This Article Abstract Full Text (PDF) Alert me when this article is cited Alert me if a correction is posted Services Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Vedhara, K. Articles by Lightman, S. Search for Related Content PubMed PubMed Citation Articles by Vedhara, K. Articles by Lightman, S. Related Collections Stress and Coping

The Role of Stressors and Psychosocial Variables in the Stress Process: A Study of Chronic Caregiver Stress

From the Department of Medicine (K.V., N.S., S.L.), Dorothy Crowfoot Hodgkin Laboratories, and the Department of Experimental Psychology (K.V., S.A.), University of Bristol, Clifton, Bristol, United Kingdom.

Address reprint requests to: Kav Vedhara, PhD, MRC HSRC, Department of Social Medicine, University of Bristol, Canynge Hall, Clifton, Bristol, BS8 2PR, United Kingdom. Email: K.Vedhara @bris.ac.uk

	ABSTRACT

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION CONCLUSIONS ACKNOWLEDGMENTS REFERENCES

 
OBJECTIVE: An investigation was conducted 1) to examine the relative importance of stressor types (ie, daily hassles, caregiving-specific stressors, and life events) on the stress response, 2) to assess the stability of relationships between psychosocial variables and stress over a 6-month period, and 3) to explore how the nature and magnitude of the contributions made by stressors and psychosocial factors to the stress process varied according to the qualitative characteristics of the stress response (ie, anxiety, depression, and stress).

METHODS: Fifty spousal caregivers of patients with dementia were recruited and asked to participate in a detailed psychosocial evaluation at 3-month intervals; the evaluation involved measurement of stressor frequency, psychosocial variables, and indices of the stress response (ie, anxiety, depression, and stress).

RESULTS: The data revealed that the effects of stressors and psychosocial factors on the stress response were considerable (accounting for 49–63% of the variance in stress response measures). Furthermore, there was some evidence of stability in the effects of the stressor and mediator variables on the stress response. Specifically, the contributions of life events and caregiver difficulties were largely consistent at both 3 and 6 months, and the psychosocial factor of "reactive coping and self-appraisal" influenced all three stress response indices at both 3 and 6 months.

CONCLUSIONS: There is some evidence of stability in the effects of stressors and psychosocial variables on the stress process over a 6-month period. However, it would also seem that the nature of the stress process differs according to the qualitative characteristics of the stress response.

Key Words: chronic stress • anxiety • life events • hassles • psychosocial variables • depression

Abbreviations: HIV = human immunodeficiency virus.

	INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION CONCLUSIONS ACKNOWLEDGMENTS REFERENCES

 
The last 50 years have witnessed an exponential increase in research dedicated to psychological stress. With this, there has been a concomitant improvement and refinement in our understanding of the phenomenon. One of the principal areas of development has been the conceptualization of stress as a process (1, 2) and the resulting distinction between stressors (events that pose a challenge to the host), psychosocial mediators (constructs that enable the host to evaluate the nature of the threat and the emotional and/or behavioral response elicited), and the stress response (typically a measure of the emotional response elicited in response to the stressor).

This clearer delineation of the stress process and its components has done much to shape the focus of recent research in this area. For example, empirical and theoretical inquiry has led to a distinction between minor stressors, such as daily hassles (3), and major challenges, such as life events (4). Research has also led to the identification of several psychosocial mediators that have been found repeatedly to be important determinants of the stress response across a variety of settings and populations; for example, social support has been isolated as a significant predictor of distress in patients with HIV infection (5), women with postpartum depression (6), and patients with cancer (7). Finally, a related development has been the recognition that these mediators do not always exert their effect independently but often combine to impart a general effect (5, 8, 9). This has led to widespread assessment of several, rather than single, mediators in studies of the nature of the stress process.

Without seeking to undermine the significance of such developments, it is evident that the one limitation peculiar to them is the absence of a "process view." Each development described has resulted from a narrow focus on only certain components of the stress process. Clearly, such focus has been critical to developing an understanding of the minutiae of the process. However, it has also resulted in a tendency to overlook the dynamic interplay between these minutiae and thus the complexion of the stress process as a whole. As a result, several key questions remain largely unanswered. The first of these concerns the relative importance of different stressor types in the stress process. Investigators routinely measure the impact of major stressors (ie, stressful life events) or minor stressors (ie, hassles). However, few examine both (10), and even fewer explore their relative contributions to the stress process.

The second consideration concerns the extent to which the effect of psychosocial mediators remains constant. Most of the mediators identified to date are moderately stable, almost traitlike, features of the psychosocial milieu (eg, coping style and social support), yet few investigators have examined the extent to which mediators identified as being salient determinants of the stress process at one time point retain their influence throughout the duration of the stressor. Indeed, much of the work in this area has been cross-sectional in nature. Such investigations led to the identification of salient mediators but do not indicate whether they will retain their influence longitudinally (11). One investigation that offers some insight into this issue examined the relationship between coping styles, social support, and adjustment to HIV infection in individuals at various stages of the disease (12). Although cross-sectional, the data implied that the salient mediators of adjustment in HIV disease varied selectively throughout the course of the disease. Such data are crucial if we are 1) to achieve an accurate understanding of the stress process as it pertains to particular stressors and 2) to develop pertinent interventions for those at risk from psychological distress.

