jurnal epidemiologid

10
POINT-COUNTERPOINT The triumph of the null hypothesis: epidemiology in an age of change Wasim Maziak Accepted 8 May 2008 Summary The recent confusion concerning the relation between hormone replacement therapy and cardiovascular disease has stirred a new wave of debate about the value and future of epidemiology. Opponents of epidemiology suggest an ever-diminishing role in an age of small risks and complex diseases, yet proponents are not in consensus about how to adapt their discipline to the challenges associated with ageing societies and changing disease patterns. While epidemiology is likely to be increasingly called upon to make sense of the risks involved with these changes, wading into this era with a mindset and tools that were derived from epidemiology’s ‘golden era’ of tackling major risk factors, has created more confusion than understanding. Increasingly, we need to downsize epidemiology to what is testable, measurable, and relevant, based on robust methodology and public health rationale. Applying an evolutionary perspective, that views health problems of modernity as a manifesta- tion of the mismatch between our ancient genes and hi-tech lifestyles, can provide guidance for a 21st century research agenda. Keywords Epidemiology, chronic disease, small risk, complex disease, evolu- tionary epidemiology Introduction The recent uncertainty surrounding the relation between hormone replacement therapy and cardio- vascular disease (HRT-CVD) has again ignited the debate about the value and future of epidemiology. 1,2 The issue this time is more serious than the optimal amount of fruits and vegetables we need to eat daily, as it involves the devastating irony that millions of healthy women have been encouraged to take a medication that may put them at risk of the same ailment they were trying to ward off. 3 Underlying this dilemma is a credibility crisis brought about by inconsistencies in the results of various epidemiol- ogical studies. 4–6 Increasingly, voices within and outside the discipline of epidemiology are calling for a total re-evaluation of its tools and paradigms, some going as far as to suggesting abandoning the field entirely. 1,7–11 One can argue whether epidemiology is to blame for this state of affairs by adopting the results of cohort studies to formulate treatment guidelines, 12 or has been the voice of reason via arguing caution about the ‘protective’ relation between HRT-CVD, 13–15 or is an innocent bystander or even pawn at the hands of mass media and cor- porate interests that manipulate public opinion about medical treatments. 1 Regardless, the unavoidable issue is the legitimate concern about the role of epi- demiology in an era of small effect, lifestyle-related risks of chronic diseases. This concern has in recent years stirred calls for major methodological and conceptual reevaluation of observational studies (e.g. case control and cohort), 1,5,8–11 as their propen- sity for subtle forms of bias and confounding can influence their value for the study of small risks of chronic disease. Yet a more suitable starting point would be to restore some of the fundamentals of epidemiological practice based on strong theoretical Corresponding author. University of Memphis, 633 Normal Street, Memphis, TN 38152, USA. E-mail: [email protected] University of Memphis, School of Public Health, Memphis, TN 38152, USA. Published by Oxford University Press on behalf of the International Epidemiological Association ß The Author 2008; all rights reserved. Advance Access publication 17 December 2008 International Journal of Epidemiology 2009;38:393–402 doi:10.1093/ije/dyn268 393

Upload: liliana-wood

Post on 15-Nov-2015

212 views

Category:

Documents


0 download

DESCRIPTION

kedoteran

TRANSCRIPT

  • POINT-COUNTERPOINT

    The triumph of the null hypothesis:epidemiology in an age of changeWasim Maziak

    Accepted 8 May 2008

    Summary The recent confusion concerning the relation between hormonereplacement therapy and cardiovascular disease has stirred a newwave of debate about the value and future of epidemiology.Opponents of epidemiology suggest an ever-diminishing role in anage of small risks and complex diseases, yet proponents are not inconsensus about how to adapt their discipline to the challengesassociated with ageing societies and changing disease patterns. Whileepidemiology is likely to be increasingly called upon to make sense ofthe risks involved with these changes, wading into this era with amindset and tools that were derived from epidemiologys golden eraof tackling major risk factors, has created more confusion thanunderstanding. Increasingly, we need to downsize epidemiology towhat is testable, measurable, and relevant, based on robustmethodology and public health rationale. Applying an evolutionaryperspective, that views health problems of modernity as a manifesta-tion of the mismatch between our ancient genes and hi-techlifestyles, can provide guidance for a 21st century research agenda.

    Keywords Epidemiology, chronic disease, small risk, complex disease, evolu-tionary epidemiology

    IntroductionThe recent uncertainty surrounding the relationbetween hormone replacement therapy and cardio-vascular disease (HRT-CVD) has again ignited thedebate about the value and future of epidemiology.1,2

    The issue this time is more serious than the optimalamount of fruits and vegetables we need to eat daily,as it involves the devastating irony that millions ofhealthy women have been encouraged to take amedication that may put them at risk of the sameailment they were trying to ward off.3 Underlying thisdilemma is a credibility crisis brought about byinconsistencies in the results of various epidemiol-ogical studies.46 Increasingly, voices within andoutside the discipline of epidemiology are calling for

    a total re-evaluation of its tools and paradigms, somegoing as far as to suggesting abandoning the fieldentirely.1,711 One can argue whether epidemiologyis to blame for this state of affairs by adopting theresults of cohort studies to formulate treatmentguidelines,12 or has been the voice of reason viaarguing caution about the protective relationbetween HRT-CVD,1315 or is an innocent bystanderor even pawn at the hands of mass media and cor-porate interests that manipulate public opinion aboutmedical treatments.1 Regardless, the unavoidableissue is the legitimate concern about the role of epi-demiology in an era of small effect, lifestyle-relatedrisks of chronic diseases. This concern has in recentyears stirred calls for major methodological andconceptual reevaluation of observational studies(e.g. case control and cohort),1,5,811 as their propen-sity for subtle forms of bias and confounding caninfluence their value for the study of small risks ofchronic disease. Yet a more suitable starting pointwould be to restore some of the fundamentals ofepidemiological practice based on strong theoretical

    Corresponding author. University of Memphis, 633 NormalStreet, Memphis, TN 38152, USA.E-mail: [email protected]

    University of Memphis, School of Public Health, Memphis, TN38152, USA.

    Published by Oxford University Press on behalf of the International Epidemiological Association

    The Author 2008; all rights reserved. Advance Access publication 17 December 2008International Journal of Epidemiology 2009;38:393402

    doi:10.1093/ije/dyn268

    393

  • guidance, proper assessment tools and clear publichealth rationale. As these elements are usually withinresearchers control, addressing them in the context ofnew directions to improve the prospects of chronicdisease epidemiology is warranted.

