Veer P. Alcohol and breast cancer results from the Netherlands cohort study. Am J. EpidemioL 1995;141:907-915. 15. Fuchs CS, Stampfer MJ, Colditz GA, et al.
Editorials and Topics
currently unknowable magnitude of risk that has led to the greater reaction to beef? Europe has had wine lakes when it produced more wine than it was ready to sell. We now face the prospect of cattle volcanoes as British farmers incinerate their older cattle. One message is clear. Scientific evidence relating exposure to harm may be necessary but is far from sufficient for actions affecting the public health. It is not appropriate to dismiss this disjunction simply as politics. It is indeed political, but the lesson extends further. In the latest turn of events, and not for the first time, the public was ahead of the politicians and distrustful of them. The real lesson is that we need a better understanding of the management and communication of risk.'8 Ol Michael Marmot Intntional Centrefor Health and Society Department ofEpidemiogy and Public Health University College London
References 1. Royal College of Physicians. A Great and Growing Evil. London, England: Tavistock; 1987:1-121. 2. Griffith Edwards et al. Alcohol Policy and the Public Good. New York, NY: Oxford University Press; 1994. 3. World Health Organization. Health in Europe: The 1993/1994 Health For All Monitoring Report. Copenhagen, Denmark: World Health Organisation; 1994. 4. Bobak M, Marmot MG. East-West mortality divide and its potential explanations: proposed research agenda. BMJ 1996;312: 421-425. 5. Rose G. 77Te Strategy ofPreventive Medicine. Oxford, England: Oxford University Press; 1992. 6. Kreitman N. Alcohol consumption and the preventive paradox. Br J Addict. 1986;81: 353-364. 7. Sinclair J, Sillanaukee P. The preventive paradox: a critical examination. Addiction. 1993;88:591-594. 8. Joint Working Group of Royal College of Physicians, Royal College of Psychiatrists, Royal College of General Practitioners. Alcohol and the Heart in PerspectiveSensible Limits Reaffirmed. London, England: Royal Colleges; 1995:1-36.
9. Marmot MG. Population science, prejudice and policy on alcohol. Addiction. 1995;98:1441-1443. 10. Committee on Medical Aspects of Food Policy. NutritionalAspects of Cardiovascular Disease. London, England: HMSO:1994;1186. 11. Sensible Drinking: The Report of an InterDepartment Working Group. London, England: Department of Health; 1995. 12. Marmot MG. A not-so-sensible drinks policy. Lancet. 1995;346:1643-1644. 13. Abel EL, Agarwal DP, Algra A, et al. Health Issues Related to Alcohol Consumption. Washington, DC: ILSI Press; 1993. 14. van den Brandt PA, Goldbohm RA, Van't Veer P. Alcohol and breast cancer results from the Netherlands cohort study. Am J EpidemioL 1995;141:907-915. 15. Fuchs CS, Stampfer MJ, Colditz GA, et al. Alcohol consumption and mortality among women. NEnglJMed. 1995;332:1245-1250. 16. Will RG, Ironside JW, Zeidler M, et al. A new variant of Creutzfeldt-Jakob disease in the UK Lancet. 1996;347:921-925. 17. Burnt by the steak. Economist. 1996:27-28. Editorial. 18. Royal Society Study Group. Risk analysis perception management. London, England: The Royal Society; 1992:1-201.
