SBS201 Final

WHO Commission definition of Social Determinants of health

"Social determinants consist of the conditions in which people are born, grow, learn, work, and age, and the wider set of forces and systems shaping the conditions of daily life. These forces and systems include economic policies and systems, development agendas, social norms, social policies and political systems" These circumstances are shaped by the distribution of money, power, and resources at global, national, and local levels.

Define Geoffrey Rose's concept of the "prevention paradox"

"a preventive measure which brings much benefit to the population offers little to each participating individual" e.g. seat belts Kawachi extension of prevention paradox: a large number of people at modest risk give rise to more cases of disease than the small number of people at big risk the majority of the disease arise within the bulk of the population that is not in the high risk tail Big risk * small cases = small population risk Small risk * large number of cases = big population risk

Why is it so challenging to engage policy makers on prevention?

1) Actual lives saved vs. statistical lives saved E.g. when donating to charity, people don't value lives consistently. Money is concentrated on a single victim even though more people would be helped if resources were dispersed. When people got the personal story and statistics, less money was given. It screwed up the effect of "individual victim" story 2) Conflation of health and health care Health care is only 1 ingredient in improving health. Even if we gave everyone health care, would only improve ½ of the population 3) Conflict between time horizon for political decision making and payoff from investments in social determinants Identifiable vs statistical lives saved Time horizons for creating policy - takes time Conflation of health and healthcare

Concavity effect (absolute income effects)

1) Concavity effect (absolute income effects) • Prediction/Theory: concave shape of the relationship between income and health predicts that, all else equal, more unequal societies have worse average health o any reduction in health as a result of taking away income for the rich is more than offset by the gain in life expectancy for the poor as a result of the transfer o in other words: a dollar earned will cause a greater absolute shift in health to someone of low income than dollar lost to someone of high income • 2 assumptions: o Shape is concave o Association between income and health is at least partly causal • How large is concavity effect? o Study by Brodish & Hakes in 2016 showed that 10% reduction in Gini approximates 4% reduction in total injury o This is equivalent to eliminating all unintentional injury deaths in the USA o Assuming full causality and no "leaky bucket"

Explain pathways through which discrimination impacts health

1) Economic and social deprivation e.g. worse hiring for jobs 2) Exposure to toxic substances/hazards e.g. flint, Michigan 3) Socially inflicted trauma e.g. racism, policy brutality 4) Targeted marketing of unhealthy products e.g billboard in poor neighborhoods 5) Inadequate healthcare 6) Health-harming responses to discrimination e.g. coping -> addiction, smoking 7) Ecosystem degradation and alienation from the land e.g. indigenous lands, gentrification

SES metrics

1) Education 2) Income 3) Occupation 4) Wealth 5) Consumption

Two reasons why intentions fail to align with actions:

1) Many behaviours skip conscious intentions - e.g. our judgements and decisions are affected by incidental emotions and mindless habits • Emotion interferes between intention to perform and actual behaviour 2) We intend to do one thing but end up doing another i.e. our stated intentions are unreliable i.e. our will gets hijacked by temptation

Describe mechanisms/ pathways through which education affects health outcomes

1) Non-causal • Always begin with non-causal • Reverse causation: childhood illness caused dropout and ill health • Confounding: o Omitted "third" variables may include: genetic disposition, innate ability, personality E.g. Negative Nancy - don't do well in school, but also have high blood pressure o By IQ or other measure of ability 2) Causal - cognitive mechanisms • Acquisition of knowledge • Health literacy (ability to read, understand and use healthcare information to make decision and follow treatment instructions) 3) Causal - "non cognitive" mechanisms • Self-regulation: patience, ability to plan for the future • Executive functioning • Early education played a role in strengthening self-regulation and executive functioning 4) Indirect mechanisms • Schooling and credentials as gateways to safer jobs, higher income • By warehousing: people who are in college are protected from harmful events, education leads to warehousing, leads to better health 5) Other • Higher prestige/status in community • More social connections ("social capital") • Improved future prospects leads to greater incentive to invest in health

Evaluate threats to causal inference in observation studies linking education to health

1) Reverse Causation Bad health compromises education attainment, not the other way around E.g. chronic conditions during childhood e.g. diabetes, ADHD, result in children missing school 2) Confounding (omitted variable bias) Association between education and health is spurious and reflects the influence of omitted "third" variables E.g. propose that IQ/SES and IQ predicts future health Therefore, IQ confounds SES/health association

Evaluate threats to causal inference in observation studies linking income to health

1) Reverse causation: bad health lowers income, not the other way around • E.g. mental illness can compromise ability to work full time, which can lower an individual's income • Leads to job loss, earnings loss and out of pocket medical expenses 2) Confounding: association between income and health is driven by something else • E.g. ability, time, preference, personality E.g. study on effect of income on health using lottery prizes. Lottery prizes used as instrument variable. Regression models adjusted for demographics, education, baseline health. Saw exogenous increase in income of 10% due to lottery winnings increased life expectancy by 5-8 weeks for the whole sample

What are propositions of Geoffrey Rose's concept of the "prevention paradox"

1) Risk of disease lies along a continuum. E.g. hypertension dx is an arbitrary distinction 2) A large number of people at modest risk give rise to more cases of disease than the small number of people at big risk Population attributable risk: depends on RR and prevalence of risk factor Even if RR is huge, PAR may not be high if prevalence is low e.g. severe HTN BUT if RR is low, PAR may be high if factor is common e.g. mild to mod HTN 3) There are two strategies of prevention: High risk strategy: target prevention to the high risk tail of the distribution OR Population strategy: attempt to shift the entire distribution of risk 4) Pursuing "high" risk strategy of prevention can lead clinicians into a "mass treatment trap" - i.e. the need to bring more and more patients under medical surveillance & treatment in an attempt to overcome the prevention paradox Cons: Diminishing benefit/risk ratio Medicalization of millions of people 5) Projected benefits of population strategy can be surprisingly large E.g. reducing avg SBP < 2 mmHg can reduce stroke risk by 15% 6) Normality and abnormality are inter-twined - they move up and down together Dichotomy of "normal" and "abnormal" allows the population as a whole to disown the tail of its distribution People don't criticize the "average weight" even in overweight populations, because "obesity is bad" 7) Abnormality is driven by norms. Since norms are features of society, need to change society to affect disease. That is, look upstream

What underlying factors that lead to employment in certain quadrants? What is not fully captured in Karasek's model?