The final issue worthy of consideration is whether and how the determinants of the stress process vary according to the qualitative characteristics of the stress response. Perhaps the most ubiquitous limitation in this field has been the heterogeneity in the conceptualization of the stress response, with investigators using indices of anxiety (13), depression (14), stress (15), or a combination of these (5). Although these variables are often highly correlated, the extent to which they measure the same domains of negative affect and share the same psychosocial determinants remains unclear.

The following study was designed to offer an initial exploration into these issues in the context of a chronic stressor, informal caregiver stress. The predicament of the informal caregiver is now widely recognized to be highly stressful across a variety of caregiving contexts (eg, HIV infection and dementia; Refs. 16 and 17). Like other chronic stressors, the demands imposed are both physical and emotional and persist without any indication of when they may cease. The population of caregivers selected for assessment were spousal caregivers of patients with dementia. This was guided by two main considerations. First, in the majority of families, the primary informal caregiver of the dementia patient is the spouse (18). This demographic bias facilitates the recruitment of a group in which the familial relationship between caregiver and care recipient can be controlled. This is less easily achieved in other populations, such as caregivers of HIV-infected individuals, where greater diversity exists (19). Second, the progression of age-associated dementing illnesses is typically protracted and not subject to sudden changes in the patient. Thus, it was possible to minimize the likelihood of sudden changes in the caregiving role on our evaluation of the stress process and to examine the extent to which stable predictive relationships existed between mediators and stress response indices in the context of a moderately stable stressor.

Thus, a study was conducted into the nature of the stress process in chronic spousal caregiver stress. Specifically, we examined the role of stressor and psychosocial variables on distress assessed at two time points. The aims of the investigation were three-fold: 1) to examine the relative importance of stressor types (ie, daily hassles, caregiving-specific stressors, and life events) on the stress response, 2) to assess the stability of relationships between psychosocial variables and stress over a 6-month period, and 3) to explore how the nature and magnitude of the contribution made by stressors and psychosocial variables to the stress process varied according to the qualitative characteristics of the stress response (ie, anxiety, depression, and stress).

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION CONCLUSIONS ACKNOWLEDGMENTS REFERENCES

 
This investigation was approved by the United Bristol Healthcare Trust and the Frenchay National Health Service Healthcare Trust Research and Ethics Committees.

Subjects
Fifty spousal caregivers of patients with dementia (24 men, 26 women) were recruited from the Bristol Memory Disorders Clinic, where their partners had previously attended and been diagnosed with dementia. The inclusion and exclusion criteria adopted at the time of recruitment ensured that all participants described themselves as their partner’s primary caregiver and stated that they cared for their partner at home. The mean age of caregivers was 72 years (±8 years). All caregivers were white, and 98% stated that they were retired. The mean household income for this group was $21,712 (±$13,899). The majority of caregivers’ spouses had received a diagnosis of probable Alzheimer’s disease (78%). The remainder had a diagnosis of multiinfarct dementia (17%) or Parkinson’s disease with progressive dementia (5%). The mean number of years that caregivers’ spouses had been diagnosed was 3.5 years (SD, 2.7 years), and the mean number of hours per day spent in the caregiving role was reported as being 13.8 hours (SD, 9.44 hours).

A control population of 67 noncaregivers was also recruited so that the psychosocial status of our caregivers could be compared with that of a cohort of noncaregiving control subjects. Noncaregivers were closely matched for age (mean ± SD, 69 ± 4 years), gender (31 males, 36 females), ethnicity (100% white), employment status (91% retired), and socioeconomic status ($22,038 ± $11,575) with the group of caregivers. Furthermore, at recruitment, all noncaregivers stated that they had partners who were alive and who did not have a diagnosis of dementia and that they did not, themselves, have any other caregiving responsibilities (eg, elderly parent or disabled child).

Procedure
All participants were followed-up at intervals of 3 months as part of a prospective investigation into the antecedents and consequences of chronic stress in spousal caregivers of dementia patients. Data from the first 6 months of the investigation (ie, 0, 3, and 6 months) are presented; during these follow-ups, a detailed psychosocial assessment was made (0 and 6 months), as was a brief psychosocial assessment (3 months). The psychosocial assessments were undertaken using established psychological questionnaires that have been demonstrated to have appropriate validity and reliability, were designed specifically for use with elderly groups, and/or have been used extensively in the stress literature. The detailed psychosocial assessments involved measurement of variables known to influence the stress process (2) and/or variables that have previously been shown to be of relevance to the stress process in spousal caregivers of dementia patients, namely, stressors (ie, life events, daily hassles, and caregiving difficulties), psychosocial mediators (ie, social support, self-esteem, coping style, and self-concept), and stress response (ie, anxiety, depression, and stress).

Frequency and Nature of Stressors
Stressful life events.
Geriatric Social Readjustment Rating Scale (20).

