Products & safety
Social determinants of health: structural conditions, causal questions and policy evaluations
On this page
Social determinants concern living conditions and the systems that distribute power, money and resources. WHO’s 2025 report places economic inequality, structural discrimination, public services and governance within that account. This guide examines how to distinguish these frameworks from evidence that a particular intervention changes health.[2]
Define the level of action before choosing an intervention
| Level | Examples and meaning | Interpretation |
|---|---|---|
| Structural determinants | Governance, social and economic policies, and the processes that distribute education, income and social position. | These shape the distribution of opportunities and exposures.[1] |
| Intermediary determinants | Material conditions, psychosocial circumstances, behaviors, biological factors and the health system. | Housing quality and care access can be pathways linking social position with health.[1] |
| Social care activities | Awareness, adjustment, assistance, alignment and advocacy within health-care organizations. | Asking about a need, adapting care and changing community resources are different activities.[3] |
Several causal directions can operate together
WHO’s conceptual framework distinguishes social causation, health-related social selection and life-course explanations. Circumstances can influence health, while illness can also affect income, employment or social position. Exposures can accumulate and interact over time. These are complementary possibilities, not a choice between blaming an individual and assuming one structural cause explains everything. A group association does not determine an individual’s future or identify the effect of a specific policy. The framework helps formulate questions; evaluation must test the proposed intervention.[1]
A link worker can help someone make a connection
In England, a social-prescribing link worker gives time to understand what matters to a person and helps connect them with community groups or practical services. Examples include social activities, debt advice and housing support. NHS England distinguishes this personalised help from signposting: giving someone information or a directory. Referral can come through several local services, and self-referral is encouraged. Ask your GP practice or local service what is available and how to access it. The role describes support with making connections; it does not establish that a particular service is available or that health will improve.[9]
Evidence & uncertainty
An income–health association is not the expected effect of a cash programme. In a randomized US study, additional income changed some outcomes without improving several physical-health measures over three years. Population, intervention, outcome and follow-up must remain part of the claim.[5]
Selected-source overview · Independent review pendingPublished questions
Are structural determinants the same as a person’s immediate living conditions?
Selected-source overview · Independent review pending. Read the evidence context.
They are related but distinct. Structural processes shape social position and resources; intermediary conditions include housing, stress, behavior and care access. Separating levels helps explain pathways without assuming that changing one factor changes everything.[1]
Evidence behind this answer
- A Conceptual Framework for Action on the Social Determinants of Health
agency conceptual framework · 2010
Selected original report PDF passages inspected through WHO regional mirror; canonical publication record inspected; no exhaustive report appraisal · Structural/intermediary categories, selection/causation/life course and model limits
This answer summarizes selected sources. An independent appraisal of the complete evidence has not been completed.
Does completing a social-needs questionnaire solve the need?
Selected-source overview · Independent review pending. Read the evidence context.
No. Awareness, care adjustment, assistance and community action are different activities. Asking or referring does not show that support was wanted, available or received. The person’s priorities and the actual service pathway matter.[3]
Evidence behind this answer
- Five Health Care Sector Activities to Better Integrate Social Care
national academies consensus report chapter · 2019
Selected original full chapter passages inspected · Chapter2: five activities, transport examples, asking alone, shared decisions and assistance pathways
This answer summarizes selected sources. An independent appraisal of the complete evidence has not been completed.
If housing stability improves, must generic quality of life improve too?
Selected-source overview · Independent review pending. Read the evidence context.
No. The Canadian trial improved stable housing days without a clear EQ-5D advantage over usual services at two years. Outcomes measure different benefits, and one combined programme’s result should not be generalized to every housing intervention.[4]
Evidence behind this answer
- Scattered-site housing with rent supplements and intensive case management: randomized trial
randomized controlled trial · 2015
Original PubMed methods/results abstract inspected; no independent full-text appraisal · 1198adults/four cities;24months; stable-housing primary and EQ-5D secondary null
This answer summarizes selected sources. An independent appraisal of the complete evidence has not been completed.
