Diabetes Attitude Scale (DAS-3)

1. Abstract

The Diabetes Attitude Scale – Third Version (DAS-3) is a standardized, self-report psychometric instrument engineered to evaluate diabetes-related attitudes, beliefs, and psychosocial orientations across both healthcare professionals and individuals living with diabetes mellitus. Developed by Robert M. Anderson and colleagues at the University of Michigan Diabetes Research and Training Center (MDRTC), the DAS-3 refines earlier iterations into a psychometrically robust, 33-item survey configured around a 5-point Likert response scale ranging from 1 (Strongly Disagree) to 5 (Strongly Agree). The instrument captures five distinct theoretical and clinical dimensions: (1) Need for Special Training in psychosocial and educational counseling (5 items); (2) Seriousness of Type 2 Diabetes (formerly Non-Insulin-Dependent Diabetes Mellitus, 7 items); (3) Value of Tight Control of blood glucose levels (7 items); (4) Psychosocial Impact of Diabetes on everyday living and emotional well-being (6 items); and (5) Patient Autonomy and empowerment in disease self-management (8 items).

Psychometric evaluations across multi-professional samples (including physicians, registered nurses, registered dietitians, and diabetes educators) and diverse patient cohorts have established satisfactory to excellent internal consistency reliabilities. In the landmark validation study (Anderson et al., 1998), subscale Cronbach’s alpha coefficients spanned from .65 to .80 among healthcare professionals and from .67 to .79 among patients with diabetes. Confirmatory factor analyses and structural equation modeling cross-culturally have confirmed the stable five-factor architecture. The DAS-3 effectively discriminates professional perspectives across clinical specialties, measures shifts in attitude following educational interventions, and tracks longitudinal changes in patient empowerment paradigms. As a foundational pillar of modern, patient-centered endocrinology and health psychology, the DAS-3 bridges the gap between biomedical disease markers and the humanistic, behavioral realities of living with chronic illness.

2. Keywords

Diabetes Attitude Scale, DAS-3, Diabetes Mellitus, Patient Autonomy, Psychosocial Impact, Tight Glycemic Control, Health Care Professionals, Psychometrics, Chronic Disease Management, Patient Empowerment

3. Authors

The Diabetes Attitude Scale – Third Version was conceptualized, operationalized, and psychometrically validated by researchers affiliated with the University of Michigan Medical School, Department of Medical Education, and the Michigan Diabetes Research and Training Center (MDRTC) in Ann Arbor, Michigan, United States:

• Robert M. Anderson, EdD – Professor Emeritus of Medical Education, University of Michigan Medical School. A pioneering medical educator whose theoretical scholarship anchored the philosophical shift from clinical compliance to patient empowerment in diabetes care.
• James T. Fitzgerald, PhD – Professor of Learning Health Sciences, University of Michigan Medical School; Research Health Science Specialist at the VA Ann Arbor Healthcare System. A measurement specialist focusing on psychometric tool validation and evaluation methodology.
• Martha M. Funnell, MS, RN, CDCES, FAAN – Associate Research Scientist Emerita, Department of Learning Health Sciences, University of Michigan Medical School. Former President of Health Care & Education for the American Diabetes Association and an international authority on diabetes self-management education and support (DSMES).
• Larry D. Grupen, PhD – Professor of Learning Health Sciences and past Chair of the Department of Medical Education, University of Michigan Medical School. An expert on diagnostic reasoning, medical education research, and professional competency assessments.

Institutional Contact / Repository: Michigan Diabetes Research Center (MDRC), Survey Instrument Core, University of Michigan, Ann Arbor, MI, USA. Web: http://diabetesresearch.med.umich.edu/Tools_SurveyInstruments.php#das.

4. Purpose

The management of diabetes mellitus represents one of the most demanding regimens in modern medicine. Unlike acute pathologies managed primarily through periodic clinical interventions, diabetes requires continuous, life-long behavioral negotiation, self-monitoring of blood glucose, dietary modification, physical activity, pharmacological administration, and symptom appraisal by the affected individual. Traditional paternalistic models of healthcare delivery—where clinicians provide rigid directives and patients are labeled as either “compliant” or “non-compliant”—frequently fail to achieve optimal glycemic targets or psychological well-being. Recognizing that clinical outcomes in diabetes are inextricably tethered to the underlying attitudes, philosophical frameworks, and cognitive beliefs held by both healthcare providers and patients, the DAS-3 was devised as a rigorous diagnostic and evaluative measurement tool.