    A historical snapshotFor some time now, epidemiologists have been debat-ing the future ability of their discipline to accommodateemerging disease patterns resulting from the ageingand lifestyle changes of modern societies.79,1622

    Mervyn and Ezra Susser identified three main historicalstages of epidemiology reflecting the main healththreats of the times and the level of knowledge aboutthem. Starting from the sanitary era with its Miasmaparadigm, to the infectious disease era accompanyingthe germ theory, to our chronic disease/risk factors erawith its so called black box paradigm, to quote PetrSkrabaneks famous metaphor.7,23 Perhaps, it is blackbox epidemiology, referring to the pursuit of exposure-outcome relations without much attention to biologicalunderstanding or inference, that has been mostproblematic.2224 The willingness of epidemiologists torun ahead of biology to influence the societal burden ofdisease is a longstanding tradition of the discipline withsome impressive successes.25,26 But while mechanisticassociations can lead to hypothesis formulation in thearea of major risk factors,26 they are unlikely to be assuccessful with small risks, given the complexity of thecausal grid. This inadequacy has paved the way for anew phase in epidemiology,9,27,28 called ecoepidemiol-ogy by the Sussers.27 The concept of ecoepidemiology isbased on a multilevel paradigm called the Chineseboxes to reinforce the importance of distal (societal),individual and microbiological interactions in diseasedevelopment.27 The ecoepidemiology concept also is anattempt to reclaim the public health edge of epidemiol-ogy, thought by many to have been lost amidst anoveremphasis on individual-level risk factors.1821,29

    Risk factor epidemiology and theimportance of guiding hypothesesObservational studies have been instrumental for theidentification of major risk factors to health (e.g.smoking, hypertension, hypercholesterolemia, malnu-trition). Yet the HRT-CVD debate has drawn attentionto the potentially high price of making unwarrantedclaims about small and interconnected associations.Epidemiologys doubters argue that the success storiesof epidemiology were all easy hits; the magnitude ofthe association between cigarette smoking and lungcancer was so large that it could be reliably observedeven with flawed study designs.30 However, when wemove to the realm of complex diseases and smallereffect sizes, bias and confounding start to creep intocohort and case control studies in a variety of

    unpredictable ways leading to their derailment inany direction.8,22,30,31 But, if we could establish majorrisk factors with crude tools, why cannot we be ableto assess small risks with better ones?

    So far, the uncertainty about epidemiological evi-dence has led to the here-today-gone-tomorrownature of medical wisdom,30 and perhaps the confu-sion we ourselves have about how to lead a healthylifestyle.3032 Notwithstanding our position on theever-changing nature of scientific knowledge and theprocesses involved, epidemiology has been hurt mostby ill conceived and conducted studies with rushedconclusions. As we witnessed, an excess of suchstudies, combined with vested interests in certainresults and media hunger for the newsworthy, hasbecome a formula for trouble.

    Epidemiologists have always been vigilant about thedanger of claiming associations that do not exist inreality by adopting the null/alternative hypothesisapproach, which emphasizes lower tolerance for sucherror (i.e. type I error) than for missing a real link(i.e. type II error).33 It is an approach similar to thejudicial system, which considers convicting an innocenta greater mistake than letting a criminal go free. Thisapproach emphasizes as well the need for research to bedriven at the outset by a sound and fully articulatedhypothesis. The wisdom of this safeguard seems tobe lost on many researchers nowadays, who like toformulate and interpret their studies by what comes outof the logistic regression grinder. Given researchersingeniousness in explaining exotic associations, and theever-expanding volume of knowledge, it is not hard tofind biological explanations for contradictory findings.For example, studies showing positive associationsbetween exposure to pets and childhood asthmaattribute this association to animal allergens (compat-ible with the allergy paradigm), while studies showingnegative associations attribute it to pet-related micro-bial products (compatible with the hygiene para-digm).34,35 There are ample examples of tailor-madepost hoc hypotheses, transforming epidemiology from arational to a ridiculous endeavour, and highlighting thegrowing importance of epidemiological studies beingguided by well-grounded a priori hypotheses.

    The asthma epidemic and theimportance of measurementIn the current age of small risk factors, the need forepidemiology to be grounded in strong inference ismatched only by the need for sensitive tools to assessexposures and outcomes. As an example, childhoodasthma is a recent global epidemic, whose secularand spatial patterns favour environmental causes.36,37

    Yet more than two decades of intensive research hasyet to yield a single target for intervention.8,38 To getout of this deadlock, researchers have been calling forlarge prospective studies, studies starting as early in

    394 INTERNATIONAL JOURNAL OF EPIDEMIOLOGY

  • life as possible and to differentiate between differentasthma phenotypes.3941 Others have argued, invokingGeoffrey Roses seminal work,42 that the etiologicalfactors involved in the asthma epidemic are likely tobe homogenously distributed within populations,indicating the need for international comparisons.43

    These attempts to shore up asthma research reflectsound epidemiological reasoning, but have yet toadvance our understanding of the etiology of theasthma epidemic, despite a big price tag.4446

    So, is this impasse in asthma research due to aninadequate conceptual framework or faulty studydesigns and tools of epidemiology? Research into theasthma epidemic was guided for the most part by thehygiene hypothesis, which relied on early observationsof the protective effect of sibship size and birth orderto suggest that the asthma epidemic could be the sideeffect of our increased hygiene and success in curbinginfectious diseases.47 Biological underpinnings of thehygiene hypothesis were soon provided based onthe preferential development of immune-allergicresponses in the absence of infectious stimuli.48,49

    Certainly, the hygiene hypothesis provided a timelymodel for an environmentally driven epidemic incor-porating changes in our lifestyles with a good dose ofevolutionary wisdom. While evidence for and againstthe hygiene model are accumulating, it has becomeapparent that it is an over-simplification of a morecomplex picture.45,5052

    Part of the reason why we still do not have a verdicton the hygiene hypothesis despite countless studies,lies in the uncertainty surrounding the measurementtools we are applying in asthma research. For the mostpart, this research relied on self-reported or parental-reported questionnaires to assess asthma symptomsand exposures. As a result, asthma was transformedgradually into wheeze, as the symptom and thedisease became synonymous in epidemiologicalstudies. Even in English, this substitution results incrude assessment of a disease that has severalphenotypic and age-related manifestations53 but inother languages, translations of wheeze can relatepoorly to asthma.54,55 Video questionnaires or objectivemarkers (e.g. bronchial hyperresponsiveness, BHR)also have been used for outcome assessment, but nonehas come close to being a gold standard for theassessment of asthma in epidemiological studies.5664

    Exposure assessment in asthma studies has not faredbetter, as questionnaires again have been the main toolto assess exposure to infection, animals and theirproducts, air pollutants, certain foods and medicines,and secondhand smoke.65,66 For example, exposure toair pollution in the International Study of Asthma andAllergies in Children (ISAAC) was assessed based onquestions asking about (i) the frequency of truckspassing on the residential street on weekdays; and(ii) the severity of traffic noise that forced participantsto close the window.67 One cannot expect suchassessment to provide credible information about this

    exposure, let alone study its relative contributionamong a variety of other exposures measuredwith similarly crude instruments. In cases wheremore objective markers of exposure were sought(e.g. endotoxin, antibodies, skin tests) they wereeither non-specific/sensitive, were poor markers ofchronic exposure relevant to the development ofasthma, or were perhaps offset by the crude measure-ment of the outcome.6873 The resultant confusion waseven greater than researchers ingenuity to explain it,to the extent that it is not uncommon to see asthmastudies conclude that factor X was associated with pastyear wheeze, but not ever-wheeze or asthma diagnosis,or with sensitization but not asthma symptoms orBHR, or for that matter any combination of a myriadof outcomes that arguably represent the same clinicalentity.7486 Such inconsistencies have yet to invoke amajor re-examination of research paradigms and toolsused to study asthma, perhaps signifying black boxepidemiology at its darkest hour.