Did the US cash study show that income never matters for health?
Selected-source overview · Independent review pending. Read the evidence context.
No. It found temporary gains and several null health outcomes in younger adults over three years. It does not test lifelong income, all age groups or every policy. Its randomized comparison is different from the cross-sectional income–health correlation.[5]
Evidence behind this answer
- Does Income Affect Health? Evidence from a Randomized Controlled Trial of a Guaranteed Income
randomized experiment working paper · 2024
Selected original investigator-hosted working-paper methods/results inspected; NBER indexed canonical record checked; no independent full-text or peer-review appraisal · August13,2026original author version;1000/2000adults;3years; temporary stress/food gains, care use, physical/biomarker nulls
This answer summarizes selected sources. An independent appraisal of the complete evidence has not been completed.
Does lower psychotropic medication use prove universal mental-health recovery?
Selected-source overview · Independent review pending. Read the evidence context.
No. Medication use is a register-based outcome, not a direct measure of every symptom or quality of care. Finland’s results were outcome- and programme-specific, with no detected effect on other health outcomes and uncertain separation of mechanisms.[6]
Evidence behind this answer
- Health effects of cash transfers: Evidence from the Finnish basic income experiment
randomized experiment register analysis · 2025
Original investigator/agency-hosted journal abstract inspected; publisher full report unavailable; no independent full-text appraisal · 2000unemployed participants;2017–2018; psychotropic medicines and other health nulls; mechanism interpretation
This answer summarizes selected sources. An independent appraisal of the complete evidence has not been completed.
Did the early-childhood study isolate education from nutrition and care?
Selected-source overview · Independent review pending. Read the evidence context.
No. The programme combined several components. Its selected adult-health findings, sex differences, small samples and missing follow-up cannot establish which ingredient caused the effect or guarantee the same result for another educational programme.[7]
Evidence behind this answer
- Early Childhood Investments Substantially Boost Adult Health
randomized trial long term followup · 2014
Selected original published report methods/results inspected through academic teaching copy; canonical indexed record checked; no independent full-text appraisal · Original111children; mid30s biomarker analysis; early package versus school-age null; sex, attrition and multiple-outcome limits
This answer summarizes selected sources. An independent appraisal of the complete evidence has not been completed.
Did the insulation trial conclusively reduce respiratory hospital admissions?
Selected-source overview · Independent review pending. Read the evidence context.
No. Reported health and several everyday outcomes improved, but the hospital-admission estimate was uncertain. Observing a favorable point estimate does not turn an inconclusive comparison into a proven reduction.[8]
Evidence behind this answer
- Effect of insulating existing houses on health inequality: cluster randomised study
cluster randomized controlled trial · 2007
Selected original published PDF methods/results inspected through investigator programme; canonical indexed record checked; no independent full-text appraisal · 1350households/4407residents; low-income respiratory-symptom setting; reported outcomes and hospital uncertainty
This answer summarizes selected sources. An independent appraisal of the complete evidence has not been completed.
Can immediate assistance and wider social action both matter?
Selected-source overview · Independent review pending. Read the evidence context.
Yes. Help with a current difficulty and policies changing how resources and opportunities are distributed address different levels. WHO’s framework supports coordinated action, while each programme still needs local evaluation of reach, outcomes and burdens.[2]
Evidence behind this answer
- World report on social determinants of health equity
agency global report · 2025
Original publication overview and selected official report-summary passages inspected; full report PDF unavailable; no full-report appraisal · Publication overview and official report summary: structural resources and four action areas
This answer summarizes selected sources. An independent appraisal of the complete evidence has not been completed.
What does a social-prescribing link worker do?
Selected-source overview · Independent review pending. Read the evidence context.