Editorial: Community Solutions to Community ProblemsPreventing Adolescent Alcohol Use This issue's first report of findings from Perry and colleagues' Project Northland could not have arrived at a more crucial time. Latest national survey results indicate that rates of cigarette smoking and illicit drug use among eighth, tenth, and twelfth graders have risen again for the third consecutive year.1 This follows a long period of decline from the highs of the late 1970s. Alcohol use rates, though not increasing, have stopped decreasing and remain level. Although there is as yet no explanation for the newly changing pattern of youthful substance use, a parallel decrease in allocation of resources and national attention paid to this vital topic accompanies this increase. We may be afflicted with a national attention deficit disorder, the inability to focus on more than one public health issue at the same time, forcing us to abandon the problem faced yesterday for one ballyhooed today. For once, however, we welcome good news. Perry and colleagues' Project Northland in northeastern Minnesota signals how far we have come in designing and evaluating prevention programs.2 By conceptualizing behavior as the product of influences acting at multiple levels of July 1996, Vol. 86, No. 7
social organization-individual, family, peer, and community-and by engaging factors at each of these levels in the prevention of underage drinking, Project Northland has succeeded not only in changing drinking behavior among youth, but also in demonstrating that communityfocused and -based interventions can be implemented and evaluated while holding to standards of scientific research. That this project has been able to report success in decreasing the use of alcohol by sixth, seventh, and eighth graders largely by delaying the onset of use can be credited to two factors: the nature of the intervention and the nature of the evaluation. In both respects, Project Northland marks an advance in our approach to the public health problem of youth alcohol use. Earlier knowledge and informationbased attempts to decrease youth alcohol use were marked by what today might be considered naivete at best.3 If we could only inform youth of the dangers of alcohol use, or make them consider use in the broader context of their beliefs and values, or enable them to say no, we could get them to abstain from or diminish their substance use. Whatever made us think
that a few hours in the classroom with a teacher, no matter how dedicated or well meaning, could change a behavior that is reinforced countless times daily by peer interactions, media messages, advertising, and the behavior of important role models, including parents? Evaluations of these simple attempts at intervention yielded uniformly negative findings. Whatever change that occurred was limited to recall of information, and even that was at best short term.4'5 Project Northland uses a much more sophisticated and broader intervention program to make its impact on alcohol use by youth. The intervention involves a mixture of specially designed participatory, action-oriented curricula; parental participation; peer leadership; and community task force activities. Interventions were tailored to the developmental capacities of youth in six, seventh, and eighth grades. Communities were explicitly engaged in the process of program design and implementation. By developing a system-based approach to community involvement, project staff sought to increase Editor's Note. See related article by Perry et al. (p 956) in this issue. American Journal of Public Health 923
Editorials and Topics
the odds the program would affect the availability and supply of alcohol to youth, in addition to their knowledge, attitudes, and drinking practices. And they appear to have succeeded. We must not underestimate the significance of the departure in prevention programming from individually focused efforts to those that conceptualize behavior as a product of a broadly construed social environment. Individuallevel prevention places the onus of use solely on the individual and ignores the influences of family, peers and community. Although we must not overlook the importance of targeting potential users, such targeting alone in a one-sided approach can impede our understanding of how social structural factors affect adolescent drinking. Without engaging in this type of thinking and without applying its results, we risk investing in educating young people for brief periods of time only to return them to an environment that facilitates underage alcohol use and abuse. Ultimately, the best way to achieve and sustain prevention is to target individuals and their environments.6 Frustration with the lack of results from individual-level interventions appears to have led to a reconceptualization of adolescent drinking and our approaches to prevention. Because our view of the problem now incorporates an understanding of behavior as shaped by a network of individual, social, and environmental factors, current models are addressing both the demand- and the supplyside determinants of alcohol use.7-1' The successful integration of demand- and supply-side strategies is evident in Project
Northland. The community model used in Project Northland follows the precedent set by Farquhar and colleagues in the Stanford Five-City Project, which used an integrated community approach to reduce a population's risk for cardiovascular disease.12 In the area of youth substance use, Pentz and colleagues conducted an integrated school-based, family-based, and community-based preventive intervention, targeting adolescents in multiple communities throughout Kansas City and Indianapolis. Using a quasi-experimental evaluation design, they compared prevalence rates in intervention to delayedintervention sites and found a sustained impact on multiple substances over 2- and 3-year periods.13 Similarly, both the Robert Wood Johnson Foundation's Fighting Back Initiative and the US Department of Health 924 American Journal of Public Health
