Experience Sampling
In spite of the agreement about the importance of environmental factors in anxiety, panic, and depression, we know little about the diurnal and temporal aspects of affect and anxiety, and their interaction in daily life. The ambulatory monitoring of patients is ideally suited to clarify these issues. Recent studies employing such methods have compared the shape of circadian rhythms of de­pressed and normal persons and demonstrated that variations in affective symp­toms are universal. In these studies the difference between well and ill groups is ore a difference of degree rather than kind. In Maastricht we use the ES tech-nique, which employs random, repeated self-report measures. Developed in the for seventies and early eighties by the author and others (4,11), the ES method has today taken its place with other substantiated assessment methods in the social and psychological sciences (8,26). ES uses a "beeper" to randomly, within 90-min periods, signal subjects to fill out self-reports 10 times per day for 1 week in the context of a person’s natural environment. This method of interrupting experience provides a representative sample of a person’s mental state. In this way we have examined the temporal characteristics, situational dependence, and time budget aspects of mental state, as well as the variability and stability of symptoms over time in a number of disorders. When the beeper signals, the respondent writes down information concerning his or her momentary situation and psychological condition. This record becomes the basic data of ES. The questionnaire form is designed to take no more than a few minutes to complete. Respondents carry all the forms in booklets. The forms used in this study had the objective of seeking a comprehensive coverage of external and internal situations as well as functioning as a repeated diagnostic instrument. Items are included that assess the time at which the form is filled out, whom the subject is with, where he or she is, and what activities he or she is involved in. Nine open-ended questions assess the social and physical context as well as the content of thoughts and activities. Further assessments of mental state are made using 25 Likert-type scales based on the mental status exam and grouped in modules for evaluating thought (17), affect (18), the presence of somatic and psychological complaints, and the level of motivation. These mod­ules have been validated by factor analysis. Added to these questions about the general properties of mental state, a pathology subsection has been added eliciting specific questions about anxiety. These questions were derived from cross-sec­tional diagnostic descriptions of anxiety, such as those offered by Spielberger et al. (27), DSM-III (1), and Marks and Mathews (23). Data were gathered at two levels. There were one time, cross-sectional measures gathered at the start of the study. These include illness severity ratings, diagnosis, demographic data, and other psychological assessments such as Zung and Spiel­berger rating scales. These diagnostic measures are in part incorporated into the Psychopathology subsection of the self-report forms. The second type of data gathered was the within-day repeated self-report measures previously described, filled in 10 times a day for 6 days. Skepticism about the application of this approach with ill persons generally arises around four central issues: (a) the adequacy of the sampling of ill persons, (b) the adequacy of the sampling of experience, (c) the experimental effects of the method, and (d) the construct validity of the self-reports. Full discussion of these methodological issues, as well as of the validity and reliability of the ap­proach, has recently been published (5,16). Most surprisingly, compliance across sampled disordered groups, ranging from pain to schizophrenia, as well as across developmental stages, from adolescents to the elderly, has been generally high and sometimes greater than that achieved with normal controls. In order to assure the acquisition of moment-to-moment information, reports are discarded if the beep is not responded to within 10 minutes. In general, the method has validity. The time allocation questions asking who, what, or where a person is favorably match reports from larger time-budget studies (29). In studies of distinct diagnostic disorders, ES data have discriminated diagnostic groups from one another and from normals (6). Reactivity and practice effects have also proved less problematic than expected, as shown by the very small differences in split week means from the end to the beginning of the week. Further pre- and posttest differences in subject self-consciousness measures, ap­plied to determine if psychological reactivity to the ES procedure takes place, have demonstrated negligible effects of the procedure (16). The research alliance is the key to increasing and maintaining compliance as well as the validity and reliability of the research. In this sense, ES is not a survey instrument, but a labor-intensive research method that requires the full coop­eration of the subject. In summary, past research suggests that the beeper method can be used with a wide array of people and that reports can be obtained on nearly all their experience. Studies also suggest that the effects of the method on what it is measuring are minimal and that the self-reports have construct validity. Studies have further shown that it can be used with adolescents, children as young as the sixth grade, and even severely psychotic individuals. It is useful to consider the method in light of traditional paper and pencil and interview meth­ods. It makes use of psychometric research instruments as they do, but unlike them, it does not rely on people’s ability to recollect and reconstruct their past experience. The ES method is capable of sensitively capturing mood and anxiety fluctu­ations over time and in different contexts in clinical populations. For example, previous studies have shown depression to co-occur with a variety of disorders when measured in daily life, such as bulimia and anorexia (7,19), and multiple personalities (22). Further, behavioral aspects of disorders have been elucidated and discussed for anxiety (12) and in schizophrenics (6,7-11). The assessment of affect by ES has correlated significantly with aspects of psychopathology, mo­tivation, and thought while still retaining its independent factor structure. For the study reported here, this means that affect and aspects of anxiety are differ-entiable by a number of the ES variables and, therefore, the study of their interplay in daily life could be successfully explored using the ES method. We thus aim to illustrate variations in time allocation, self-report of mental states, and temporal properties of anxious and depressed subjects. (more…)