1) SES: People with higher SES may have higher levels of education, which gives them access to lower-strain and higher-control jobs. Generally, high decision latitude is associated with high SES; low decision latitude is associated with low SES 2) Institutionalized discrimination: People who are not white are more likely than white counterparts to hold high-stress/low-control jobs. Consider also the effects of institutionalized sexism, heterosexisim, transphobia, xenophobia ect. 3) Non-standard Work & Job Insecurity: Part-time, temporary, contract, under-the-table employment are often linked to SES and institutional discrimination and are additional sources of stress not fully captured in this model. 4) Work/Life Balance & Gender Differences: Many women work outside the home and then come home and do the majority of housework/childcare (i.e., the "double shift."), and this is not captured in Karasek's model. Women may also have less decision latitude in the home, which can impact experience of stress and health. 5) Experience: This model does not take into account tenure and years of experience in a position

describe three mechanisms through which social networks promote health

1) Social influence/behavioural regulation 2) Social engagement/participation and 3) Social support

What are the 3 theories linking income inequality to population health

1. Concavity effect (absolute income effects) 2. Social comparisons (relative) 3. Contextual→ equal society Summary: • Three theories are not mutually exclusive, they may all be correct, or none may be correct • Of the 3 theories linking income inequality to population health, the absolute income effect is the easiest to subscribe to o Just need to agree with two assumptions: a) income is causal and b) shape is concave • Simulations suggest lowering income equality by 10% lowers overall deaths by ~ 4%

If you were a wealthy person in the top 0.01% what type of society would you be better off in according to each theory?

1. Concavity effect (absolute)→ unequal society a. Because you might experience a (small) loss in health if you had to share your money to live in a more equal society 2. Social comparisons (relative)→ unequal society a. Because you have the relative satisfaction of being at the top & comparing yourself to everyone "below" you 3. Contextual→ equal society a. Because even wealthy can't escape the pollution effects! b. E.g Zika - arises from growing poverty. Mosquitoes will bite everyone equal.

What are the unintended consequences of population paradox?

1. Shifting distributions might increase concentration of risk at one end of the curve (the highest risk) and concentration of benefits at the other (low risk). 2. Public health guidelines might be wrong 3. Inequality paradox: population strategies can widen inequalities

Discuss what is meant by the "dark side" of social capital.

All capital can be good or bad. Cohesive community can: • Gather signatures to enact moratorium on fast food outlets in neighborhood, repeal a soda tax • Social capital is agnostic with respect to goals Dark Side • Excessive obligation to help others • Conformity and downleveling of norms - "everyone should be the same, don't stand out" • Exclusion of "out-group" members - discriminate against members; bullying, targeting, people different from majority

Major cognitive biases in health decision making

Bandwidth tax (scarcity) Inter-temporal choice Affect heuristic

Why do we measure SES?

Because SES predicts every exposure and every outcome, and therefore confounds everything. Need to know how to measure it so we can control for it.

Mechanisms linking social capital and health

Collective Efficacy The ability to mobilize to undertake collective action. People that can organize can work to address situations that are health diminishing (foreclosure, discrimination, work policies, etc.). e.g. belief in ability to clean up after disaster e.g. protest closure of neighborhood school Social cohesion Communities that are trusting of each other and know that their neighbors/other members of the group would act in the event of an emergency or need. This can reduce stress but also have direct benefits (bidirectional exchange of social support). e.g. drop mail off for a neighbour; trust, reciprocity, bidirectional Social organization Groups of people in a community can improve health by acting as a voice to advocate for and generate structural, environmental, and policy changes. They can also be resources for social support e.g. neighborhood walk organization, PTA organization. unit that would have collective efficacy and interest and goals and social cohesion (think of collective efficacy as the ABILITY; social organization - dedicated organization) Social interactions Similarly, social interactions that take place between people in a community can serve to spread information (social contagion), reduce stress, mitigate violence, etc. e.g. social block party Maintaining social norms The ability of people in a community to maintain social order; to step in and intervene when they notice deviant behavior (aka informal social control). e.g. norms around cat calling in a community

Describe threats to causal inference in studies that assess the potential influence of neighborhood contexts on health.

Confounding • Example: presence of fast food outlets in neighbourhood à obesity • Omitted "third" variable might include individual preferences/tastes Reverse Causation (Y à X) • E.g. neighborhood with clean air ß asthma o More like to move to neighborhood with cleaner air 1) People choose where to live • Food conscious people move closer to supermarkets • Physically active people move to communities with more athletic facilities 2) Services choose where to location • Fast food outlets move in where there is consumer demand The BIMBE problem in built environment • Neighborhood walkability à physical activity is confounded by preference for living in a neighborhood with good walkability • E.g. study on whether built environments explain physical activity o People with low preference for walkable neighborhood walk much less than people with high preference - regardless of the environment o BUT regardless of preference for living in a walkable neighborhood, living in a low- walkability neighborhood is associated with less walking • Solution: collect additional data on confounding variables and control for them The BIMBE problem in local food environment • Access to healthy food à healthy diet is confounded by resident food preferences • Does access = utilization? o Can argue yes - if you built it ppl will come o Can argue no - its about personal tastes • E.g. new neighborhood store increased awareness of food access but didn't alter habits o But there are limits to what we can learn from N=1 studies o Insufficient follow up duration 6 mo to 1 yr o Building a new store may not be sufficient to change dietary behaviour • Solution: instrumental variable o Proximity to fast food outlet à diet, confounded by food preferences ▪ Z = coin toss, o Studies show no association and partial association ▪ Partial: greater access to fast food is positively associated with obesity in non white rural residents (though not white residents) 3rd solution is to randomize X - cluster randomized trial, randomize residents to move • Study where housing voucher given to move to a low poverty <10% neighborhood vs. no geographic restrictions or control group • Those who moved to low poverty had statistically significantly less obesity than control group. Those who moved without restriction had a diff but no statistical significant difference between traditional and control

Critically evaluate the evidence on work/life balance and employee health.

