Somatic Symptom Disorder, Medically Unexplained Symptoms, Somatoform Disorders, Functional Neurological Disorder: How DSM-5 Got It Wrong

“Truth is the daughter of time, not of authority.”

Francis Bacon

One in five individuals seeking primary medical care in any part of the world complains of physical symptoms which, after very thorough assessment and investigation, cannot be found to be caused by any identifiable general medical condition.^1,2 The way in which these patients are understood, and their conditions classified, shapes the treatment they are offered, the course of their symptoms, and their future quality of life.

There is a very long history of medicine and psychiatry working to develop an understanding of conditions characterized by “medically unexplained symptoms” (MUS).^3,4,5 Largely by virtue of historical medical discipline subdivision, these symptoms have been separated into two groups: those suggestive of neurological disease (such as weakness, convulsions, or disturbance of vision) and those suggestive of nonneurological disease (such as gastrointestinal disturbance, pain, or fatigue). There are theoretical and empirical reasons for suggesting that the two groups be collapsed into one, as many patients with these conditions develop both neurological and nonneurological symptoms.^6,7

Unexplained neurological symptoms retain classification that is partly true to their nature under the American Psychiatric Association’s DSM-5.⁸ Conversion disorder remains, gaining the new co-title functional neurological symptom disorder. Here, the clinician still has to judge that symptoms are “medically unexplained” in deeming that “clinical findings provide evidence of incompatibility between the symptom and recognized neurological or medical conditions.” ⁸ There is, however, no longer the need to determine that “psychological factors are judged to be associated” (as was the case in DSM-IV), a change that we believe to be in essence incorrect.

Unexplained nonneurological symptoms are abandoned as a distinct group by DSM-5 and are dealt with in a new and fundamentally different fashion. They are “lumped in” with conditions where demonstrable general medical conditions cause excessive emotion, behavior, or cognitions in the patient. This change represents an important and unfortunate error, one that is destined to mislead clinicians and researchers in the field for the foreseeable future. Under DSM-5, most patients with this group of symptoms will not receive the type of clinical approach that gives them the best chance of positive outcomes, and our understanding of the field will not advance as well as it could.

DSM-IV Needed Improvement

DSM-III had done away with the term “hysteria” in relation to these conditions, a welcome step.⁹ DSM-IV, published in 1994, described somatoform disorders as a group of syndromes that had in common the “presence of physical symptoms that suggest a general medical condition” that were “not fully explained by a general medical condition.”¹⁰ Despite the face validity and clinical experience supporting this group of disorders, the way DSM-IV dealt with somatoform conditions muddied the waters. For example, there were significant threshold and sensitivity problems with the major subdivisions. In a study looking at 119 patients in family practitioners’ waiting rooms, somatization disorder criteria were met in just one patient, while 94 met criteria for undifferentiated somatoform disorder.¹¹ Thus, criteria were too tight for one diagnostic entity and too loose for the other, with neither serving as a useful instrument for separating the majority of individuals with significant somatoform symptoms from the general population. There was also an unclear basis for separating the syndromes, as similar sets of symptoms could exist across many of the defined diagnoses. DSM-IV did not promote a valid or clinically useful understanding of somatoform concepts, and it discouraged clinicians from identifying these disorders. A 2009 survey of physicians revealed that DSM-IV definitions of somatoform disorders and its subgroups were unclear to many.¹² Not surprisingly, clinicians simply didn’t use the classification system and, although known to have high prevalence, these disorders were severely underdiagnosed.¹³ There was clearly room for improvement.

DSM-5 Abandons Nonneurological MUS

DSM-5 “reconceptualized” somatoform disorders,¹⁴ changing the category name to somatic symptom and related disorders, a group of disorders all characterized by the presence of physical symptoms. The most profound change was the creation of a newly defined core disorder, somatic symptom disorder (SSD), a syndrome characterized by persistent and clinically significant somatic complaints accompanied by excessive and disproportionate health-related thoughts, feelings, and behaviors regarding the physical symptoms. Patients who met DSM-IV criteria for a nonneurological somatoform condition would almost all currently meet SSD criteria. Further, SSD now includes most individuals previously diagnosed with hypochondriasis.⁸ In addition, and crucially, SSD now includes individuals with demonstrable physical disease with “excessive thoughts, feelings, or behaviors” about their physical symptoms. Thus, SSD groups together patients with demonstrable peripheral pathology and those with no such pathology.

