This Article Abstract Figures Only 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 Request Permissions Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Ciechanowski, P. S. Articles by Russo, J. E. Search for Related Content PubMed PubMed Citation Articles by Ciechanowski, P. S. Articles by Russo, J. E. Related Collections Health Policy Somatoform Stress and Coping

Attachment Theory: A Model for Health Care Utilization and Somatization

From the Department of Psychiatry and Behavioral Sciences, University of Washington, Seattle, WA.

Address reprint requests to: Paul Ciechanowski, MD, MPH, Box 356560, Department of Psychiatry and Behavioral Sciences, University of Washington, Seattle, WA 98195. Email: pavelcie{at}u.washington.edu

	ABSTRACT

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION ACKNOWLEDGMENTS REFERENCES

 
OBJECTIVE: Attachment theory proposes that cognitive schemas based on earlier repeated experiences with caregivers influence how individuals perceive and act within interpersonal relationships. We hypothesized that medical patients with two types of insecure attachment—preoccupied and fearful attachment—would have higher physical symptom reporting compared with those with other attachment styles, but that preoccupied attachment would be associated with higher and fearful attachment with lower primary care utilization and costs.

METHODS: In a large sample of adult female primary care health maintenance organization patients (N = 701), we used analysis of covariance and Poisson regressions to determine whether attachment style was significantly associated with 1) symptom reporting based on questions from the somatization section of the Diagnostic Interview Schedule and 2) contemporaneous automated utilization and cost data.

RESULTS: Attachment style was significantly associated with symptom reporting (p = .02), with patients with preoccupied (p = .03) and fearful (p = .003) attachment having a significantly greater number of physical symptoms compared with secure patients. There were no significant differences in medical comorbidity between attachment groups. Attachment was also significantly associated with primary care visits and costs. Patients with preoccupied attachment had the highest primary care costs and utilization, whereas patients with fearful attachment had the lowest.

CONCLUSIONS: These results suggest that attachment style is an important factor in assessing symptom perception and health care utilization. Despite being on opposite ends of the utilization spectrum, patients with preoccupied and fearful attachment have the highest symptom reporting. These data challenge the observation that increased symptom reporting is uniformly associated with increased utilization in medical patients.

Key Words: attachment theory • somatization • health care utilization • health care costs • doctor-patient • negative affect

Abbreviations: CDS = Chronic Disease Score;; GHC = Group Health Cooperative;; HMO = health maintenance organization;; MHI-5 = Mental Health Inventory-5;; NIMH = National Institute of Mental Health;; RQ = Relationship Questionnaire;; RSQ = Relationship Scales Questionnaire.

	INTRODUCTION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION ACKNOWLEDGMENTS REFERENCES

 
The frequently observed lack of correlation between symptoms and the degree of pathophysiologic findings has always fascinated physicians. Given a physiologic signal, patients are likely to vary along a continuum of symptom perception ranging from symptom amplification at one end of the spectrum to symptom dampening at the other. Most modern-day studies have focused on symptom amplifiers (1), but recent research on silent cardiac ischemia has also provoked interest about symptom dampeners or minimizers (2, 3).

Symptom perception is linked to use of medical services, but studies have often emphasized that behaviors associated with medical utilization are complex. Most studies have found that researchers can only predict about 10% to 15% of the variance in medical utilization (4). Medical symptoms and disorders as well as psychiatric disorders often increase a patient’s perception of personal vulnerability. Patients respond to this sense of vulnerability with behaviors ranging from scheduling health visits immediately and frequently to delaying health care seeking (5).

In this article, we will describe the results of a study aimed at increasing the understanding of symptom perception and health care seeking behavior. We propose that the concept of attachment behavior is a useful model for understanding symptom perception, health care utilization, and their interaction. John Bowlby, who first developed attachment theory, proposed that individuals internalize earlier experiences with caregivers, forming enduring cognitive schemas of relationships that influence whether they perceive themselves as worthy of care (model of self) and whether others can be trusted to provide care (model of other) (Figure 1) (6, 7). These cognitive schemas or "internal working models" influence the kinds of interactions individuals have with others and their interpretations of these interactions throughout life (6). Two traditions of measuring attachment have developed (8). In the first, the "nuclear family" tradition, "states of mind" regarding patterns of attachment in parent-child relationships are assessed in adults using a semistructured interview developed by Main and colleagues, known as the Adult Attachment Interview (unpublished manuscript, 1985, University of California, Berkeley). In the second, the "peer/romantic partner" tradition, Hazan and Shaver (42), Bartholomew and Horowitz (7), and many others have developed various questionnaires and interviews to measure how adults with different attachment histories perceive and behave in close relationships.