Data on job stress and health are sparse and less consistent for women • E.g. women with low job control have a higher RR of CHD (Whitehall II study) • E.g. study on CHD by gender and control at HOME o Women with low control have RR 2.6. o Men with low control at home have RR 0.7 • E.g. psychological distress level by % of housework performend o Married female average 70% o Married male average 36%

Apply the concepts of the twin strategies of prevention to analyze alternative approaches to deal with public health problems (e.g. obesity prevention)

Obesity: High risk: individualized weight loss program based on BMI cut off Population Targeting the entire population: educating the public - nutrition labeling and education act altering incentives for healthy behaviour - taxing sugary sodas and subsidizing healthy food options legislation/regulations such as banning sales of "super-size" sugary soda; giving FTC authority to restrict marketing of junk food to kids

contextual effect

Difference in neighborhood health due to characteristics of the environment • E.g. if risk doubled if you move somewhere, then it would be contextual effect • Group characteristic • E.g. access to unhealthy foods, neighborhood pollution- something about the place makes you unhealthy

Discuss threats to causal relationship between social ties and health.

Does smoking cessation really spread via social influence - is it causal? Between social network à obesity; what is the confounder? 1. Confounding: • clustering is explained by shared environmental exposures • An apparent connection between two individuals (starting or stopping) smoking at the same time simply reflect the influence of shared environmental exposure e.g. billboard • But if smoking behaviour is driven by environmental exposure, the billboard doesn't care who named who as a friend • According to the Framingham study, the probability a subject will quit smoking according to type of relationship with a contact who quits smoking -> higher for mutual friend and subject perceived friend 2. Reverse Causation • Lack of social network ß obesity • Homophily: "birds of a feather flock together" o E.g. two smokers became friends when they met outside during a work break i.e. their smoking caused the friendship to develop • Friends à smoking

Define what is meant by the term "ecological fallacy".

Drawing individual inferences from grouped (aggrate level) data e.g. national chocolate consumption and Nobel Prize winners • Nothing about eating chocolate that will people any individual person a Nobel Prize winner

Institutional examples

E.g. at height of Great Depression, banks were reluctant to make housing loans. In 1934, President Roosevelt established Federal Housing Administration to encourage new home building and to create jobs for unemployed construction works o FHA-guaranteed mortgage program made it more possible to afford hoes o To assist the banks in making house loans, the Home Owner's Loan Corporation (HOLC) was established to create an inventory of all residential areas in the nation o Several hundred secret maps of "residential security" ▪ They assigned grades. First grade was free of black and foreign-born white residents. Second grade - lenders were advised to make loans 10-15% below maximum. Fourth grade - often refused loans. o The assumption was that if minorities moved into a nice neighborhood, property values would suffer • E.g. labour markets o Discrimination @ level of hiring and interviewing o E.g. Unemployment rates by race worse for black than white, irrespective of recession or not o E.g. identical resumes sent out differing only by name. White sounding names were 50% more likely to be invited. ▪ Racial gap rises with credential level of applicants ▪ No evidence of gender discrimination within racial groups - men and wome equally likely to be invited (or not) • Summary: Race à unequal tx by institutions (schools, workplaces, access to consumer credit) à unequal SES outcomes à health disparities

Policy proposals [work/life balance and employee health]

Eliminate discrimination against non-standard workers a. Adopt legislation like the EU part time direction i. Pay equity, pension benefits, training/promotion opportunities, bargaining rights 2. Improve work life balance for working families a. Expand eligibility for paid parental and sick leave b. Subsidize day care and child care e.g. Sweden - parental leave at 80% of salary, parents can use 390 days of paid leave; at least 2 months of leave for fathers. Now 85% of Swedish fathers take parental leave.

Personally-mediated - Interpersonal examples

Encounters in typical day - more specific/overt • E.g. looking for a job, looking for housing, getting health care

Discuss the process of agenda-setting to address health disparities through public policy.

Every country that has succeeded in addressing inequalities has done these 3 things: 1) Commissioning independent inquires • e.g. Black Report, RWJ commission to build a healthier America 2) Setting national goals and targets • e.g. healthy people 2020 - tobacco use goal 3) Communicating to the public • e.g. in polls, white people think African americans' LE is the same as them. Over half whites think no diff by race. • Therefore, documentaries such as Unnatural Causes are important

Distinguish between compositional and contextual effects

Example From lab e.g.: T stop - do people walk to work because of who they are (compositional effect) or because of the characteristics of the environment (walkability)?

Identify different forums and domains of discrimination

Forms/ Levels: institutional, interpersonal, internalized Domains: Housing, healthcare, labor market, justice system, education, consumer credit

Describe how framing can influence health behaviors.

Framing is important in health communication • E.g. obesity metaphors are "sinful" "eating disorder" "food addiction" o Perhaps we would frame as "time crunch" "industry manipulation" or "Toxic food environment" • When framing obesity as sin - interventions look like higher insurance premiums for the overweight, decreased antidiscrimination protections, decreased tax credit for gym membership • When framing obesity as toxic environment - interventions look like increased funding for healthier school lunches, increased zoning laws, increased food labeling, increased restrictions on junk food advertising, increasing junk food tax

Why are some countries healthy and others not

Health Care But access to health care by itself cannot explain health inequalities between countries E.g. those with highest income in USA have an equal or higher prevalence of diabetes than those with the lowest income in England (2002 study) Lifestyle Lifestyle differences may not matter as much either E.g. both England and US have similar # of current smokers. England has more % heavy drinkers (30% vs 14%) BUT obesity in USA = 31% vs 23% in England. Why? Social Environment Immigration effect: Based on JAMA 2004 paper on obesity among US immigrants by duration of residence, the longer people live in the US, the more they gain weight. "America-10" is similar to "Freshman 15" Thus: American environment is "obesogenic" Everybody is at risk We should stop blaming individuals. Start thinking about curing environment. Physical Environment Because of upstream policy decisions about public transport, USA favours automobiles. USA has a car culture. In Europe, public transport, bikes, and pedestrians are preferred. They make it difficult for cars e.g. raising taxes, making it difficult for cars.