This scale measures the occurrence of major life events and consists of items that were judged by a panel of gerontologists to be of particular relevance to the life experiences of elderly individuals. Previous research has demonstrated the validity of the scale in that scores on this inventory are significantly related to measures of physical and mental health (20, 21). Scores pertaining to the total number of events and their severity were calculated. However, only the former were used in the analyses presented in this article because the latter provide a measure of stressor impact and thus are more akin to a measure of the stress response.

Daily hassles.
Hassles and Uplifts Scale (3).

This instrument consists of two subscales designed to assess minor events of a negative nature (ie, hassles) and minor events of a positive nature (ie, uplifts). Both subscales provide a measure of the frequency of events and their severity. However, as with the life events scale, only the frequency measure from the hassles subscale was used in this study. Criterion validity was established through significant correlations between measures of negative affect and the hassles severity measure and by the absence of a significant association between the hassles severity measure and a measure of the impact of life events. Test-retest reliability over a 9-month period for the hassles subscale was high (0.79) for frequency scores and moderate (0.48) for severity scores.

Caregiving difficulties.
Burden Interview (22).

This scale was designed to assess the challenges faced by family caregivers of elderly and disabled individuals. The Cronbach value for the scale’s internal consistency is between 0.88 and 0.91, test-retest reliability is 0.71, and correlations between the scale and the Brief Symptom Inventory (23) and global ratings of burden confirm the scale’s validity. Two scores were calculated from this scale, a measure of the total number of caregiving-related difficulties experienced and the total amount of distress due to caregiving-related difficulties. As per the data obtained from the life events and hassles scales, only the former score was used in the analyses presented in this article.

Psychosocial Mediators
Social support.
Significant Others Scale (24).

The short form of this scale measures actual and ideal levels of emotional and practical support and discrepancies between the two. The authors report a test-retest reliability, measured over 6- and 9-month follow-up periods, of 0.52 to 0.85, which indicates stability over time, and appropriate concurrent validity statistics (based on correlations with the General Health Questionnaire; Ref. 25) and construct validity based on the results of their factor analysis (24). Summary scores of actual and ideal support were used in this investigation.

Self-esteem.
Culture-Free Self-Esteem Inventory (26).

This scale measures subjective feelings of self-worth. The version of the scale used in this study was reduced from the original 40-item inventory (which examines general, personal, and social self-esteem and includes a lie subscale) to 16 items that measure personal and social self-esteem. The lie subscale was excluded because it was not considered to be relevant to the aims of this study. The general self-esteem subscale was omitted because the personal and social self-esteem subscales were expected to yield more precise information on self-esteem. The test-retest correlations for the entire scale and the two subscales are respectable at 0.81 (whole scale), 0.56 (personal self-esteem), and 0.78 (social self-esteem). The Cronbach values for internal consistency of the subscales are reported to be 0.72 and 0.57 for the personal and social subscales, respectively. Finally, the content validity of the scale is apparent from correlations with a measure of depression. The measures of personal and social self-esteem were used in the analyses presented in this article.

Coping style.
Ways of Coping Scale (27).

This scale measures eight approaches to coping with particular stressful experiences: confrontational, distancing, self-controlling, seeking social support, accepting responsibility, escape/avoidance, planful problem-solving, and positive reappraisal. Participants were asked to indicate how they respond to "difficulties they experience when caring for their husband or wife." The Cronbach values for these subscales range from 0.61 to 0.79, indicating strong internal consistency. Similarly, the intercorrelations between the subscales are low (ranging from 0.01 to 0.39), indicating the orthogonal nature of the subscales. All eight subscales were used in the present investigation.

Self-concept.
Savage Personality Screening Scale (28).

This scale was designed to measure self-concept, introversion, anxiety, and depression. Factor analysis of the inventory has demonstrated the independence of the subscales (28) and a Kuder-Richardson reliability coefficient for the self-concept subscale of 0.48. The author notes that self-concept is a construct of considerable relevance to personal adjustment in the elderly because advancing age typically demands a reassessment of one’s view of oneself. Thus, this variable was selected for measurement in this study.

Subjective Evaluations of the Stress Response
Anxiety and depression.
Savage Personality Screening Scale (28).

The reliability data for these subscales of the inventory revealed Kuder-Richardson reliability coefficients of 0.47 and 0.76 for depression and anxiety, respectively. Total anxiety and depression scores were generated, and both were used in the analyses presented here.

Stress.
Global Measure of Perceived Stress (29).

This inventory measures self-reported levels of stress in the previous month. Cronbach’s coefficients for the scale range from 0.84 to 0.86. Furthermore, the scale correlates with life event scores, depressive symptomology, and social anxiety. The inventory yields a total stress score, which was used in the analyses presented in this article.

The brief postal follow-up questionnaire (administered at 3 months) assessed only the frequency of stressors (ie, the measures of life events, hassles, and caregiving difficulties) and subjective evaluations of the stress response (ie, anxiety, depression, and stress). The remaining variables were excluded because they were previously shown to be more stable (5) and thus were not expected to fluctuate significantly over intervals of 3 months.

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION CONCLUSIONS ACKNOWLEDGMENTS REFERENCES

 
Psychosocial Status of the Cohort
Initially, descriptive statistics were calculated for each psychosocial index measured in this study to ascertain the psychosocial status of our cohort of spousal caregivers of dementia patients compared with that of a cohort of closely matched, noncaregiving control subjects.

