Open Access

Crisis and chaos in behavioral nutrition and physical activity

International Journal of Behavioral Nutrition and Physical Activity20063:27

DOI: 10.1186/1479-5868-3-27

Received: 15 August 2006

Accepted: 14 September 2006

Published: 14 September 2006

Abstract

Resnicow & Vaughn challenged the field of behavioral nutrition and physical activity to conduct research in new ways. They challenged the predictiveness of our models, sensitivity to initial conditions, factors predisposing to change and measurement procedures. While the predictiveness of our models will reflect the sophistication of our thinking and research, and the sensitivity to initial conditions is subsumed under the sophistication of our models, research on conditions predisposing to change (e.g. epiphanies), more longitudinal designs, refined measurement procedures and testing of critical issues can only enhance the quality of our research. Improved research quality should lead to enhanced efficacy and effectiveness of our interventions, and thereby our making meaningful contributions to mitigating the chaos in our field and the crisis from the rising epidemic of obesity.

Background

Our field of behavioral nutrition and physical activity should be operating in crisis mode. The prevalence of obesity and overweight (an essentially nutrition and physical activity problem in its etiology and control) continues to increase at alarming rates in all age, demographic and gender groups in the US [1], Europe [2], and many other parts of the world [3, 4]. There is concern that this will reverse the recent advances in chronic disease control [5]. In the face of this encroaching epidemic, obesity treatment programs have tended to have weak effects mostly for short periods of time [6]; and review after review have shown that obesity prevention programs also tend not to work [79]. Furthermore, using the mediating variable model (see Fig. 1) as a structured framework, it is not clear we know what changes in diet or physical activity behavior have led to the current problems and thereby provide the best behavioral targets for change [link A in Fig. 1] [10, 11]; nor what mediating variables are most strongly related to these behaviors and thereby provide the best mechanisms for change [link B in Fig. 1] [12]; nor how best to manipulate the mediating variables to obtain behavior change and lower obesity [link C in Fig. 1] [13]. This is a frightful state of affairs. We should all be doing innovative theoretically guided, but high risk, research to quickly build a stronger knowledge base from which more effective interventions could be crafted. Yet, most of us appear to be acting in our usual way of doing things: "same old, same old,"
https://static-content.springer.com/image/art%3A10.1186%2F1479-5868-3-27/MediaObjects/12966_2006_Article_60_Fig1_HTML.jpg
Figure 1

Mediating variable model for obesity.

In this context, Resnicow & Vaughn [14] challenged our "same old" way of thinking about our field. They correctly specified the assumption of linearity in our predictive models, and proposed Chaos and Dynamic Systems Theories as alternative nonlinear models. They did not throw out all our theories per se, but challenged how we interrelated the variables, how we related them to behaviors, and offered some new variables predisposing to change. While Glass & McAtee [15] recently pointed out deficiencies in the social dimensions of our research, Resnicow & Vaughn targeted our thinking about behavior change. Some of the issues they raised are non-issues, but others deserve that we morph our basic methods to test the new ideas.

Discussion

Nonissues

Since we use statistical methods, all our models are by definition probabilistic, rather than deterministic (unless we could account for 100% of the variance, which will not happen in our lifetimes).

A key issue in their article was the predictiveness of our current models. They correctly identified the very limited predictiveness of the current models. The key issue, however, is the level of predictiveness that could possibly be achieved in predicting behavior. For example, our biological research colleagues are not satisfied unless their models account for 90% or more of the variance in their phenomena of interest. We are well below that [16]. Resnicow & Vaughn have not taken into account the emerging research on environmental influences, e.g. home availability [17], neighborhood characteristics [18, 19]; biological influences, e.g. genes [20], sensitivity to tastes [21], the hormone rages of adolescent development [22, 23]; emotional influences [23, 24]; nor the likely interrelationships and interactions among these variables and our more usual psychosocial and behavioral predictors [25, 26]. The higher the predictiveness of our models, the more we can engage in our logical approach to designing interventions based on these models. The larger number of and more diverse variables incorporated into these models, the more complex our interventions will need to be to address components of the model. And the interventions will need to both segment the population for differing types of interventions to different gender, age, ethnic, socioeconomic, and/or neighborhood groups, and tailor the intervention to individual characteristics within these groups [27]. At this time, we need to build and test the more comprehensive models. This is a daunting, but exciting, challenge.