View larger version (91K): [in this window] [in a new window]   Fig. 1. Attachment style categories and hypothesized relationship of model of self and other with primary health care utilization and symptom reporting. Adapted from Bartholomew and Horowitz (7).

Bartholomew (11) has suggested that these individual differences in attachment patterns may have implications for communication in interpersonal relationships, including, we propose, the patient-provider relationship. For example, individuals with a negative perception of others, or with a negative other model (dismissing and fearful attachment), have learned that people are likely to ignore or reject their attempts to gain support and thus compared with those with a positive other model, are less likely to seek support and are generally thought to have less self-disclosure and openness in relationships. Individuals with a negative self model (preoccupied and fearful attachment), on the other hand, may generally be more likely to report somatic symptoms as a consequence of their tendency to focus on negative affect (13). This corresponds with research suggesting that a focus on negative affect or negative emotionality is correlated with subjective health complaints even in the absence of objective evidence of disease (4, 14, 15). Individuals with dismissing attachment characteristically do not focus on negative affect, and as adults, they may be less likely to report physical symptoms.

In this study, we hypothesized that female medical patients with preoccupied and fearful attachment (negative self model of attachment) would have significantly higher physical symptom reporting compared with patients who have secure or dismissing attachment, even after accounting for severity of medical illness. We hypothesized that patients with preoccupied attachment would have the highest rates of primary care utilization and resulting costs due to their tendency toward high symptom reporting as well as their tendency to overly rely on others for a sense of self-esteem. We also hypothesized that patients with fearful attachment would have the lowest levels of primary care utilization and costs as a result of their fear of intimacy, despite their tendency to have high symptom reporting, and that patients with dismissing attachment would have lower primary care utilization based on their compulsive self-reliance (Figure 1).

	MATERIALS AND METHODS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION ACKNOWLEDGMENTS REFERENCES

 
Study Population
This study involved sending questionnaires in 1999 to female subjects who were previously enrolled in a large NIMH-funded study (16), which explored health care utilization, perceived health status, prior childhood maltreatment, and functional disability among female HMO members. The original study took place in 1995 to 1996 among the enrolled membership of GHC, a large staff model HMO that provides primary and specialty medical and mental health services to approximately one-half million individuals in the Puget Sound area of Washington State. The current study entailed administration of a three-page questionnaire, which assessed the participating subjects’ attachment style. These results were combined with self-report and automated utilization data from the original study of 1225 women. Based on the respondent pool of women aged 18 to 65 at the time of the original study for whom current addresses could be obtained, 1119 women in the current study were contacted with an approach letter, which described the purpose of the current study and of the enclosed questionnaire assessing interpersonal relationships. Subjects received a $3 token of appreciation for their time to participate in the current study. The questionnaire was approved by the Human Subjects Committees of GHC and the University of Washington.

Self-Report Measures
Attachment style.
Two related instruments measuring attachment style were utilized. Both the 30-item RSQ (17) and the four-item RQ (17) are valid and reliable instruments that determine attachment style of respondents. RSQ and RQ results were combined by averaging z-transformed continuous data, from which attachment style was determined categorically for each subject based on the attachment category with the highest score (18). Attachment style is considered to be a trait that has been shown to be stable from 4 to 25 years (19– 22).

Index of depression from the SF-36 (MHI-5).
In the primary study, psychological distress was assessed using the five-item MHI-5 from the mental health domain of the SF-36 (23), which performs well in screening primary care patients for psychiatric disorders. The MHI-5 has been shown to have good reliability and construct validity (24). We created a dichotomous depression variable, using a cutoff score of 16/17 which has been shown to have a high positive predictive value to detect major depression based on receiver operating characteristic analysis (24).