The data presented in Table 1 are from the first stage of the study (ie, 0 months). The mean scores obtained for caregivers and noncaregivers suggested between-group differences on several measures. The results from a series of one-way analyses of variance revealed that although the groups did not differ in the frequency of stressors experienced, significant differences were evident for the following psychosocial mediators: self-controlling coping (F = 10.75, p = .001), accepting responsibility coping (F = 4.72, p = .032), personal self-esteem (F = 7.91, p = .006), ideal support (F = 57.95, p < .0001), and self-concept (F = 6.11 ± 0.02). The nature of these differences indicated that caregivers had a propensity toward less self-controlling coping and accepting responsibility coping, lower personal self-esteem, the desire for greater levels of support, and poorer self-concept. Similarly, the psychological morbidity data indicated significantly greater levels of anxiety (F = 10.58, p = .002), depression (F = 16.00, p < .0001), and stress (F = 10.94, p = .001) among caregivers.

The results revealed that all three stress response variables were significantly correlated with the psychosocial variables (Table 2) of personal self-esteem, self-concept, confrontational coping, and escape/avoidance coping. However, there was also evidence of associations between the variables that were exclusive to the stress response in question. In particular, the measure of anxiety, but not depression or stress, correlated with self-controlling coping (r = .430). Similarly, the measure of depression, but not anxiety or stress, was found to correlate with social self-esteem (r = -.291), planful problem-solving coping (r = -.366), and positive reappraisal coping (r = -.312). Furthermore, the measure of stress, but not anxiety and depression, correlated with seeking social support coping (r = .292).

Nature of the Stress Process in Chronic Caregiver Stress
Factor Analysis of Psychosocial Mediator Variables.
The psychosocial mediators identified as significant correlates of the three stress response indices were subjected to principal components analysis. This was conducted because many of the mediators were likely to be correlated (and thus could not be entered as independent predictors in the regression analyses) and to reduce the number of statistical comparisons conducted on the data, which were from a modestly sized sample. The results of the principal components analysis revealed the presence of three independent factors that, together, accounted for 66.4% of the variance in the original data (Table 3). These factors were extracted according to their eigen values (ie, all factors with eigen values >1 were extracted) and implementation of Cattell’s scree test (30). The factor loadings of the original variables revealed that factor 1 was characterized by a propensity toward self-controlling, confrontational, and escape/avoidance coping combined with poor self-concept and low personal self-esteem. This factor was therefore labeled "reactive coping and self-appraisal." In contrast, factor 2 was characterized by a propensity toward positive reappraisal and planful problem-solving coping combined with high social self-esteem. This factor was therefore labeled "proactive coping and self-appraisal." Finally, factor 3 was characterized by a propensity toward seeking social support and was accordingly labeled "seeking social support." The factor scores from these three factors were used in subsequent regression analyses.

The second anxiety regression analysis was similar to the first. However, the dependent variable in this equation was the anxiety score from the 6-month follow-up. Similarly, the independent stressor variables, entered into the first block, were the frequency scores for life events, hassles, and caregiving difficulties at the 6-month follow-up. The psychosocial mediators entered into the second block were identical to those used in the first regression equation (ie, the factor scores for reactive coping and self-appraisal, proactive coping and self-appraisal, and seeking social support). The results from both regression analyses are presented in Table 4.

The percentage of variance accounted for by the psychosocial mediators was found to be comparable (ranging from an additional 18% to 32%), and, as with the stressor variables, the mediators that influenced anxiety levels fluctuated. Specifically, reactive coping and self-appraisal was isolated as a significant predictor at both 3 and 6 months, and seeking social support was a significant predictor at 6 months only. A post hoc evaluation of the ß scores for the psychosocial mediator factors and their standard errors revealed that the ß scores did not differ significantly for any of the psychosocial mediator factors. This indicates that the magnitude of the significant contribution made by reactive coping and self-appraisal and the nonsignificant contribution of proactive coping and self-appraisal was comparable at 3 and 6 months. However, contrary to the probability values obtained, the magnitude of the contribution made by seeking social support was comparable at 3 and 6 months.

Antecedents of depression.
Two additional regressions were conducted in which depression scores at 3 and 6 months were entered as dependent variables. The regression equations were similar in nature to those constructed for the analyses using anxiety scores, such that the regression involving depression scores at 3 months included the stressor variables from the 3-month follow-up (first block) and the psychosocial mediator factors from the 0-month follow-up (second block). Similarly, the regression involving depression scores at 6 months included the stressor variables from the 6-month follow-up (first block) and the psychosocial mediator factors from the 0-month follow-up (second block). The results from these regression analyses are presented in Table 5.

The psychosocial mediator factors were found to account for a marginally greater proportion of the variance in levels of depression (accounting for an additional 27–41%). According to the probability values, the factors of reactive coping and self-appraisal and proactive coping and self-appraisal were isolated as significant predictors at both 3 and 6 months. The post hoc evaluation of the ß scores for the psychosocial mediator factors revealed that the ß scores did not differ significantly for any of the psychosocial mediator factors. Thus, the magnitude of the significant effect exerted by reactive coping and self-appraisal and proactive coping and self-appraisal and the nonsignificant effect of seeking social support was comparable at both 3 and 6 months.

