Several speakers at various times during the workshop agreed that most youth who are obese will probably remain so for their rest of their lives. The question then arises of whether it is possible for overweight or obese children, adolescents, and adults to have healthy metabolic profiles. In the final session of the first day, moderated by Cedric Bryant, speakers addressed that question.
First, Andrea Kriska discussed results of the landmark National Institutes of Health (NIH)-funded Diabetes Prevention Program (DPP) and its 10-year follow-up in high-risk adults. Among other findings, the researchers noted a decreased incidence of diabetes in the lifestyle intervention arm, with physical activity being a critical component of the intervention. This decrease in the incidence of diabetes was similar in both men and women and appeared to hold across all age and ethnic/racial groups. Kriska remarked that researchers are still analyzing data from an expanded community-wide lifestyle intervention program translated from the successful DPP to suit the community setting. Thus far they have observed increased physical activity, reduced weight, and significant improvements in diabetes and cardiovascular risk factors. For Kriska, the evidence indicates that increasing physical activity levels in adults who are overweight and showing signs of prediabetes and/or metabolic syndrome is doable, and that improvement in physical activity levels may have a significant impact on diabetes prevention and other health outcomes. She echoed calls made
earlier in the workshop to increase understanding of the impact of all intensities of physical activity, not just that of moderate to vigorous intensity.
Shifting the focus back to children, Gabriel Shaibi examined evidence demonstrating that, independent of any effect on weight, physical activity can both improve health status and reduce risk factors for disease in children who are overweight or obese. He cautioned that researchers need to think carefully about the outcomes they are measuring to determine the success of interventions in youth who are overweight. He suggested that, instead of weight, cardiometabolic risk factors may be more relevant indicators of health improvement for children and adolescents. Among other findings, researchers have reported that exercise training increases cardiac function in children who are obese, essentially normalizing mycocardial dysfunction, even without weight loss. Shaibi also stressed thinking about the kind of exercise, or physical activity, in which youth like to engage and suggested that incentives to get them moving may be worth considering.
If obesity is not part of the discussion on the effects of physical activity on health outcomes, John Jakicic cautioned, “we will miss something.” He discussed evidence demonstrating that, indeed, weight matters. For example, data from the cross-sectional Look AHEAD trial show that while physical activity predicts hypertension, weight has an independent effect as well. Studies on adults with knee osteoarthritis illustrate this same point—that is, exercise improves outcomes, but exercise combined with weight loss improves those outcomes even more. “Yes, we can get the effect [with exercise alone],” he said. “But how do we maximize the effect?”
In the question-and-answer period following Jakicic’s talk, a key topic of discussion was identifying weight loss interventions that work in the long run. The challenge, in Jakicic’s opinion, is not to determine which dose of physical activity to prescribe, but how to get more people to adopt what researchers know works. “I think that’s where the action really should be,” he said.
The Diabetes Prevention Program
The strength of the DPP, Andrea Kriska began, was in the diversity of its participants, in terms of age (25 years and up), ethnicity and race, and geography (27 sites across the United States). A total of 3,234 individuals,
1 This section summarizes information and opinions presented by Andrea M. Kriska, Ph.D., M.S., University of Pittsburgh, Pennsylvania.
all with high weight and prediabetes, were randomly assigned to one of three groups: lifestyle intervention, drug intervention (i.e., metformin), or placebo.
Kriska focused her presentation on differences between the lifestyle and placebo groups. The minimum goals of the lifestyle intervention were for participants to lose 7 percent of body weight and engage in the nationally recommended 150 minutes per week of moderate-intensity physical activity, similar to a brisk walk. Over the course of the 3-year study, participants randomized to the lifestyle intervention reported significantly greater physical activity levels relative to participants in the other two randomized arms, as determined by the past-year version of the Modifiable Activity Questionnaire (MAQ). Lifestyle intervention participants also had significantly greater weight loss than participants in the other two groups by the end of the study (Diabetes Prevention Program Research Group, 2002). Interestingly, at the end of the first year, secondary analyses showed a significant decrease in weight—about 7 to 8 percent—for all racial and ethnic groups in the lifestyle intervention group with the exception of black women, who lost significantly less than that (4.5 percent) (West et al., 2008).