Self-reported history of physical symptoms.
In the primary study, participants were asked to rate the degree that they were bothered during the prior 6 months by the 15 most common physical symptoms derived from the somatization section of the Diagnostic Interview Schedule used in the Epidemiological Catchment Area study (25). These symptoms have been shown to account for a significant number of visits to primary care physicians. The participants were asked to rate the degree of bother on a five-point Likert scale from "none of the time" (one) to "all of the time" (five). Subjects with ratings of three or more were considered to have significant symptoms in subsequent analyses.

Demographics.
Age, marital status, income, race, and education level were obtained from questionnaire data.

Automated Data
Primary care utilization and costs.
The automated cost accounting system of the HMO was used to determine mean annual primary care costs and utilization corresponding temporally to administration of the original questionnaire, which took place in four waves 4 months apart. Costs and utilization data were available in 6-month increments during 1995 and 1996. Thus, two consecutive 6-month increments were chosen for each patient to most closely correspond to administration of the original questionnaire, so that a full year of cost and utilization data were available for each subject. This ensured that data on symptom reporting, which covered the 6 months before the time of the questionnaire, coincided as closely as possible with cost and utilization data. Cost and utilization data were collected using the GHC automated data system. Cost estimates reflect direct costs of professional services and overhead costs to GHC for providing different forms of service rather than a schedule of charges. Each primary care clinic allocates its actual monthly costs (eg, facilities, payroll, and supplies) over the total number of outpatient visits provided, thus fully allocating all overhead costs to each patient care department. The perspective taken in assessing the primary care costs is from the perspective of the HMO and is reported in dollars adjusted for inflation to the year of the last data collection (1999).

Chronic Disease Score.
Ratings of chronic medical comorbidity at initial assessment were obtained from the CDS, which is an index derived from automated pharmacy data of medications used to treat chronic medical conditions. The CDS has been shown to correlate with physician ratings of physical disease severity and to predict mortality and hospital utilization (26).

Statistical Analyses
Data were analyzed using SPSS 10.0 for Windows. Two-tailed t tests or chi-square tests with continuity corrections were used to compare respondents and nonrespondents on age, ethnicity, marital status, income, level of education, and primary care utilization. Demographic and clinical variables in respondents were then compared between attachment groups using analysis of variance or chi-square tests. For significant results, Tukey post hoc tests or chi-square tests were performed. Any demographic or clinical characteristics that were significantly different between the attachment groups in these bivariate analyses were included as covariates in the analyses assessing symptom reporting and the cost and utilization analyses.

Analyses of covariance with planned contrasts were used to determine whether number of somatic symptoms and primary care costs varied as a function of attachment style. The three degrees of freedom attachment style main effect was partitioned into three 1-degree of freedom planned comparisons between attachment groups. To satisfy the requirements of a normal distribution for the outcomes, health care cost data, which are heavily skewed to the right, were log-transformed (27). Poisson regression in which we used dummy variables for the four attachment style groups with preoccupied attachment style as the reference category was used to determine whether primary care utilization varied as a function of attachment style. For descriptive purposes, the adjusted means for cost data, before log-transformation, were reported.

To examine the relationships between continuous measures of the four attachment styles within each subject (secure, fearful, preoccupied, and dismissing) and primary care costs (log-transformed), utilization, and somatization (number of somatic symptoms), we conducted three regression analyses, in which we controlled for covariates. Ordinary least squares linear regression was used for cost and somatization data, and Poisson regression was used for utilization data.

	RESULTS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION ACKNOWLEDGMENTS REFERENCES

 
Of the 1119 surveys mailed, 701 were completed for a final return rate of 63%. There were no significant differences between women who completed the surveys and nonrespondents with respect to mean number of annual primary care visits or marital status. However, respondents were slightly older (43.4 ± 10.8 years vs. 40.5 ± 11.9 years, t = 2.92, df = 1116, p < .001) (mean ± SD), more likely to have at least 1 year of college (90% vs. 87%, ² = 4.255, df = 1, p = .04), more likely to be white (81% vs. 72%, ² = 13.88, df = 1, p < .001), and to have higher income (² = 31.21, df = 5, p < .001). The difference in income was a result of a smaller proportion of the respondents, compared with nonrespondents, having an income between $0 and $10,000 (3.2% vs. 10.1%). Despite these demographic differences, participants still closely reflect demographic characteristics of the HMO population from which the sample was drawn and are typical of the Puget Sound area of Washington State.