Antecedents of stress.
The final two regressions concerned the measures of stress obtained at 3 and 6 months. As before, the regression equations were similar to those constructed for the analyses with anxiety scores, such that the regression involving stress scores at 3 months included the stressor variables from the 3-month follow-up (first block) and the psychosocial mediator factors from 0 months (second block). Similarly, the regression for stress scores at 6 months included the stressor variables from the 6-month follow-up (first block) and the psychosocial mediator factors from 0 months (second block). The results of these regression analyses are presented in Table 6.

In contrast to the data obtained for anxiety and depression, the percentage of variance accounted for by the psychosocial mediator factors was appreciably smaller (ranging from 10% to 15%). The probability values revealed that only the factor of reactive coping and self-appraisal significantly predicted stress scores, emerging as a significant predictor at both 3 and 6 months. The post hoc assessment of ß scores and standard errors also indicated that the magnitude of the contribution made by each psychosocial mediator factor was comparable at both 3 and 6 months (ie, significant for reactive coping and self-appraisal and nonsignificant for proactive coping and self-appraisal and seeking social support).

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION CONCLUSIONS ACKNOWLEDGMENTS REFERENCES

 
This investigation was designed to further our understanding of the stress process in chronic stress. In particular, we examined three key issues in the context of caregiver stress in spousal caregivers of dementia patients: 1) the relative importance of stressor types on the stress response, 2) the stability of relationships between psychosocial mediators and stress over a 6-month period, and 3) the nature of the stress process (ie, whether it varies according to the qualitative characteristics of the stress response). The implications of our data for each of these issues is discussed in this section.

Relative Importance of Stressor Types on the Stress Response
The results of the regression analyses highlight two important issues related to the role of stressors in the stress process: that stressors were influential determinants of the stress response, with the proportion of variance accounted for by the stressor indices (which ranged from 20% to 53%), and that the different stressor types exhibited differing relationships with the outcome measures. In particular, the contributions of life events and caregiving difficulties (independent of their significance) were largely consistent between 3 and 6 months. In contrast, the contribution of hassles was found to fluctuate during the same period.

Together, these results support the findings of previous investigations of the importance of stressors in the stress process (31, 32). However, our data and the results of those investigations cannot help to elucidate why the different stressor types "behave" differently during the stress process. This is due, in part, to the fact that the effects of multiple stressor types have not been considered in most previous studies. Also, when more than one stressor type has been examined, the predictive influence of the stressors has been examined at only one time point. For example, Fernandez and Sheffield (33) examined the predictive influence of hassles and life events on headache occurrence and observed that only hassles emerged as a significant predictor. In contrast, Lau et al. (34) examined the role of hassles and life events on patients with recurring atypical chest pain. Their results revealed that only life events distinguished patients with atypical chest pains from healthy control subjects. The absence of additional assessment points in both studies prevent us from determining whether, as in the present study, the effects of these stressors would have fluctuated. It is clear, therefore, that further prospective investigations with multiple assessment points are required before we can determine whether stressors have both cogent and fluctuating roles in the stress process. In addition, more detailed analysis of the qualitative nature of reported stressors may help to determine the circumstances in which one stressor type assumes prominence over another.

Stability of Relationships Between Psychosocial Mediators and Stress Over a 6-Month Period
The second aim of this study was to assess the predictive stability of psychosocial mediators over a 6-month period. We chose to focus on a stressor that was unlikely to fluctuate markedly during the study period to ascertain whether, in the context of such stability, the predictive influence of the mediator variables would also remain stable. The results of the regression analyses revealed stable predictive relationships between the mediators factors and the stress response indices. Specifically, reactive coping and self-appraisal and proactive coping and self-appraisal predicted depression at 3 and 6 months. Similarly, reactive coping and self-appraisal predicted stress at 3 and 6 months. Finally, anxiety was predicted by reactive coping and self-appraisal at 3 and 6 months and by seeking social support at 6 months.

These data indicate that in the context of a moderately invariable stressor, considerable stability is evident in the psychosocial mediators that influence the stress process. This observation is also supported by the post hoc assessment of the ß coefficients and standard errors, which demonstrated that the magnitude of the contribution made by each psychosocial mediator factor was comparable at both 3 and 6 months for all three stress response indices. It should, however, be acknowledged that the follow-up period in this study was limited to 6 months. Thus, the stability of these mediators over a longitudinal time frame has not been established. Furthermore, these results do not indicate whether such stability would also occur during stressors of a more dynamic nature. The results from a recent study suggest that this is unlikely. McColl et al. (9) examined the nature of coping and social support in a cohort of spinal injury patients over a 12-month period and observed that although the nature of coping did not change over time, the nature of social support did, moving from informational to emotional support. These data indicate that the psychosocial mediators that influence the stress process may indeed fluctuate. Additional research into how and why they fluctuate is imperative if we are to achieve an accurate understanding of the processes in chronic stress and to develop suitable interventions for those at risk from psychological distress.

