Health Disparities

Health & Healthcare Disparities — Deep Dive

Part 1 of 4 — Social Determinants & the Social Gradient in Health

The Central Finding of Medical Sociology

Health tracks social position. Not as a rich-vs-poor binary, but as a social gradient: at EVERY step up the SES ladder, average health improves and mortality falls — executives outlive their deputies, who outlive their clerks, even when all have salaries, housing, and healthcare. The famous civil-servant cohort studies established this within a single employer with universal health coverage, which is the point: access to doctors cannot be the whole story.

Social Determinants of Health — The Causal Inventory

The social determinants of health are the conditions in which people are born, live, work, and age:

Fundamental cause theory explains a puzzle: specific diseases and risk factors change across history, yet the SES-health association persists. Because SES embodies flexible resources — money, knowledge, power, beneficial connections — advantaged groups redeploy those resources against WHATEVER the current health threat is (from cholera to cigarettes to novel screening technology). Prediction: disparities are LARGEST for diseases we know how to prevent or treat, and smallest where medicine is helpless — a signature the exam loves to test with data.

• Fundamental cause theory was proposed by the sociologists Bruce Link and Jo Phelan.

Stress Biology — Where Sociology Meets the Endocrine Lesson

Chronic social adversity gets under the skin through the stress axes (connect to the endocrine/nervous lesson):

• Repeated or unrelenting stressors — economic insecurity, discrimination, unsafe neighborhoods — keep the HPA axis and sympathetic system chronically engaged.
• Allostatic load: the cumulative physiological wear (blood pressure, cortisol dysregulation, inflammation, metabolic markers) from prolonged adaptation; it rises with lower SES and mediates part of the gradient.
• The weathering hypothesis: chronic exposure to discrimination and disadvantage accelerates biological aging in marginalized groups — proposed to explain why some racial disparities (e.g., birth outcomes) WIDEN with maternal age and persist at every education level.

The Vocabulary of Difference vs. Injustice

• Health disparity: any measurable difference in health between groups.
• Health inequity: the subset of disparities that are avoidable and unjust — produced by social arrangements rather than biology or free informed choice. The terms are not synonyms; exam items test the distinction.
• Disparities are documented across SES, race/ethnicity (persisting after controlling SES — pointing at discrimination, segregation, and care differences), gender (women live longer but report more morbidity; the mortality-morbidity paradox), and geography (rural hospital closures, urban environmental burdens).

Life-Course Effects

Early environments program later health: childhood poverty predicts adult cardiovascular and metabolic disease even among the upwardly mobile (latency effects), while disadvantage also accumulates across life (pathway effects). Critical-period logic will feel familiar from embryology — the sociology version simply runs on housing, nutrition, and stress instead of teratogens.

Worked Example — Testing Fundamental Cause Theory Against Data

Passage-style problem. Researchers assemble mortality data for Country X:

1. In 1950, deaths from Disease A (no known prevention or treatment then or now) show almost no SES gradient. In 2020, Disease A mortality is still SES-flat.
2. Disease B was untreatable in 1950, with SES-flat mortality. A cheap screening test and effective treatment arrived in 1980. By 2020, Disease B mortality has fallen 60 percent overall — but the decline is 85 percent in the top SES quintile and 35 percent in the bottom, opening a wide gradient where none existed.
3. Disease C's major risk behavior was distributed EVENLY across classes in 1950; after mass publicity of the risk in 1970, the behavior declined fastest among the educated, and Disease C mortality now shows a steep gradient.

Reading finding 1. Where medicine and prevention are powerless, flexible resources have nothing to purchase — no gradient forms. This is fundamental cause theory's negative control, and its presence in a passage is never decorative: expect a question contrasting it with Disease B.

Reading finding 2. The technology arrived neutral, but deployment followed resources: knowledge of the test, insurance, clinic access, follow-up capacity. A life-saving innovation WIDENED the disparity — the theory's core prediction that new health-relevant technology is captured first and best by the advantaged. Note the trap answer "the treatment doesn't work for low-SES patients": the data show slower diffusion, not differential biology.

Reading finding 3. Same logic, behavioral channel: information is also a resource, and the educated converted the publicity into behavior change faster. Risk behaviors that were once class-neutral became class-patterned — which is why TODAY'S behavioral gradients (smoking) are historically recent products, not timeless class traits.