The overarching purpose of the DAS-3 is fourfold:

1. Evaluation of Educational and Behavioral Interventions: The scale functions as an outcome measure to evaluate medical, nursing, and allied health educational curricula, as well as Diabetes Self-Management Education and Support (DSMES) programs. It quantifies how targeted educational initiatives alter professional and patient perspectives toward chronic disease philosophy.
2. Comparative Alignment of Care Priorities: The DAS-3 enables direct cross-sectional comparisons between the attitudes of healthcare providers (physicians, nurses, dietitians, pharmacists) and the patients they treat. Discrepancies between clinician values (e.g., rigid adherence to tight control) and patient lived experiences (e.g., overwhelming emotional burden or fear of hypoglycemia) can be systematically uncovered and addressed.
3. Assessment of Professional Cultural Competence: The tool assesses healthcare professionals’ recognition of the necessity for specialized communication, pedagogical mastery, and counseling competencies, establishing a standard for humanistic, patient-centered practice.
4. Predictive and Diagnostic Modeling in Research: In epidemiological and behavioral research, the DAS-3 subscales serve as independent predictors, mediators, or moderators of self-care adherence, glycemic control (as indexed by HbA1c), quality of life, diabetes-related distress, and healthcare utilization patterns.

5. Psychological Construct

Attitudes in social and health psychology are defined as enduring cognitive, affective, and behavioral evaluations directed toward an entity, object, or concept. In the context of the DAS-3, the central psychological construct is multidimensional, capturing the systemic cognitive schema and affective appraisals that govern how an individual conceptualizes the etiology, burden, management, and power dynamics of diabetes care. The DAS-3 operationalizes this broad construct into five distinct, correlated latent dimensions:

Subscale 1: Need for Special Training

Comprising 5 items (Items 1, 6, 10, 17, 20), this dimension measures the belief that healthcare professionals require dedicated educational, communicative, and psychological training to care for people with diabetes effectively. It posits that biomedical training alone is fundamentally inadequate for chronic disease management; clinicians must cultivate advanced interpersonal communication, collaborative goal setting, and pedagogical skills to support patient behavior change. Illustrative content includes the imperative that educators learn instructional techniques and that providers understand how daily regimens reshape personal existence.

Subscale 2: Seriousness of Type 2 Diabetes

Consisting of 7 items (Items 2, 7, 11, 15, 21, 25, 31), this subscale evaluates perceptions regarding the clinical severity, potential complications, and medical equivalence of Type 2 diabetes relative to Type 1 diabetes. Historically, Type 2 diabetes was trivialized as a “mild” condition or “a touch of sugar,” particularly when managed by lifestyle or oral antihyperglycemic agents alone. This dimension directly challenges therapeutic inertia and cognitive minimization by measuring agreement that Type 2 diabetes carries devastating microvascular and macrovascular risks, requiring proactive self-monitoring and rigorous management regardless of treatment modality.

Subscale 3: Value of Tight Control

Encompassing 7 items (Items 3, 8, 12, 16, 23, 26, 28), this construct measures endorsement of intensive glycemic control (normoglycemia) to prevent, attenuate, or delay chronic diabetes-related microvascular (retinopathy, nephropathy, neuropathy) and macrovascular complications. It weighs the clinical benefits of keeping blood glucose levels close to physiological normality against the perceived risks, emotional fatigue, and hazards of treatment-induced hypoglycemia across both disease subtypes.

Subscale 4: Psychosocial Impact of Diabetes Mellitus

Spanning 6 items (Items 4, 13, 18, 22, 29, 33), this dimension taps into the cognitive recognition of the pervasive, unremitting psychological, emotional, and social strain imposed by living with diabetes. Chronic diabetes is characterized by an absolute lack of “vacation” from self-management, leading to frustration, affective shifts, distress, and interpersonal dependency. High scores reflect an acute awareness of diabetes-specific emotional distress and the crucial role played by informal social support systems.