Active job

High decision latitude High psychological demand e.g. ER doctor natural scientist Pros: Time off, more prestige Apply skills, acquire new kills Cons: Anxiety, CHD, positive mental health/self esteem, occupational injuries, resources

low strain job

High decision latitude Low psychological demand e.g. natural scientist Pros Can dictate own hours, flexibility Apply skills, acquire new skills Few adverse outcomes, positive self-esteem, resources to overcome problems

What are caveats of Geoffrey Rose's concept of the "prevention paradox"

High risk and population strategies are not mutually exclusive. We need to do both! Prevention paradox not applicable to every public health problem

Explain the distinction between "high risk" and "population" strategies of prevention

High risk strategy: target prevention to the high risk tail of the distribution Seeks to identify high risk individuals and offers them individual protection Produces interventions appropriate to the particular individuals Population strategy: attempt to shift the entire distribution of risk Targeting the entire population: educating the public - nutrition labeling and education act altering incentives for healthy behaviour - taxing sugary sodas and subsidizing healthy food options legislation/regulations such as banning sales of "super-size" sugary soda; giving FTC authority to restrict marketing of junk food to kids Protecting the entire population via: adding fluoride to water, iodizing salt, fortifying flour with folate Seeks to control the determinants of incidence in the population as a whole, to lower the mean level of risk factors, to shift the whole distribution of exposure in a favourable direction

1. Describe the theory of "dual processing" in shaping health behaviors.

Human cognition is made of dual systems that operate simultaneously. The dual processing refers to a person's use of two separate cognitive systems in decision making: System 1: Intuition - emotional, fast, automatic, effortless, associative, , parallel System 2: Reasoning - slow, controlled, effortful, rule-governed, neutral, serial • Monitors the activities of system 1 • But this takes effort so much of the time, our judgements and choices are automatically made by system 1 • BUT most public health behavioural interventions are focused on system 2 e.g. targets of public health interventions aimed at smoking cessation are: • Increase perceived susceptibility to disease, build self-efficacy to quit BUT all competitors strategies are appealing to system 1. E.g. emotion inducing tobacco ad - system 1

High Strain Job

Low decision latitude High psychological demand e.g. service work (e.g. waitress) Little flexibility Cons: unhealthy eating, smoking & drinking, anxiety and depression, hazardous occupational condition

Passive Job

Low decision latitude Low psychological demand e.g. truck driver/ toll booth worker Loss of skills/psychological atrophy Gradual atrophy of previously learned skills Little flexibility Monotony, disengagement Can result in negative learning Cons: low physical activity, more likely to have stimulant drug abuse/self medication, anxiety/depression

Further policy implications of education and health?

Marshmallow test: o 4 year olds who waited 5 min longer score 300 pts higher on SATs o longer delay in gratification at age 4 associated with lower BMI 3 yrs later Successful development is contingent on ability to invest for the future e.g. doing homework, resist temptation. Early education builds these skills. Affluent families already provide enriching environment for their pre school children. Disadvantaged children often miss out. ​Policy Implications: • Early education is one of the BEST BUYS in the social determinants of health o The most cost effective strategy for education is investing in social and cognitive environments of children who are disadvantaged, beginning as early in life as possible. Estimated rate of return per dollar of cost for early education is > 17%, far higher than the standard rate of return on stock market equity. • It can improve health and even narrow the health gap for disadvantaged families • Young children viewed as "blameless" even by many conservatives • Why is it hard to get it done? o Children don't vote! o Conflict between "now" (political election cycle) versus "later" (pay-off from investing in early education)

Describe the validity and reliability of methods used to measure the dimensions of socio-economic status (SES) [consumption]

Measure: Common variables measured - Durables: bicycle, motorcycles, sewing machine, refrigerator, TV, radio, tractor, fan, animals - Housing: type of floor & roof, type of drinking water and sanitation, type of cooking & lighting fuel etc Pros: Construction of an asset index: • Used in low SES; exercise in data reduction • Run principal components analysis on index variables. This reduces the number of variables to a smaller number of principal components which accounts for most of the variance.

Describe the validity and reliability of methods used to measure the dimensions of socio-economic status (SES) [wealth]

Measure: Income = incidence measure; wealth = prevalence measure • Wealth e.g. home ownership, liquid assets (checking and savings accounts, stocks and bonds). Liabilities (credit card debt, unpaid car loans, home mortgage). People can have negative wealth Cons: •Wealth is difficult to measure. Actual $ value may change from time to time. If we used wealth, we could overwhelm the measurement error.

Describe the validity and reliability of methods used to measure the dimensions of socio-economic status (SES) [education]

Measure: 2 Questions: 1) Years of schooling 2) Credentials gained Pros 1. Short and simple 2. Accurately reported (vs. income or occupation which are likely to be misreported) 3. Fairly stable beyond early adulthood - least susceptible to reverse causation Cons 1. Most restricted range 2. Varies by birth cohort 3. Tricky to asses quality of education 4. Unequal incomes given the same level of education according to race and gender

Describe the validity and reliability of methods used to measure the dimensions of socio-economic status (SES) [income]

Measure: At minimum questions must ask: 1) Unit: individual vs household 2) Time: past month, past year, life time 3) Sources: wages, social security, interest, rent, alimony 4) Gross vs. disposable Pros: 1. Consistency ("dollar is a dollar") 2. Continuous and captures very fine-grained difference in the standard of living from the depths of poverty to extreme wealth Cons: 1. High refusal 2. False reporting and errors 3. Unstable 3. Susceptible to reverse causation 4. Need to take account of "equivalence" in welfare across households of different sizes and composition. That is, same income will yield diff standard of living depending on # of people in the household. Can't just divide by # of people bc there are economies of scale for heating, lighting

Describe the validity and reliability of methods used to measure the dimensions of socio-economic status (SES) [occupation]

Measure: Requires 1 screener + 3 open ended questions 1) Are you in active labour force. If yes then: a) which industry do you work for? b) what kind of work do you do? c) what do you call yourself (job title)? Pros: 1. Standard Occupational Classification (SOC) can be mapped to hierarchical SES scales that in turn represent multiple dimensions of socioeconomic experience e.g. prestige or power & supervisory authority Cons: 1. Needs professional coder 2. Title inflation ("Vice President" bias) 3. Problematic to unemployed, home-makers, retirees 4. Susceptible to reverse causation

Explain Social networks

Structural o Size: how many friends do you have? o Frequency: how often do you see your close contacts o Diversity: a) marital status b)presence of close friends/confidants c) membership in social groups d) church membership

Distinguish between absolute vs. relative approaches to measuring health disparities.