The integrative question to expect: "Which policy would fundamental cause theory predict REDUCES the Disease B gradient most?" Ranked by the theory: interventions that do not depend on personal flexible resources — automatic screening defaults, universal outreach, treatment brought to neighborhoods — beat information campaigns, which the advantaged metabolize faster. Any answer that only adds MORE technology without changing who can deploy it preserves or widens the gap.

Method footnote. All three findings are ecological time-series — no individual-level causation is licensed; the theory survives because its distinctive PATTERN (gradients tracking treatability) appears where rivals predict none.

Social Determinants & the Gradient 🎯

Key Takeaways — Part 1

• Health follows a fine-grained social GRADIENT (every step matters), persisting even with universal care — access alone cannot explain it
• Social determinants — income, education, occupation, neighborhood, connection — are the upstream causes; isolation and low job control are mortality-grade exposures
• Fundamental cause theory: flexible resources (money, knowledge, power, connections) get redeployed against each era's threats — so disparities concentrate where disease is preventable/treatable, and new technology widens gaps until deployment stops depending on personal resources
• Chronic adversity becomes biology via stress axes: allostatic load tallies the wear; weathering extends it to discrimination-accelerated aging
• Disparity = any group difference; inequity = the avoidable, socially produced subset; life-course effects (latency + accumulation) mean childhood conditions echo in adult disease

• Low-yield extras: fundamental cause theory's authors (Link and Phelan)

Health & Healthcare Disparities — Deep Dive

Part 2 of 4 — Healthcare as an Institution: Access, Encounters & Unequal Treatment

Illness Is Social Twice Over

Medicine distinguishes disease (the biological pathology), illness (the person's subjective experience of being unwell), and sickness (the social role and status of the unwell person). The same pathology yields different illness experiences and different social treatment by class, gender, and culture — the reason a purely biomedical model under-predicts what happens to patients.

• Illness experience: how people notice symptoms, interpret them (cultural schemas decide what counts as "sick enough"), narrate them, and manage identity through chronic illness — chronic disease often forces biographical revision ("Who am I now?"), and stigmatized diagnoses (from the deviance lesson) add stigma management to treatment burden.
• The sick role (structure lesson, Part 2) formalized sickness as a sanctioned deviance with rights and duties; its known failures — chronic illness, contested diagnoses, unequal access to the role's protections — set this part's agenda: WHO gets believed, excused, and treated?

The Access Cascade — Where Care Is Lost

Access is not one barrier but a cascade; disparities compound at each step:

Underinsurance matters as much as uninsurance on modern exams: coverage that shifts costs to patients suppresses BOTH unnecessary and necessary care — patients cannot tell which is which (the classic insurance-experiment finding), so cost barriers cut preventive and essential use, disproportionately among the poor and sick.

The Clinical Encounter — Micro-Sociology with Vital Signs

• Paternalistic model: physician decides, patient complies; informative/consumer model: physician supplies facts, patient decides alone; shared decision-making: deliberation toward a choice consistent with patient values — the contemporary ideal. Know them as ideal types the exam asks you to classify from dialogue.
• Power asymmetries structure the encounter: knowledge gaps, time rationing, institutional scripts (the 15-minute visit), and interruption patterns — recorded encounters show physicians redirect patient openings within seconds, and redirection is patterned by patient class, race, and gender.
• Implicit bias in providers: automatic group associations predicting differences in pain-treatment decisions, referral rates, and communication warmth, even among providers with egalitarian explicit attitudes (the audit-study logic from the groups lesson, applied to clinicians — vignette studies vary only patient race/gender and observe treatment recommendations).
• Documented treatment disparities survive controls for insurance and disease severity: fewer cardiac catheterizations, less adequate analgesia, and later diagnoses for minority patients — placing part of the disparity INSIDE the clinic, not just upstream of it.
• Cultural competence / humility: institutional (interpreters, community health workers) and interpersonal (eliciting the patient's explanatory model of illness) remedies; "humility" flags that the skill is ongoing self-scrutiny, not a mastered checklist.

The System Level

• Second sickness: the aggravation of illness produced by social injustice and by the healthcare system itself — medical debt, iatrogenic harm concentrated where care is thinnest, discharge to conditions that caused the disease.
• Safety-net institutions (community health centers, emergency departments as care of last resort) absorb the uninsured; federal emergency mandates guarantee stabilization, not treatment or follow-up — why ED reliance is both a symptom of access failure and a costlier, discontinuous substitute for primary care.