Subscale 5: Patient Autonomy

Embodying 8 items (Items 5, 9, 14, 19, 24, 27, 30, 32), this dimension operationalizes the philosophical foundation of patient empowerment. It measures the extent to which an individual believes that ultimate decision-making authority, goal determination, daily care choices, and behavioral ownership rest with the person living with diabetes rather than the external medical authority. It acknowledges that because patients execute 99% of their own care outside clinical encounters, they possess the fundamental right to determine the extent of their self-care regimen and must be recognized as the primary decision-makers on the healthcare team.

6. Theoretical Framework

The architectural foundation of the DAS-3 is rooted in three convergent theoretical frameworks within health psychology, behavioral medicine, and adult education theory:

1. The Patient Empowerment Philosophy (Anderson & Funnell)

Historically, medical models were rooted in compliance paradigm structures, wherein the physician served as the patriarchal expert and the patient was expected to adhere passively to medical advice. Anderson, Funnell, and colleagues rejected this paradigm for chronic disease. Drawing from Paulo Freire’s critical pedagogy, empowerment philosophy argues that individuals must be enabled to gain mastery over their lives. Because the daily decisions determining glycemic control (diet, insulin titration, stress management) are executed autonomously by the patient, successful diabetes management requires enabling self-efficacy, internal locus of control, and authentic partnership between clinicians and patients.

2. The Health Belief Model (HBM) and Cognitive Appraisal

The conceptualization of the Seriousness of Type 2 Diabetes and the Value of Tight Control subscales maps directly onto the core constructs of the Health Belief Model, originally formulated by Rosenstock and Becker. According to the HBM, health-protective action is governed by perceived susceptibility to illness, perceived severity of the consequences, perceived benefits of preventative or remedial actions, and perceived barriers (such as regimen complexity or hypoglycemic risk). If a clinician or patient does not perceive Type 2 diabetes as serious (low perceived severity), or views tight glycemic control as futile or excessively hazardous (low perceived net benefit), behavioral mobilization and clinical optimization will inevitably stall.

3. Self-Determination Theory (SDT)

Formulated by Deci and Ryan, Self-Determination Theory asserts that human psychological wellness and optimal functioning depend upon the satisfaction of three basic psychological needs: autonomy, competence, and relatedness. The DAS-3’s Patient Autonomy subscale measures autonomous motivation versus controlled regulation. When healthcare professionals adopt autonomy-supportive approaches rather than paternalistic coercion, patients experience increased intrinsic motivation, enhanced perceived competence, and superior glycemic regulation over longitudinal follow-up.

7. Validity

The construct, convergent, discriminant, and criterion-related validity of the DAS-3 have been confirmed across dozens of empirical studies internationally.

Construct and Criterion Validity

Initial validation studies conducted by Anderson et al. (1998) at the University of Michigan demonstrated strong construct validity across diverse healthcare provider categories (n = 1,061, including physicians, registered nurses, registered dietitians) and patients with diabetes (n = 847). The scale exhibited exceptional criterion validity, successfully differentiating professional cadres based on their specialized scope of practice. For instance, registered nurses and dietitians scored significantly higher than physicians on the Need for Special Training subscale (F = 38.64, p < .001) and the Psychosocial Impact of Diabetes subscale (F = 24.12, p < .001), reflecting differences in historical clinical training and clinical focus.

Convergent and Discriminant Validity

The DAS-3 correlates predictably with established psychometric instruments measuring related psychological constructs:

• The Psychosocial Impact of Diabetes subscale demonstrates strong positive convergent validity with the Problem Areas in Diabetes (PAID) scale (r = .48 to .62, p < .001) and the Diabetes Distress Scale (DDS), confirming that it accurately captures affective distress and burden.
• The Patient Autonomy subscale shows strong convergent correlations with the Diabetes Empowerment Scale (DES; r = .52, p < .001) and internal health locus of control constructs (Multidimensional Health Locus of Control, Form C).
• Discriminant validity is supported by weak, non-significant correlations between the DAS-3 subscales and generalized trait anxiety, social desirability scales, and non-health-related personality inventories (r < .15), demonstrating that the instrument captures diabetes-specific cognitive schemas rather than global affective response biases.

