The Need for Social Epidemiology: Now as Much as Ever

Abstract

MD, MSPHWhen Rudolf Virchow was sent to Upper Silesia in 1848 toinvestigate the source of a typhus epidemic raging there, heconcludedthattherootcauseswerepoorhousing,hazardousworking conditions, poor diet, and lack of sanitation (1).When John Cassel studied mill workers in North Carolina,he concluded that social disorganization, poor accultura-tion, and lack of psychosocial resources were importantdeterminants of disease (2). When Syme and Berkmanstudied the epidemiology of disease in Alameda County,California, they concluded that social isolation was a domi-nant risk to health (3, 4).In the August 2011 issue of Annals of Epidemiology,Kucharska-Newton et al. (5) describe an associationbetween socioeconomic indicators and the incidence ofcoronary heart disease (CHD) in two cohorts similar inage and sex. One cohort, the Atherosclerosis Risk inCommunities Study (ARIC), is in the United States, andthe other, FINRISK, is in Finland. Despite the fact thatFinland has universal health insurance, higher high schoolgraduation rates, and less disparity of income, the authorsobserved inverse gradients of association between levels ofeducation and incident CHD events in Finland as well asin the United States. An inverse association betweenincome and incident CHD events was also present in bothcohorts. Unfortunately, the structure of the data did notpermitthe investigatorstotest thehypothesis thattheasso-ciations were stronger or weaker in the United Statescompared with Finland.An assumption that Finland is a country free of socialdisparitiesisincorrect.Despitethefactthatthesocialgradi-entsinFinlandaremuchlesssteepthanintheUnitedStatesand that reducing health inequalities has been an objectivein Finnish health policy programs since 1986, significantgradientsdoexist(6).Long-termillnessesareapproximately50%morecommonamongthelowesteducationalandothersocioeconomic groups than among the highest groups, anda social gradient exists for both life expectancy and healthylife expectancy in Finland. The socioeconomic differencesin health-related behaviors are also great. Smoking, heavyalcohol use, and binge drinking are more common in thelower socioeconomic groups. At the same time, followingdietary and physical activity recommendations is morecommon in the higher socioeconomic groups. As is truefor the ARIC and FINRISK cohorts, socioeconomic gradi-entsarepresentforbiologicalriskfactorsintheFinnishpop-ulation as a whole, and there are socioeconomic differencesin the use of health services that do not fully correspond tothe estimated need for care. These gradients probablycontribute to the observed CHD gradient to some extent.Both in the time of Virchow and today, the wealthiertend to be the healthier (7). However, comparing themagnitude of disparities among nations and within the 50states of the United States, Wilkinson and Pickett (7)have concluded that the magnitude of social disparities isa much better predictor of population health than is theabsolute level of wealth, and disparities burden entire soci-etiesdnot just the poor. Countries and states with largerdisparities in income and education have more mentalillness, drug addiction, and alcoholism; lower life expec-tancy and greater infant mortality; more obesity; poorereducational performance of children; more teenage births;more homicides; greater imprisonment rates; lower levelsof trust; and lower social mobility.Despite advances in health care since the time of Virch-ow,‘‘thecausesofthecauses’’remainthesame:beingonthelower end of the social ladder; stress; achallenging early lifeexperience; social exclusion; stress at work; unemployment;lack of social support; addiction; lack of access to high-quality food; and, lack of safe transportation (8). Althoughit clearly has an impact on the health of individuals, healthcareisbelievedtocontributetoonly20%,andforperhapstoas little as 10%, of the health of populations (9!11).Social epidemiology can contribute to the task ofoptimizing the health of individuals and improving thehealth of populations in several ways. One is to arm withdata those who are already taking public health action.This contribution requires no further explanation.A second contribution of social epidemiology would bestrengthening the evidence that it is in the self-interest ofbusinesses to promote population health. This is necessarybecause, as observed by Webber and Mercure at theNationalBusinessCoalitiononHealth,‘‘Moralresponsibilityand doing the right thing are not dominant factors in corporatedecision making [emphasis added]. Investment decisions aremade by building a business case that an investment today