Alterations in the Stress Process According to Qualitative Characteristics of the Stress Response
Our final aim was to explore how the nature and magnitude of the contribution made by stressors and mediators to the stress process varied according to the qualitative characteristics of the stress response (ie, anxiety, depression, and stress). The regression analyses revealed several interesting findings. First, with regard to the stressor variables, only stress was found to be consistently influenced by caregiving-related difficulties. In contrast, all three indices were influenced by the measures of hassles and life events, although the significance and magnitude of the predictive influence exerted by the hassles measure fluctuated. These data suggest that the role of hassles and life events in the stress process is not affected by the qualitative characteristics of the stress response, whereas caregiving-related difficulties seem to be associated exclusively and consistently with stress.

The apparent irrelevance of caregiving-related stressors to depression and anxiety is worthy of further comment. The initial descriptive assessment of the caregivers revealed that despite comparable numbers of stressors, caregivers reported significantly greater anxiety, depression, and stress than the noncaregiving control subjects (Table 1). These results suggest that the feature that distinguished the two groups (ie, the caregiving role) in some way contributed to the increased psychological morbidity observed in the caregivers. This contention was, however, only partially supported by the regression data, with caregiving difficulties, but not anxiety or depression, isolated as a significant determinant of stress at 3 and 6 months. Although our data do not permit more detailed analysis of these anomalous results, several interpretations can be proffered. For example, the results may indicate that anxiety and depression are not relevant in the context of caregiving difficulties. Conversely, the data may highlight limitations in the measure of caregiving difficulties (eg, the scale may not offer a comprehensive measure of caregiving difficulties, and/or the use of frequency scores may not provide a cogent measure of caregiving-specific stressors). Alternatively, the data may highlight the prominence of psychosocial mediators in influencing depression and anxiety. Indeed, the regressions revealed that stressor and mediator variables accounted for comparable levels of variance in anxiety and depression, whereas stressor variables assumed prominence for levels of stress. The absence of comparable relationships between the measures of depression and anxiety and caregiving-specific stressors clearly warrants further investigation. In addition, as noted above, inquiry into the qualitative characteristics of stressor types is necessary if we are to determine 1) why particular stressor types are more influential in particular stress response contexts and 2) the circumstances under which the effects of stressor types fluctuate within the same stress response context.

A second observation is that unlike the stressor variables, all three stress response indices were predicted by a single mediator factor, reactive coping and self-appraisal. Furthermore, this factor influenced all three indices at both 3 and 6 months. However, the measure of anxiety was also influenced by seeking social support at 6 months, and the measure of depression was also influenced by proactive coping and self-appraisal at both 3 and 6 months. These data indicate that some psychosocial mediators seem to influence the stress process regardless of the qualitative characteristics of the stress response, although the relative independence of anxiety, depression, and stress is demonstrated by the heterogeneity of the additional mediators found to exert an effect. These results clearly demonstrate that the mediators that influence anxiety, depression, and stress states differ.

One final issue addressed in this study concerned the extent to which the stressor and mediator variables influenced the stress response measures. The regressions revealed that, together, stressors and mediators accounted for a large proportion of the variance: anxiety, 49% to 56%; depression, 52% to 61%; and stress, 61% to 63%. However, the data also demonstrated that the relative importance of stressors and mediators to the stress response varied according to the qualitative characteristics of the stress response. For anxiety and depression, the percentage of variance accounted for by the stressor and mediator variables was comparable. However, the reverse was true for stress, with the percentage of variance accounted for by the stressor variables being appreciably higher than that accounted for by the mediators. These data suggest that levels of anxiety and depression experienced by our caregivers were influenced to a comparable degree by both the number of stressors and the psychosocial mediators they brought to bear on them. In contrast, their levels of stress were influenced to a greater extent by the sheer frequency of stressors they faced.

Together, the data pertaining to the effects of stressor types, variations in the influence of psychosocial mediators, and the relative influence of stressors and mediators reveal that considerable variability seems to exist in the nature of the stress process according to the qualitative characteristics of the stress response. Additional research is clearly required to establish the integrity of our observations. Nonetheless, it is clear that our data caution against the interchangeable use of measures of anxiety, depression, and stress in this field and suggest that drawing parallels between investigations that adopt different stress response measures may be wholly inappropriate.

In conclusion, it is our hope that the current investigation has offered a preliminary explication of issues pertinent to our understanding of the stress process. However, it is evident that several limitations exist with the present study, all of which limit the conclusions that can be drawn and thus warrant consideration in future investigations. First, the modest size of our cohort necessitated the use of factor analysis in our assessment of psychosocial mediators. Although the factor analysis revealed the presence of three mediator factors, all of which were found to influence at least one of the stress response indices, the use of factor scores rather than data from the original variables prevented us from conducting a finely detailed analysis of the role of mediators in the stress process. Second, the moderately stable context of chronic caregiver stress was chosen specifically because it enabled assessment of the predictive stability of psychosocial mediators to be undertaken in an immutable context. However, this fact also necessarily limits the extent to which our results can be generalized to stressors of a more dynamic nature. Third, our reliance on frequency scores for each of the stressor indices, rather than an assessment of the qualitative nature of the stressors, prevented us from exploring the circumstances in which one stressor type assumed prominence over another. Such qualitative assessments are required if we are to delineate further the dynamic role of stressors in the stress process. Finally, our cohort of caregivers consisted of individuals who were caring for spouses with a range of dementing illnesses. It is possible that heterogeneity in the illness characteristics of the spouses with dementia may have influenced our findings. In particular, the nature of the illness may have influenced the caregiver’s appraisal of and response to the caregiving situation. It may therefore be advisable in future investigations to control for the nature of the disease experienced by the care recipient.