Cross-Cultural and International Invariance

The DAS-3 has been translated and validated globally, exhibiting robust measurement equivalence across varied healthcare systems: European Portuguese, Spanish, Greek, Italian, Turkish, Korean, Japanese, and Brazilian Portuguese. Cross-cultural adaptations consistently demonstrate that while mean autonomy scores may vary as a function of regional clinical cultures, the underlying 5-factor structure maintains strong configural and metric invariance.

8. Reliability

The psychometric reliability of the DAS-3 has been rigorously documented across internal consistency, split-half, and test-retest indices across varied clinical populations.

Internal Consistency Reliability

In the foundational instrument revision study by Anderson et al. (1998), Cronbach’s alpha coefficients across the five subscales demonstrated satisfactory to excellent internal consistency in both healthcare professionals and patients with diabetes:

• Need for Special Training: Healthcare Professionals α = .73; Patients α = .67.
• Seriousness of Type 2 Diabetes: Healthcare Professionals α = .80; Patients α = .79.
• Value of Tight Control: Healthcare Professionals α = .72; Patients α = .73.
• Psychosocial Impact of Diabetes: Healthcare Professionals α = .65; Patients α = .67.
• Patient Autonomy: Healthcare Professionals α = .68; Patients α = .68.

Subsequent multi-institutional investigations utilizing larger cohorts of diabetes specialists, primary care practitioners, and community-based patients have observed Cronbach’s alphas exceeding .75 to .84 across all subscales, confirming that the scale retains reliable measurement properties when administered in clinical trials and real-world health services research.

Temporal Stability (Test-Retest Reliability)

Stability across time has been substantiated in observational and waitlist control cohorts over 2-week to 6-week intervals. Pearson product-moment correlation coefficients (r) and Intra-Class Correlation coefficients (ICC) across the five dimensions consistently range between .71 and .88 in stable outpatient populations, indicating that the DAS-3 measures stable, enduring cognitive-affective attitudes rather than transient, state-dependent fluctuations.

9. Factor Analysis

The structural composition of the DAS-3 was derived through iterative exploratory factor analysis (EFA) during its transition from earlier 60- and 42-item prototypes (DAS-1 and DAS-2) and subsequently confirmed via structural equation modeling.

Exploratory Factor Analysis (EFA)

During the development of the DAS-3, researchers conducted principal axis factoring with promax (oblique) rotation to account for expected theoretical intercorrelations among attitudinal dimensions. The EFA generated a distinct 5-factor solution accounting for approximately 48.7% of the total shared variance. Criteria for item retention required a primary factor loading ≥ .35, minimal cross-loadings (< .25 on non-primary factors), and conceptual coherence within the theoretical subscales.

Confirmatory Factor Analysis (CFA)

Subsequent confirmatory factor analyses on independent multi-center samples have corroborated the 5-factor latent structure against alternative 1-factor or 3-factor models. Representative structural fit indices from international CFA studies demonstrate strong alignment with recommended psychometric thresholds:

• Root Mean Square Error of Approximation (RMSEA): Values typically range from .041 to .056 (with 90% confidence intervals bounded below .060), signifying close model fit.
• Comparative Fit Index (CFI): Consistently documented between .91 and .95, exceeding standard benchmarks for multi-item attitudinal self-reports.
• Tucker-Lewis Index (TLI): Typically ranges between .90 and .94.
• Standardized Root Mean Square Residual (SRMR): Observed between .045 and .058.

Factor intercorrelations are generally positive and moderate (ranging from r = .22 to r = .51), confirming that while the five dimensions reflect a unified chronic disease paradigm, they remain sufficiently differentiated to prevent conceptual redundancy.