The DSM-5 Working Group, 10 specialists in the field, worked on the changes from 2000 until publication of DSM-5 in 2013. Through these years, the classification of these disorders was much debated in the literature. The main areas of discussion were the wisdom of using MUS as part of the criteria, the validity of subdivisions, and the nomenclature. Some argued for a fundamental change in the category^15,16,17,18; others for essentially retaining the “somatoform” concept while making other refinements.^19,20,21,22 The Working Group shared their reasons for the fundamental changes, both during the development process,²³ and on publication of DSM-5.¹⁴ In response to appeals for input, the authors shared a critique of the intended changes in a letter to the Working Group in 2011.²⁴

The DSM-5 Rationale for the Eradication of “MUS” as a Central Idea for Nonneurological Symptoms and Signs

Let us consider the reasons given for this fundamental change, and some counterarguments:

The reliability of assessing whether or not there is an explanation for somatic symptoms is notoriously poor.¹⁴

On the contrary, after a thorough clinical assessment, including special investigations as necessary, a clinician can draw the conclusion that symptoms are medically unexplained with a similar degree of certainty as one would have in making the diagnosis in many general medical conditions and other psychiatric diagnoses.^25,26 Furthermore, any limitations in diagnostic precision need to be weighed against the risks of not making a diagnosis where one may reasonably be made, and the consequent risk of potentially treatable syndromes going untreated.

Some MUS are not so much “Unexplained” as “Unexamined.”¹⁴

This is a criticism of incomplete assessment, not specific to MUS. This same criticism could be made regarding approaches to any other psychiatric or general medical condition. A thorough assessment is always essential.

A diagnosis built upon a foundation of MUS is perilous because it reinforces mind/body dualism.¹⁴

It is thereby argued that to say that physical symptoms are “medically unexplained” may lead some patients and clinicians to conclude that the mind and body are distinct and separable, and, presumably, that the symptoms may be due to some “nonphysical” process. This dualistic thinking is incorrect. All neuropsychiatric conditions are mediated by brain function. All aspects of somatoform conditions are products of complex brain function and dysfunction, the specific nature of which we do not yet understand. It is vital that we address the challenge presented by any tendency to dualistic thinking directly. The answer is not to abandon the concept of somatoform conditions out of fear that some may misunderstand but rather to wrestle with the challenge, to engage the field in discussion, and to educate clinicians and patients about the nature of these conditions. They do indeed involve complex interactions between what most of us would consider to be the mind and the body. But all of these interactions are manifestations of various levels of physiological function and dysfunction. Indeed, a thorough understanding of the somatoform process serves as an integrative force and proves supportive to arguments against dualism.

…the MUS approach is not well accepted by patients who feel that MUS implies that their symptoms are inauthentic and “all in your head.” This is a poor basis for a therapeutic alliance with patients who are suffering distressing somatic complaints.¹⁴

A similar challenge exists when helping all individuals suffering other psychiatric and psychological conditions. This challenge is partly related to the stigma regarding mental illness that persists in our society and partly due to the complex emotions (including fear, puzzlement, guilt, anger) that an individual may experience when they have the thought that something may be awry with their mind. The only way forward is to understand these conditions for what they are and, on the foundation of thorough assessment and empathic therapeutic alliance, to help those suffering from them. A clinician should, in a supportive fashion, share the truth about the condition with the patient. These disorders are, indeed, all mediated by brain function and dysfunction, so they are indeed literally based in large part “in one’s head”; however, from a figurative perspective, they are not “imagined,” so this distinction must be made clear for the patient. The unconscious and involuntary nature of the process must be emphasized, and the nature of the condition plainly described. To avoid the truth because it may be unpalatable does not serve patients well at all, as optimal management is predicated on a valid understanding of the disorder. It is also poor science.

It bases a diagnosis on a negative….²³

…a medical diagnosis does not usually define a disorder based on the absence of something.¹⁴

Clinicians go through a logical process of elimination whenever they consider the differential diagnosis for any set of symptoms and signs. Exclusion of known diseases is part of every diagnostic process. If, at the end of such a process, the clinical picture suggests MUS, then they should be called that. That judgment is not qualitatively very different from what we are doing with many other disease entities. It is further implied that making the diagnosis of a MUS is somehow a “negative” step in a broader sense, in that clinicians and patients see it as the “taking away” of something rather than a “positive” diagnosis. Many diagnoses in general medicine embrace the fact that there are unknown components to the pathogenesis. For instance, after ruling out hypertension secondary to renal, endocrine, or other general medical conditions, the hypertension is labeled “idiopathic.” Patients don’t complain that they have lost something or had something “taken away” in that process. We would argue that making a somatoform disorder diagnosis is as proactive a diagnostic step as the making of any other diagnosis, and we believe that we in the field should work to actively frame it as such. The patient should be informed that their syndrome is the result of complex brain processes, not demonstrable brain or peripheral tissue pathology. This has positive implications for treatment, and we should help the patient understand that. Indeed, we would suggest that it is the clinician’s duty to make such judgments and to offer to treat the patient accordingly. It is only with such a judgment call that the disorder can be understood, and appropriate management instituted.