The worker listens to what matters to you and supports connection with community groups or practical services. This can involve more help than simply giving you a directory. Local arrangements in England vary.[9]
Evidence behind this answer
- Social prescribing – frequently asked questions
official service policy explanation · Date not stated
Relevant complete original sections read by child and parent through fresh Firecrawl retrieval after native verification shell. No clinical effectiveness, historical workforce-target achievement or universal access claim adopted. Main social-prescribing page conservatively shares this work allowance; its native shell was not a body read. · Complete original role, referral routes, community activities/services and signposting explanations
This answer summarizes selected sources. An independent appraisal of the complete evidence has not been completed.
A housing outcome and a health outcome can differ
A Canadian trial randomized 1,198 adults experiencing homelessness, with mental illness and moderate support needs, to rent-supplemented scattered-site housing plus intensive case management or usual available services. Over 24 months, the intervention increased the percentage of days stably housed. Generic quality of life measured with EQ-5D did not improve significantly relative to the comparison group. Housing stability is a valuable outcome even when one broad health measure does not clearly change. This combined programme does not isolate the contribution of housing versus case management or apply to every population experiencing homelessness.[4]
One cash trial found temporary gains and several null health results
The August 2026 working-paper version of a US randomized study compared 1,000 low-income adults aged twenty-one to forty receiving monthly cash for three years with 2,000 receiving a much smaller payment. It reported early improvements in stress and food security that faded, and increased hospital, emergency-department and dental use. Self-reported physical health and clinical biomarkers did not improve; mental-health benefits did not persist after the first year. More care use is not itself proof of better health or of harm. The young population, pandemic period and three-year horizon limit wider inference.[5]
Another income experiment measured medication and service use
A 2025 analysis of Finland’s 2017–2018 experiment used register data for 2,000 randomly selected unemployed adults offered guaranteed income. The programme changed benefit rules and increased average income. Psychotropic medication use fell by about eight to eleven percent; other health outcomes did not show detected effects. Reduced medicine use and a possible decline in specialist mental-health visits are different endpoints from a symptom scale or universal recovery. The investigators attributed the findings mainly to higher income because many participants continued claiming unemployment benefits. That mechanism interpretation is not separately randomized.[6]
An early-childhood package had selected effects decades later
The Carolina Abecedarian study originally randomized 111 children from families facing disadvantage to an intensive early-childhood programme or control. Education, caregiving, nutrition and pediatric care were combined. A 2014 analysis in participants’ mid-thirties found lower selected cardiovascular and metabolic risk factors, with particularly strong evidence among men. The later school-age intervention did not show comparable adult-health effects. Small samples, missing follow-up data and many outcomes required statistical adjustments. These results concern an intensive historical programme, rather than proving that any educational policy prevents every adult disease or identifying one effective ingredient.[7]
Changing a material exposure can be evaluated directly
A New Zealand cluster-randomized study included 1,350 households and 4,407 people in low-income communities, with at least one resident reporting respiratory symptoms. Retrofitted insulation made homes somewhat warmer and drier, while improving reported health and reducing wheeze and time away from school or work. Respiratory hospital admissions were lower, but that estimate was not statistically conclusive. Participants could see that insulation had been installed, so subjective outcomes need that context. The findings evaluate this housing package and setting; they do not establish an identical benefit from every retrofit or relocation.[8]
Policy frameworks and local programmes answer different questions
WHO’s 2025 agenda includes investment in social infrastructure and public services, action on structural discrimination, and governance supporting local resources and community engagement. These are policy directions, not a single tested package. Clinical assistance can address an immediate difficulty while wider change addresses its distribution. Neither replaces the need to examine who is reached, what changes, who bears burdens and which outcomes remain uncertain. A programme can improve a material condition without rapidly changing every clinical endpoint, and a null endpoint does not erase an improvement in another outcome.[2]
References
- A Conceptual Framework for Action on the Social Determinants of Health
Agency conceptual framework
Orielle Solar and Alec Irwin; World Health Organization · Content/publication date: 2010
Accessed: 2026-10-14 UTC · Selected original report PDF passages inspected through WHO regional mirror; canonical publication record inspected; no exhaustive report appraisal
Cited sections: Structural/intermediary categories, selection/causation/life course and model limitsCanonical WHO page listsJuly13,2010; report printed2010 withOctoberforeword. Year precision retained rather than treating page release as exact publication of all contents. Original mirror is the same report, not an additional source allowance.
