Whether self-esteem measures foretell the onset of mania is less certain. Psychologist Sheri Johnson and her colleagues at the University of Miami have reported a small study in which they found that low self-esteem in remitted patients predicted later depression but not mania.30 However, Jan Scott and Marie Pope in Newcastle, in a similar study, recently measured positive and negative self-esteem separately, finding that the negative component predicted both future depression and future mania.31
The psychology of mania
Very little psychological research has been conducted with actively manic patients, almost certainly because of the difficulty in getting them to sit still for long enough to complete a set of tests. However, Helen Lyon, Mike Startup and I decided to repeat my earlier Stroop experiment with bipolar patients who were attending psychiatric clinics in North Wales.32 Consistent with my earlier findings, we found slowed colour-naming for depression-related but not euphoria-related words in both bipolar-depressed and bipolar-manic patients (see Figure 11.2).
Perhaps we should not have been surprised by this finding. After all, as we saw earlier, euphoria is usually evident only in the earliest stages of manic episodes, and dysphoria predominates in the later stages. Stronger evidence for the manic-defence hypothesis would require the demonstration that manic patients show depressive-like responding on some measures (preferably non-obvious or implicit measures, which might be relatively unaffected by defensive processes) while, at the same time, responding in a way that seems to be incompatible with depression on other measures (preferably obvious or explicit measures, which would be expected to be affected by defensive processes). In fact, this kind of evidence emerged from the same study. In the last chapter, I described the performance of bipolar patients on the Attributional Style Questionnaire (see Figure 10.1, p. 244), noting that patients suffering a manic episode appeared to make fairly normal attributions, in contrast to depressed patients, who made pessimistic (internal, global and stable) attributions for negative events. Figure 11.3 shows the performance of the same patients on Winters and Neale’s Pragmatic Inference Task. Note that the performance of the normal and depressed participants on the PIT was concordant with their performance on the ASQ. The healthy controls showed a self-serving bias on both measures whereas the depressed patients showed a pessimistic bias, again on both measures. In contrast, the performance of the manic patients seemed to vary dramatically between the two tests. Whereas their performance on the ASQ showed a normal self-serving bias, their attributions on the PIT were highly pessimistic. In fact, on the implicit measure, little difference can be seen between their scores and the scores of the depressed patients.
Figure 11.2 Stroop data reported for bipolar-manic, bipolar-depressed and normal participants by Lyon et al. (1999). The columns show interference indices (the extent to which the participants were slowed when attempting to colour-name emotionally salient words, calculated by subtracting the amount of time required to colour-name 50 emotionally neutral words from the time required to colour-name 50 salient words). This measure is shown for both depression-related (for example, dread, rejected, suicide, depressing) and euphoria-related words (for example, wonderful, glorious, joy, jubilant). Note that the two clinical groups show substantial interference for depression-related but not euphoria-related words.
Further evidence relevant to the manic-defence hypothesis can be obtained by looking at measures of self-representation. In the last chapter, I discussed a study of the self-discrepancies of bipolar patients conducted by Kerry Manson, Peter Kinderman and myself, in which we observed an abnormal lack of self–ideal discrepancies in manic patients. This finding is consistent with their apparent grandiosity and suggests that, on the surface at least, they harbour no doubts whatsoever about their attributes and talents, a finding that has been reported by other researchers using less complex measures of self-esteem.33
Evidence of a less positive view of the self in mania emerged from an additional measure included in the study that I carried out with Helen Lyon and Mike Startup. We exploited the self-reference effect, the tendency to recall information that is specifically relevant to the self. As I mentioned in the last chapter, many studies have shown that depressed people, when given a list of negative and positive words, say that more of the negative words describe their own personalities and, more importantly, recall more of the negative words if given a surprise memory test afterwards.34 We therefore administered this kind of test to our participants. Consistently with previous research, our bipolar-depressed participants endorsed and recalled more negative words than our well controls. However, our manic patients endorsed more positive words than negative words but recalled more negative words than positive words (see Figure 11.4).
Figure 11.3 Internality data for manic, depressed and normal participants assessed using the Pragmatic Inference Task. These scores should be contrasted with the ASQ data from the same patients, shown in Figure 10.1, p. 244.
Figure 11.4 Left panel shows the number of positive and negative personality traits endorsed as ‘true’ by manic, depressed and normal participants. On the right is shown how many of the items were recalled by the participants when they were given a surprise recall test. From Lyon, Startup and Bentall (1999).
Clearly, more psychological research must be carried out with manic patients. However, the evidence available to date supports at least the weak version of the manic-defence hypothesis, which assumes that mania arises from patients’ attempts to avoid negative mood.