    From static to dynamic andrelevant epidemiologyThe transformation needed in small risk epidemiol-ogy involves not only applying better guiding modelsand sensitive markers, but perhaps as importantly, adeparture from a static perception of the relationbetween exposure and disease, reflecting the era ofquestionnaires and categorical variables, to a moredynamic, context-relevant and life-course appro-ach.87,88 As such, snapshot exposure assessmentshould be replaced with modelling of exposures basedon information from a variety of sources/levels (GIS,mobility patterns, biosensors), the use of sensitivemarkers of long-term exposure (e.g. hair/nail nicotinefor exposure to ETS, or glycosolated haemoglobin forcontrol of blood sugar), and the incorporation of macro-level attributes (e.g. area level characteristics).8994 Thiscan potentially sharpen our ability to approximate thedynamic nature of our interaction with our ever-changing environments. Similar transformation isneeded in our perception of risk to reflect its dynamicnature across peoples life roles, past experiences andcurrent behaviors and contexts. For example, many ofus will fall under some risk category for several chronicdiseases only by virtue of our age, which may lead us toseek controversial treatments that can be themselvesrisky (e.g. vitamin/mineral supplements).9597

    This inner conceptual dynamism should be supple-mented with an outer public health orientation.Without that, epidemiology may dwell in the realmof the ideal rather than deal with the constraints ofreality. For example, even if efficacy trials confirm thebenefits of a school-based programme or five portions/day of fruits in obesity prevention, this can be uselessif such results are not translatable into society-widesustainable interventions.98 Another form of thisreductionist approach is for epidemiological studies

    THE NULL HYPOTHESIS AND FUTURE OF EPIDEMIOLOGY 395

  • to statistically control for socioeconomic deprivation,zooming in on more proximal behavioral determi-nants, while relegating an apparently major deter-minant of chronic disease risk to the status of anuisance variable.20 Both approaches shift focus fromthe society to the sufferers,17,20 who are arguablyblamed for failing to follow the health recommenda-tions we have imparted to them for years, and, in theprocess, stripping epidemiology of its public healthessence. This is not to say that public health cannot beadvanced by purely medical interventions (e.g. vacci-nation), or that we need to structure peoples livesaround some health ideals, but to desensitize epide-miology to a major determinant of chronic disease,such as socioeconomic deprivation has perhapscontributed to the widening of health inequalities,particularly in the USA.99 Classifications such associal epidemiology, clinical epidemiology, infectiousdisease or cardiovascular epidemiology, may havefacilitated this selectiveness in how we treat differentvariables, by creating a false perception of compart-mentalization of different disease processes withinindividuals and populations.

    The opposite can be true as well, as some studiesof distal-level determinants lack clear publichealth orientation. For example, what are the publichealth implications of demonstrating that lowneighbourhood-level educational attainment is detri-mental to cognitive functions of elderly residents,100

    as one can argue that intervening on older residentsto improve their cognitive ability is more practicalthan increasing the average education level of wholegeographic areas. Finally, many relations sought byepidemiologists are fly-by-night endeavours embody-ing no foreseeable public health rationale or follow-upplan, an example of which is the finding of a positiveassociation between husbands participation in house-work and wives psychosocial health.101 One canargue that such a study proves the obvious, canneither be free of residual confounding (husbandswho help with housework are likely to be a differentbreed from those who do not), nor has a clear publichealth message (who should use this information andhow?). For epidemiology to stay relevant, associationsmust not only be driven by statistical empiricism, butmust have a clear public health rationale.

    Epidemiology in the age ofcomplex lifestyle-related diseasesIt has been hard for epidemiologists to adapt tochanging patterns of diseases and their risks inmodern societies, partly because of the pace of suchchanges. New epidemics, such as obesity, asthma anddepressive disorders are evolving rapidly and creatingpressure for evidence-based solutions. In response,inspired by past successes with major risk factors,epidemiologists rushed to the scene with their usualtools. As it turned out, the task this time was more

    difficult, casting some serious doubt on epidemiol-ogys ability to respond to current and future threatsto health.8,9,30 While only time can tell the fate ofepidemiology, one major lesson to be learned from theHRT-CVD story is that observational epidemiologycan never be free of residual confounding whencomparisons involve health-oriented behaviours.102,103

    Yet, observational studies can be valuable for thestudy of long-term side effects of drug treatments, asthese are mostly unintended and unpredictable,therefore are not confounded by indication; meaningthat they are usually not associated with the treat-ment decision.104106 A similar valid scenario forobservational epidemiology is the study of potentialrisk/protective factors that are unknown to the public,as this reduces the probability of associative-selectionbias based on differential health awareness betweenthe comparison groups.107

    In reality, as we continue to monitor the health ofpopulations, epidemiology will likely be called uponincreasingly to make sense of the risks involved in thedramatic changes in our lifestyles and environments.But to continue doing epidemiological studies usingthe same tried and failed approaches is not an option;the asthma example is a clear indicator of howepidemiology can turn into an absurd exercise whenaims are put ahead of tools and concepts. Realizingthe current crisis of credibility, epidemiologists haveresponded by calling for larger and longer studies,coupling of epidemiology with molecular genetics(e.g. genetic epidemiology and biobanks), stronginference, incorporating multilevel attributes and forgreater attention to residual confounding.8,108114

    They have also mobilized to improve the reportingof observational studies (STROBE guidelines) to alloweditors, reviewers and consumers of epidemiologicaldata to make informed judgement on the quality ofreported studies.115 All these approaches can poten-tially improve epidemiologys ability to zoom in onsmall risks, yet the defining feature of the small risk/chronic disease era has so far been the triumph of thenull hypothesis (i.e. no effect).8,9,116

    Such responses reflect healthy self-criticism on thepart of epidemiologists, but a clearer articulation ofpriorities is needed in the face of mounting criticism.Bigger samples and longer studies, for example, canlead to the magnification of errors, loss to follow up,and increase in costs, while applying multilevelapproaches without a sound theoretical frameworkruns the risk of becoming another form of black boxepidemiology.117 Genetic epidemiology on the otherhand, is not free of the problems of observationalepidemiology,8 yet recent years have witnessedsome promising advances in this field, especially theapplication of Mendelian randomization to controlfor environmental confounders in observationalstudies.118 Mendelian randomization utilizes therandom distribution of genetic alleles at the time ofgamete formation to identify genetic variants that

    396 INTERNATIONAL JOURNAL OF EPIDEMIOLOGY

  • robustly predict environmentally modifiable exposuresand uses them as un-confounded proxies (instru-ments) for those exposures. While the promise of thisapproach is still unfolding,119 there are currently notmany conditions that can be studied this way, i.e. inwhich we have a clear understanding of the geneticbasis of suspected etiological exposures.

