Duke Social Support and Stress Scale (DUSOCS)

Abstract

The Duke Social Support and Stress Scale (DUSOCS) is a multidimensional self-report psychometric instrument designed to measure both perceived social support and personal social network-related stress within ambulatory medical and psychiatric settings. Developed by George R. Parkerson Jr. and colleagues at the Duke University Medical Center Department of Community and Family Medicine in 1989 and formally validated in 1991, the instrument directly addresses a persistent methodological limitation in behavioral medicine: the confounding or omission of interpersonal conflict within perceived social support measurement. The DUSOCS consists of two parallel sections comprising a total of 24 item prompts: Part I evaluates perceived emotional and instrumental support across 10 distinct social network categories, accompanied by qualitative confidant identification items, while Part II assesses perceived personal stress generated across the identical 10 social network categories, accompanied by primary interpersonal stressor identification items. Responses are recorded on a 3-point categorical rating format (“None” = 0, “Some” = 1, “A lot” = 2; with an “There is No Such Person” option coded as 0), which can be normalized onto 0–100 metric scores across four core subscales: Family Support, Non-Family Support, Family Stress, and Non-Family Stress, alongside summary composite scores. Psychometric investigations indicate adequate to high internal consistency (Cronbach’s alpha coefficients typically ranging between 0.60 and 0.80 across heterogeneous network subscales), robust test-retest reliability (intraclass correlation coefficients between 0.65 and 0.82), and demonstrated construct, convergent, and predictive validity in relation to the Duke Health Profile (DUKE), psychological symptom severity, healthcare utilization metrics, and physical functional capacity. This article provides a comprehensive psychometric review of the DUSOCS, analyzing its historical development, theoretical frameworks, dimensional structure, empirical validity, scoring algorithms, and clinical applications.

Keywords

Duke Social Support and Stress Scale, DUSOCS, social support, interpersonal stress, family support, primary care psychometrics, psychosocial epidemiology, health-related quality of life, functional health status, network-related stress

Authors

The DUSOCS was developed and validated by a multidisciplinary team of physician-scientists, epidemiologists, and biostatisticians within the Department of Community and Family Medicine at Duke University Medical Center, Durham, North Carolina, USA:

• George R. Parkerson Jr., MD, MPH: Professor Emeritus and former Chair of the Department of Community and Family Medicine, Duke University School of Medicine. Dr. Parkerson led the development of the Duke family of health assessment tools, including the Duke Health Profile and Duke Case-Mix System.
• W. E. Broadhead, MD, PhD: Associate Professor, Department of Community and Family Medicine, Duke University Medical Center; an authority on psychosocial determinants of clinical outcomes and social support measurement methodology in primary care.
• C. K. J. Tse, MPH: Biostatistician and epidemiological researcher, Department of Community and Family Medicine, Duke University Medical Center.
• J. L. Michener, MD: Professor and clinical investigator, Department of Community and Family Medicine, Duke University Medical Center.
• L. R. Wu, MD: Physician and health services researcher, Department of Community and Family Medicine, Duke University Medical Center.

Purpose

The clinical and epidemiological objective of the Duke Social Support and Stress Scale is to provide a brief, psychometrically sound, and easily administered measurement of the two distinct interpersonal dynamics that influence human health: perceived social support and perceived social stress. Historically, psychosocial epidemiology frequently examined social support as a singular, protective buffer against physical morbidity and mortality, treating the absence of support as the primary interpersonal hazard. However, clinical family medicine and social psychological research repeatedly documented that social networks operate as dual-edged swords: the very relationships that provide emotional security, material assistance, and social integration are often the most potent sources of emotional turmoil, chronic conflict, unrealistic demands, and behavioral strain.