Becoming Manic
The account of the psychology of mania that I have given so far has focused on the remitted state, and on mania itself. However, a workable theory of mania requires a further vital ingredient: a mechanism that enables patients to pass from the former state to the latter.
We will begin considering this problem by examining two accounts of the processes that lead to mania which, at first sight, seem to be at odds with the manic-defence hypothesis. The first, which might be called the excitability hypothesis (a term which, so far as I know, is not used by any of its proponents), suggests that people who are vulnerable to mania are in some way hyper-excitable. The most carefully worked out version of this idea has been put forward by Richard Depue, of Cornell University in Ithaca, New York,35 who based his model on a neurophysiological theory of personality proposed by British psychologist Jeffrey Gray.36 On the basis of animal studies, Gray argued that two emotional systems could be identified in the brain, the behavioural inhibition system (BIS), which responds to stimuli associated with punishment or frustration, and the behavioural activation system (BAS), which responds to stimuli associated with reward. (Activation of the BIS is associated with negative mood whereas activation of the BAS is associated with positive mood, so the theory maps on to Watson and Tellegren’s two-dimensional model of emotion, which we discussed in Chapter 9.) According to Depue, manic symptoms are a manifestation of a dysregulation of the behavioural activation system, which is why patients are highly sociable, desire excitement, and show high levels of motor activity, as if they are interminably over-reacting to a reward.
Sheri Johnson has reported two findings that fit this theory rather well. First, she demonstrated that hypomanic people score highly on a questionnaire designed to measure the subjective experience of BAS activation (items included, ‘When good things happen to me, it affects me strongly’; ‘When I want something I usually go all-out to get it’; ‘I will often do things for no other reason than that they might be fun’).37 She also reported that patients with a diagnosis of bipolar disorder experience an increased risk of manic symptoms following the achievement of an important life goal.38 This second finding can be interpreted in more ways than one, as Johnson also observed that positive events in general did not lead to an increase in manic symptoms (as might be expected from Depue’s model), and as goal attainment often carries with it the threat of later failure (examples of goal attainments given in Johnson’s paper include gaining admission to graduate
school and obtaining a new job). A more important limitation of Depue’s excitability hypothesis is that it does not seem to explain the close association between depression and mania.
The second apparent alternative to the manic-defence hypothesis implicates sleep disturbances. Many people feel energetic after a night without sleep, only to feel tired after the following night. David Healy has argued that, because of a disorder of the circadian clock, bipolar patients experience a magnification of this effect, so that they rapidly spiral into mania following a period of sleep deprivation. According to this account, the euphoria experienced by manic patients, and their grandiose ideas, are reactions to the high energy levels that they suddenly experience.39 In an intriguing extension of this sleep deprivation hypothesis, Charles Raison and his colleagues at the University of California have recently suggested that old superstitions linking the full moon to insanity may be explained in this way.40 Before the development of modern lighting, the moon was the main source of nocturnal illumination. People living before the development of electricity would therefore have been especially likely to experience sleep difficulties when the moon was full.
It is commonly observed that a period of sleep loss precedes an episode of mania.41 For example, Thomas Wehr and his colleagues at the US National Institute of Mental Health have described a number of cases of people being provoked into a manic episode by a single sleepless night:
For the past 2 years, Mr. B, a 28-year-old bipolar patient, drove 800 miles to spend the Christmas holidays with his family. Impatient to return home and unable to find a motel room, he drove through the night without stopping. By the time of his arrival, Mr. B, who had been euthymic [that is, in a normal mood], became hypomanic. The first year the mania escalated and required hospitalisation; the following year it was possible to attenuate the severity of his mania through early treatment with adjunctive neuroleptic medications.42
Nocturnal physiological measurements have confirmed that loss of sleep is a characteristic of mania.43 Several studies,44 including one reported by Thomas Wehr’s research group,45 have also noted correlations between sleep duration and manic symptoms the following day. Taking this idea further, Susan Malkoff-Schwartz, Ellen Frank and their colleagues at the University of Pittsburgh School of Medicine have investigated whether manic episodes are typically preceded by what they term social rhythm disruption (SRD) events.46 In their most extreme form, these events (for example, starting night work, the arrival of a new baby) completely disrupt sleep whereas, in their milder forms (for example, travelling between time zones, staying up to revise for an examination), the sleep-wake cycle is only partially disrupted. When bipolar and unipolar patients were compared, it was found that about two thirds of patients who became manic had experienced an SRD event in the eight weeks preceding their mania, whereas no such effect was found for episodes of depression, either in the bipolar patients or in the unipolar patients.