    As epidemiologists relish the promise of ecoepide-miology,28,111 this concept can come with its owncaveats. Namely, a comprehensive hypothesis basedon this approach can involve many-to-many factorsto account for,8,120 as the range and possibilitiesof geneenvironment interactions and pathwaysinvolved in chronic disease form a vast causaluniverse. As such, a vision that lies between blackbox and Chinese boxes may be needed, wherebycertain causal pathways can be identified and workedupon without the need to unravel the whole complex-ity of the relation between exposures, genes, andchronic diseases. For example, if we can establish aneffective way to influence obesity in the society (e.g.through policy that rewards/supports active transpor-tation), we can potentially influence a good deal ofCVD, and perhaps a variety of other illnesses, withoutneeding to have all the information about the factorsand mechanisms involved. Ecological models of majorhealth risks are emerging, which can harness candi-date pathways for study and intervention (Figure 1).In parallel, methodological paradigms and analyticaltools are being developed to accommodate this newdirection (e.g. complexity theory, multilevel model-ling, pathway analysis).98,115,121124 For example, theuse of multilevel analysis for the study of obesity andCVD risk has helped to shift the focus from personalbehaviours to encompass wider environmental influ-ences (e.g. neighbourhood adversities) that maybe more amenable to sustainable interventionsthrough policy.125129 The widening of health inequal-ities in the US, driven mainly by non-linear socio-economic and political dynamics,99 underscores theimportance of these new trends in epidemiology.

    An evolutionary perspectivefor epidemiologyGiven the methodological constraints imposed onepidemiological research by the complex nature ofproblems we are increasingly facing, epidemiology canbenefit from some guidance as to what risk factorsrepresent good targets for exploration, and how tointerpret inconsistent epidemiological data. An evolu-tionary perspective can help provide a guiding frame-work for the epidemiology of chronic diseases, beingviewed as a result of the immense adaptive pressuresbrought about by the mismatch between our genes stilllingering at the hunter-gatherer era and our hi-techlifestyles.130 Accordingly, exposures relevant to ourhealth are those that either underwent a rapid changewithin a short time, or represent an obvious diversionfrom the environments that prevailed during most ofour evolution. In this sense, our eating, mobility,recreation, socialization and communication patterns,as well as our increasingly indoor existence should berelevant to the development of chronic diseases suchas obesity, asthma, CVD and depression.131

    As broad as this perspective can be, its applicationcan help to sift through the tides of confusing healthinformation we are bombarded with each day. Forexample, examining the HRT-CVD relation under theevolutionary lens would have raised a big red flagabout a potential for harm; women did not evolve tohave lifelong active ovaries, despite the clear repro-ductive advantage such a trait would have conferred.The same would apply to the long-held protectiverelation between low fat diet and CVD risk132136,which was recently refuted in a randomized clinicaltrial.137 As huntergatherers, humans until 500 gen-erations ago consumed mainly wild and unprocessedfood foraged and hunted from their environment andrich in fats, fibre, vitamins and minerals.138140 So,the low-fat mania perhaps does not tell the wholestory about the dietCVD risk relation, especiallywhen corporate voices join the choir by promoting

    Influences Mediators Behaviours Outcomes

    Increased sedentary pastime; TV/PC/games

    Over consumption ofjunk food-soft drinks

    Individual-FamilyStructure (parenthood, siblings, pets), SES,

    perceptions/attitudes (safety, healthy nutrition & behaviour), time structure

    SchoolFood/PA policy, PA equipment, food sales

    and sponsorships

    Area Walkability, safety (crime-traffic), social

    networks, deprivation, access to fast food orquality foods, recreational facilities

    - Increased time-indoors- Increased access to food - Decreased opportunity of PA - Increased TV/screen time- Increased exposure to food advertisement

    Decreased unstructured PA En

    ergy

    imba

    lanc

    e/at

    risk

    of

    obe

    sity/

    obes

    ity

    Society Food policies and regulations, urban

    planning, energy policies

    Figure 1 Simplified multilevel approach to the study of environmental influences on obesity98

    THE NULL HYPOTHESIS AND FUTURE OF EPIDEMIOLOGY 397

  • low-fat options, potentially creating a complacencythat can even lead to increased consumption.98

    Another benefit of the evolutionary perspective is tohelp understand the difficulties facing genetic epide-miology, based on the fact that the direction ofevolution of traits is from phenotype to genotype andnot the opposite.141 This can underlie the existence ofmultiple genetic pathways to each outcome compli-cating the genetic study of chronic diseases. To addanother layer of complexity is the fact that phenotypefitness is not a simple measure, but represents acomplex fitness landscape involving the trade-off ofnumerous traits, each of which is the result of evengreater genetic diversity.142 For complex traits there-fore, considerable genetic variance can be maintainedwithin the population even in the face of strongselection forces.143 Such complexity undermines thehopes that genetic epidemiology will provide readyanswers to the puzzles of chronic diseases.

    Finally, an evolutionary perspective can guideintervention research. Specifically, rather than basingnutritional and physical activity recommendations onobservational data that may well reflect a systematicdifference in the way we assemble our comparisongroups (e.g. those who eat healthy are likely to beengaged in other less-measurable health behaviours),we may need to inform such recommendations withknowledge of our ancestral dietary and activitypatterns.144,145 In addition, the appreciation that wehave a genetically hard-wired taste for energy-densefood can help us understand why most dietary-basedobesity interventions fail in the long run.98 So insteadof trying to work more forcefully against our instinctsby increasing the intensity and length of interven-tions, we may be better off trying to manipulate foodtaste and energy content,146 to imitate the Paleolithicdiet but without the excess calories.

