Studies indicate that between 6% and 41% of patients with psychogenic symptoms display an apparent lack of concern about their symptoms (9, 10, 11, 12), suggesting that it is far from universal in these conditions. This wide variation in prevalence estimates probably reflects the different criteria used to select patients in these studies, and the use of subjective physician ratings of la belle indifférence without a clear operational definition of this phenomenon. Indeed, psychophysiologic research shows that physicians are actually quite poor at judging patients’ attitude to their symptoms. Lader and Sartorius (13), for example, found that conversion disorder patients who were identified by their physicians as indifferent to their symptoms showed more physiologic symptoms of anxiety than did organic controls, in line with patients’ own self-reports; a similar pattern of findings was obtained by Meares and Horvath (14) in patients with chronic conversion symptoms. Rice and Greenfield (15) also found that patients who were rated as indifferent to their symptoms showed more physiologic reactivity to emotional and illness stimuli than did organic controls, and there were few psychophysiologic differences between the two groups at rest. Thus, it may be that physician ratings of la belle indifférence reflect a confirmational bias resulting from a traditional psychiatric training in the features of psychogenic illness, rather than an accurate appraisal of patient behavior.

Other studies have shown that an apparent lack of concern about symptoms is also observed in patients with general medical conditions (16) and may be just as common in organic controls as in psychogenic illness (9,11). This clearly undermines the use of la belle indifférence as a diagnostic indicator in unexplained illness (17) and casts doubt on the theoretical significance of the phenomenon for these conditions.

With the exception of case reports, only two published studies could be found that have empirically assessed
whether unexplained symptoms solve an emotional conflict being faced by the sufferer (9,18). In a blinded, prospective study of this issue (9), it was found that conversion symptoms were significantly more likely to be rated as offering a solution to an ongoing conflict than organic symptoms; in addition, patients with conversion symptoms were more likely to report a history of previous unexplained symptoms occurring in response to psychosocial stressors than were organic controls. Although these findings seem encouraging, no information about how these ratings were made was included in the published report, making it difficult to confirm the validity of this study or to replicate its findings. A similar criticism can also be leveled at the only other study in this area (18), which apparently found evidence of symptoms serving a “defensive and expressive function” in patients with psychogenic illness. This study was also limited by its retrospective case-note design, an approach that would be particularly sensitive to physician biases in record-keeping and which might explain any recorded differences between the groups. The conflict resolution hypothesis is clearly difficult to assess, however, as demonstrated by the paucity of studies in this area.

Other studies have attempted to investigate whether unexplained symptoms are precipitated by psychosocial stressors. Evidence suggests that many unexplained symptoms develop at times of stress (9,11,12,18, 19, 20), although clinically it is often difficult to establish a clear etiologic link between such stressors and symptom onset (21,22). This is particularly problematic given that psychosocial stress is reported by a similar proportion of patients with established organic conditions (11,19,23). Although case-control studies of this sort do not rule out the possibility that psychosocial stress is an important factor in the development of unexplained symptoms, they clearly invalidate its role as a differential indicator in the diagnosis of these conditions. It remains a challenge for future researchers to develop meaningful but ethical methods for assessing the causal role of stress in the development of psychogenic illness.

The idea that unexplained symptoms are maintained by external benefits or secondary gains has been addressed in a number of studies. Although most studies have concluded that such gains are important in many cases of unexplained illness, the same appears to be true for general medical conditions (9,12,16,18,24,25). This pattern of findings appears to be more consistent with a learning theory account of illness behavior per se rather than a psychodynamic model of unexplained illness.

On the face of it, the available empirical evidence concerning la belle indifférence, conflict resolution, psychosocial precipitation, and secondary gain in psychogenic illness is limited and raises important questions about the validity of conversion theory. Most clinicians working in this area have, however, encountered cases of unexplained illness where the role of psychodynamic and psychosocial factors seems unquestionable, and it remains an important task to address these factors in the assessment and treatment of these patients. Nevertheless, future researchers should aim to establish how, and to what extent, these factors are specific to psychogenic illness as opposed to emotional disturbance more generally.

A substantial number of studies have investigated the occurrence of childhood abuse in patients with unexplained symptoms, providing probably the strongest support for the conversion concept. Table 16.1 summarizes cross-sectional studies that have directly compared the retrospective abuse histories of patients with psychogenic symptoms and those without unexplained complaints. All of the studies in this area have addressed the early Freudian hypothesis that childhood sexual abuse is an important etiologic factor in the development of psychogenic illness; a proportion of these studies have also addressed early physical and emotional abuse. Abuse in a range of unexplained syndromes has been considered, with the majority of studies addressing patients with chronic pelvic pain, functional gastrointestinal complaints, or nonepileptic attacks. With very few exceptions, these studies have shown that a history of childhood sexual abuse is reported significantly more often by patients with unexplained symptoms than by organic controls. A smaller proportion of studies have found that physical abuse histories are also more common in patients with psychogenic conditions, although the smaller number of studies in this area makes it more difficult to confirm the reliability of this difference. Other studies have reported a relationship between abuse history and the occurrence of unexplained symptoms, without reporting a direct comparison of patients with and without unexplained symptoms (50, 51, 52, 53, 54, 55, 56, 57, 58).

As most studies in this area have included an organic illness rather than a psychiatric control group, it is unclear whether the occurrence of childhood victimization is more common in psychogenic illness than in other emotional disorders. Moreover, many of these studies have measured abuse using unstandardized and poorly described clinical interviews, which are of questionable reliability and validity, or questionnaires that tend to underestimate the occurrence of abuse (59). Also problematic is the dichotomous coding of abuse as either present or absent, which is employed in most studies in this area. Such a practice precludes analyses concerning the role of abuse severity, frequency, and chronicity in the development of unexplained symptoms (60). Furthermore, although sexual and physical victimization are commonly reported by patients with unexplained symptoms, much less is known about exposure to
early emotional abuse in these patients. One study in this area used a single question to assess psychological abuse (“When you were a child did an older person insult, humiliate, or try to make you feel guilty?”) (36), which is unlikely to provide an adequate measure of this phenomenon. Exposure to insults or criticism from an older sibling is unlikely to be as damaging as similar treatment from a parent, for example, a difference that is obscured by the single-question approach. More meaningful information has been obtained in two studies that have used questionnaire measures of perceived parental care (33,34), both finding that emotionally rejecting parents may contribute to the later development of unexplained symptoms. Nevertheless, further information about the role of emotional abuse in childhood is clearly required to assess which aspects of early victimization are most relevant to unexplained illness. It is possible, for example, that the relatively high prevalence of childhood sexual trauma reflects a pathogenic early environment more generally for patients with unexplained illness rather than the etiologic importance of sexual abuse per se (61). Other dimensions of childhood experience may also be important in unexplained illness, including early losses and parental neglect.