Resnicow & Vaughn proposed the principal of sensitivity to initial conditions, as if this were a new idea. All of our models of longitudinal relationships (as equations) have built into them sensitivity to initial conditions, i.e. the initial values of the variables. How diverse the outcomes depends on the nature of the relationships. As our models become more comprehensive and complex, fairly similar initial conditions could lead to quite divergent outcomes. In part this is a function of the sophistication of our knowledge base. We need to build more sophisticated predictive models.

The idea of a tipping point or when it might occur, is not well defined [28]. In some ways it reifies a change, as if there is something intrinsic to or magical about the change process. If a tipping point is nothing more than a critical point on a variable beyond which change occurs, it is not clear the concept adds much, but identifying those points would be helpful.

Issues deserving intensive research

Investigators could take away from the Resnicow & Vaughn message that change is random and cannot be predicted, and thereby cannot be understood by our usual research methods on behavior or its change. This would be very unfortunate. Resnicow & Vaughn will have made a major contribution, only if it leads to innovative research and new insights. Even in the vast complexities of molecular science, investigators are hammering away at delineating linear and nonlinear patterns to better understand the biology. Chances are we can do the same in behavior research.

Resnicow & Vaughn proposed that change does not occur in a linear "persuasion slowly overcoming resistance" manner, but rather in what they characterized as "quantum leaps," i.e. an epiphany or "aha!" event occurs from which the person decides to change. This is an interesting idea and should be testable. Innovative methods will be needed to identify people soon after the aha! experience to learn more about it. Perhaps interviewing new recruits to Weight Watchers™ or to fitness centers would accomplish this? A related issue would be what could we do to encourage aha! experiences? Are they a response to an overload of information (probably not, since we have done a lot of this already)? To repeated thinking about the issues (we could program prompts to thinking)? To setting off some emotional experience related to the behavior (we might be able to tailor messages to issues people found emotionally charged)? Resnicow & Vaughn invoked the concept of "cues" from the Health Belief Model. There has been some research on cues [2931], but this has not as yet led to substantial insights. Relating cues to aha! experiences could be an important avenue for research. Developing valid and reliable retrospective methods to identify and recall aha! experiences would be necessary to make much progress.

Resnicow & Vaughn correctly pointed out the cross sectional nature of most of our research. Dynamic Systems modeling proposes that dynamic research be done, and this would be focused on change over time which requires longitudinal designs [32]. The importance of longitudinal designs was emphasized when Nigg [33] found that physical activity predicted ensuing self efficacy, but not the other way around. If self efficacy is really caused by physical activity, but doesn't cause physical activity, it doesn't make sense to try to increase self efficacy in interventions. While it is challenging to recruit and maintain longitudinal cohorts, such cohorts are required to address issues of direction of causality and thereby which variables should be targets for change in intervention programs. While ten year cohorts may not be necessary, perhaps 3 mo or 6 mo cohorts would provide tests for the changes we need. Longitudinal dynamic systems research has been initiated in other fields [34, 35], which should provide a guide for our further development.

Whether behavior change can only be understood in retrospect instead of prospectively is an empirical issue. In part this is a function of how much variance our models will ultimately predict. Perhaps a few retrospective analyses will be necessary, perhaps using qualitative methods, to map out the processes occurring? But predictive science should be where we are headed, since predictive relationships clearly demonstrate what we know.

Resnicow & Vaughn correctly identified our current approaches to measurement as providing severe limitations to how we could understand our phenomena of interest. There have been limits on the extent to which existing measurement methods (e.g. classical test theory) have been used and reported [36], and limits on the predictiveness of existing measurement models [37]. One innovation in measurement theory that has recently drawn attention is Item Response Theory (IRT) [38]. IRT fits latent variables to items (and respondents) which identifies portions of the underlying variable being poorly measured [39], and assesses reliability across the range of the underlying variable [39, 40]. Having items measuring specific locations on the underlying variable permits an assessment of whether the measures work differently after participation in an experiment [41]; differ by ethnic, gender or other groupings; and permit more efficient multidimensional modeling of the variable [42]. Use of IRT offers great promise for better understanding and minimizing the problems due to measurement of our constructs, and deserves much wider use.