In addition to the limitations outlined above, it is apparent that this investigation was also limited by two main issues that were not addressed. First, the study did not include an assessment of the relationship between the stressor and mediator indices. Although these analyses were beyond the aims of this initial investigation, the relationship between these two indices is an integral feature of the stress process and should therefore be considered in future investigations. Second, we focused exclusively on the emotional outcomes of caregiving. There is, however, the burgeoning literature on the health effects of this chronic stressor indicates that negative health effects are ubiquitous, although the magnitude of these effects can vary (35, 36). Thus, health outcomes are clearly an area worthy of assessment in future research.

	CONCLUSIONS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION CONCLUSIONS ACKNOWLEDGMENTS REFERENCES

 
We aimed in this investigation to undertake a process approach to understanding the dynamics of the stress process in chronic caregiver stress. Our results have highlighted several issues, many of which warrant additional investigation. Our observations concerning the relative influence of stressor types led us to conclude that although stressors often play a significant role in the stress process, the type of stressor that influences the process can fluctuate. With regard to the stability of relationships between psychosocial mediators and stress, our data suggested that in the context of a moderately stable stressor, certain psychosocial mediators seem to consistently influence the stress process. Finally, we examined the nature of the stress process in the context of three qualitatively different stress responses (ie, anxiety, depression, and stress). Our results demonstrated that the interactions between the various components of the stress process differed according to the qualitative characteristics of the stress response, indicating that the tendency for researchers to consider these variables as transposable may be misguided.

	ACKNOWLEDGMENTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION CONCLUSIONS ACKNOWLEDGMENTS REFERENCES

 
The work presented in this article was made possible by a grant from the Neuroendocrinology Charitable Trust. The authors thank Professor Gordon Wilcock (Department of Care of the Elderly, Frenchay Hospital, Bristol, UK) for allowing us to recruit study participants through his clinic.

Received for publication March 31, 1999.

Revision received October 5, 1999.

	REFERENCES

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION CONCLUSIONS ACKNOWLEDGMENTS REFERENCES

 