The DUSOCS was explicitly engineered to address the following theoretical and clinical mandates:

• Decoupling Support and Stress: Traditional assessment tools frequently treated support and stress as polar opposites on a unidimensional continuum. The DUSOCS treats social support and social stress as independent, orthogonal constructs that can simultaneously coexist within the same individual, family, and wider social ecology. A spouse, child, or parent may provide immense emotional solace while simultaneously generating substantial psychological stress.
• Differentiating Family versus Non-Family Spheres: Interpersonal relationships differ fundamentally across kinship boundaries. Family ties often involve obligatory, non-elective bonds characterized by structural interdependence, genetic relatedness, and shared living environments, whereas non-family networks (friends, co-workers, neighbors, church members) are typically more voluntary and elective. The DUSOCS permits granular disaggregation of support and stress into discrete family and non-family components.
• Identification of Confidants and Key Stressors: Beyond general network ratings, the DUSOCS captures qualitative data on whether an individual possesses a singular trusted confidant (a known clinical buffer against severe depression and somatic decompensation) and conversely, whether a single individual serves as the predominant driver of interpersonal stress.
• Clinical Feasibility in Primary Care: Long, exhaustive social network batteries often prove impractical in routine primary care consultations, epidemiological field surveys, and longitudinal clinical trials. The DUSOCS was designed to be completed in under 5 minutes, placing minimal cognitive burden on patients while producing clinically actionable, standardized scores that correlate with health service utilization, medication adherence, and health-related quality of life.

Psychological Construct

The DUSOCS operationalizes the social environment through two multidimensional, overarching psychological constructs: Perceived Social Support and Perceived Social Stress, each mapped systematically onto the structural architecture of the respondent’s current social network.

1. Perceived Social Support

Within the DUSOCS framework, perceived social support reflects an individual’s subjective cognitive appraisal of the adequacy, availability, and quality of emotional, instrumental, and informational aid provided by designated relational ties. Rather than measuring objective social network size (e.g., counting the gross number of living relatives or acquaintances), the scale prioritizes functional perceived support:

• Family Support: Reflects the degree of perceived warmth, assistance, and dependability derived from primary biological or legal relationships, including spouses or domestic partners, children and grandchildren, parents and grandparents, siblings, other blood relatives, and in-laws or relatives by marriage.
• Non-Family Support: Reflects perceived supportive transactions arising from community and institutional social circles, specifically neighbors, workplace colleagues, religious or church affiliates, and general friends.
• Confidant Support: Represents a distinct qualitative protective dimension defined as the availability of a single, deeply trusted individual to whom one can disclose personal difficulties and vulnerabilities without fear of rejection or abandonment.

2. Perceived Social Stress

Perceived social stress encapsulates the negative, antagonistic, or draining properties inherent in interpersonal relationships. It encompasses friction, interpersonal conflict, excessive caregiving demands, criticism, boundary violations, and emotional disappointment generated by relational members:

• Family Stress: Represents the subjective burden, friction, or distress elicited by immediate or extended family members. Because family relationships are difficult to terminate due to legal, moral, and emotional constraints, family stress represents a potent source of chronic, unremitting distress that can trigger endocrine and autonomic dysregulation.
• Non-Family Stress: Quantifies the emotional wear, interpersonal hostility, or social strain encountered across broader communal environments, such as occupational disputes, neighborhood conflicts, or discord within faith-based organizations.
• Primary Stressor: A categorical indicator capturing whether an individual is experiencing an acute focal disruption due to a specific person who acts as the primary interpersonal stressor in their life.

Theoretical Framework

The conceptual architecture of the DUSOCS is grounded in three foundational theoretical traditions within health psychology and psychiatric epidemiology:

1. The Stress-Buffering and Main-Effect Hypotheses

Pioneered by Sheldon Cohen and Thomas A. Wills (1985), social support theory posits two pathways linking social ties to health. The main-effect model proposes that social integration and supportive networks exert generalized beneficial effects on psychological wellbeing and neuroendocrine stability regardless of environmental stress levels. Conversely, the stress-buffering model asserts that social support acts primarily as a protective cushion during times of acute crisis or chronic adversity, attenuating toxic cognitive appraisals and dampening autonomic arousal. The DUSOCS integrates both hypotheses by evaluating ambient supportive resources across regular life domains while determining whether critical buffers (such as an intimate confidant) exist.

2. Rook’s Model of Negative Social Interactions

Karen S. Rook (1984) demonstrated that negative interpersonal interactions exert a disproportionately more powerful negative impact on psychological wellbeing and somatic symptom presentation than the positive effects of supportive interactions. By incorporating a parallel stress inventory mirroring the support network categories, the DUSOCS embodies Rook’s paradigm, recognizing that measuring support without assessing concurrent relational conflict leads to systematic mischaracterizations of an individual’s true psychosocial burden.