An issue Resnicow addressed in his oral presentation in Boston (but not in his paper) was the falsifiability of a theory, and whether our current cognitive models are really theories. He correctly stated that in our current approach to research, no theories have been discarded (which would be considered a sign of progress and development in a field). Our best current research fits multivariate models to sets of variables [43], and determines which variables were significantly related to other variables in the model. This is useful for assessing the predictiveness of particular variables in certain situations, but does not necessarily address the usefulness of the larger theory. To move our field forward we need more attention to theoretical issues in our research, tests of clearer more specific predictions from theory applied to particular issues [44, 45], and delineation of "critical issues" where two theories would make different predictions or model fitting research would need to test the fit of competing models [24], where the alternative models were predicated on different theories. More highly controlled experimental research on critical issues will also be necessary. Accumulation of findings across "critical" studies would enable the field to find more comprehensive and more predictive theoretical frameworks, and capitalize upon them in more likely to be effective interventions. There has been a distaste for theory in our field [46], and some have proposed continuing conducting intervention research until randomly hitting on intervention procedures that work [47]. Alternatively, I believe highly predictive theory should guide the design of effective interventions. In a complicated set of many possible variables and relationships, a random search may never result in finding effective change techniques, and even if it did, we wouldn't have the conceptual framework to understand why it happened in order to exploit it.

Conclusion

Resnicow & Vaughn challenged the field of behavioral nutrition and physical activity to conduct research in new ways. While the predictiveness of our models will reflect the sophistication of our thinking and research, and the sensitivity to initial conditions is subsumed under the sophistication of our models, research on conditions predisposing to change (e.g. epiphanies), more longitudinal designs, refined measurement procedures and testing of critical issues can only enhance the quality of our research. Improved research quality should lead to enhanced efficacy and effectiveness of our interventions, and thereby our making meaningful contributions to mitigating the chaos in our field and the crisis from the rising epidemic of obesity.

Abbreviations

US: 

United States

IRT: 

Item Response Theory

Declarations

Acknowledgements

This work is in part a publication of the United States Department of Agriculture (USDA/ARS) Children's Nutrition Research Center, Department of Pediatrics, Baylor College of Medicine, Houston, Texas, and was funded in part with federal funds from the USDA/ARS under Cooperative Agreement No. 58-6250-6001. The contents of this publication do not necessarily reflect the views or policies of the USDA, nor does mention of trade names, commercial products, or organizations imply endorsement from the U.S. government.

Authors’ Affiliations

(1)
Children's Nutrition Research Center, Department of Pediatrics, Baylor College of Medicine