1. Lazarus RS, Folkman S. Stress, appraisal, and coping. New York: Springer; 1984.
2. Paterson RJ, Neufeld RWJ. The stress response and parameters of stressful situations. In: Neufeld RWJ, editor. Advances in the investigation of psychological stress. New York: Wiley; 1989. p. 7–42.
3. Kanner AD, Coyne JC, Schaefer C, Lazarus RS. Comparison of two modes of stress measurement: daily hassles and uplifts versus major life events. J Behav Med 1981; 4: 1–39.[Medline]
4. Holmes TH, Rahe RH. The social readjustment rating scale. J Psychosom Med 1967; 11: 213–8.
5. Vedhara K, Nott KH. Psychosocial vulnerability to stress: a study of HIV positive homosexual men. J Psychosom Res 1996; 41: 255–67.[Medline]
6. Beck CT. A meta-analysis of predictors of postpartum depression. Nurs Res 1996; 45: 297–303.[Medline]
7. Hann DM, Oxman TE, Ahles TA, Furstenberg CT, Stuke TA. Social support adequacy and depression in older patients with metastatic cancer. Psychooncology 1995; 4: 213–21.
8. Blaney NT, Goodkin K, Morgan RO, Feaster D, Millon C, Scapocznik J, Eisdorfer CA. Stress moderator model of distress in early HIV-1 infection: concurrent analysis of life events, hardiness and social support. J Psychosom Res 1991; 35: 297–305.[Medline]
9. McColl MA, Lei H, Skinner H. Structural relationships between social support and coping. Soc Sci Med 1995; 41: 395–407.
10. Nyklicek I, Vingerhoets AJJM, VanHeck GL, VanLimpt MCAM. Defensive coping in relation to casual blood pressure and self-reported daily hassles and life events. J Behav Med 1998; 21: 145–61.[Medline]
11. Passik SD, Newman ML, Brennan M, Tunkel R. Predictors of psychological distress, sexual dysfunction and physical functioning among women with upper extremity lymphedema related to breast cancer. Psychooncology 1995; 4: 255–63.
12. Pakenham KI, Dadds MR, Terry DJ. Relationships between adjustment to HIV and both social support and coping. J Consult Clin Psychol 1994; 62: 1194–203.[Medline]
13. Lutgendorf SK, Antoni MH, Ironson G, Starr K, Costello N, Zuckerman M, Klimas N, Fletcher MA, Schneiderman N. Changes in cognitive coping skills and social support during cognitive behavioral stress management intervention and distress outcomes in symptomatic human immunodeficiency virus (HIV)-seropositive gay men. Psychosom Med 1998; 60: 204–14.[Abstract/Free Full Text]
14. Masel CN, Terry DJ, Gribble M. The effects of coping on adjustment—re-examining the goodness of fit model of coping effectiveness. Anxiety Stress Coping 1996; 9: 279–300.
15. Zautra AJ, Hoffman J, Potter P, Matt KS, Yocum D, Castro L. Examination of changes in interpersonal stress as a factor in disease exacerbations among women with rheumatoid arthritis. Ann Behav Med 1997; 19: 279–86.[Medline]
16. Claxton RPR, Catalan J, Burgess AP. Psychological distress and burnout among buddies: demographic, situational and motivational factors. AIDS Care 1998; 10: 175–90.[Medline]
17. Pot AM, Deeg DJH, VanDyck R. Psychological well-being of informal caregivers of elderly people with dementia: changes over time. Aging Ment Health 1997; 1: 261–8.
18. Zarit SH, Ord ND, Zarit JM. The hidden victims of Alzheimer’s disease: families under stress. New York: New York University Press; 1985.
19. McCann K, Wadsworth E. The role of informal carers in supporting gay men who have HIV related illness: what do they do and what are their needs? AIDS Care 1992; 4: 25–34.[Medline]
20. Amster LE, Krauss HH. The relationship between life crises and mental deterioration in old age. Int J Aging Hum Dev 1974; 5: 51–5.[Medline]
21. Cutrona CE, Russell D, Rose J. Social support and adaptation to stress by the elderly. Psychol Aging 1986; 1: 47–54.[Medline]
22. Zarit SH, Zarit JM. The memory and behavior problems checklist and the burden interview. Pennsylvania: Gerontology Center, Pennsylvania University; 1990.
23. Derogatis LR, Lipman RS, Covi L, Richels K, Uhlenhuth ER. Dimensions of outpatient neurotic pathology: comparison of a clinical versus an empirical assessment. J Consult Clin Psychol 1970; 34: 164–71.[Medline]
24. Power MJ, Champion LA. The Significant Others Scale. In: Milne D, editor. Assessment: a mental health portfolio. Windsor, United Kingdom: NFER-Nelson; 1992. p. 18–29.
25. Goldberg DP. The detection of psychiatric illness by questionnaire. London: Oxford University Press; 1972.
26. Battle J. Culture free SEI self esteem inventories for children and adults. Seattle (WA): Special Child Publications; 1981.
27. Folkman S, Lazarus RS. The Ways of Coping Questionnaire. Palo Alto (CA): Consulting Psychologists Press; 1988.
28. Savage RD. Savage Personality Screening Scale (SPSS-20) [interim manual notes]. Perth, Western Australia: Murdoch University; 1990.
29. Cohen S, Kamarack T, Mermelstein R. A global measure of perceived stress. J Health Soc Behav 1983; 24: 385–96.[Medline]
30. Cattell RB. The scientific use of factor analysis in the behavioral and life sciences. New York: Plenum Press; 1978.
31. Sorbi MJ, Maassen GH, Spierings ELH. A time series analysis of daily hassles and mood changes in the 3 days before the migraine attack. Behav Med 1996; 22: 103–13.[Medline]
32. Steptoe A, Lipsey Z, Wardle J. Stress, hassles and variations in alcohol consumption, food choice and physical exercise: a diary study. Br J Health Psychol 1998; 3: 51–63.
33. Fernandez E, Sheffield J. Psychosocial stressors predicting headache occurrence—the major role of minor hassles. Headache Q Curr Treatment Res 1995; 6: 215–20.
34. Lau GKK, Hui WM, Lam SK. Life events and daily hassles in patients with atypical chest pain. Am J Gastroenterol 1996; 9: 2157–62.
35. Shaw WS, Patterson TL, Semple SJ, Ho S, Irwin MR, Hauger RL, Grant I. Longitudinal analysis of multiple indicators of health decline among spousal caregivers. Ann Behav Med 1997; 19: 101–9.[Medline]
36. Schulz R, Newsom J, Mittelmark M, Burton L, Hirsch C, Jackson S. Health effects of caregiving: the caregiver health effects study: an ancillary study of the cardiovascular health study. Ann Behav Med 1997; 19: 110–6.[Medline]

This article has been cited by other articles:

				J. A. Zauszniewski, S. J. F. Picot, B. L. Roberts, S. M. Debanne, and M. L. Wykle Predictors of Resourcefulness in African American Women J Aging Health, October 1, 2005; 17(5): 609 - 633. [Abstract] [PDF]

				L. L. Davis, M. Weaver, E. Zamrini, A. Stevens, D.-H. Kang, and C. R. Parker Jr Biopsychological Markers of Distress in Informal Caregivers Biol Res Nurs, October 1, 2004; 6(2): 90 - 99. [Abstract] [PDF]

				P.-F. Tsai A Middle-Range Theory of Caregiver Stress Nurs Sci Q, April 1, 2003; 16(2): 137 - 145. [Abstract] [PDF]

This Article Abstract Full Text (PDF) Alert me when this article is cited Alert me if a correction is posted Services Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Vedhara, K. Articles by Lightman, S. Search for Related Content PubMed PubMed Citation Articles by Vedhara, K. Articles by Lightman, S. Related Collections Stress and Coping