• The US emergency mandate is EMTALA (1986): emergency departments must screen and stabilize every patient regardless of ability to pay.

Worked Example — Finding the Disparity's Location

Passage-style problem. Researchers investigate why Group M patients in a metropolitan area die of Disease D at 1.8 times Group N's rate. Four studies:

1. Insurance-claims analysis: after matching on insurance plan, disease stage at diagnosis, and comorbidities, Group M patients remain 30 percent less likely to receive the guideline-recommended procedure.
2. Vignette experiment: 400 physicians receive identical case files for a hypothetical Disease D patient, randomized to carry Group M or Group N names and photos; recommended procedure rates are 25 percent lower for Group M files. Physicians' self-reported egalitarian attitudes do not predict their recommendations.
3. Geographic audit: hospitals capable of the procedure average 4 miles from Group N neighborhoods and 11 miles from Group M neighborhoods; two Group M-area hospitals closed in the past decade.
4. Interview study: Group M patients report prior dismissive encounters, describe delaying care until symptoms are severe, and are more often uncertain which symptoms warrant a visit.

Study 1 localizes a within-clinic disparity. Matching on plan, stage, and comorbidity removes the standard "sicker and uninsured" explanations; what remains occurs AFTER patients reach comparable clinical situations. But observational matching cannot prove provider behavior causes it — unmeasured differences (patient preferences?) survive matching. That residual doubt is what Study 2 exists to close.

Study 2 is the causal lever. Random assignment of group signal to IDENTICAL files isolates the patient's perceived group as the cause of differential recommendations — and the null result for explicit attitudes locates the mechanism at the implicit/automatic level (discrimination without self-reported prejudice; the audit logic from the groups lesson). Together, 1 and 2 place part of the disparity inside clinical judgment.

Study 3 adds the institutional layer: distance and closures are structural access barriers — no biased individual required (institutional discrimination's signature). Note it cannot explain Study 2's file-based gap; the studies partition the disparity rather than competing.

Study 4 supplies the demand-side pathway, sociologically read: dismissive prior care erodes trust → delayed presentation → later stage → worse outcomes. Warning: as an explanation it is downstream of PRIOR institutional behavior — an exam distractor will reframe Study 4 as "patient noncompliance," converting an institutionally produced pattern into a victim-blaming trait attribution. The interviews also show illness-interpretation barriers (uncertainty about symptom significance = health literacy as a navigation barrier).

Synthesis question to expect: "Which intervention addresses the mechanism in Study 2 specifically?" Not new hospitals (Study 3's fix), not outreach (Study 4's) — but decision-level redesign: standardized protocols that reduce discretionary judgment, decision support triggered by clinical criteria, and audit feedback to clinicians. Match each remedy to the layer whose data support it.

Access, Encounters & Treatment Disparities 🎯

Key Takeaways — Part 2

• Disease (pathology) ≠ illness (experience) ≠ sickness (social role); they dissociate, and legitimacy — who is believed and excused — is unequally granted
• Access is a cascade (coverage → reach → afford → navigate → stay); underinsurance deters necessary and unnecessary care alike, harming the poor and chronically ill most
• Encounter models: paternalistic / informative / shared decision-making — classify from dialogue; power asymmetries (time, interruptions, scripts) are patterned by patient class, race, gender
• Vignette and audit designs localize part of the disparity inside clinical judgment: implicit bias operates despite egalitarian explicit attitudes; structural barriers (closures, distance) are institutional discrimination needing no biased individual
• Second sickness: the system's own operation (emergency-only care, medical debt, discontinuity) can amplify the illness it treats — match every proposed remedy to the layer whose data implicate it

• Low-yield extras: the US emergency mandate's name and date (EMTALA, 1986)

Health & Healthcare Disparities — Deep Dive

Part 3 of 4 — Measuring Disparities: Epidemiologic Tools, Segregation & Discrimination as Exposure

The Measurement Toolkit

Disparities are quantitative claims; know the units they are made of:

• Incidence: NEW cases per population at risk per time period — measures the rate of becoming sick. Prevalence: ALL existing cases at a point (or period) — measures the burden of being sick. The linking approximation: prevalence rises with incidence AND with disease duration. A treatment that extends life with a disease RAISES prevalence while incidence is unchanged — the classic trap.
• Morbidity (disease burden) vs. mortality (death); life expectancy summarizes mortality across ages; infant mortality is the standard cross-group sensitivity indicator because it responds quickly to social conditions and care access.
• Comparing groups: absolute difference (rate A minus rate B) vs. relative ratio (rate A over rate B). A rare disease can show a dramatic ratio (3x) with a trivial absolute difference; a common disease can show a modest ratio hiding an enormous absolute burden. Passages exploit the mismatch — always ask WHICH comparison a conclusion needs.
• Age adjustment: groups with different age structures (from the demographics part) cannot be compared on crude rates; standardization removes the age confound. An older population can have a higher crude death rate while being healthier at every age.

Residential Segregation — A Fundamental Cause with Coordinates

Segregation is not merely correlated with health; it ORGANIZES the exposure package:

• Concentrates poverty, and with it: food deserts (low geographic access to affordable nutritious food), pollution sources and hazardous land uses (environmental injustice — siting follows the path of least political resistance), housing quality hazards (lead, mold, crowding), violence exposure and its chronic-stress sequelae, under-resourced schools (looping back through education → SES → health), and provider shortages/hospital closures.
• Because segregation bundles exposures, naive models that "control for" each downstream factor one at a time can make segregation's coefficient vanish while its causal role remains — you cannot adjust away the variable that ASSIGNED the other variables. Recognizing this over-adjustment logic is a tested skill.

Discrimination as a Measurable Exposure

Modern disparities research treats discrimination the way toxicology treats a toxin — with dose, duration, and mechanism:

• Self-report instruments (everyday discrimination scales) predict hypertension, depression, and biomarker profiles prospectively; the chronic-vigilance pathway runs through the stress axes to allostatic load (Part 1).
• Weathering predictions: disparities in outcomes like preterm birth WIDEN with maternal age (cumulative exposure) and persist at high SES — because higher-SES minority individuals often carry MORE cross-race exposure and its vigilance costs, SES adjustment does not erase these disparities.
• Race is a social category in these data: group differences in health reflect the distribution of exposures (segregation, discrimination, care quality), not innate biology — cross-national comparisons (the same ancestry groups showing different disparities under different social systems) and immigrant-generation studies (health often DETERIORATING across generations in the new society despite rising income — the immigrant/acculturation paradox) are the standard evidence.
• Controlling for SES is necessary but insufficient: SES measures are not equivalent across groups (same income buys different neighborhoods and wealth), so "adjusted" residual gaps understate or misplace the social causation rather than proving biological difference.

Gender, Intersectionality & the Paradoxes

• The mortality-morbidity paradox: women outlive men in nearly every society yet report higher morbidity and healthcare use — explained by cause-of-death profiles (men's higher external-cause and cardiovascular mortality), help-seeking norms (masculinity discouraging care), and differential diagnosis patterns.
• Intersectionality (from the stratification lesson) applied to health: axes of disadvantage interact multiplicatively — the disparities of, say, low-income minority women are not the sum of three main effects; designs must include interaction terms or stratified analyses to see them.

Worked Example — Auditing a Disparities Data Table

Passage-style problem. A health department reports on Disease Z in City Q's two districts (North, historically segregated and low-income; South, affluent):

A columnist concludes: "South's higher prevalence proves the disease burden falls on the affluent, so resources should flow south."

Step 1 — reconcile the crude and adjusted rates. Adjustment WIDENED the gap (90 vs. 60 became 110 vs. 45): South's population is older (older populations accumulate deaths, inflating its crude rate), so at any given age, North residents die of Disease Z at roughly 2.4 times South's rate. Crude rates understated the disparity; a question asking "why did adjustment increase North's relative rate?" is answered by age structure, not data error.

Step 2 — decode the incidence/prevalence inversion. Incidence is nearly equal — people GET the disease at similar rates. Prevalence is 2.25x higher in the South because prevalence tracks incidence times DURATION: South's average survival is roughly 2.4 times longer, so its residents LIVE with the disease (raising prevalence), while North's residents die of it quickly (truncating prevalence). High prevalence here is a marker of BETTER outcomes — the columnist has read a survival advantage as a burden.