It is important to remember that SRD events may be stressful for reasons other than their effect on sleep. For this reason, the best way to establish whether sleep deprivation per se leads to mania is to ask patients with a history of the disorder to volunteer to go without sleep. In general, the results of experiments of this sort suggest that only a minority of patients may be as sensitive to sleep deprivation as Thomas Wehr’s case studies suggest. In perhaps the best study of this kind, reported by Christina Colombo, Francesco Benedetti and their colleagues at the University of Milan, over 200 bipolar-depressed patients underwent three nights of total sleep deprivation, interspersed with nights of normal sleep (it was hoped that this would benefit their depression). About 5 per cent of the patients became manic and a similar number became hypomanic. Only four of the patients became ill enough to require drug treatment.47 Overall, then, the available evidence suggests that disruption to sleep can precipitate mania, but that this effect is not inevitable.
The excitability hypothesis and the sleep deprivation hypothesis should not be regarded as mutually exclusive. As we have seen, both are to some extent supported by research. However, unlike the manic-defence hypothesis, both seem to assume that the events leading up to mania are largely beyond patients’ own control. We can incorporate both mechanisms within the framework of the manic-defence hypothesis if we assume that patients, faced with the threat of negative mood, sometimes choose to pursue activities that either cause them intense excitement or lead to sleep loss.
To see how this might happen we must return to the reaction-maintenance principle and some of the methods that people use to cope with negative mood, which we considered in the last chapter (see p. 262). Recall that, according to studies carried out by Susan Nolen-Hoeksema, people tend to react to being depressed in one of four ways. Some people ruminate about their feelings, others launch into attempts to solve the problems that they believe have led to their depression, some try to distract themselves and a few indulge in dangerous activities. Nolen-Hoeksema’s research established that rumination leads to periods of depression that are deeper and more enduring.48
In a recent study, my Ph.D. student Justin Thomas administered Nolen-Hoeksema’s questionnaire to a large group of students, who also completed questionnaire measures of depression and hypomanic traits.49 Consistent with Nolen-Hoeksema’s findings, we observed a strong correlation between rumination and depression. However, although we also found a weak association between rumination and hypomania, hypomania was more strongly predicted by the distraction and dangerous activities coping styles.
I think that anyone who has spent any time talking to bipolar patients will recognize these coping strategies (which include working around the clock, intense socializing, indulging in alcohol and drugs, and generally seeking novel experiences) from the patients’ own accounts. Unfortunately, at the time of writing, only one study has systematically asked bipolar patients about their coping styles. Dominic Lam and Grace Wong at the Institute of Psychiatry in London divided patients into those who seemed to cope well with their symptoms, and who seemed to be able to control them, and those who did not. Many of the strategies reported by the poor copers (‘continue to move about and take on more tasks’, ‘go out and spend money’, and ‘find more to do to fill out the extra minutes of the day’) obviously involved some kind of distraction.50
Towards a Psychology of Mania
Less psychological research has been conducted into mania than into any other major psychiatric complaint. In this chapter, I have tried to weave together an account of the pathway from incipient negative mood to manic episodes that makes best sense of the limited evidence. This pathway, from distraction and risk-taking, to excitement, sleep loss and eventually mania, is shown in Figure 11.5, which is a modification of the final model of depression that we arrived at in the last chapter.
It is important to note that this tentative model seems to explain better some aspects of mania than others. Earlier, I explained that mania consists of five separable components: dysphoria, increased hedonic functioning (euphoria), psychomotor pressure, psychosis and irritability. Dysphoria is accounted for by those components of the model that are shared with our earlier model of depression. It is possible that psychomotor pressure can also be accommodated within this framework if we assume that it is a consequence of hyper-excitability. Irritability may arise from the inevitable conflict with other people that manic behaviour provokes. It is possible that euphoria can be partially accounted for also, especially if feelings of excitement and boundless energy lead the individual to believe that he is capable of extraordinary feats, thereby affecting current beliefs about the self, as indicated by the dotted arrow on the diagram. The psychotic elements of the condition, however, are not so easily explained. In later chapters, we will see that these symptoms require different models.
Figure 11.5 The dysphoria model, modified to include mania (unique features shaded).
12
Abnormal Attitudes
I am absolutely certain that in this regard I command exp
eriences which – when generally acknowledged as valid – will act fruitfully to the highest possible degree among the rest of mankind… You like other people may be inclined at first to see nothing but a pathological offspring of my imagination in this; but I have an almost overwhelming amount of proof of its correctness.
Daniel Schreber1
It remains for the future to decide whether there is more delusion in my theory than I should like to admit, or whether there is more truth in Schreber’s delusion than other people are yet prepared to believe.
Madness Explained Page 34