    Concluding remarksApparently, the golden era of major risk factor epidemi-ology is giving way to a less glorious, but certainly morecomplex and perhaps more important one of studyingsmall and interconnected risk factors related to ourever-changing lifestyles and environments. We perhapshave just scratched the surface of what epidemiologycan achieve and how it can help us understand howunfavourable environments are shaping our behaviourand health. Wading into this era with a major riskfactor mentality and instruments has created confu-sion. While endorsing new developments in epidemiol-ogical research, the era of chronic disease epidemiologymandates more than ever the need to rely on soundtheoretical models as well as accurately measurableoutcomes and exposures. So in contrast to the calls forlarger, longer and wider-reach epidemiology, what isadvocated here is to downsize epidemiological researchto what is testable, measurable and relevant. Anevolutionary perspective of the dynamic interaction ofhumans with their environments can help guide such aresearch agenda. In the age of publish or perish, vestedinterests, publication bias, scarce funding, mediashunger for hit news and web publishing, epidemiologycan best navigate these rough waters by being anchoredin a clear sense of its inner methodological constraintsand outer public health thrust.

    FundingNIDA (R01DA024876-01 to W.M.).

    AcknowledgementsWasim Maziak thankfully acknowledges Dr KennethWard for critically reading and editing this article.

    Conflict of interest: None declared.

    KEY MESSAGES

    There is a crisis of credibility facing epidemiology today, brought about by the barrage of studies withless than optimal methods and conflicting results.

    As epidemiology enters the era of chronic disease and small risk, it becomes more critical forepidemiological studies to be guided at the inception by well-grounded hypotheses, a dynamicperception of the relation between exposure-outcome and to utilize accurate assessment tools.

    Novelty or methodological precision should not substitute for public health relevance when evaluatingepidemiological studies.

    New conceptual (e.g. multilevel ecoepidemiology) and methodological (e.g. Mendelian randomization)advances should be embraced in light of the need to downsize epidemiology to what is testable,measurable and relevant.

    An evolutionary and dynamic understanding of our interactions with our changing environments canprovide a guiding context for epidemiological research.

    398 INTERNATIONAL JOURNAL OF EPIDEMIOLOGY

  • References1 Lawlor DA, Davey Smith G, Ebrahim S. Commentary: the

    hormone replacement-coronary heart disease conundrum:is this the death of observational epidemiology? Int JEpidemiol 2004;33:46467.

    2 Stampfer M. Stampfer. Commentary: Hormones andheart disease: do trials and observational studies addressdifferent questions? Int J Epidemiol 2004;33:45455.

    3 Harrington DM, Howard TD. From presumed benefits topotential harmhormone therapy and heart disease.N Engl J Med 2003;349:51921.

    4 Whittmore AS, McGuire V. Observational studiesand randomized studies of hormone replacementtherapy: what can we learn from them? Epidemiology2003;14:810.

    5 Michels KB. Hormone replacement therapy in epidemio-logic studies and randomized clinical trialsare wecheckmate? Epidemiology 2003;14:35.

    6 Taubes G. Epidemiology faces its limits. Science1995;269:16469.

    7 Susser M, Susser E. Choosing a future for epidemiology:I. Eras and paradigms. Am J Public Health 1996;86:66873.

    8 Buchanan AV, Weiss KM, Fullerton SM. Dissectingcomplex disease: the quest for the Philosophers Stone?Int J Epidemiol 2006;35:56271.

    9 Davey Smith G, Ebrahim S. Epidemiologyis it time tocall it a day? Int J Epidemiol 2001;30:111.

    10 Pritchard KI. Should observational studies be a thing ofthe past? J Natl Cancer Inst 2008;100:45152.

    11 Le Fanu J. Rise and fall of modern medicine. Lancet1999;354:518.

    12 Barton S. Which clinical studies provide the bestevidence? The best RCT still trumps the best observationalstudy. BMJ 2000;321:25556.

    13 Petitti DB, Perlman JA, Sidney S. Noncontraceptiveestrogens and mortality: long-term follow-up of womenin the Walnut Creek Study. Obstet Gynecol 1987;70(3 Pt 1):28993.

    14 Barrett-Connor E, Stuenkel CA. Hormone replacementtherapy (HRT)risks and benefits. Int J Epidemiol2001;30:42326.

    15 Vandenbroucke JP. How much of the cardioprotectiveeffect of postmenopausal estrogens is real? Epidemiology1995;6:2078.

    16 Rothman KJ. Commentary: Epidemiology still ascendant.Int J Epidemiol 2007;36:71011.

    17 Pearce N. Commentary: The rise and rise of corporateepidemiology and the narrowing of epidemiologys vision.Int J Epidemiol 2007;36:71317.

    18 Pearce N. Traditional epidemiology, modern epidemiol-ogy, and public health. Am J Public Health 1996;86:67883.

    19 Savitz DA, Poole C, Miller WC. Reassessing the role ofepidemiology in public health. Am J Public Health1999;89:115861.

    20 McMichael AJ. Prisoners of the proximate: loosening theconstraints on epidemiology in an age of change. Am JEpidemiol 1999;149:88797.

    21 Krieger N. Epidemiology and social sciences: towards acritical reengagement in the 21st century. Epidemiol Rev2000;22:15563.

    22 Davey Smith G. Reflections on the limitations toepidemiology. J Clin Epidemiol 2001;54:32531.

    23 Skrabanek P. The emptiness of the black box.Epidemiology 1994;5:55355.

    24 Susser M. Does risk factor epidemiology put epidemiologyat risk? Peering into the future. J Epidemiol CommunityHealth 1998;52:60811.

    25 Savitz DA. In defense of black box epidemiology.Epidemiology 1994;5:55052.

    26 Greenland S, Gago-Dominguez M, Castelao JE. The valueof risk-factor (black-box) epidemiology. Epidemiology2004;15:52935.

    27 Susser M, Susser E. Choosing a future for epidemiology: II.From black box to Chinese boxes and eco-epidemiology.Am J Public Health 1996;86:67477.

    28 Poole C, Rothman KJ. Our conscientious objection to theepidemiology wars. J Epidemiol Community Health 1998;52:61314.

    29 Shy CM. The failure of academic epidemiology: witnessfor the prosecution. Am J Epidemiol 1997;145:47984.

    30 Taubes G. Do We Really Know What Makes Us Healthy? TheNew York Times, September 17, 2007.

    31 Mayes LC, Horwitz RI, Feinstein AR. A collection of56 topics with contradictory results in case-controlresearch. Int J Epidemiol 1988;17:68085.

    32 Taubes G. Nutrition. The soft science of dietary fat. Science2001;291:253645.

    33 Schulz KF, Grimes DA. Bias and causal associations inobservational research. Lancet 2002;359:24852.

    34 Apter AJ. Early exposure to allergen: is this the catsmeow, or are we barking up the wrong tree? J Allergy ClinImmunol 2003;111:93846.

    35 Heinrich J, Gehring U, Douwes J et al. INGA-StudyGroup. Pets and vermin are associated with high endo-toxin levels in house dust. Clin Exp Allergy 2001;31:183945.