Clinical and Demographic Characteristics by Attachment Group
Thirty-four percent of subjects reported having secure attachment, whereas 21%, 22%, and 23% reported having fearful, preoccupied, and dismissing attachment, respectively. Attachment style was significantly associated with age, marital status, household income, ethnicity, and MHI-5 depression (Table 1). Although patients with secure attachment were significantly different from two or all of the insecure attachment groups in marital status, household income, and ethnicity, there were also significant differences between insecure attachment categories. Dismissing individuals were older than fearful and preoccupied individuals, and they were less likely to be white compared with preoccupied individuals. Preoccupied individuals were significantly more likely to be married than fearful individuals. There were no significant differences between attachment categories with regard to medical comorbidity scores.

Primary Care Costs and Utilization
Ninety-four percent of patients had two consecutive 6-month periods of primary care cost and utilization data that were contemporaneous with administration of the original questionnaire (ie, ensuring that the 6-month symptom reporting period overlapped fully with the cost and utilization data for each patient). Cost and utilization analyses were limited to these subjects.

Poisson regression revealed that attachment style was significantly associated with number of primary care visits, after adjusting for age, marital status, income, ethnicity, and depression (Figure 2). Compared with the mean annual number of primary care visits of subjects with preoccupied attachment (3.95, SD = 3.08), subjects with fearful attachment (2.84, SD = 3.11; p < .001), dismissing attachment (3.41, SD = 3.13; p = .017), and secure attachment (3.32, SD = 3.13; p = .003) all had significantly fewer visits.

View larger version (34K): [in this window] [in a new window]   Fig. 2. Primary care visits and attachment categories.

View larger version (29K): [in this window] [in a new window]   Fig. 3. Primary care costs and attachment categories.

	DISCUSSION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION ACKNOWLEDGMENTS REFERENCES

These observations were predicted by attachment theory. For example, the link between preoccupied and fearful attachment and higher symptom reporting is based on a common denominator—a tendency to have low self-esteem and to focus on negative affect—an association that remained significant even after adjusting for concurrent depression. Individuals with secure and dismissing attachment, on the other hand, do not characteristically have negative self-esteem or focus on negative affect and therefore would not be predicted to have high numbers of somatic symptoms. Our results neither support nor refute an association between dismissing attachment and lower symptom reporting.

Individuals with preoccupied attachment tend to be overly dependent on others based on earlier inconsistent caregiving relationships, and this together with their tendency to report more somatic symptoms may lead to their becoming high medical utilizers and possibly dependent on their providers. Although the association between higher utilization and higher symptom reporting has been described by previous researchers (4, 5), the novel finding that proportionately as many individuals with high symptom reporting (those with fearful attachment) are less likely to schedule primary care visits may be surprising to clinicians, but is nevertheless congruent with attachment theory. It is very likely that individuals with fearful attachment report more physical symptoms as a result of their increased distress and focus on negative affect, yet because of their cognitive schema of attachment, they are also more likely to engage inconsistently in any one mode of health care. These individuals may avoid more intimate forms of health care such as regularly scheduled primary care visits due to their fear of intimacy and may opt instead for irregular visits with various providers in several different systems of care, such as emergency rooms, walk-in clinics, and inpatient admissions. This pattern of health care utilization may put these individuals at higher risk of delayed medical care for significant medical symptoms or for obtaining needed medical treatment inconsistently.

A larger proportion of the American population is now covered under managed care, and primary care clinicians are increasingly caring for medical conditions that were once largely the domain of varied specialists. As a result, it is increasingly important to identify factors that may help providers and systems of care to systematically understand why certain groups of patients cannot engage with primary care providers. Effective treatment of chronic medical and mental illness requires collaborative management between provider and patient (28). It is in the interest of patients, providers, and researchers to improve collaboration among a potentially sizable segment of the health care population that may have insecure attachment. Measuring patient attachment styles may help significantly in this process.