The DSM-5 Rationale for the Change in Nomenclature From “Somatoform Disorders” to “Somatic Symptom and Related Disorders”

With the move away from MUS, it is not surprising that DSM-5 moved away from the “somatoform” label. Nonetheless, an additional set of reasons was offered for removing the term, namely, it “has been difficult to understand,” it was “a neologism, blending Latin and Greek,” and because it was “often confused with somatization disorder.” ¹⁴ Various terms have been suggested as the best label to refer to the group of conditions based on MUS, including “functional,” “psychogenic,” “psychosomatic,” “psychophysiological,” “bodily distress,” “somatic,” “somatization,” and “somatoform.” “Somatoform” is precisely the word that most elegantly captures the essence of these conditions. It suggests a condition that is “in the form of” the body while simultaneously implying that there is more to it than that. The fact that “somatoform” is a Greek–Latin hybrid puts it in the good company of many other hybrid medical terms including “neuroscience,” “neurotransmitter,” and “hypertension.”²⁷ “Somatoform” has no more potential for being confused with “somatization” than does “somatic symptom.” This is a matter of educating clinicians and patients and being clear and consistent in our use of the term. “Somatic symptom disorder” is arguably a confusing term because on the face of it, it could easily be misunderstood to describe any disorder that involves a physical (“somatic”) symptom. In this sense, it is potentially hobbled by its breadth and lack of inherent specificity, as was “pain disorder” in DSM-IV.

Core Concerns and Implications

Overall, DSM-5 argues that the core problems with DSM-IV somatoform disorders were based on confusing nomenclature, and numerous untenable problems relating to identifying MUS. However, the problem with DSM-IV was not the “somatoform” name, nor the concept itself, but rather the fact that the classification system did not promote a clear understanding of somatoform conditions, and the main disorders did not identify and separate out patients with somatoform disorders in a clinically useful or valid fashion. Thus, the system was hardly used, and under it, the majority of individuals with somatoform disorders were not identified or assisted.

Empirical evidence and clinical experience still supports the idea of somatoform disorders as a valid diagnostic entity, embracing both neurological and nonneurological symptom types. Granted, a small percentage of these patients may be suffering rare or as yet unidentified medical conditions, and all patients with medically unexplained physical symptoms require very thorough ongoing monitoring for the possible emergence of physical disease. But there is strong evidence that the vast majority do not have peripheral tissue pathology but are suffering the results of direct or indirect expression of underlying emotional distress, through the process referred to as “somatization.” Evidence for the validity of the somatization process includes but is not limited to: the nonphysiological patterns of symptoms or signs (they do not follow the patterns seen in well-described diseases); the patterns of symptoms based on patient beliefs; the atypical course of the condition; the high rates of association of these symptoms with overt emotional distress; the many persuasive clinical examples of “conversion” of mental distress into physical syndromes; the longitudinal natural history of these conditions (with only very small percentages revealing some other causative underlying tissue pathology over time); and the positive responses to psychological and psychiatric therapies when a somatization model is assumed.

The crucial flaw in the approach of DSM-5 is around the clinician’s judgment regarding MUS. DSM-5 and others are advocating for “acceptance of etiological neutrality about those symptoms that are not clearly associated with a general medical condition.” ¹⁷ One can see the attraction of the apparent logical rigor and prudence suggested by the DSM-5 approach: How can we ever be sure that there is not some as-yet-unidentified medical condition causing the symptoms and signs? And, as we can’t be certain, let us rather withhold judgment. But diagnoses in complex neurobehavioral conditions are seldom made with certainty. In making judgment calls about MUS and the nature of somatoform disorders, the clinician has to weigh the risk of error in diagnosis against the risk of not making a diagnosis where one can reasonably be made. The latter risks include potentially treatable illnesses not being treated.

Figure.

Diagram demonstrating the differences between DSM-5 somatic symptom disorder and the recommended somatoform disorders classification. The authors maintain that it is valid and advisable to make the division between “nonneurological psychogenic symptoms and signs” and “general medical conditions with excessive affect, behaviors, (and/or) cognitions.”

There is a more robust and straightforward approach to understanding and classifying these conditions, as some have suggested.^20,21,28,29 The idea and label of “somatoform disorders” should be resurrected. A somatoform disorder should be diagnosed in any clinical situation where physical symptoms and signs are judged, after thorough assessment, to be the result of underlying emotional distress rather than primary physical disease. The disorder can then be further characterized by listing the psychogenic physical symptoms and signs, and giving an account of the nature of the underlying dysphoria (be it a psychiatric syndrome and/or related to psychological factors and stressors).²⁹

This understanding naturally leads to the planning of effective management for the majority of patients, an approach that addresses physical MUS as well as underlying dysphoria. Management is readily customized to the specific needs of each patient and can also take advantage of the specific skill set of the treating clinician.²⁹ Using this approach, prognosis is often surprisingly good and largely depends on the prognosis of the cause of the underlying dysphoria.³⁰