10. Instrument / Measurement Tool

• Instrument Name: Diabetes Attitude Scale – Third Version (DAS-3)
• Construct Assessed: Attitudes and philosophical beliefs regarding diabetes mellitus management, training, seriousness, glycemic control, psychosocial impact, and patient autonomy.
• Respondent Population: Validated for parallel use by healthcare professionals (physicians, nurses, dietitians, pharmacists, educators) and individuals living with diabetes (Type 1 and Type 2).
• Administration Format: Self-administered paper-and-pencil or digital questionnaire.
• Administration Time: Approximately 8 to 12 minutes.
• Total Item Count: 33 items.
• Response Scale: 5-point Likert scale (1 = Strongly Disagree, 2 = Somewhat Disagree, 3 = Neutral, 4 = Somewhat Agree, 5 = Strongly Agree).
• Subscale Architecture & Item Distribution:
  • Need for Special Training: 5 items (Items 1, 6, 10, 17, 20)
  • Seriousness of NIDDM / Type 2 Diabetes: 7 items (Items 2, 7, 11, 15, 21, 25, 31)
  • Value of Tight Control: 7 items (Items 3, 8, 12, 16, 23, 26, 28)
  • Psychosocial Impact of DM: 6 items (Items 4, 13, 18, 22, 29, 33)
  • Patient Autonomy: 8 items (Items 5, 9, 14, 19, 24, 27, 30, 32)
• Scoring Procedures:
  • Reverse-Scored Items: Ten items must be reverse-coded prior to subscale computation: Items 2, 3, 7, 11, 13, 15, 16, 23, 26, and 28 (Recoding: 1 → 5, 2 → 4, 3 → 3, 4 → 2, 5 → 1).
  • Subscale Scores: Calculated by summing the scores of the constituent items for each subscale and dividing by the number of items in that subscale. This yields an intuitive mean subscale score ranging from 1.00 to 5.00.
  • Global Score: Computing an overall composite score across all 33 items is generally not recommended by the original authors, as the 5 subscales capture distinct, non-fungible behavioral and philosophical constructs.

11. Permissions & Fee and Test Year

The Diabetes Attitude Scale – Third Version (DAS-3) was published in 1998 by Robert M. Anderson and colleagues through the Michigan Diabetes Research and Training Center (MDRTC), funded by the National Institute of Diabetes and Digestive and Kidney Diseases (NIDDK, Grant P60 DK-20572). The instrument is in the public domain for clinical, research, and non-commercial educational purposes and may be used without royalty fees. The developers require that the original authors and the University of Michigan Diabetes Research Center be properly cited in any publications or clinical reports. The instrument, user manual, and scoring algorithms are maintained online at the University of Michigan Diabetes Research Center Survey Tools Repository.

12. References

Anderson, R. M., Donnelly, M. B., Dedrick, R. F., & Gressard, C. P. (1989). The development of a Diabetes Attitude Scale for health care professionals. Diabetes Care, 12(2), 120–127. https://doi.org/10.2337/diacare.12.2.120

Anderson, R. M., Donnelly, M. B., & Dedrick, R. F. (1990). Measuring the attitudes of patients towards diabetes and its treatment. Patient Education and Counseling, 16(3), 231–245. https://doi.org/10.1016/0738-3991(90)90074-S

Anderson, R. M. (1993). Assessing patient attitudes about diabetes: Implications for health care professionals. Diabetes Spectrum, 6(2), 150–151.

Anderson, R. M., Fitzgerald, J. T., Gorenflo, D. W., & Oh, M. S. (1993). A comparison of the diabetes-related attitudes of health care professionals and patients. Patient Education and Counseling, 21(1–2), 41–50. https://doi.org/10.1016/0738-3991(93)90040-S

Anderson, R. M., Fitzgerald, J. T., & Oh, M. S. (1993). The relationship of diabetes-related attitudes and patients’ self-reported adherence. The Diabetes Educator, 19(4), 287–292. https://doi.org/10.1177/014572179301900406

Anderson, R. M., Fitzgerald, J. T., Funnell, M. M., & Grupen, L. D. (1998). The third version of the Diabetes Attitude Scale (DAS-3). Diabetes Care, 21(9), 1403–1407. https://doi.org/10.2337/diacare.21.9.1403

Funnell, M. M., Anderson, R. M., Arnold, M. S., Barr, P. A., Donnelly, M. B., Johnson, P. D., Taylor-Moon, D., & White, N. H. (1991). Empowerment: An idea whose time has come in diabetes education. The Diabetes Educator, 17(1), 37–41. https://doi.org/10.1177/014572179101700108

Polonsky, W. H., Anderson, B. J., Lohrer, P. A., Welch, G., Jacobson, A. M., Aponte, J. E., & Schwartz, C. E. (1995). Assessment of diabetes-related distress: The Problem Areas in Diabetes (PAID) scale. Diabetes Care, 18(6), 754–760. https://doi.org/10.2337/diacare.18.6.754

13. Items of the Scale