Based on our findings and on prior research, we propose that health care utilization may be positively associated with the ability to trust caregivers (the other model) and that symptom reporting may be positively associated with one’s self-esteem and tendency to experience distress (the self model)—two orthogonally related dimensions (Figure 1). Patients who have a positive view that they can receive comfort and support from others (secure and preoccupied attachment) are more likely to utilize health care, whereas patients with negative perceptions of the support they can receive from others (fearful and dismissing attachment) are less likely to seek health care. We believe that interventions can be tailored to improve engagement of individuals with negative other models in health care settings. For example, it may be important to arrange for patients with a negative other model, who are less apt to form a trusting relationship with an individual provider, to work together with a cohesive group of health care providers as opposed to their working with a single provider. In other words, the "clinic" rather than an individual professional may become the provider. Additionally, for individuals with dismissing attachment and "compulsive self-reliance," a focus on empowerment and a respect for their need for autonomy may be appropriate clinical responses, whereas scheduling multiple visits with the clinician may be counterproductive. Population-based use of automated tracking of appointments, increased communication through telephone calls, and use of proactive contacts and mailed reminders may be necessary for patients with dismissing or fearful attachment. High utilizers of health care with high numbers of somatic complaints—such as patients with preoccupied attachment—typically require regularly scheduled frequent brief visits with providers (29). Such visits may be carried out more cost-effectively by nurses, nurse practitioners, or social workers given their increased availability and lower cost.

Training health care professionals and students to understand the developmental framework of attachment may lead to greater empathy and less frustration among providers who find themselves in difficult patient-provider relationships. It is important for providers to recognize that the prevalence of secure attachment in the general population, as found in the National Comorbidity Survey, is only 50% to 60% (30). In our sample, even a smaller proportion (34%) had secure attachment, possibly due to the fact that this was a clinical sample, due to use of a different measurement instrument than was used in the National Comorbidity Survey, or due to possible response bias (ie, women with abuse histories, interpersonal difficulties, somatic symptoms, and distress who may also have more insecure attachment may have been more inclined than women without those issues to respond to the original questionnaire dealing with childhood maltreatment and medical care or to our secondary questionnaire that dealt with relationship issues). In general, a significant proportion of providers’ patients may have health behaviors dictated, in part, by insecure attachment behavioral patterns. Understanding that attachment characteristics do not always fit perfectly into prototypes (31), so that even securely attached individuals can have "shades" of attachment insecurity, may reduce resistance by providers to reflect not only on their patients but also on their own role in unsatisfying or noncollaborative patient-provider relationships. Dozier et al. (32) have found, for example, that secure case managers of chronically mentally ill patients were more likely to attend and respond to their clients’ underlying needs, whereas those who were insecure responded to the most obvious presentation of needs.

There are several important limitations in this study. This study is limited to female HMO patients, and researchers have found that there are differences in symptom reporting and patterns of health care use between genders. Thus, these data may not generalize to male populations. Also limiting generalizability was the fact that subjects responding to the second questionnaire showed minor but statistically significant differences in age, education, ethnicity, and income compared with nonrespondents. Furthermore, despite data suggesting that attachment is a relatively stable trait over numbers of years (19– 22), it is possible that the attachment style of a proportion of subjects changed during the 3 years between the two stages of this study. Such changes in attachment style during the 3-year period would be most likely to occur in the presence of significant life changes (eg, serious negative life events such as significant trauma), however we did not collect data to assess this in the second questionnaire in 1999. An important limitation is also the cross-sectional characteristic of this analysis. As a result, we are unable to comment on causal inference; however, because attachment status has been found to be a stable trait over many years, it is very likely that attachment style precedes symptom reporting or health utilization patterns. Another limitation was that in the current study, we did not have provider identifiers and could not assess the impact of providers on patient health behaviors. Recent studies, however, have shown that relatively little of the variation in resource utilization or outcomes (eg, HbA1c levels in diabetic patients, number of visits in acute phase of depression treatment, or percentage of depressed patients refilling antidepressant medications) is due to individual physician practice style differences (33, 34). Nevertheless, in looking at health care utilization from an attachment perspective, which is ultimately concerned with dyadic relationships, it will be important in future studies to assess provider factors. Our research team is currently initiating research that uses quantitative and qualitative methodologies to better understand the patient-provider relationship as it pertains to health care utilization and outcomes in diabetic patients. Finally, symptom reporting was obtained using a self-reported checklist derived from a structured interview, and we could not determine whether these symptoms were somatoform. However, both high numbers of somatoform symptoms as well as symptoms physicians felt could be due to physical illness have been found to be associated with a high likelihood of having psychiatric illness (35). We also had only the self-reported MHI-5 as a covariate for depression, and future studies may benefit from using structured interviews for determining DSM-IV depression and somatization diagnoses.