- World report on social determinants of health equity
Agency global report
World Health Organization · Content/publication date: 2025
Accessed: 2026-10-14 UTC · Original publication overview and selected official report-summary passages inspected; full report PDF unavailable; no full-report appraisal
Cited sections: Publication overview and official report summary: structural resources and four action areasReport-summary team page represents the same report. No prevalence percentage or causal ranking of all determinants inferred. Recommendations are policy framework, not an independently assessed single intervention.
- Five Health Care Sector Activities to Better Integrate Social Care
National academies consensus report chapter
National Academies of Sciences, Engineering, and Medicine · Content/publication date: 2019
Accessed: 2026-10-14 UTC · Selected original full chapter passages inspected
Cited sections: Chapter2: five activities, transport examples, asking alone, shared decisions and assistance pathwaysWeb chapter4 is the genuine authored Chapter2 of report25467, not an arbitrary URL variant. A dated consensus framework and selected evidence discussion, not an exhaustive2026comparative intervention review.
- Scattered-site housing with rent supplements and intensive case management: randomized trial
Randomized controlled trial
Stergiopoulos and colleagues; JAMA · Content/publication date: 2015
Accessed: 2026-10-14 UTC · Original PubMed methods/results abstract inspected; no independent full-text appraisal
Cited sections: 1198adults/four cities;24months; stable-housing primary and EQ-5D secondary nullHousing plus intensive case management for moderate support needs; other AtHome/ChezSoi publications can share participants and are not additional independent replications.
- Does Income Affect Health? Evidence from a Randomized Controlled Trial of a Guaranteed Income
Randomized experiment working paper
Miller, Rhodes, Bartik, Broockman, Krause and Vivalt; NBER Working Paper32711 · Content/publication date: 2024
Accessed: 2026-10-14 UTC · Selected original investigator-hosted working-paper methods/results inspected; NBER indexed canonical record checked; no independent full-text or peer-review appraisal
Cited sections: August13,2026original author version;1000/2000adults;3years; temporary stress/food gains, care use, physical/biomarker nullsOriginal paperJuly2024; inspected author versiondatedAugust13,2026. NBER indexed PDF can reflectJanuary2026revision. One report/one budget; author copy is not added as independent evidence. Author page lists conditional AER acceptance, not completed publication.
- Health effects of cash transfers: Evidence from the Finnish basic income experiment
Randomized experiment register analysis
Hämäläinen, Simanainen and Verho; Journal of Public Economics; Kela · Content/publication date: 2025
Accessed: 2026-10-14 UTC · Original investigator/agency-hosted journal abstract inspected; publisher full report unavailable; no independent full-text appraisal
Cited sections: 2000unemployed participants;2017–2018; psychotropic medicines and other health nulls; mechanism interpretationOriginal journal onlineSeptember17,2025 verified via VATT bibliographic record; Journal of Public Economics250. Lower drug use is not equated to universal symptom recovery. Conditionality and income mechanisms not separately randomized.
- Early Childhood Investments Substantially Boost Adult Health
Randomized trial long term followup
Campbell and colleagues; Science · Content/publication date: 2014
Accessed: 2026-10-14 UTC · Selected original published report methods/results inspected through academic teaching copy; canonical indexed record checked; no independent full-text appraisal
Cited sections: Original111children; mid30s biomarker analysis; early package versus school-age null; sex, attrition and multiple-outcome limitsAcademic copy https://dmessinger.nyc3.cdn.digitaloceanspaces.com/www/c_c/rsrcs/rdgs/emot/1478.full.pdf is this same report, not another source or allowance. Historical setting and intervention package; missing biomedical follow-up not ignored.