Step 3 — name the disparity correctly. The injustice-relevant numbers are age-adjusted mortality and post-diagnosis survival: equal risk of getting sick, radically unequal survival once sick — pointing at detection timing and treatment access (Part 2's cascade), not at differential disease susceptibility. Fundamental cause logic (Part 1) predicts exactly this pattern for a treatable disease.

Step 4 — anticipate the over-adjustment trap. Suppose a regression controls for insurance, stage at diagnosis, and hospital quality, and North's survival deficit "disappears." That does NOT acquit segregation — those controls are the very PATHWAYS segregation works through (mediators, not confounders). Adjusting for mediators answers "how does the disparity operate?", never "is there a disparity?"

Verdict on the columnist: wrong measure (prevalence for burden-of-injustice claims), wrong direction (high prevalence reflecting long survival), and a resource recommendation that would widen the actual mortality disparity.

Measurement, Segregation & Discrimination 🎯

Key Takeaways — Part 3

• Incidence = new cases (risk of becoming sick); prevalence = existing cases (burden), rising with duration — treatments that extend survival raise prevalence with flat incidence
• Compare groups with the RIGHT arithmetic: absolute differences vs. relative ratios answer different questions; never compare crude rates across different age structures — adjustment can flip or widen gaps
• Segregation bundles exposures (food deserts, pollution siting, housing hazards, provider shortage); its downstream variables are mediators — adjusting for them explains the mechanism, never refutes the cause
• Discrimination is a dosed exposure: vigilance → stress axes → allostatic load; weathering predicts age-widening, SES-resistant disparities; immigrant-generation and cross-national patterns pin health gaps on environment, not ancestry
• Women's mortality-morbidity paradox and intersectional (interactive, not additive) disadvantage require stratified reading of any disparities table

Health & Healthcare Disparities — Deep Dive

Part 4 of 4 — MCAT Integration: Theories at the Bedside & the Disparities Passage Playbook

The Paradigms Round on Medicine

Every disparities passage can be theory-tagged, and questions ask you to do it:

• Functionalist: medicine maintains a healthy workforce; the sick role regulates legitimate withdrawal from obligations; disparities appear as system "dysfunctions" to be corrected. Weakness the exam cites: treats existing arrangements as consensual.
• Conflict: health is a resource distributed by power — care rationed by ability to pay, medical knowledge gatekept, pharmaceutical agendas set by profit, and the "second sickness" produced by injustice. Predicts disparities are STRUCTURAL outputs, not accidents.
• Interactionist: illness meanings negotiated in encounters — labeling (diagnosis as label, stigma as spoiled identity), the self-fulfilling prophecies of provider expectations, and how "compliance" talk assigns blame. Micro data (recorded consultations, interviews) signal this lens.
• Fundamental cause (the field's own theory): flexible resources track every new health technology — the through-line connecting all four parts of this lesson.

The Intervention Ladder — Match Remedy to Level

A signature MCAT question format: "Which intervention targets the mechanism identified in Study N?" Build the map before test day:

Two exam-grade principles: (1) interventions relying on personal resources (information campaigns, opt-in programs) are captured fastest by the advantaged and can WIDEN gaps (fundamental cause logic); default/automatic designs narrow them. (2) "Downstream" fixes do not discharge "upstream" causes — a navigation program does not answer a hospital-closure finding.

The Disparities Passage Playbook

1. Inventory the studies: for each, note design (ecological, cohort, vignette experiment, audit, interview), level (structural/institutional/encounter/individual), and the measure used (incidence vs. prevalence, crude vs. adjusted, absolute vs. relative).
2. Assign each finding a causal privilege: randomized designs (vignette/audit) license causal claims about their manipulated variable; matched observational studies localize but do not prove; ecological data suggest and never transfer to individuals (ecological fallacy); interviews supply mechanism and meaning, not magnitude.
3. Watch the four recurring traps:
   • Prevalence read as burden when it reflects survival duration
   • Mediator adjustment read as refutation (the over-adjustment fallacy)
   • Group-difference findings read as fixed biology when the design cannot separate exposure from ancestry
   • Institutionally produced patterns re-described as individual choice ("noncompliance," "lifestyle")
4. For "which additional finding would strengthen..." items: the right answer completes the causal chain at its weakest link — temporal order for cross-sectional claims, individual-level data for ecological claims, a dose-response or natural experiment for confounded claims.
5. For ethics-flavored stems (community trust, research in disadvantaged groups): answers weighting community partnership, transparency, and returning benefit outrank efficiency-first options — the sociology section's ethics questions reward structural awareness, not paternalism.