References

  1. Ogden C, Carroll M, Curtin L, McDowell M, Tabak C, Flegal K: Prevalence of overweight in the United States, 1999–2004. JAMA. 2006, 295 (13): 1549-55. 10.1001/jama.295.13.1549.View ArticleGoogle Scholar
  2. Bostrom G, Eliasson M: Chapter 5.3: Major public health problems – overweight and obesity. Scan J Public Health Suppl. 2006, Jun (67): 69-77. 10.1080/14034950600676941.View ArticleGoogle Scholar
  3. Cheng TO: Obesity is a global challenge. Am J Med. 2006, 119 (6): e11-10.1016/j.amjmed.2006.04.006.View ArticleGoogle Scholar
  4. Malecka-Tendera E, Mazur A: Childhood obesity: a pandemic of the twenty-first century. Int J Obes (Lond). 2006, S1-3. 10.1038/sj.ijo.0803367. 30 Suppl 2
  5. Jones AP, Homer JB, Murphy DL, Essien JD, Milstein B, Seville DA: Understanding diabetes population dynamics through simulation modeling and experimentation. Am J Public Health. 2006, 96 (3): 488-94. 10.2105/AJPH.2005.063529.View ArticleGoogle Scholar
  6. Avenell A, Broom J, Brown TJ, Poobalan A, Aucott L, Stearns SC, Smith WC, Jung RT, Campbell MK, Grant AM: Systematic review of the long-term effects and economic consequences of treatments for obesity and implications for health improvement. Health Technol Assess. 2004, 8 (21): iii-iv. 1–182View ArticleGoogle Scholar
  7. Flynn MA, McNeil DA, Maloff B, Mutasingwa D, Wu M, Ford C, Tough SC: Reducing obesity and related chronic disease risk in children and youth: a synthesis of evidence with 'best practice' recommendations. Obes Rev. 2006, 7-66. 10.1111/j.1467-789X.2006.00242.x. 7 Suppl 1
  8. Baranowski T, Cullen KW, Nicklas T, Thompson D, Baranowski J: School-based obesity prevention: A blueprint for taming the epidemic. Am J Health Behav. 2002, 26 (6): 486-93.View ArticleGoogle Scholar
  9. Flodmark CE, Marcus C, Britton M: Interventions to prevent obesity in children and adolescents: a systematic literature review. Int J Obes (Lond). 2006, 30 (4): 579-89. 10.1038/sj.ijo.0803290.View ArticleGoogle Scholar
  10. Bachman CM, Baranowski T, Nicklas TA: Is there an association between sweetened beverages and adiposity?. Nutr Rev. 2006, 64 (4): 153-74. 10.1301/nr.2006.apr.153-174.View ArticleGoogle Scholar
  11. Nicklas T, Yang SJ, Baranowski T, Zakeri I, Berenson GS: Eating patterns and obesity in children: The Bogalusa Heart Study. Am J Prev Med. 2003, 25 (1): 9-16. 10.1016/S0749-3797(03)00098-9.View ArticleGoogle Scholar
  12. Baranowski T, Anderson C, Carmack C: Mediating variable framework in physical activity interventions. How are we doing? How might we do better?. Am J Prev Med. 1998, 15 (4): 266-97. 10.1016/S0749-3797(98)00080-4.View ArticleGoogle Scholar
  13. Baranowski T, Lin LS, Wetter DW, Resnicow K, Hearn MD: Theory as mediating variables: why aren't community interventions working as desired?. Annals of Epidemiology. 1997, 7: S89-S95.View ArticleGoogle Scholar
  14. Resnicow K, Vaughn : A chaotic view of behavior change: a quantum leap for health promotion. IJBNPA.
  15. Glass TA, McAtee MJ: Behavioral science at the crossroads in public health: extending horizons, envisioning the future. Soc Sci Med. 2006, 62 (7): 1650-71. 10.1016/j.socscimed.2005.08.044.View ArticleGoogle Scholar
  16. Baranowski T, Cullen KW, Baranowski J: Psychosocial correlates of dietary intake: Advancing intervention. Annual Review of Public Health. 1999, 19: 17-40.Google Scholar
  17. Jago R, Baranowski T: Home food availability: A microenvironmental mediating variable?. Public Health Nutrition. 2006,Google Scholar
  18. van der Horst K, Oenema A, Ferreira I, Wendel-Vos W, Giskes K, van Lenthe F, Brug J: A systematic review of environmental correlates of obesity-related dietary behaviors in youth. Health Educ Res. 2006Google Scholar
  19. Jago R, Baranowski T, Baranowski JC: Observed, GIS, and self-reported environmental features and adolescent physical activity. Am J Health Promot. 2006, 20 (6): 422-8.View ArticleGoogle Scholar
  20. Rankinen T, Bouchard C: Genetics of food intake and eating behavior phenotypes in humans. Annu Rev Nutr. 2006, 26: 413-34. 10.1146/annurev.nutr.26.061505.111218.View ArticleGoogle Scholar
  21. Goldstein GL, Daun H, Tepper BJ: Adiposity in middle-aged women is associated with genetic taste blindness to 6-n-propylthiouracil. Obes Res. 2005, 13 (6): 1017-23.View ArticleGoogle Scholar
  22. Himes JH, Obarzanek E, Baranowski T, Wilson DM, Rochon J, McClanahan BS: Early sexual maturation, body composition, and obesity in African American girls. Obes Res. 2004, 64S-72S. 12 Suppl
  23. Richardson LP, Garrison MM, Drangsholt M, Mancl L, LeResche L: Associations between depressive symptoms and obesity during puberty. Gen Hosp Psychiatry. 2006, 28 (4): 313-20. 10.1016/j.genhosppsych.2006.03.007.View ArticleGoogle Scholar
  24. Taylor SD, Bagozzi RP, Gaither CA: Decision making and effort in the self-regulation of hypertension: testing two competing theories. Br J Health Psychol. 2005, 10 (Pt 4): 505-30. 10.1348/135910704X22376.View ArticleGoogle Scholar
