013) Attention-deficit/hyperactivity disorder: are we helping or harming? BMJ, 347(7932). https://doi.org/10.1136/bmj.f6172

12. Raskin, J. D., & Gayle, M. C. (2016). DSM-5: Do Psychologists Really Want an Alternative? Journal of Humanistic Psychology, 56(5), 439–456. https://doi.org/10.1177/0022167815577897

13. Gazzaniga, M., Irvy, R. B., Mangun, G. R. (2019). Cognitive Neuroscience: The Biology of the Mind, Fifth Edition. W. W. Norton & Company.

14. Nugent, A. & Furey, M. (2016). Neuroimaging Promises and Caveats: Methodological Issues and Implications for Research in Psychological Disorders and Treatments. The Oxford Handbook of Treatment Processes and Outcomes in Psychology: A Multidisciplinary, Biopsychosocial Approach. https://doi.org/10.1093/oxfordhb/9780199739134.013.41

15. Takano, H. (2018). Cognitive Function and Monoamine Neurotransmission in Schizophrenia: Evidence From Positron Emission Tomography Studies. Frontiers in Psychiatry, 9, 228. https://doi.org/10.3389/fpsyt.2018.00228

16. Nielsen, A.K., Möller, I.R., Wang, Y. et al. (2019). Substrate-induced conformational dynamics of the dopamine transporter. Nature Communications, 10(2714). https://doi.org/10.1038/s41467-019-10449-w

17. Pizzagalli, D. A., Berretta S., Wooten D., et al. (2019). Assessment of Striatal Dopamine Transporter Binding in Individuals With Major Depressive Disorder: In Vivo Positron Emission Tomography and Postmortem Evidence. JAMA Psychiatry, 76(8), 854–861. https://doi.org/10.1001/jamapsychiatry.2019.0801

18. Filkowski, M.M. and Haas, B.W. (2017). Rethinking the Use of Neutral Faces as a Baseline in fMRI Neuroimaging Studies of Axis‐I Psychiatric Disorders. J Neuroimaging, 27(3), 281-291. https://doi.org/10.1111/jon.12403

19. Weyandt, L., Swentosky, A., Gudmundsdottir, B.G. (2013) Neuroimaging and ADHD: fMRI, PET, DTI Findings, and Methodological Limitations. Developmental Neuropsychology, 38(4), 211-225, https://doi.org/10.1080/87565641.2013.783833

20. Amen, D. (2010). Brain SPECT imaging in clinical practice. The American Journal of Psychiatry, 167(9), 1125. https://doi-org.libproxy.temple.edu/10.1176/appi.ajp.2010.10060814