and Human Services' Center for Substance Abuse Prevention Community Partnership demonstration programs were organized to target the entire community as the locus and mechanism of change. Both these efforts explicitly mandated broad-based community involvement, direction by representative community task forces, careful needs assessments, and documentation of program content and implementation. Recognizing that individual behavior is most malleable when reinforced by messages from multiple community sectors, and that programs are most efficient when existing resources are applied to shared priorities, these programs brought together representatives and resources from across community sectors, including businesses, tenant and housing authorities, religious institutions, schools, parent associations, health institutions and hospitals, and local government agencies. Recognizing further that behavior change is sustained when the norms of a community change, these programs set out to change the environmental conditions and practices that shape and reinforce norms. Although these efforts are excellent examples of innovative social programming, they are only beginning to demonstrate, with quasi-experimental methods, their impact on youthful substance use. Project Northland provides evidence that this type of integrated community trial is worthwhile. By providing a vehicle for community participation in prevention programming, these trials develop consensus and create multiplier effects for disseminating public health messages. No doubt this is crucial to changing the supply-side characteristics of intervention sites. That said, findings from community trials are not necessarily free from ambiguity; nor do all multilevel trials appear to work equally well across different populations, settings, and target behaviors.-4 Findings of Project Northland may reflect a synergy of timing with topic. The project targeted primarily younger healthy adolescents and paid attention to life course and developmental factors in its intervention design. Indeed, results suggest that the program was less effective for early maturers or youth already engaged in multiple problem behaviors. Changing a social system requires a sustained investment of resources and effort. Linking systemic changes to behavioral changes and health outcomes challenges our patience, our methods, and the political and economic context in which research is conducted. Ultimately, change
of this order requires participation from all levels of community. Government and philanthropic initiatives require the support of professional and lay publics, a reality acknowledged by analysts at the forefront of broader social policy. Coleman15,16 and Putnam17 have theorized how "social capital"-pattems of mutual obligation, trust, and support characteristic of communities' networks of formal and informal relationships-may determine health and development. We hypothesize that it is the accrual and expenditure of social capital through sharing responsibility, resources, and roles to achieve reductions in youth substance use that will, in the end, achieve sustained reductions in the extent of use and the progression to abuse. It may well be that the more we conceive of the community as both locus and mechanism for change, and the more we explicitly target communities in addition to their individual residents, the greater will be the odds of our success. In this fashion, we return environment to its place with agent and host in the classical public health paradigm. How this thinking is translated into effective programming remains to be more fully articulated, applied and evaluated. We look forward to future reports by Perry and colleagues and others who take the difficult next step of determining how intervention programs affect communities and how community characteristics and resources can be used to modify individual behavior. O Henry Wechsler Elissa R. Weitzman Departfent of Health and Socioi Behavior Harvard School of Public Health Boston, Mass
References 1. Johnston LD, O'Malley PM, Bachman JG. National Survey Results on Drug Use from the Monitonng the Future Study, 1975-1994. Rockville, Md: National Institute on Drug Abuse; 1995. 2. Perry CL, Williams CL, Veblen-Mortenson S, et al. Project Northland: outcomes of a communitywide alcohol use prevention program during early adolescence. Am J
Public Health. 1996;86:956-965. 3. Klitzner MD. Report to Congress on the Nature and Effectiveness of Federal, State, and Local Drug Prevention/Education Programs. Part 2: An Assessment of the Research on School-Based Prevention Programs. Prepared for the US Department of Education, Office of Planning, Budget, Evaluation. 4. Moskowitz JM. The primary prevention of alcohol problems: a critical review of the research literature. JStudAlcohol. 1989;50:
5. Rundall TG, Bruvold WH. A metaanalysis of schcol-based smoking and alco-
July 1996, Vol. 86, No. 7
Editorials and Topics hol use prevention programs. Health Educ Q. 1988;15:317-334. 6. Gardner SE, Green PF, Marcus C, eds. Signs of Effectiveness II: PreventingAlcohol, Tobacco, and Other Drug Use: A Risk Factor/Resiliency-Based Approach. Rockville, Md: Center for Substance Abuse Prevention; 1994. 7. Hawkins JD, Catalano RF, Miller JY. Risk and protective factors for alcohol and other drug problems in adolescence and early adulthood: implications for substance abuse prevention. Psychol Bul. 1992;112:64-105. 8. Institute of Medicine. Broadening the Base of TreatmentforAlcohol Problems. Washington, DC: National Academy Press; 1990:23-41.
9. Walsh DC. The shifting boundaries of alcohol policy. Health Aff 1990(summer): 47-62. 10. Giesbrecht N, Krempulec L, West P. Community-based prevention research to reduce alcohol-related problems. Alcohol Health Res World. 1993;17:84-88. 11. Gruenewald PJ, Millar AB, Treno AJ. Alcohol availability and the ecology of drinking behavior. Alcohol Health Res World. 1993;17:3945. 12. Farquhar JW, Fortmann SP, Maccoby N, et al. The Stanford Five-City Project,Am J Epideniol. 1985;122:323-334. 13. Pentz MA, Dwyer JH, MacKinnon DP, et al. A multicommunity trial for primary
prevention of adolescent drug abuse: effects on drug use prevalence. JAMA4. 1989;261:3259-3266. Susser M. The tribulations of trialsintervention in communities. Am J Public Health. 1995;85:156-158. Coleman JS. Social capital in the creation of human capital. Am J SocioL 1988; 94(suppl):S95-S120. Coleman JS. 7he Foundations of Social 7Theory. Cambridge, Mass: Harvard University Press; 1990:300-321. Putnam RD. Bowling alone: America's declining social capital. J Democr. 1995;6: 65-79.