    36 Kay AB. Allergy and allergic diseases. First of two parts.N Engl J Med 2001;344:3037.

    37 Eder W, Ege MJ, von Mutius E. The asthma epidemic.N Engl J Med 2006;355:222635.

    38 von Mutius E. Allergies, infections and the hygienehypothesisthe epidemiological evidence. Immunobiology2007;212:43339.

    39 Sheikh A, Strachan DP. The hygiene theory: fact orfiction? Curr Opin Otolaryngol Head Neck Surg 2004;12:23236.

    40 Rebordosa C, Kogevinas M, Srensen HT, Olsen J. Pre-natal exposure to paracetamol and risk of wheezing andasthma in children: A birth cohort study. Int J Epidemiol2008;37:58390.

    41 Wright AL. Epidemiology of asthma and recurrentwheeze in childhood. Clin Rev Allergy Immunol2002;22:3344.

    42 Rose G. Sick individuals and sick populations. Int JEpidemiol 1985;14:3238.

    43 Lewis S. ISAACa hypothesis generator for asthma?International Study of Asthma and Allergies inChildhood. Lancet 1998;351:122021.

    44 Beasley R, Ellwood P, Asher I. International patterns ofthe prevalence of pediatric asthma the ISAAC program.Pediatr Clin North Am 2003;50:53953.

    THE NULL HYPOTHESIS AND FUTURE OF EPIDEMIOLOGY 399

  • 45 Maziak W. The asthma epidemic and our artificialhabitats. BMC Pulm Med 2005;5:5.

    46 Buchanan AV, Weiss KM, Fullerton SM. On stones,wands, and promises. Int J Epidemiol 2006;35:59396.

    47 Strachan DP: Hay fever, hygiene, and household size.BMJ 1989;299:125960.

    48 Martinez FD, Holt PG. Role of microbial burdenin aetiology of allergy and asthma. Lancet 1999;354(Suppl 2):SII1215.

    49 Cookson WO, Moffatt MF. Asthma: an epidemic in theabsence of infection? Science 1997;275:4142.

    50 Ramsey CD, Celedon JC. The hygiene hypothesis andasthma. Curr Opin Pulm Med 2005;11:1420.

    51 Effros RM, Nagaraj H. Asthma: new developmentsconcerning immune mechanisms, diagnosis and treat-ment. Curr Opin Pulm Med 2007;13:3743.

    52 van Schayck CP, Knottnerus JA. No clinical evidence baseto support the hygiene hypothesis. Prim Care Respir J2004;13:7679.

    53 Stein RT, Martinez FD. Asthma phenotypes in childhood:lessons from an epidemiological approach. Paediatr RespirRev 2004;5:15561.

    54 Pekkanen J, Remes ST, Husman T et al. Prevalence ofasthma symptoms in video and written questionnairesamong children in four regions of Finland. Eur Respir J1997;10:178794.

    55 Hong SJ, Kim SW, Oh JW et al. The validity of the ISAACwritten questionnaire and the ISAAC video questionnaire(AVQ 3.0) for predicting asthma associated withbronchial hyperreactivity in a group of 1314 year oldKorean schoolchildren. J Korean Med Sci 2003;18:4852.

    56 Remes ST, Pekkanen J, Remes K, Salonen RO, Korppi M.In search of childhood asthma: questionnaire, tests ofbronchial hyperresponsiveness, and clinical evaluation.Thorax 2002;57:12026.

    57 Giovannangelo M, Gehring U, Nordling E et al. Childhoodcat allergen exposure in three European countries:The AIRALLERG study. Sci Total Environ 2006;369:8290.

    58 Crane J, Mallol J, Beasley R, Stewart A, Asher MI.International Study of Asthma and Allergies in ChildhoodPhase I study group. Agreement between written andvideo questions for comparing asthma symptoms inISAAC. Eur Respir J 2003;21:45561.

    59 Pizzichini MM, Rennie D, Senthilselvan A, Taylor B,Habbick BF, Sears MR. Limited agreement betweenwritten and video asthma symptom questionnaires.Pediatr Pulmonol 2000;30:30712.

    60 Michel G, Silverman M, Strippoli MP et al. Parentalunderstanding of wheeze and its impact on asthmaprevalence estimates. Eur Respir J 2006;28:112430.

    61 Braun-Fahrlander C, Gassner M, Grize L et al.Comparison of responses to an asthma symptomquestionnaire (ISAAC core questions) completed byadolescents and their parents. SCARPOL-Team. SwissStudy on Childhood Allergy and Respiratory Symptomswith respect to Air Pollution. Pediatr Pulmonol 1998;25:15966.

    62 Gharagozlou M, Khalili S, Hallajmofrad M,Mohammadzadeh R, Mousavi G, Golkari H. Gendersimilarity in low agreement between written and videoISAAC asthma questionnaires. Monaldi Arch Chest Dis2006;65:18488.

    63 Hederos CA, Hasselgren M, Hedlin G, Bornehag CG.Comparison of clinically diagnosed asthma with parentalassessment of childrens asthma in a questionnaire.Pediatr Allergy Immunol 2007;18:13541.

    64 Rahimi Rad MH, Hejazi ME. Agreement between writtenand video asthma symptoms questionnaires in schoolchildren in Urmia, Iran. Iran J Allergy Asthma Immunol2007;6:2125.

    65 Asher MI, Keil U, Anderson HR et al. International Studyof Asthma and Allergies in Childhood (ISAAC): rationaleand methods. Eur Respir J 1995;8:48391.

    66 Burney PG, Luczynska C, Chinn S, Jarvis D. TheEuropean Community Respiratory Health Survey. EurRespiratory J 1994;7:95460.

    67 Behrens T, Taeger D, Maziak W et al. Self-reported trafficdensity and atopic disease in children. Results of theISAAC Phase III survey in Muenster, Germany. PediatrAllergy Immunol 2004;15:33139.

    68 Maziak W. Endotoxin and asthma. N Engl J Med2003;348:17174.

    69 Rylander R. Endotoxin and asthma. Am J Respir Crit CareMed 2006;173:1177.

    70 Lynch NR, Hagel IA, Palenque ME et al. Relationshipbetween helminthic infection and IgE response in atopicand nonatopic children in a tropical environment.J Allergy Clin Immunol 1998;101(2 Pt 1):21721.

    71 Obihara CC, Beyers N, Gie RP et al. Respiratory atopicdisease, Ascaris-immunoglobulin E and tuberculin testingin urban South African children. Clin Exp Allergy2006;36:64048.

    72 Ponte EV, Lima F, Araujo MI, Oliveira RR, Cardoso LS,Cruz AA. Skin test reactivity and Der p-induced inter-leukin 10 production in patients with asthma or rhinitisinfected with Ascaris. Ann Allergy Asthma Immunol2006;96:71318.