The Course-Wide Web

This lesson is the sociology section's hub: stratification supplies SES and capitals (Part 1's gradient), culture/socialization supplies norms and illness meanings, groups/interaction supplies bias mechanisms and encounter dynamics, demographics supplies the rates and pyramids, and the biology lessons supply the stress-axis wiring (HPA, allostatic load) that turns social structure into pathology. A full-length MCAT will cross these freely; so should your review.

Worked Example — A Capstone Passage, Played by the Book

Passage. A state introduces automatic (opt-out) colorectal screening mailers for all residents aged 45 or more, replacing the previous physician-referral system. Researchers report:

• Study 1 (before/after ecological): statewide screening completion rises from 55 to 74 percent; the rise is 12 points in the top income quartile (70 to 82 percent) and 27 points in the bottom quartile (42 to 69 percent), narrowing the screening gap from 28 to 13 points — by more than half.
• Study 2 (cohort): among screened patients, time-to-treatment after a positive result remains 2.5 times longer for uninsured patients.
• Study 3 (interviews): low-income patients describe the mailed kit as "the first time the system came to me," while several report discarding kits due to instructions written at college reading level.
• A commentator objects: "Screening gaps closed because poor residents finally took responsibility."

Question 1 — Why did an opt-out design narrow the gradient when decades of awareness campaigns widened it? Fundamental cause logic: campaigns are information — a flexible resource the advantaged deploy faster. Defaults require NO personal resources; they equalize by removing the resource-dependent step (remembering, scheduling, transportation to a referral visit). The differential gain (27 vs. 12 points) is the theory's fingerprint in reverse — remove resource dependence and the disadvantaged gain most because the barrier was never motivation.

Question 2 — What does Study 2 add? The intervention fixed ONE stage of the access cascade (initiation) and left another (post-screening treatment) untouched: insurance still gates treatment speed. Expect the item "which intervention addresses Study 2's finding?" — the answer lives at the structural/coverage level, not more mailers, not navigation alone. Disparities relocate to the stages a policy does not reach.

Question 3 — Evaluate the commentator. The claim converts a DESIGN effect into a trait explanation ("responsibility") — the individual-choice trap. The data directly contradict it: the same individuals under the referral system screened less; what changed was the system's architecture, not residents' character. This is also the interactionist point about "compliance" language assigning moral blame for structural facts.

Question 4 — What do the interviews contribute, and what can't they? Mechanism and meaning: the "system came to me" narrative explains WHY defaults work here (they reverse the institutional burden of navigation), and the reading-level complaint identifies a health-literacy refinement (a real equity leak inside a successful program). As interview data, they quantify nothing — using them to estimate effect sizes would be the wrong-tool answer.

Question 5 — Which additional finding would MOST strengthen the causal claim for the mailer program? Study 1 is ecological and uncontrolled — secular trends could contribute. Strongest strengthener: a neighboring comparable state without the program showing flat screening over the same period (difference-in-differences logic), or staggered county-level rollout with screening rising county-by-county as the program arrives. Both supply the missing counterfactual; "larger sample" answers do not.

Integration & Passage Strategy 🎯

Key Takeaways — Part 4

• Theory-tag medicine passages: system maintenance (functionalist), power and profit (conflict), negotiated meaning and labels (interactionist), flexible resources chasing technology (fundamental cause)
• Match intervention to mechanism level — structural, institutional, encounter, individual; resource-dependent designs (opt-in, information) widen gaps, defaults narrow them; fixing one cascade stage relocates disparities to the next
• The four traps: prevalence-as-burden, mediator-adjustment-as-refutation, exposure-vs-ancestry conflation, and structure relabeled as individual choice
• Strengthen-the-claim answers complete the weakest causal link: counterfactual comparisons (difference-in-differences, staggered rollout) for ecological claims, individual longitudinal data for level jumps
• Ethics stems reward community partnership, transparency, and returned benefit — and the whole sociology section rewards connecting this hub lesson to stratification, interaction, demographics, and the stress biology that embodies them