This study is one of several recent papers and studies targeting attachment schemas and behavior patterns as a focus of clinical assessment and interventions (32, 36–40, 43– 46). One of its strengths is that the results discern health behaviors unique to patients with specific subtypes of insecure attachment. As attachment research data continue to accrue in clinical settings, investigators will hopefully be able to use attachment theoretical models to make "rational" (41) certain health behaviors and aspects of the patient-provider relationship that have challenged clinicians for centuries.

	ACKNOWLEDGMENTS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION ACKNOWLEDGMENTS REFERENCES

 
This study was supported by a grant from the National Institutes of Mental Health K-20 MH-01106. This work was carried out at Group Health Cooperative of Puget Sound, Seattle, WA.

Received for publication August 28, 2000.

	REFERENCES

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION ACKNOWLEDGMENTS REFERENCES

 

1. Katon W, Sullivan M, Walker E. Medical symptoms without identified pathology: relationship to psychiatric disorders, childhood and adult trauma, and personality traits. Ann Intern Med 2001; 134: 917–25.[Abstract/Free Full Text]
2. Barsky AJ, Hochstrasser B, Coles NA, Zifsein J, O’Donnell C, Eagle KA. Silent myocardial ischemia: is the person or the event silent? JAMA 1990; 264: 1132–5.[Abstract/Free Full Text]
3. Lumley M, Rowland L, Torosian T, Bank A, Ketterer M, Pickard S. Decreased health care use among patients with silent myocardial ischemia: support for a generalized rather than a cardiac-specific silence. J Psychosom Res 2000; 48: 479–84.[CrossRef][Medline]
4. Russo J, Katon W, Lin E, Von Korff M, Bush T, Simon G, Walker E. Neuroticism and extraversion as predictors of health outcomes in depressed primary care patients. Psychosomatics 1997; 38: 339–48.[Abstract/Free Full Text]
5. Barsky AJ. Patients who amplify bodily sensations. Ann Intern Med 1979; 91: 63–70.
6. Bowlby J. Attachment and loss. Vol II: Separation: anxiety and anger. New York: Basic Books; 1973.
7. Bartholomew K, Horowitz LM. Attachment styles among young adults: a test of a four-category model. J Pers Soc Psychol 1991; 61: 226–44.[CrossRef][Medline]
8. Simpson JA, Rholes WA. Attachment in adulthood. In: Simpson JA, Rholes WS, editors. Attachment theory and close relationships. New York: Guilford Press; 1998. p. 3–21.
9. Ainsworth MS, Blehar MC, Waters E, Wall S. Patterns of attachment: a psychological study of the Strange Situation. Hillsdale (NJ): Erlbaum; 1978.
10. Bowlby J. The making and breaking of affectional bonds. Br J Psychiatry 1977; 130: 201–10.[Abstract/Free Full Text]
11. Bartholomew K. Avoidance of intimacy: an attachment perspective. J Soc Pers Relations 1990; 7: 147–78.[Abstract]
12. Bartholomew K. From childhood to adult relationships: attachment theory and research. In: Duck S, editor. Learning about relationships: understanding relationships processes series. Vol 2. Newbury Park (CA): Sage Publications; 1993.
13. Feeney J, Ryan S. Attachment style and affect regulation: relationships with health behavior and family experiences of illness in a student sample. Health Psychol 1994; 13: 334–45.[CrossRef][Medline]
14. Costa PT, McCrae RR. Neuroticism, somatic complaints, and disease: is the bark worse than the bite? J Pers 1987; 55: 299–316.[CrossRef][Medline]
15. Watson D, Pennebaker JW. Health complaints, stress, and distress: exploring the central role of negative affectivity. Psychol Rev 1989; 96: 234–54.[CrossRef][Medline]
16. Walker EA, Gelfand A, Katon WJ, Koss MP, Von Korff M, Bernstein D, Russo J. Adult health status of women with histories of childhood abuse and neglect. Am J Med 1999; 107: 332–9.[CrossRef][Medline]
17. Griffin DW, Bartholomew K. The metaphysics of measurement: the case of adult attachment. Adv Pers Relations 1994; 5: 17–52.
18. Ognibene TC, Collins NL. Adult attachment styles, perceived social support and coping strategies. J Soc Pers Relations 1998; 15: 323–45.[Abstract]
19. Kirkpatrick LA, Hazan C. Attachment styles and close relationships: a four-year prospective study. Pers Relations 1994; 1: 123–42.
20. Hamilton CE. Continuity and discontinuity of attachment from infancy through adolescence. Child Dev 2000; 71: 690–4.[CrossRef][Medline]
21. Waters E, Merrick SK, Treboux D, Crowell J, Albersheim L. Attachment security in infancy and early adulthood: a twenty-year longitudinal study. Child Dev 2000; 71: 684–9.[CrossRef][Medline]