    73 Ota MO, van der Sande MA, Walraven GE et al. Absenceof association between delayed type hypersensitivity totuberculin and atopy in children in The Gambia. Clin ExpAllergy 2003;33:73136.

    74 Magnusson LL, Olesen AB, Wennborg H, Olsen J.Wheezing, asthma, hayfever, and atopic eczema inchildhood following exposure to tobacco smoke in fetallife. Clin Exp Allergy 2005;35:155056.

    75 Schachter LM, Salome CM, Peat JK, Woolcock AJ.Obesity is a risk for asthma and wheeze but not airwayhyperresponsiveness. Thorax 2001;56:48.

    76 Yarnell JW, Stevenson MR, MacMahon J et al. Smoking,atopy and certain furry pets are major determinants ofrespiratory symptoms in children: the International Studyof Asthma and Allergies in Childhood Study (Ireland).Clin Exp Allergy 2003;33:96100.

    77 Arshad SH, Kurukulaaratchy RJ, Fenn M, Matthews S.Early life risk factors for current wheeze, asthma, andbronchial hyperresponsiveness at 10 years of age. Chest2005;127:5028.

    78 Bolte G, Schmidt M, Maziak W et al. The relation ofmarkers of fetal growth with asthma, allergies and serumimmunoglobulin E levels in children at age 57 years. ClinExp Allergy 2004;34:38188.

    79 Annus T, Bjorksten B, Mai XM et al. Wheezing in relationto atopy and environmental factors in Estonian andSwedish schoolchildren. Clin Exp Allergy 2001;31:184653.

    400 INTERNATIONAL JOURNAL OF EPIDEMIOLOGY

  • 80 Davey G, Venn A, Belete H, Berhane Y, Britton J. Wheeze,allergic sensitization and geohelminth infection inButajira, Ethiopia. Clin Exp Allergy 2005;35:3017.

    81 Omenaas E, Fluge O, Buist AS, Vollmer WM, Gulsvik A.Dietary vitamin C intake is inversely related to cough andwheeze in young smokers. Respir Med 2003;97:13442.

    82 Hermann C, De Fine Olivarius N, Hst A, Begtrup K,Hollnagel H. Prevalence, severity and determinants ofasthma in Danish five-year-olds. Acta Paediatr2006;95:118290.

    83 Lau S, Illi S, Platts-Mills TA et al. Multicentre AllergyStudy Group. Longitudinal study on the relationshipbetween cat allergen and endotoxin exposure, sensitiza-tion, cat-specific IgG and development of asthma inchildhoodreport of the German Multicentre AllergyStudy (MAS 90). Allergy 2005;60:76673.

    84 Carter PM, Peterson EL, Ownby DR, Zoratti EM,Johnson CC. Relationship of house-dust mite allergenexposure in childrens bedrooms in infancy to bronchialhyperresponsiveness and asthma diagnosis by age 6 to 7.Ann Allergy Asthma Immunol 2003;90:4144.

    85 Cullinan P, MacNeill SJ, Harris JM et al. Early allergenexposure, skin prick responses, and atopic wheeze atage 5 in English children: a cohort study. Thorax2004;59:85561.

    86 Harris JM, Williams HC, White C et al. Early allergenexposure and atopic eczema. Br J Dermatol2007;156:698704.

    87 Kuh D, Ben-Shlomo Y, Lynch J, Hallqvist J, Power C. Lifecourse epidemiology. J Epidemiol Community Health2003;57:77883.

    88 Lawlor DA, Davey Smith G, Patel R, Ebrahim S. Life-coursesocioeconomic position, area deprivation, and coronaryheart disease: findings from the British Womens Heart andHealth Study. Am J Public Health 2005;95:9197.

    89 Clougherty JE, Levy JI, Kubzansky LD et al. Synergisticeffects of traffic-related air pollution and exposure toviolence on urban asthma etiology. Environ Health Perspect2007;115:114046.

    90 Buxton DB. Nanotechnology in the diagnosis andmanagement of heart, lung and blood diseases. ExpertRev Mol Diagn 2007;7:14960.

    91 Srensen M, Bisgaard H, Stage M, Loft S. Biomarkers ofexposure to environmental tobacco smoke in infants.Biomarkers 2007;12:3846.

    92 Frank LD, Sallis JF, Conway TL, Chapman JE,Saelens BE, Bachman W. Many pathways from landuse to health associations between neighborhood walk-ability and active transportation, body mass index, andair quality. J Am Plann Assoc 2006;72:7587.

    93 O Campo P. Invited commentary: Advancing theory andmethods for multilevel models of residential neighbor-hoods and health. Am J Epidemiol 2003;157:913.

    94 Maziak W. Asthma and the exposure-disease tenet. J ClinEpidemiol 2002;55:73740.

    95 Bolland MJ, Barber PA, Doughty RN et al. Vascularevents in healthy older women receiving calcium supple-mentation: randomised controlled trial. BMJ 2008;336:26266.

    96 Buijsse B, Feskens EJ, Kwape L, Kok FJ, Kromhout D.Both {alpha}- and -Carotene, but Not Tocopherolsand Vitamin C, Are Inversely Related to 15-Year

    Cardiovascular Mortality in Dutch Elderly Men. J Nutr2008;138:34450.

    97 Heinen MM, Hughes MC, Ibiebele TI, Marks GC,Green AC, van der Pols JC. Intake of antioxidantnutrients and the risk of skin cancer. Eur J Cancer2007;43:270716.

    98 Maziak W, Ward KD, Stockton MB. Childhoodobesity: are we missing the big picture? Obes Rev 2008;9:3542.

    99 Unnatural causes. Is inequality making us sick.www.unnaturalcauses.org (Accessed March 12, 2008).

    100 Wight RG, Aneshensel CS, Miller-Martinez D et al. Urbanneighborhood context, educational attainment, andcognitive function among older adults. Am J Epidemiol2006;163:107178.

    101 Khawaja M, Habib RR. Husbands involvement inhousework and womens psychosocial health: findingsfrom a population-based study in Lebanon. Am J PublicHealth 2007;97:86066.

    102 Navas-Acien A, Bleys J, Guallar E. Selenium intake andcardiovascular risk: what is new? Curr Opin Lipidol2008;19:4349.

    103 Lawlor DA, Davey Smith G, Kundu D, Bruckdorfer KR,Ebrahim S. Those confounded vitamins: what canwe learn from the differences between observationalversus randomised trial evidence? Lancet 2004;363:172427.

    104 Vandenbroucke JP. When are observational studies ascredible as randomised trials? Lancet 2004;363:172831.

    105 Smeeth L, Douglas I, Hubbard R. Commentary: we stillneed observational studies of drugsthey just need tobe better. Int J Epidemiol 2006;35:131011.