Policy intervention A: Absolute difference saves 10 lives per 1000 live births but relative difference increases by 0.5 deaths per 1000 live births • This represents health maximization - greatest good for greatest number. This may widen the health gap because if the health of the better off improves faster • This also represent utilitarianism Policy intervention B: absolute difference saves 9 lives per 1000 live births but relative difference decreases by 0.5 deaths per 1000 live births • This represents egalitarianism • Gave priority to less well off even at the expense of holding back health improvements for the better off No right or wrong answer for policy choice. Both are ethically defensible. It's important to decide what you are trying to achieve.

Health behaviours can be hard to change due to:

Profit motive o Tobacco industry spends 50x the budget of CDC. They spend in 1 day what CDC spends in 1 yr. • Want/should conflict o Many behaviours skip conscious intentions - e.g. our judgements and decisions are affected by incidental emotions and mindless habits o Emotion interferes between intention to perform and actual behaviour • Scarcity and present bias

What are the pros and cons of population strategies?

Pros Big population benefit from small individual change Radical: addresses root causes Pervasive (wide reach) Easier to change with norms Behaviourally appropriate (e.g. if non smoking becomes normal, will be less necessary to keep persuading individuals Cons Small individual benefit (e.g. saving statistical lives - seat belt laws) Can lead to unintended consequences e.g. cut fat, but increases carbs (substitution effect) Inequality paradox: can widen health disparities e.g. cervical ca screening adopted more readily by high income than low income women -Population strategies that rely on mass ed campaign will amplify pre-existing disparities. Those based on legislation/regulation or targeted incentives to change behaviour will tend to flatten health disparities Not necessarily cheaper or cost effective compared to high risk strategies. May be higher start up costs, delayed benefit. Poor motivation of subject and physician

What are the pros and cons of high risk strategies?

Pros Large individual benefit Modest population benefit Subjective motivation Physician motivation No interference with others Cost-effective use of resources Benefit:risk ratio favourable Diminish health disparity Cons Small population impact Can lead to a mass tx trap Diminishing risk/benefit ratio Medicalization; stigma, non-normative behaviours required Difficulties and cons of screening Need to have access to intervention Palliative and temporary - not radical Limited potential for a) individual and b) population Behaviourally inappropriate

Understand the mechanisms through which social capital contributes to disaster resilience.

Resilience - ability to recover from shocks including natural disasters - come from social connections to others and not from physical infrastructure or disaster kits. • Most important factor in disaster resilience is not material resources like food - it is social capital Japanese study on cognitive dysfunction after disaster • If housing destroyed, 3x the risk of getting dementia o Live in communal spaces, toilet communal. Move to trailer homes after 8 months. • Destruction of home = proxy for loss of social capital • Prior to 2011, residents were encouraged to conduct disaster preparedness in own districts • Results: o The more individuals participated, the less the decline o For large X - even if people don't socialize, if live in a neighborhood where others do, there is a contextual effect. Analogous to herd immunity. • Cognitive disability increased with severity of housing damage o However, was higher for those who had low social participation and lower for high social participation Policy implications: • When people relocate by lottery: decreased social cohesion and social participation • When relocate as a group: social cohesion is same and participation increases

Define the term "SES gradient in health", and discuss alternative explanations for it.

SES: individual's position in the social hierarchy based on prestige and access to resources. characterized along 3 dimensions - education, income, and occupation "SES gradient in health": In general, the lower an individual's SES position, the worse his/her health outcome. Essentially, this theory proposes that SES health Socioeconomic inequalities in health status are not just a threshold effect of poverty. There is a "gradient" in health across the SES hierarchy such that the higher the level of household income, wealth, education, or occupational ranking, the lower the risks of morbidity and mortality. The gradient of health is observed throughout the range of SES so that the middle class have better health than the poor and wealthy have better health than the middle class Alternative Explanation The relationship between SES and health is reciprocal and dynamic across the life course

bandwidth tax

Scarcity is associated with uncertain environmental context for e.g. short of money or time • This context imposes a tax on our ability to plan for the future • People develop tunnel vision and become present focused and focus on the problems immediately in front of them • People are not born short-sighted. Anyone in that situation could develop same set of behavioural responses • The bandwidth tax falls heavily on our ability to plan for the future, that is our executive functions, including self-regulation • E.g. disasters often associated with scarcity - japan study o Survivors who experienced housing damage exhibit increased present bias. Present bias in turn is associated with more unhealthy behaviours (snacking, drinking) o Obesity second largest problem during relocation after cognitive decline o Significant evidence of present bias with more scarcity (major damage or destroyed house) o Key preliminary findings: ▪ People whose homes were destroyed increased their alcohol consumption and frequency of dining out. Dining & eating out are strong predictors of weight gain. Implication • Low SES is a condition of scarcity • Socioeconomic disadvantage produces more occasions for negative emotions eg. Job loss, foreclosure, debt • Scarcity and sadness trap together can explain why SES gradients in health behaviour can be difficult to reverse via health education alone Policy Implication • Financial support for disaster victims • Need for careful planning of where to build temporary shelters • Pre-commitment devices should be introduced to help recipients of financial aid to save for the future o E.g. opening up saving accounts at same time so they can save instead of blow it on gambling

Define the "Gini coefficient and describe how it is derived.

The measurement of the extent to which income is distributed across a society • summary measure of income distribution derived from the Lorenz curve. It plots the proportion of aggregate income (on vertical axis) earned by each segment of the population, ranked from poorest to wealthiest households on the horizontal axis. • Range from 0.0 (perfect equality) to 1.0 (perfect inequality) • Given by the ratio of the area between the line of equality and Lorenz curve (A) over the total area of under the line of equality (A+B) o Gini = A/(A+B) • In perfect equality, each % of popn contributes relevant % of their income Hypothetically, if every household in a country earned the same income, we would have 45 degree line representing perfect equality. The greater the area between the 45 degree line of perfect equality and the Lorenz curve, the greater the degree of income inequality.