22. Klohnen EC, Oliver JP. Working models of attachment: a theory-based prototype approach. In: Simpson JA, Rholes WS, editors. Attachment theory and close relationships. New York: The Guilford Press; 1998. p. 115–40.
23. Ware JE. SF-36 Physical and Mental Health Summary Scales: a user’s manual. Boston: The Health Institute, New England Medical Center; 1994.
24. Berwick DM, Murphy JM, Goldman PA, Ware JE, Barsky AJ, Weinstein MC. Performance of a five-item mental health screening test. Med Care 1991; 29: 169–76.[CrossRef][Medline]
25. Robins LN, Helzer JE, Croughan J, Ratcliff KS. National Institute of Health Diagnostic Interview Schedule: its history, characteristics and validity. Arch Gen Psychiatry 1981; 38: 381–9.[Abstract/Free Full Text]
26. Clark DO, Von Korff M, Saunders K, Baluch WM, Simon GE. A chronic disease score with empirically derived weights. Med Care 1995; 33: 783–95.[CrossRef][Medline]
27. Diehr P, Yanez D, Ash A, Hornbrook M, Lin DY. Methods for analyzing health care utilization and costs. Annu Rev Public Health 1999; 20: 125–44.[CrossRef][Medline]
28. Von Korff M, Gruman J, Schaefer J, Curry SJ, Wagner EH. Collaborative management of chronic illness. Ann Intern Med 1997; 127: 1097–102.[Abstract/Free Full Text]
29. Smith GR, Rost K, Kashner TM. A trial of the effect of a standardized psychiatric consultation on health outcomes and costs in somatizing patients. Arch Gen Psychiatry 1995; 52: 238–43.[Abstract/Free Full Text]
30. Mickelson KD, Kessler RC, Shaver PR. Adult attachment in a nationally representative sample. J Pers Soc Psychol 1997; 73: 1092–106.[CrossRef][Medline]
31. Bartholomew K. Adult attachment processes: individual and couple perspectives. Br J Med Psychol 1997; 70: 249–63.
32. Dozier M, Cue KL, Barnett L. Clinicians as caregivers: role of attachment organization in treatment. J Consult Clin Psychol 1994; 62: 793–800.[CrossRef][Medline]
33. Katon W, Rutter CM, Lin E, Simon G, Von Korff M, Bush T, Walker E, Ludman E. Are there detectable differences in quality of care or outcomes of depression across primary care providers? Med Care 2000; 38: 552–61.[CrossRef][Medline]
34. Hofer TP, Hayward RA, Greenfield S, Wagner EH, Kaplan SH, Manning WG. The unreliability of individual physician "report cards" for assessing the costs and quality of care of a chronic disease. JAMA 1999; 281: 2098–105.[Abstract/Free Full Text]
35. Kroenke K, Spitzer RL, Williams JB, Linzer M, Hahn SR, deGruy FV III, Brody D. Physical symptoms in primary care: predictors of psychiatric disorders and functional impairment. Arch Fam Med 1994; 3: 774–9.[Abstract/Free Full Text]
36. Dozier M. Attachment organization and treatment use for adults with serious psychopathological disorders. Dev Psychopathol 1990; 2: 47–60.
37. Ciechanowski PS, Katon WJ, Russo JE, Walker EA. The patient-provider relationship: attachment theory and adherence to treatment in diabetes. Am J Psychiatry 2001; 158: 29–35.[Abstract/Free Full Text]
38. Taylor RE, Mann AH, White NJ, Goldberg DP. Attachment style in patient with unexplained physical complaints. Psychol Med 2000; 30: 931–41.[CrossRef][Medline]
39. Taylor RE, Mann AH. Somatization in primary care. J Psychosom Res 1999; 47: 61–6.[CrossRef][Medline]
40. Adshead G. Psychiatric staff as attachment figures: understanding management problems in psychiatric services in light of attachment theory. Br J Psychiatry 1998; 172: 64–9.[Abstract/Free Full Text]
41. Hohmann AA. A contextual model for clinical mental health effectiveness research. Ment Health Serv Res 1999; 1: 83–91.[CrossRef]
42. Hazan C, Shaver P. Romantic love conceptualized as an attachment process. J Pers Soc Psych 1987; 52: 511–24.[CrossRef]
43. Hunter JJ, Maunder RG. Using attachment theory to understand illness behavior. Gen Hosp Psychiatry 2001; 23: 177–82.[CrossRef][Medline]
44. Maunder RG, Hunter JJ. Attachment and psychosomatic medicine: developmental contributions to stress and disease. Psychosom Med 2001; 63: 556–67.[Abstract/Free Full Text]
45. Ciechanowski PS, Hirsch IB, Katon WJ. Interpersonal predictors of HbA₁C in patients with type 1 diabetes. Diabetes Care 2002; 25: 731–6.[Abstract/Free Full Text]
46. Ciechanowski PS, Katon WJ, Russo JE, Dwight-Johnson MM. Association of attachment style to lifetime medically unexplained symptoms in patients with hepatitis c. Psychosomatics 2002; 43: 206–12.[Abstract/Free Full Text]