Editorial: Understanding and Treating Obesity Obesity can be defined as an excess of body fat. Although on the surface this definition seems simple enough, in reality it is very vague. What is an excess of body fat? More than the average? More than what is considered cosmetically pleasing? More than the amount necessary to feel healthy and physically fit? Or is obesity present when the amount of body fat reaches a level that increases a person's risk for serious illness and significantly shortens life expectancy? It is this last definition that carries with it the most serious public health consequences. Numerous studies have shown that when the level of body fat reaches 10 to 15% above the "ideal" amount, the risks for developing coronary artery disease, hypertension, diabetes, and perhaps certain cancers significantly
increase.' What is this ideal amount of body fat that can be used to calculate whether a person is obese? It is the amount of body fat that is associated with the greatest life expectancy. It will differ depending on the person's sex (women usually have a higher percentage of body fat) and height (the taller the person, the greater the absolute amount of body fat). Thus the ideal amount of body fat will vary with each person depending on sex and height, and tables have been derived that can be used to determine one's ideal amount of body fat. Unfortunately, these tables are of limited use either to the general public or to investigators involved in population studies of obesity. Determining total body fat directly is a tedious and complicated process, and although a number of relatively accurate ways are available, most are limited to a laboratory setting. However, one simple measurement that has July 1996, Vol. 86, No. 7
been validated against the laboratory measurements is skinfold thickness. This simple procedure is easily performed in the field and can be used to measure body fat directly. It should be used routinely along with measurements of height and weight. The most common way to measure body fat is to use body weight. In our society, heavier people at any given height are generally fatter. While this may not be true for certain athletes, it is usually true for the average man or woman. Thus tables exist (the one most commonly used is the current Metropolitan Life Insurance table) that list ideal weights for men and women at any given height. It is these tables that have been used to determine relative risks of obesity. About 30 years ago, a series of brilliant studies by Jules Hirsch and associates began to define obesity at a cellular level.2 These investigators showed that fat may be packaged in the body in two ways: either in small droplets in many fat cells or in large droplets in a few fat cells. Thus they distinguished two types of obesity: one in which there were too many fat cells, each containing a relatively normal amount of fat (hyperplastic obesity), and one in which there were a normal number of fat cells, each containing too much fat (hypertrophic obesity). Weight loss always reduced the amount of fat per cell, but never the number of fat cells. Thus the person who had hypertrophic obesity was returned to normal when he or she lost weight. There were a normal number of fat cells each containing a normal amount of fat. By contrast, the person who had hyperplastic obesity was made doubly abnormal with weight reduction. He or she had too many too small fat cells. In this type of person,
weight gain and the return of obesity were almost inevitable. These studies generated a great deal of excitement. What determined that an obese individual would have the hyperplastic or hypertrophic form of the disease? Was one form or the other or both associated with the increased health risks of obesity? Should treatment for one be different from treatment for the other? A number of studies designed to answer these questions were performed over the next 10 to 15 years.2 Childhood obesity was invariably hyperplastic, and it almost always progressed to adult hyperplastic obesity. Adult onset obesity, however, could be either hypertrophic or hyperplastic or, most commonly, both. While sustained weight reduction was more difficult in people with hyperplastic obesity, recidivism was very common in all three types. And as for health risks, the form of obesity didn't seem to matter. Finally, the treatment for obesity, no matter what the cellular form, remained the same: decreasing caloric intake and increasing caloric expenditure. Beginning in the late 1970s and continuing through the present, a new emphasis has been developing. From a physiologic standpoint, body fat is deposited in excess amounts when caloric intake exceeds caloric expenditure. Most of our calories are expended in a resting state. Do obese people have a lower resting metabolism than do lean people? Are they able to drive the everyday machinery of the body more efficiently than their leaner counterparts? Does this reduced
Note. See related articles by Jeffery and French (p 1005) and Kawachi et al. (p 999) in this issue.
American Journal of Public Health 925