3. Allostatic Load and Biological Embedding

From a psychosomatic perspective, sustained interpersonal conflict activates the hypothalamic-pituitary-adrenal (HPA) axis and sympathetic nervous system, accelerating allostatic load (McEwen, 1998). Chronic family stress within clinical populations accelerates vascular inflammation, worsens glycemic control in diabetics, and exacerbates functional disability. The DUSOCS operationalizes interpersonal friction into quantified metrics capable of entering biomedical and epidemiological multivariate regression modeling.

Validity

The psychometric validity of the DUSOCS has been extensively evaluated in ambulatory primary care cohorts, family medicine clinics, and community epidemiological samples.

1. Construct and Factorial Validity

Construct validity was initially established by Parkerson et al. (1989, 1991) during clinical validation studies at Duke University Medical Center. Confirmatory and exploratory factor analyses revealed that items measuring support and items measuring stress loaded on separate, orthogonal latent factors rather than clustering as opposing poles of a single factor. This provided definitive empirical confirmation that interpersonal support and interpersonal stress represent functionally distinct behavioral domains.

2. Convergent and Discriminant Validity

The DUSOCS subscales correlate in theoretically expected directions with established clinical instruments, including the Duke Health Profile (DUKE), the Symptom Checklist-90 (SCL-90), and the Functional Health Status indices:

• Family Support and Non-Family Support: Correlate positively with perceived mental health ($r = 0.25$ to $0.42, p < 0.001$), self-esteem ($r = 0.28$ to $0.38, p < 0.001$), and social functional capacity, while correlating negatively with psychological anxiety and depressive symptomatology.
• Family Stress and Non-Family Stress: Correlate strongly and positively with generalized anxiety ($r = 0.35$ to $0.48, p < 0.001$), depressive affect ($r = 0.32$ to $0.45, p < 0.001$), somatic symptoms ($r = 0.22$ to $0.36, p < 0.01$), and functional disability. Discriminant validity is supported by the relatively modest intercorrelations observed between the Support composite score and the Stress composite score ($r = -0.08$ to $-0.18$, often not statistically significant), demonstrating that social stress is not merely the statistical inverse of social support.

3. Predictive and Criterion-Related Validity

Longitudinal investigations demonstrate that baseline DUSOCS scores predict meaningful downstream medical and behavioral outcomes. In clinical trials evaluating primary care interventions (Parkerson et al., 1989b, 1995), elevated baseline Family Stress was a statistically significant predictor of increased ambulatory healthcare utilization, elevated somatic distress over a 12-month follow-up, and slower post-operative recovery times. Patients reporting both low family support and elevated family stress exhibited the highest risks of persistent functional impairment and depression recurrence.

Reliability

The reliability of the DUSOCS has been demonstrated across diverse demographic cohorts, including socioeconomically disadvantaged populations, rural primary care patients, and elderly clinical samples.

1. Internal Consistency

Because the DUSOCS measures social network members that are naturally heterogeneous (an individual may have a highly supportive spouse but unsupportive distant relatives), traditional internal consistency metrics such as Cronbach’s alpha are expected to be moderate rather than excessively high. Validation studies by Parkerson, Broadhead, and Tse (1991) reported the following coefficients:

• Family Support Subscale: Cronbach’s $\alpha = 0.68\text{ to }0.77$
• Non-Family Support Subscale: Cronbach’s $\alpha = 0.62\text{ to }0.72$
• Family Stress Subscale: Cronbach’s $\alpha = 0.64\text{ to }0.74$
• Non-Family Stress Subscale: Cronbach’s $\alpha = 0.58\text{ to }0.70$
• Total Social Support Composite: Cronbach’s $\alpha = 0.75\text{ to }0.82$
• Total Social Stress Composite: Cronbach’s $\alpha = 0.72\text{ to }0.80$

These values demonstrate solid reliability for composite indices comprising heterogeneous environmental indicators.