  25. Kremers SP, de Bruijn GJ, Visscher TL, van Mechelen W, deVries NK, Brug J: Environmental influences on energy balance-related behaviors: A dual-process view. Int J Behav Nutr Phys Act. 2006, 3: 9-10.1186/1479-5868-3-9.View ArticleGoogle Scholar
  26. Cullen KW, Baranowski T, Owens E, Marsh T, Rittenberry L, de Moor C: Availability, accessibility and preferences for fruit, 100% juice and vegetables influence children's dietary behavior. Health Education & Behavior. 2003, 30: 615-626. 10.1177/1090198103257254.View ArticleGoogle Scholar
  27. Kroeze W, Werkman A, Brug J: A systematic review of randomized trials on the effectiveness of computer-tailored education on physical activity and dietary behaviors. Ann Behav Med. 2006, 31 (3): 205-23. 10.1207/s15324796abm3103_2.View ArticleGoogle Scholar
  28. Gladwell M: The Tipping Point – How Little Things Can Make a Big Difference. 2002, New York, NY: Back Bay BooksGoogle Scholar
  29. Jones T, Fowler MC, Hubbard D: Refining a tool to measure cues to action in encouraging health-promoting behavior-The CHAQ. American Journal of Health Promotion. 2000, 14: 170-173.View ArticleGoogle Scholar
  30. Strychar IM, Champagne F, Ghadirian P, Bonin A, Jenicek M, Lasater TM: Impact of receiving blood cholesterol test results on dietary change. American Journal of Preventive Medicine. 1998, 14: 103-110. 10.1016/S0749-3797(97)00022-6.View ArticleGoogle Scholar
  31. Kip KE, McCreath HE, Roseman JM, Hulley SB, Schreiner PJ: Absence of risk factor change in young adults after family heart attack or stroke. American Journal of Preventive Medicine. 2002, 22: 258-266. 10.1016/S0749-3797(02)00416-6.View ArticleGoogle Scholar
  32. Howe ML, Lewis MD: The importance of dynamic systems approaches for understanding development. Developmental Review. 2005, 25 (3–4): 247-251. 10.1016/j.dr.2005.09.002.View ArticleGoogle Scholar
  33. Nigg CR: Explaining adolescent exercise behavior change; A longitudinal application of the transtheoretical model. Annals of Behavioral Medicine. 2001, 23: 11-20. 10.1207/S15324796ABM2301_3.View ArticleGoogle Scholar
  34. van Geert P, Steenbeek H: Explaining after by before: Basic aspects of a dynamic systems approach to the study of development. Developmental Review. 2005, 25 (3–4): 408-442. 10.1016/j.dr.2005.10.003.View ArticleGoogle Scholar
  35. Smith LB: Cognition as a dynamic system: Principles from embodiment. Developmental Review. 2005, 25 (3–4): 278-298. 10.1016/j.dr.2005.11.001.View ArticleGoogle Scholar
  36. Baranowski T, Klesges LM, Cullen KW, Himes J: Measurement of outcomes, mediators and moderators in behavioral obesity prevention. Preventive Medicine. 2004, 38: S1-S13. 10.1016/j.ypmed.2003.12.029.View ArticleGoogle Scholar
  37. Wilson M, Allen D, Li H: Comparing item response modeling with the Classical Test Theory approach in measurement in the behavioral sciences. Health Educ Res. 2007,Google Scholar
  38. Wilson M: Constructing Measures: An Item Response Modeling Approach. 2005, Mahwah, N.J.: ErlbaumGoogle Scholar
  39. Watson K, Thompson D, Baranowski T: Item response theory evaluation of the children's fruit and vegetable self efficacy scale. Health Education Research. 2007,Google Scholar
  40. Wilson M, Allen DD, Li JC: Improving measurement in behavioral sciences using item response models: comparison with the classical test theory approach. Health Educ Res. 2006:Jul 31 [Epub]
  41. Baranowski T, Allen D, Masse L, Wilson M: Does participation in an intervention affect responses on self report questionnaires?. Health Educ Res. 2006,Google Scholar
  42. Watson K, Baranowski T, Thompson D, Jago R, Baranowski J, Klesges L: Innovative Application of a Multidimensional Item Response Model In Assessing the Influence of Social Desirability on the Pseudo-Relationship between Self-Efficacy and Behavior. Health Educ Res.
  43. Motl RW, Dishman RK, Saunders RP, Dowda M, Pate RR: Perceptions of Physical and Social Environment Variables and Self-Efficacy as Correlates of Self-Reported Physical Activity Among Adolescent Girls. J Pediatr Psychol. 2006:May 17 [Epub]
  44. Resnicow K, Baranowski T: Are pre-contemplators less likely to change their dietary behavior? A prospective analysis. Health Education Research. 2003, 18 (6): 693-705. 10.1093/her/cyf052.View ArticleGoogle Scholar
  45. Weinstein ND, Sutton S, Rothman AJ: Stage theories of health behavior: Conceptual and methodological issues. Health Psychology. 1998, 17: 290-299. 10.1037/0278-6133.17.3.290.View ArticleGoogle Scholar
  46. Jeffery RW: How can Health Behavior Theory be made more useful for intervention research?. Int J Behav Nutr Phys Act. 2004, 1 (1): 10-10.1186/1479-5868-1-10.View ArticleGoogle Scholar
  47. Robinson TN, Sirard JR: Preventing childhood obesity: a solution-oriented research paradigm. Am J Prev Med. 2005, 28 (2 Suppl 2): 194-201. 10.1016/j.amepre.2004.10.030.View ArticleGoogle Scholar

Copyright

© Baranowski; licensee BioMed Central Ltd. 2006

This article is published under license to BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

Advertisement