    106 Avorn J. In defense of pharmacoepidemiologyembracing the yin and yang of drug research. N Engl JMed 2007;357:221921.

    107 Vandenbroucke JP. Observational research, randomisedtrials, and two views of medical science. PLoS Med2008;5:e67.

    108 Weed DL. Commentary: rethinking epidemiology.Int J Epidemiol 2006;35:58386.

    109 Coggon D. Commentary: Complex diseaserespondingto the challenge. Int J Epidemiol 2006;35:58183.

    110 Smith GD, Ebrahim S. Mendelian randomization:prospects, potentials, and limitations. Int J Epidemiol2004;33:3042.

    111 Krieger N. Postmenopausal hormone therapy. N Engl JMed 2003;348:236364.

    112 Schwartz S, Susser E. Commentary: What can epide-miology accomplish? Int J Epidemiol 2006;35:58790.

    113 Millikan RC. Commentary: The human Genome: philo-sophers stone or magic wand? Int J Epidemiol 2006;35:57881.

    114 Merikangas KR, Low NCP, Hardy J. Commentary:understanding sources of complexity in chronic diseases:the importance of integration of genetics and epidemiol-ogy. Int J Epidemiol 2006;35:59092.

    115 von Elm E, Altman DG, Egger M, Pocock SJ,Gtzsche PC, Vandenbroucke JP. STROBE Initiative.The Strengthening the Reporting of ObservationalStudies in Epidemiology (STROBE) statement: guide-lines for reporting observational studies. Epidemiology2007;18:8004.

    THE NULL HYPOTHESIS AND FUTURE OF EPIDEMIOLOGY 401

  • 116 Ioannidis JP. Why most published research findings arefalse: authors reply to Goodman and Greenland. PLoSMed 2007;4:e215.

    117 Galea S, Ahern J. Invited Commentary: considerationsabout specificity of associations, causal pathways, andheterogeneity in multilevel thinking. Am J Epidemiol2006;163:107982.

    118 Smith GD, Lawlor DA, Harbord R, Timpson N, Day I,Ebrahim S. Clustered environments and random-ized genes: a fundamental distinction betweenconventional and genetic epidemiology. PLoS Med2007;4:e352.

    119 Lawlor DA, Timpson NJ, Harbord RM et al. Exploring thedevelopmental overnutrition hypothesis using parental-offspring associations and FTO as an instrumentalvariable. PLoS Med 2008;5:e33.

    120 Poole C, Rothman KJ. Our conscientious objection to theepidemiology wars. J Epidemiol Community Health1998;52:61314.

    121 Egger G, Swinburn B. An ecological approach to theobesity pandemic. BMJ 1997;315:47780.

    122 Pearce N, Merletti F. Complexity, simplicity, andepidemiology. Int J Epidemiol 2006;35:51519.

    123 Diez Roux A. Multilevel analysis in public healthresearch. Ann Rev Public Health 2000;21:17192.

    124 Royston P. A useful monotonic non-linear model withapplications in medicine and epidemiology. Stat Med2000;19:205366.

    125 Krull JL, MacKinnon DP. Multilevel modeling ofindividual and group level mediated effects. MultivariateBehav Res 2001;36:24977.

    126 Lin G, Spann S, Hyman D, Pavlik V. Climate amenityand BMI. Obesity (Silver Spring) 2007;15:212027.

    127 Singh GK, Kogan MD, van Dyck PC. A multilevelanalysis of state and regional disparities in childhoodand adolescent obesity in the United States. J CommunityHealth 2008;33:90102.

    128 Joshu CE, Boehmer TK, Brownson RC, Ewing R.Personal, neighbourhood and urban factors associatedwith obesity in the United States. J Epidemiol CommunityHealth 2008;62:20208.

    129 Diez Roux AV, Merkin SS, Arnett D et al. Neighborhoodof residence and incidence of coronary heart disease.N Engl J Med 2001;345:99106.

    130 Eaton SB, Cordain L, Lindeberg S. Evolutionary healthpromotion: a consideration of common counterargu-ments. Prev Med 2002;34:11923.

    131 Maziak W. The hygiene hypothesis and the evolutionaryperspective of health. Prev Med 2002;35:41518.

    132 Oh K, Hu FB, Manson JE, Stampfer MJ, Willett WC.Dietary fat intake and risk of coronary heart disease inwomen: 20 years of follow-up of the Nurses HealthStudy. Am J Epidemiol 2005;161:67279.

    133 Liu S, Stampfer MJ, Hu FB et al. Whole grainconsumption and risk of coronary heart disease: resultsfrom the Nurses Health Study. Am J Clin Nutr1999;70:41219.

    134 Liu S, Manson JE, Lee I-M et al. Fruit and vegetableintake and risk of cardiovascular disease: the WomensHealth Study. Am J Clin Nutr 2000;72:92228.

    135 Fung TT, Stampfer MJ, Manson JE, Rexrode KM,Willett WC, Hu FB. Prospective study of major dietarypatterns and stroke risk in women. Stroke2004;35:201419.

    136 Kris-Etherton PM, Harris WS, Appel LJ. American HeartAssociation Nutrition Committee. Fish consumption, fishoil, omega-3 fatty acids, and cardiovascular disease.Circulation 2002;21:274757.

    137 Howard BV, Van Horn L, Hsia J et al. Low-fat dietarypattern and risk of cardiovascular disease: the womenshealth initiative randomized controlled dietary modifica-tion trial. JAMA 2006;295:65566.

    138 OKeefe JH Jr, Cordain L. Cardiovascular disease result-ing from a diet and lifestyle at odds with our Paleolithicgenome: how to become a 21st-century hunter-gatherer.Mayo Clin Proc 2004;79:1018.

    139 Milton K. The critical role played by animal sourcefoods in human (Homo) evolution. J Nutr 2003;133(11 Suppl 2):S388692.

    140 Cordain L, Eaton SB, Sebastian A et al. Origins andevolution of the Western diet: health implications for the21st century. Am J Clin Nutr 2005;81:34154.

    141 Weiss KM, Buchanan AV. Evolution by phenotype: abiomedical perspective. Perspect Biol Med 2003;46:15982.

    142 Nettle D. The evolution of personality variation inhumans and other animals. Am Psychol 2006;61:62231.

    143 Houle D. How should we explain variation in the geneticvariance of traits? Genetica 1998;102103:24153.

    144 Eaton SB. The ancestral human diet: what was it andshould it be a paradigm for contemporary nutrition? ProcNutr Soc 2006;65:16.

    145 Eaton SB, Eaton SB. An evolutionary perspective onhuman physical activity: implications for health. CompBiochem Physiol A Mol Integr Physiol 2003;136:15359.

    146 Ackroff K, Sclafani A. Energy density and macronutrientcomposition determine flavor preference conditioned byintragastric infusions of mixed diets. Physiol Behav2006;89:25060.

    402 INTERNATIONAL JOURNAL OF EPIDEMIOLOGY