2. Test-Retest Reliability

Test-retest stability was evaluated across intervals ranging from one to four weeks among clinically stable primary care outpatients. The resulting intraclass correlation coefficients (ICCs) and Pearson product-moment correlations were:

• Family Support: $r = 0.72\text{ to }0.82$
• Non-Family Support: $r = 0.65\text{ to }0.78$
• Family Stress: $r = 0.68\text{ to }0.80$
• Non-Family Stress: $r = 0.61\text{ to }0.74$

These coefficients illustrate that the scale reliably captures stable perceptions of the interpersonal social environment while remaining adequately sensitive to acute interpersonal transitions or life crises.

Factor Analysis

Structural evaluations of the DUSOCS using both exploratory factor analysis (EFA) with oblique rotation and confirmatory factor analysis (CFA) have substantiated a four-dimensional primary structural model, confirming the independence of the structural and emotional domains:

Factor Architecture

1. Factor 1: Family Support: Marked by strong factor loadings (0.55 to 0.82) on items referencing spouse/partner, children/grandchildren, parents/grandparents, and siblings.
2. Factor 2: Non-Family Support: Dominated by loadings (0.48 to 0.76) on neighbors, co-workers, church members, and other friends.
3. Factor 3: Family Stress: Exhibiting salient factor loadings (0.52 to 0.79) across familial relationship categories when queried regarding stress induction.
4. Factor 4: Non-Family Stress: Marked by loadings (0.45 to 0.74) across extra-familial social sectors.

Goodness-of-fit parameters from confirmatory structural equation modeling indicate acceptable model fit for this four-factor correlated solution (Comparative Fit Index [CFI] > 0.90; Tucker-Lewis Index [TLI] > 0.88; Root Mean Square Error of Approximation [RMSEA] ≤ 0.058). Competing unidimensional models (forcing support and stress into a single bipolar spectrum) or two-factor models (aggregating family and non-family into single support and stress factors) display substantially poorer fit, validating the theoretical necessity of retaining both network locus and affective valence distinctions.

Instrument / Measurement Tool

• Instrument Name: Duke Social Support and Stress Scale (DUSOCS)
• Test Format: Self-administered paper-and-pencil or digital questionnaire; also adaptable for structured clinician-administered interview.
• Target Population: Adults and adolescents aged 15 years and older presenting in primary care, psychiatric, community, or epidemiological settings.
• Item Inventory: 24 prompts divided into two mirrored halves:
  • Part I: People Who Give Personal Support: Items 1–6 (Family Members: spouse/partner, children/grandchildren, parents/grandparents, siblings, other blood relatives, relatives by marriage); Items 7–10 (Non-Family Members: neighbors, co-workers, church members, other friends); Item 11 (Confidant binary question); Item 12 (Confidant category designation).
  • Part II: People Who Cause Personal Stress: Items 1–6 (Family Members: spouse/partner, children/grandchildren, parents/grandparents, siblings, other blood relatives, relatives by marriage); Items 7–10 (Non-Family Members: neighbors, co-workers, church members, other friends); Item 11 (Primary Stressor binary question); Item 12 (Primary Stressor category designation).
• Response Categories and Scoring Weights:
  • “There is No Such Person” = 0
  • “None” = 0
  • “Some” = 1
  • “A lot” = 2
  • Binary Questions (Part I Item 11, Part II Item 11): “No” = 0; “Yes” = 2
• Subscale Score Computations (Normalized 0–100 Metrics):
  • Family Support Score: Sum of raw scores for Part I items 1–6 divided by maximum possible raw score (12), multiplied by 100: $\left(\frac{\sum \text{Part I Items 1 to 6}}{12}\right) \times 100$
  • Non-Family Support Score: Sum of raw scores for Part I items 7–10 divided by maximum possible raw score (8), multiplied by 100: $\left(\frac{\sum \text{Part I Items 7 to 10}}{8}\right) \times 100$
  • Social Support Summary Score: Average of Family Support and Non-Family Support scores, or sum of items 1–10 plus Confidant item (Item 11, scored 0 or 2) divided by maximum raw total (22), multiplied by 100.
  • Family Stress Score: Sum of raw scores for Part II items 1–6 divided by maximum possible raw score (12), multiplied by 100: $\left(\frac{\sum \text{Part II Items 1 to 6}}{12}\right) \times 100$
  • Non-Family Stress Score: Sum of raw scores for Part II items 7–10 divided by maximum possible raw score (8), multiplied by 100: $\left(\frac{\sum \text{Part II Items 7 to 10}}{8}\right) \times 100$
  • Social Stress Summary Score: Average of Family Stress and Non-Family Stress scores, or sum of items 1–10 plus Primary Stressor item (Item 11, scored 0 or 2) divided by maximum raw total (22), multiplied by 100.
• Administration Time: Approximately 3 to 5 minutes.