This article has been cited by other articles:

				P. Oliveira and M. E. Costa Interrelationships of Adult Attachment Orientations, Health Status and Worrying among Fibromyalgia Patients J Health Psychol, November 1, 2009; 14(8): 1184 - 1195. [Abstract] [PDF]

				C. Meisinger, U. Kandler, and K.-H. Ladwig Living Alone Is Associated With an Increased Risk of Type 2 Diabetes Mellitus in Men but Not Women From the General Population: The MONICA/KORA Augsburg Cohort Study Psychosom Med, September 1, 2009; 71(7): 784 - 788. [Abstract] [Full Text] [PDF]

				M. D. Sullivan, P. S. Ciechanowski, J. E. Russo, L. A. Soine, K. Jordan-Keith, H. H. Ting, and J. H. Caldwell Understanding Why Patients Delay Seeking Care for Acute Coronary Syndromes Circ Cardiovasc Qual Outcomes, May 1, 2009; 2(3): 148 - 154. [Abstract] [Full Text] [PDF]

				R. Noyes Jr., S. P. Stuart, and D. B. Watson A Reconceptualization of the Somatoform Disorders Psychosomatics, February 1, 2008; 49(1): 14 - 22. [Abstract] [Full Text] [PDF]

				A. Wearden, K. Perryman, and V. Ward Adult Attachment, Reassurance Seeking and Hypochondriacal Concerns in College Students J Health Psychol, December 1, 2006; 11(6): 877 - 886. [Abstract] [PDF]

				R. J. Waldinger, M. S. Schulz, A. J. Barsky, and D. K. Ahern Mapping the Road From Childhood Trauma to Adult Somatization: The Role of Attachment Psychosom Med, January 1, 2006; 68(1): 129 - 135. [Abstract] [Full Text] [PDF]

				A. Tan, C. Zimmermann, and G. Rodin Interpersonal processes in palliative care: an attachment perspective on the patient-clinician relationship Palliative Medicine, March 1, 2005; 19(2): 143 - 150. [Abstract] [PDF]

				P. Ciechanowski, J. Russo, W. Katon, M. Von Korff, E. Ludman, E. Lin, G. Simon, and T. Bush Influence of Patient Attachment Style on Self-care and Outcomes in Diabetes Psychosom Med, September 1, 2004; 66(5): 720 - 728. [Abstract] [Full Text] [PDF]

This Article Abstract Figures Only 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 Request Permissions Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Ciechanowski, P. S. Articles by Russo, J. E. Search for Related Content PubMed PubMed Citation Articles by Ciechanowski, P. S. Articles by Russo, J. E. Related Collections Health Policy Somatoform Stress and Coping