- Effect of insulating existing houses on health inequality: cluster randomised study
Cluster randomized controlled trial
Howden-Chapman and colleagues; BMJ · Content/publication date: 2007
Accessed: 2026-10-14 UTC · Selected original published PDF methods/results inspected through investigator programme; canonical indexed record checked; no independent full-text appraisal
Cited sections: 1350households/4407residents; low-income respiratory-symptom setting; reported outcomes and hospital uncertaintyOriginal PDF datedFebruary26,2007online; March2007issue. Investigator programme copy is this same report, not an additional source allowance. Subjective outcomes unblinded; later table6editorial correction identified, no corrected physician-row values extracted.
- Social prescribing – frequently asked questions
Official service policy explanation
NHS England · Content/publication date: not stated
Accessed: 2026-10-20 UTC · Relevant complete original sections read by child and parent through fresh Firecrawl retrieval after native verification shell. No clinical effectiveness, historical workforce-target achievement or universal access claim adopted. Main social-prescribing page conservatively shares this work allowance; its native shell was not a body read.
Cited sections: Complete original role, referral routes, community activities/services and signposting explanations
Scope & review
- Population
- Individuals and populations affected by social conditions; selected Canadian housing, US young-adult cash/early-life and Finnish unemployed-adult examples
- Exposure or intervention
- Structural and intermediary conditions; health-sector social care; evaluated housing, income, early-childhood and insulation packages
- Comparison
- Source-specific randomized usual services, smaller cash payment or control conditions; frameworks are not trials
- Outcomes
- Housing stability, medication/service use, subjective and clinical health, early-life biomarkers, material conditions and uncertainty
- Timeframe
- Months, two or three-year programmes and adult follow-up decades after childhood; source versions checked October 14, 2026
- Jurisdiction
- Global WHO concepts; dated US consensus framework and original setting-specific interventions; eligibility, benefit systems and delivery differ locally
Selected-source educational explanation. Official guidance, original study findings and editorial interpretation are distinguished. No exhaustive search, independent systematic synthesis or qualified medical review has been completed.
Sources checked on their individual access dates. Selected sources were checked; ongoing surveillance and qualified review have separate statuses on the coverage page.
Revision history
- 2026-10-14 UTC: Added a bounded equity and community foundation guide; qualified clinical, lived-experience and methods review remains pending.
- 2026-10-20 UTC: Depth19 additive practical-scope expansion after full old/neighbor body audit and selected original-source access; all old section, FAQ and source objects preserved. Qualified review remains pending.
What this article still leaves open
- Which structural interventions produce sustained health gains and for whom?
- How do resources, service quality and implementation change intervention effects?
- How can assistance respect people’s priorities while measuring whether needs were actually met?
Suggest a correction
Submit a content correction to the editorial queue, or download a local note. Leave out personal medical details. Submitted text is stored privately with a browser receipt session and is visible to this browser and the editor. Clearing cookies removes access to your submission history. This is not an account or a medical service.
Your submitted corrections for this article
Explore the topic map
Browse the health domains connected to this article. The map also identifies topics whose articles remain unwritten.
Identifying a need is different from meeting it
The National Academies describes awareness, adjustment, assistance, alignment and advocacy. For transport, those can mean asking about access, reducing unnecessary journeys, helping obtain transport, coordinating community resources, or supporting wider infrastructure changes. A referral is not evidence that a person obtained the service. People may not want the offered assistance, and available resources differ. Shared decisions should respect what matters to the person. The 2019 committee also found limited evidence for asking about social risks in isolation and discussed possible unintended consequences and opportunity costs.[3]