Diurnal patterns in anxiety and depression scores in the highly and medium depressed groups were explored applying a linear regression technique to the entire sequence of beeps per subject and collapsed over subjects per group. For the moderately depressed group, no daily pattern was found for anxiety or depression symptoms, but for the highly depressed group, a within-day significant fluctuation was found for depression (t = -2.06, p < 0.04). This significant effect was produced by one subject who demonstrated a strong late in the day mood upswing (t = -3.93, p < 0.001). This finding was supported by a trend in the other four subjects. Interday differences per subject were also detected, but at the group level this effect disappeared. We next asked if we could separate the two groups on other temporal dimen­sions such as the recovery rate or decay time from high anxiety or low mood states. To demonstrate this, we tracked the flow of the symptom within the day from a point minimally one SD below that individual’s mean mental state on anxiety or depression. The "symptom event" was followed over subsequent sig­nals that day to manufacture a recovery or decay curve of intense emotional anxiety episodes for each group. Figures 2 to 4 show these recovery curves in the course of the day in relation to the group mean. Figure 2 depicts temporal pattern of recovery from anxiety states in medium and highly depressed subjects. On this graph we see that the high depression group recovers from anxiety more slowly. We also note that at the end of the day, a rebound or a diurnal decrease in anxiety is reported by the high depression group. This finding is greater than may be expected from the regression to the mean shown by the moderately depressed group. In Fig. 3, the recovery from a depressive episode is depicted for both high and medium groups. In contrast to anxiety, a more rapid recovery is noted in both groups as well as a more erratic recovery pattern that does not distinguish the groups from each other. Since the two groups are differentiated by the degree of depression scores, it is interesting to note that the temporal moment-to-moment recovery from a depressive episode is not altered by increased severity. On the other hand, it is anxiety that recovers more slowly in the highly depressed group in addition to showing a late-day rebound. This suggests at least an interplay in daily life between the effects of anxiety and affect. Affect perhaps predisposes the individual to problems in the experience of and coping with anxiety.
(more…)

While frequency distributions clearly differentiated the two groups on ecological and time allocation variables, the psychological reactions to similar situations, such as being alone or being at home, did not differentiate the groups significantly. A difference was found for the entire group, however, on thought congruence, a self-rating of thoughts, as focused on the activity that one is involved in. This measure differentiated both groups from other ES samples of mental disorders. For example, in contrast with schizophrenics, who are markedly more incon-gruent when they are alone (17), both anxiety groups were less congruent when they were away from home. The anxiety group then, interestingly, reports more disorganization when they are with people, whereas the schizophrenics report more unfocused thought when they are alone than when they are with company, In summary, for self-ratings of mental state, daily mean measures of affect, mo­tivation, and reactions to situations did not differentiate the medium and high groups.

Department of Social Psychiatry, University of Limburg, 6200 MD Maastricht, The Netherlands

(more…)

Between these subtly defined groups of high anxiety with medium depression and high anxiety with high depression as defined by Zung score, characteristic differences in daily activities and temporal aspects of anxiety and depression may be found using ES. This is particularly true for data at the behavioral level representing activities selected by the subjects in each group, and at the temporal level in terms of each group’s differential capacity to recover and rebound from high anxiety and depressed mood episodes. ES separated the groups on self-report scores of thought content, psychopa-thology, and time allocation of activities. Mean self-report scores of moment-to-moment mental states did not separate the groups nor did the situations in which they occurred. This finding suggests that anxious depression is relatively context independent. Further, behavioral changes seem a stronger indicator of changes in mental state than average scores of changes in the self-perception of state. Although much of the data presenting temporal relationships in this chapter are exploratory, we may tentatively conclude from the two findings, the slow decay of anxiety in the high depression group and the predictive trend of mood for anxiety, that depression in a day-to-day sense renders an individual more sensitive to anxiety. The basic pathological process in the phenomenology of anxious depression may therefore be depression. Although it may be a mea­surement artifact inherent in the Zung or Spielberger ratings, it is interesting to note in the context of this volume that in the sequential sample of 15 anxiety admissions reported here, none had a low depression score. The findings presented, when pursued further and in large samples, could lead us to a more comprehensive description of the phenomena of depression and anxiety in temporal and contextual terms. Such findings may not only be useful in solving current diagnostic controversies, but may also be useful in providing new avenues for treatment based on the frequency of the occurrence of the actual illness phenomenon, its context, and its temporal recovery properties. Such in-formation should be of significance to the clinician and the psychopharmacolo-gist alike.

REFERENCES (more…)

This report is based on a group of patients who presented seeking relief from a primary complaint of anxiety in two ambulatory settings for the treatment of anxiety. They represent a convenience sample gathered from sequential admissions,
(more…)