NIH defines a health disparity as a largely preventable and measurable difference in health that exists between specific population groups. This is a rigorously derived and data-driven determination operationalized by quantifiable metrics, rather than a statement of attribution. It is a scientific starting point that identifies where health outcomes diverge, not a predetermined conclusion about why. Scientific investigations to identify or understand the root causes are separate and distinct scientific questions, which are necessarily multifaceted and complex.
Health disparities may be observed in the risks, prevalence, or problems resulting from specific behaviors, as well as the incidence, prevalence, and mortality from conditions, diseases, and/or disorders.
Understanding and Addressing Health Disparities
On average, low socioeconomic status (SES), rural, and specific racial and/or ethnic minority populations have higher rates of most chronic diseases, medical comorbidities, and other health problems relative to the overall population. Health disparities are also observed among people living with a disability (broadly defined) or who are attracted romantically or sexually to people of the same or both sexes.
SES is a social construct with various indicators, such as educational attainment, employment, and income. There is a need for integrative and holistic research that provides explanations for the mechanisms by which multilevel factors contribute to disease etiology and health outcomes.
Race and ethnicity are also social constructs that should not be applied as a proxy for human genetic variation. The use of race and ethnicity as proxies for human genetic similarity can lead to conflation with genetic heritage.
Addressing health disparities in a meaningful way requires a comprehensive view of how health is maintained, improved, or worsened via modifiable influences, which include but are not limited to:
- Factors including access to high-quality health care and adverse environmental and macro-level exposures (physical, chemical, and/or community-related elements) that may impact individual level factors such as:
- Biological factors and reactions (e.g., age, inherited conditions, stress hormones, metabolic perturbations).
- Gene expression (e.g., social epigenetics).
- Lifestyle and behavioral factors (e.g., tobacco use, physical activity, food consumption, decision-making).
See the NIMHD Research Framework for more information.
What Are Metrics of Health and Health Care Disparities?
Health disparities can be identified based on significantly greater or disproportionate morbidity or premature mortality that is largely preventable on one or more of the following measures:
- Incidence and/or prevalence, including earlier onset of disease or higher prevalence of preclinical disease/biomarkers.
- Premature or excessive mortality from specific conditions.
- Population health disease metrics, such as life expectancy, quality-adjusted life years, or health-related quality of life.
- Condition-specific symptoms on validated self-reported measures that reflect daily functioning in physical, cognitive, or socio-emotional domains.
- Prevalence of short-term and/or long-term preventable complications.
- Prevalence of modifiable risks, health risk behaviors, and adverse clinical outcomes.
- Variations in the access, utilization, availability, and quality of health care services.
How Do Scientists Select an Appropriate Reference Population/Group in Health Disparities Research*?
- The choice of a reference population is important for benchmarking health outcomes and evaluating health disparities. In the case of racial and/or ethnic health disparities science, the reference population in U.S. health disparities research has often utilized the racial majority group (i.e., White persons) as the reference group.*
- The traditional conceptual and analytic practices of using White persons as the reference, and thus, the standard, may perpetuate stereotypes through deficit-based hypotheses and/or interpretations. Moreover, White persons often do not have significantly better outcomes than all other populations.
- The selected reference group in health disparities research must be chiefly guided by the scientific questions and potential for generalizable results. Reference group options include, but may not be limited to:
- The majority population/group in the geographic context.
- The population group with the best health outcome, lowest disease incidence/prevalence, lowest prevalence of risk factors, and/or lowest disease prevalence.
- The population group with the largest sample size (in the study), which is also referred to as the majority-referenced approach.
- Setting absolute targets for outcomes based on population-level goals, such as Healthy People 2030 goals, across all demographic groups, rather than focusing on variation across groups.
- Using a positive deviance approach, which focuses on people within population groups or subgroups who are thriving or showing positive health-related outcomes despite social, economic, and/or environmental challenges, rather than those who are not.
*Similar reference group considerations apply to health disparities research focused on low socioeconomic status groups, limited-resource rural communities, sexual minority groups, and people with disabilities.
Page published February 25, 2026