Permissions & Fee and Test Year

The Duke Social Support and Stress Scale was formulated in 1989 and formally validated in 1991 by George R. Parkerson Jr. and associates at Duke University. The instrument is copyrighted by the Department of Community and Family Medicine, Duke University Medical Center, Durham, North Carolina.

The DUSOCS is accessible without licensing fees for non-profit academic research, university teaching, and clinical primary care implementation, provided that appropriate bibliographic citation is extended to the original authors and Duke University. Commercial utilization, automated digital incorporation into proprietary Electronic Health Record (EHR) platforms, or for-profit pharmaceutical trials may require formal written authorization from the Duke University Office of Licensing and Ventures or the Department of Community and Family Medicine. Detailed user instructions, scoring guidelines, and normative tables are published in the User’s Guide for Duke Health Measures (Parkerson, 2002).

References

• Cohen, S., & Wills, T. A. (1985). Stress, social support, and the buffering hypothesis. Psychological Bulletin, 98(2), 310–357. https://doi.org/10.1037/0033-2909.98.2.310
• McDowell, I. (2006). Measuring Health: A Guide to Rating Scales and Questionnaires (3rd ed.). Oxford University Press. https://doi.org/10.1093/acprof:oso/9780195165678.001.0001
• McEwen, B. S. (1998). Stress, adaptation, and disease: Allostasis and allostatic load. Annals of the New York Academy of Sciences, 840(1), 33–44. https://doi.org/10.1111/j.1749-6632.1998.tb09546.x
• Parkerson, G. R., Jr. (2002). User’s Guide for Duke Health Measures. Durham, NC: Duke University Department of Community and Family Medicine.
• Parkerson, G. R., Jr., Broadhead, W. E., & Tse, C. K. J. (1991). Validation of the Duke Social Support and Stress Scale. Family Medicine, 23(5), 357–360. https://pubmed.ncbi.nlm.nih.gov/1884931/
• Parkerson, G. R., Jr., Broadhead, W. E., & Tse, C. K. J. (1992). Quality of life and functional health of primary care patients. Journal of Clinical Epidemiology, 45(11), 1303–1313. https://doi.org/10.1016/0895-4356(92)90170-U
• Parkerson, G. R., Jr., Broadhead, W. E., & Tse, C. K. J. (1995). Perceived family stress as a predictor of health-related outcomes. Archives of Family Medicine, 4(3), 253–260. https://doi.org/10.1001/archfami.4.3.253
• Parkerson, G. R., Jr., Michener, J. L., Wu, L. R., Finch, J. N., Muhlbaier, L. H., Magruder-Habib, K., Knesper, D. J., & Wagner, E. H. (1989a). Associations among family support, family stress, and personal functional health status. Journal of Clinical Epidemiology, 42(3), 217–229. https://doi.org/10.1016/0895-4356(89)90059-3
• Parkerson, G. R., Jr., Michener, J. L., Wu, L. R., Finch, J. N., Muhlbaier, L. H., Magruder-Habib, K., Knesper, D. J., & Wagner, E. H. (1989b). The effect of a telephone family assessment intervention on the functional health of patients with elevated family stress. Medical Care, 27(7), 680–693. https://doi.org/10.1097/00005650-198907000-00002
• Rook, K. S. (1984). The negative side of social interaction: Impact on psychological well-being. Journal of Personality and Social Psychology, 46(5), 1097–1108. https://doi.org/10.1037/0022-3514.46.5.1097

13. Items of the Scale (Questionnaire)