Social comparisons (relative income effects)

Theory: income inequality creates a bigger gap between your income and the incomes of others you compare yourself to • Prediction: more unequal societies have worse average health • Why? Because people make social comparisons and wider income gap leads to stress and frustration • Wilkinson/Pickett argument from their book "The Spirit Level" o Toxic effect of income equality is based on social comparisons o Conservative pundits dismiss this as envy but in many instances it's based on people correctly understanding the unfair distribution of opportunity o Feeling unfairly treated leads to psychosocial effects - anger, frustration, despair • Neoliberal argument is that income inequality is necessary to spur more effort - to encourage greater growth and mobility o But the reverse is true - US states that have the highest income inequality also have the worst economic mobility • How does limited social mobility affect health outcomes? o Americans believe they live in a land of equal opportunity (the Horatio Alger myth) o But the reality is that social mobility in America is very limited ▪ The US has high intergenerational earning elasticity indicating that it has a lower extent of mobility o Failure to attain the American Dream leads to self-blame ▪ When expectations of upward mobility are frustrated, this can lead to "deaths of despair" e.g. increasing suicide, alcohol use, poisonings ▪ "deaths of despair" are attributed to stagnant economic mobility ▪ Although people think about economic mobility as separate from effects of income inequality, roughly 2/3 of stagnant economic mobility is caused by income inequality (remainder due to slowing pace of economic growth) • Summary: relative income theory is bolstered by U.S. evidence indicating a dramatic rise in the "deaths of despair". But theory is not limited to toxic psychosocial harms resulting from upward social comparisons (frustration and hopelessness). Downward social comparisons also result in rich people becoming more selfish, less generous and more entitled and scornful

Contextual "pollution" effects

Theory: income inequality erodes social cohesions o When the incomes of the top 1% pull away from the rest, they cause a variety of pollution effects on the quality of life of the bottom 99% o Rent-seeking behaviour by the top 0.1% creates a variety of "pollution effects" on the quality of life of the bottom 99.9% • Prediction: more unequal societies have worse average health including the wealthy • Why: erosion of social cohesion leads to disinvestment in social capital i.e. the rich use their power to stop paying taxes - life gets worse for everyone • Income inequality and social cohesion o The rich pay for their own children's education, health services, security o The 0.1% use their financial power to lower taxes on themselves (rent-seeking behaviour) o Loss of tax revenue threatens the quality of life for everybody else • Testing for presence of contextual effect of income inequality o Controlling for income, does living in an unequal place = worse health status? o Meta-analysis of multi-level studies of income inequality and health of 60 million subjects: excess risk of total mortality per 0.05 unit increase in Gini • This is the most hotly contested theory o Inequality resulting from selfish behaviour of top 0.01% o The top 0.01% prefer to misdirect our attention because we might raise their taxes • Also difficult to prove given complexity of hypothesized pathways - erosion of social social cohesion and capture of political priorities by top 0.01% • Greatest threat to society is apathy - the "haves" in society coming to view inequality as natural, normal, inevitable, and deserved

Define "social cohesion"

Through contact with friends and family and participation in voluntary activities, life acquires a sense of coherence, meaningfulness, and interdependence.

Measurement of SES in Global Setting

Unemployment and under employment common • Education is a poor predictor of living standards because of lack of jobs • People don't earn steady income - engaged in informal sector or seasonal work • Solution: measure consumption

What is meant by 'upstream' vs 'downstream' health interventions?

Upstream: "Action taken on the environment (social or physical) via policy, changing social norms or physical modification" Focus is on society factors such as transport policy, urban planning play a role. Upstream social factors influence health via a cascading sequence of "downstream" events and exposures. Interventions include: Educating public e.g. nutrition labeling and education act 1990 Altering incentives for healthy behaviour: e.g. taxing sugary sodas and subsidizing healthy food options Legislation/regulations: municipal ban on sales of large-size sugary beverages by restaurants, food carts, movie theatres, and delis (May 2012, NYC) Downstream Focus is on individual factors such as diet, physical activity, and genes

"hierarchy of interventions" available to address health disparities.

Weakest to strongest-> information, incentives, regulation

what are propensity scores?

What is probability of something happening to someone vs not happening • Good tool for social epidemiologist to rule out threats to causal inference o Along with instrumental variables and stratification

Define "job strain"

When physical and emotional demands become so overwhelming that the worker is unable to function in healthy fashion

describe the Karasek demands/control model of stress

When physical and emotional demands become so overwhelming that the worker is unable to function in healthy fashion Work stress is caused by interaction of 2 factors: 1) Psychological demands • Quantitative o Allows for enough time to get job one o Involves excessive amount of work • Qualitative o Job requires working fast o Proportion of work performed under pressure o Worker free from conflicting demands from others 2) Decision latitude (control) • Composed of skill development and utilization + decision authority • Skill utilization: o Job allows for variety of different tasks vs repetitive tasks o Requires learning new things o Involves opportunities to develop special skills; occasions for creativity o Involves high level of skill • Decision authority o Allows worker to make own decisions o Worker ability to influence what happens on the job According to the Karasek model, job strain occurs when there is high psychological demand and low decision latitude.

Modeling intertemporal choice:

o Delay discounting - refers to how large a premium an individual will place on enjoyment nearer in time over delaying gratification o Internal discount rate: how much our enjoyment of eating the marshmallow would drop as result of being forced to wait o The higher your internal discount rate, the more impatient you are o BUT the classical model of delay discounting implies dynamically consistent discount rate i.e. short run trade offs are assumed to be the same as long run trade offs • Preference reversal occurs when choice is switched between later vs even later to NOW vs later o e.g. when pre-ordering from Amazon, customers have more "should" items vs. want items when delivery date is further away in the future

Relative deprivation

may have sufficient income (in absolute sense) for basic needs but be lacking in the means to purchase goods and services that others in the community are able to afford • the relative income hypothesis is closely tied to concept of relative deprivation. • We can roughly say that a person is relatively deprived of X (in this instance, X being income) when (i) he does not have X, (ii) he sees some other persons as having X, (iii) he wants X, and (iv) he sees it as feasible that he should have X". Not every individual is sensitive to social comparisons that generate a sense of relative deprivation.

Explain gender differences in social support and health

men receive more support from spouse than women o women receive more support from friends than men • e.g. after the Trier social stress test alone vs. with stranger vs with partner, men with partners experienced the lowest stress based on salivary cortisol

Affect heuristic

mental short cut" allows people to make a decision on an emotional reaction or state (system 1) rather than a logical calculation of risks and benefits (system 2) • E.g. TGIF evokes feelings of release of stress

Explain Social support

assistance that an individual receives from other members of their social network. • Functional measures o Instrumental - provision of financial assistance, material goods, or services e.g money, labour in kind o Information - provision of advice, information, guidance o Emotional support - offering sympathy, care, concern, understanding, esteem, affection, love, trust, acceptance o Appraisal support - giving appropriate feedback to help decision making, constructive criticism (not in lecture, don't focus on this)

Explain intergenerational social support

o daughters provide most of the needed assistance for widowed parents o sons more likely to help with finances, funeral arrangements, yard work, home repair, dealing with bureaucracies o having at least one daughter matters for 3 kinds of support: phoning, visiting ,ad helping o key to receiving help is having at least one daughter o for living with children, size matters (3+ better)

Examples of Social influence

behavioural regulation • The 3 degrees of influence rule • E.g. Framingham study on smoking cessation - believed to be contagious. Stopping smoking ripples through the network, having an impact on our friends (1 degree), our friends' friends (2 degrees) and our friends' friends' friends (3 degrees) • E.g. if all friends are teachers, maybe you wouldn't use vape

"Halo Effect"

having a virtuous ingredient alongside sinful foods. People underestimate calories if celery present next to a burger

Relative rank

individual's positional location (or relative rank) in the income hierarchy is a determinant of their health status • But unclear whether dominant or subordinates suffer more • Whether the adverse physiological effects of subordinate rank can explain a part of the relation between income inequality and health remains an open question; the proposition is difficult to test empirically, not least because of the high collinearity between absolute income and rank. • From an intervention perspective, moreover, if relative rank is what turns out to confer health advantage or disadvantage, then it implies a limited role for social policy in affecting population health. The reason is because policies such as income taxes can change people's absolute and relative incomes, but they will usually preserve rank.

How to measure intertemporal choice

marshmallow test

Define "social capital"

resources to which individuals and groups have access to through their social networks • For individuals (x): social support • From the group (X): helping, doing, belonging (social cohesion, informal social control/ maintenance of norms, collective efficacy ) o Helping: norms of reciprocity ▪ Neighbours constantly helping each other. Will they help check in on neighbour who hasn't been seen in a while? o Doing: collective efficacy ▪ Organize to undertake collective action E.g. gathering signatures for petition o Belonging: solidarity, trust, attachment to the group (social cohesion) ▪ Informally socialize e.g. weekly cookout ▪ Trust is a moral resource • In this course, involves more group level perspectives

Define discrimination

socially structured and sanctioned phenomenon, justified by ideology and expressed in interactions, among and between individuals and institutions, that maintains privileges for members of dominant groups at the cost of deprivation of others.

Absolute deprivation

the inability to fulfill the basic human needs of food, shelter, and clothing

income inequality

the unequal distribution of income across a society • In US, the incomes of the top 0.01% grew by 76.2% during the decade between 2002 and 2012 (i.e., straddling the Great Recession), while the incomes of the bottom 90% fell by 10.7% in inflation-adjusted terms.

Intertemporal choice

the way in which individuals assign differential value to two possible consequences (positive or negative) of a decision, based on the amount of time it takes for these consequences to be realized • Investment goods: when pain occurs now and payoff occurs later o E.g. exercise for weight loss, flossing • Sinful goods: when pleasure occurs now and cost is incurred later e.g. o E.g. high calorie foods, smoking, drugs, alcohol, unprotected sex • Arises in situations when costs and benefits of a decision are separated across time, resulting in two kinds of self-control problems: o Procrastination o Temptation • Happens regardless of knowledge about thebehaviour • Want/should conflicts are apt to occur in situations when costs and benefits of a decision are separated across time o E.g. exercise, eating healthy, drinking in moderation, getting decent nights sleep • 2 problems of self control: o Play now - pay later o Pain today - gain later

What are policy implications of education and health?

• 1: Skill development o skill development and brain maturation are hierarchical processes. Basic functions must be established before higher-level functions can be added. Skills beget skills. Therefore, preschoolers need to learn basic skills. • 2: Early experiences o early experiences have a uniquely powerful influence on the development of cognitive and social skills and on brain architecture. o There are developmentally sensitive periods for optimal learning e.g. language acquisition. Will be unable to learn certain skills after some time o Early education strengthens executive function and self-regulation. Process is a slow one that begins in infancy, continues into early adulthood and is shaped by our experiences. • 3: ROI better with early intervention o Early intervention lowers the cost of later investment o Later remediation efforts are less effective e.g. adult literacy services, prisoner rehab programs, & education programs for disadvantaged adults yield lower economic returns compared to early intervention • 4: sustained investments o Everything is not over by age 4! Early interventions are sustained when followed by continued high quality learning experiences o Early investments must be followed by later investments to recoup maximum value o Nobody is against boosting high school graduation rates or expanding college education o Point is: playing field already tilted by the time kids reach compulsory schooling age

Describe mechanisms/ pathways through which income affects health outcomes

• Ability to afford better housing e.g. heating, air conditioning • Ability to purchase healthy behaviours (fresh vegetables, membership of health clubs) • Psychological sense of security • Ability to afford (quality) healthcare

Multi-Level Studies

• Allow us to disentangle composition from context • If we find an "effect" of neighborhood poverty (X) on health of resident (y), this does not necessarily prove causality. o X neighborhood poverty à y CHD incidence at individual level • There could be compositional confounding by an individual level variable x e.g. individual income

Principles of Implementing a Population Strategy

• Cure the environment not just the individuals o E.g. collaborate with urban planners to enhance walkability of neighborhoods • Regulate supply side not just the demand side o E.g. reduce salt/sugar content in packaged food • Nudge where you can, regulate when necessary

compositional effect

• Differences in neighborhood health attributable to the characteristic of the people who live there • E.g. who are the ppl that live in the neighborhood? • E.g. if you move and risk of asthma doesn't change, then 100% compositional effect • E.g. my income predicts my asthma rate

Examples of engagement/participation

• E.g. participation in sports and prevention of functional disability in older Japanese o Adjusted HR of incident functional disability: ▪ Active (1.0), sedentary (1.65), lone exerciser (1.29), and passive participation (1.16) o Exercising alone doesn't cut it - it's not as good as working out AND doing it as part of a group o Social participation captures most of the benefit from belonging to a sports club • E.g. social engagement as a mechanism not just the activity of playing soccer

"hierarchy of interventions" {information}

• Education o E.g. nutrition labeling on packaged foods o E.g. warning labels on cigarette packs • Advantages: provision of information • Drawbacks: 1) studies show that health conscious people tend to read labels. Those who aren't, don't pay attention. Since more educated people tend to be more health conscious, net result - widening inequalities ▪ E.g smoking prior to 1960s - no disparities. After surgeon generals warning, increased disparities between educated and not ▪ Plus, adolescents don't pay attention at all - "immortality bias" o 2) risk of "capture" by producers ▪ E.g. in hindsight, cigarette warning labels were more advantageous to tobacco companies than for public health • Tobacco never lost a case bc label told you about the harm to health. This has saved them from class action lawsuits.

"hierarchy of interventions" {incentives}

• Examples o E.g. cigarettes exercise tax o E.g. fat taxes and subsidies on fruits/vegetables • Advantages o more effective and reliable in terms of changing behaviour than providing information o Taxes also reduce health inequalities because low income individuals are price sensitive o Taxes often discourage harmful consumption by children who are also more price sensitive • Disadvantages 1. Potentially regressive - unless matched by subsidies on healthy, attractive alternatives 2. Substitution effects - ban on soda, ppl drink more fruit juice 3. Taxes are a blunt instrument - they don't distinguish between bad and good consumption. Coke is an easy target, but even cheeseburger provides nutrients. ▪ E.g. saturated tax in Denmark but it got complicated so had to abandon it

"hierarchy of interventions" {regulation}

• Examples: o Restricting smoking in public places o Removing vending machines in schools • Advantages: Changing norms o Besides intended action (e.g. protecting non smokers health), regulations can change norms ▪ By reducing prevalence of public exposure to second had smoker, non smokers react even more negatively when it occurs -> causes it to happen even less commonly, further increasing social disapproval. New norm emerges - don't smoke in private spaces o In fact, change of norms is often the main story behind successful regulations e.g. restricting indoor smoking ▪ Laws change public norms! # of lives saved from restricting indoor smoking lower than # of lives from smokers stopping to smoke • Disadvantages o Most paternalistic of government actions. Therefore most likely to meet with consumer and industry resistance o Costs need to be weighed against benefits (e.g. loss of school revenue as a result of removing vending machine contracts) ▪ What's worse? Not drinking coke or not having after school program?

The affect heuristic impacts health though:

• Harmful health behaviours based on emotions rather than logic • Industry exploitation

Describe how we can intervene to build community social capital

• Social policies o E.g. policies that promote work family balance, require community service participation • Community service programs that promote service work o e.g. experience corps, peace corp, americorp ▪ Intergenerational social capital • Urban design and architecture o Wide enough sidewalks so people will interact o Architecture as it relates to social interaction o Neighborhood parks, public library • Local investment - programs that boost social participation o E.g. case study - 2007 - campaign to promote social capital and healthy aging. Opened community centers for seniors called salons. Using instrumental variable of distance from salon, instrumented salon participation led to better health for participants than non participants. o E.g. ppl donate paint each summer so community can paint their homes o E.g. community well installation

Internalized examples

• Summary: Race à perceptions of unfair treatment à stress & maladaptive coping responses à health disparities • Discrimination is so ingrained, people internalize those views and harm themselves • E.g. stereotype threat and intellectual test performance of African Americans o Stereotype threat refers to existence of a negative stereotype about a group o Threat can be disruptive enough to impair cognitive performance o E.g. 24 black and 23 white students randomized to take ½ hr GRE test. Half asked to report race, other half not. Test scores adjusted for SAT score. ▪ Stereotype threat affected individuals! Race primed black students did half as well. No race prime - black kids did slightly better • E.g. stereotype threat

Evaluate the causal association between job insecurity and poor health.

• The demand for more labour flexibility in the global economy has resulted in more out-sourcing, subcontracting, more non-standard work, including part-time causal, temporary work • More job security = precarious work Pros • Flexibility allows employers to cut labour costs during slack times • Employers can screen workers before hiring them permanently (decrease training costs) • Workers may be able to control their schedules Cons • Non standard jobs are often "bad jobs" - poor pay, lack of health insurance and pension benefits • Lack of protection from unions and labour laws Are non-standard jobs damaging to worker's health? • Healthy workplaces attract healthy workers • E.g. study of 4th wave of Korean labour and income panel - part-time, temporary and contigent work à outcome of poor self-rated health Causal inference problem • Workers select into different types of work • Several of these selection factors e.g. education, marital status, pre-existing health - also determine health. • Therefore, comparison of precarious and non-precarious workers may be biased

Examples of Social support

• e.g. social support (e.g. friend coming to lab) reduces cardiovascular reactivity to psychological challenge • e.g. 6 month survival after heart attack affected by sources of support. The more support, the higher chance of survival

Discrimination is cumulative over:

• life course of individuals • across generations • across domains (discrimination in education can affect outcomes in labour market even if no further discriminations occurs in latter domain)

Present Bias

• observed when: o negative emotions are involved (Sadness trap) o system 2 is overloaded (Stress) o scarcity and uncertainty is involved • observed when immediately available rewards have a disproportionate impact on preferences relative to more delayed rewards o e.g. most smokers have the same long run preferences as non smokers - they prefer to stop smoking. o But their problem is their long run preferences are not the same as their short run preferences o There is a link between sadness and present bias ▪ Sadness induces people to be more present focused • Sad photo on cig pack can increase smoking related behaviour o "sadness trap" helps to explain robust correlation between depression and smoking o Making smokers feel sad might increase their craving for the next nicotine hit - an unintended effect of PSAs appealing to sadness ▪ People who feel sad - smoke more per puff! They measured this