In terms of diabetes prevention, the lifestyle intervention worked, Kriska said, as those participants showed a 58 percent greater decrease in diabetes incidence relative to the placebo group. The intervention worked across all subgroups, including age, sex, baseline body mass index (BMI), and ethnicity and race (see Figure 4-1). The decrease in diabetes development, which was demonstrated across all of these subgroups, was a very important finding, in Kriska’s opinion, with respect to using the DPP as a model for lifestyle intervention in diverse community settings. Additionally, the lifestyle intervention group showed a 41 percent decrease in the incidence of metabolic syndrome relative to the placebo group.
Early in the study, the DPP investigators, including Kriska, examined the separate impacts of weight and physical activity on the risk of developing diabetes among lifestyle intervention participants only (Hamman et al., 2006). Although the study was not designed to separate out the effect of each lifestyle goal individually, these secondary results suggested that change in weight from baseline significantly predicted reduced diabetes incidence, but change in reported physical activity levels from baseline did not (although activity itself was shown to predict weight loss). Yet if physical activity was handled categorically instead, with regard to whether participants met the physical activity goal, those who did achieve that goal had a 46 percent reduced incidence of diabetes.
After the DPP results had been presented, the investigators were funded to conduct a 10-year follow-up (the Diabetes Prevention Program Outcome Study or DPPOS), with all participants being offered the lifestyle intervention. Over the course of the DPPOS, participants from the original lifestyle
FIGURE 4-1 Diabetes incidence rates by ethnicity at the end of the 3-year Diabetes Prevention Program.
SOURCE: Presented by Andrea Kriska on April 14, 2015 (Diabetes Prevention Program Research Group, 2002).
intervention group continued to maintain an advantage over those from the original placebo group, Kriska said, in the form of a 34 percent reduced incidence of diabetes (Diabetes Prevention Program Research Group et al., 2009).
In the last 2 years of the DPPOS, an ancillary accelerometer study was conducted across nearly all of the DPPOS sites to advance the investigation of the importance of physical activity in preventing diabetes. Although still being analyzed, objective accelerometry data, coupled with data from the past-year MAQ (administered every year since baseline), suggest that physical activity may be an independent predictor of diabetes incidence, even after controlling for weight. In summary, results from the DPP and its follow-up DPPOS indicate that diabetes can be prevented with lifestyle intervention and that physical activity is a critical component of such intervention.
Community Translation of the DPP Lifestyle Intervention
DPP investigators in Pittsburgh modified the DPP lifestyle intervention to develop a more community-friendly program called Group Lifestyle Balance, a 1-year program that entailed 16 sessions during the first half and
monthly sessions thereafter. The researchers were funded by the National Institute of Diabetes and Digestive and Kidney Diseases (NIDDK) to test the program in three very different types of community settings: three community senior centers of varying socioeconomic status, a worksite, and a military base. Participants had to be 18 years of age or older, have a BMI of 24 or greater, and have prediabetes or metabolic syndrome but no reported history of diabetes. The mean age of participants was 58 years; 63 percent were female.
According to 6-month attendance data, median attendance across all sites was 14 of the first 16 sessions, with 75 percent of participants attending 12 or more sessions. The researchers are still analyzing weight loss data from the 18-month follow-up, but the combined data from all sites at both 6 and 12 months demonstrated a 5-6 percent weight loss, representing approximately 12-13 pounds. Of equal importance is that the combined physical activity data from all sites showed a highly significant increase in physical activity from baseline to both 6 and 12 months, even after adjusting for changes in activity due to season (Eaglehouse et al., 2015).
This evidence of significant improvement in physical activity levels in these community translation intervention efforts is an important finding, according to Kriska, considering that nearly half of existing translation studies provide no information on change in physical activity (Eaglehouse et al., 2015). Moreover, none of those studies adjust their results for seasonal variation, which, she noted, can be a considerable factor in intervention studies that occur over several months in locations that experience substantial changes in weather/season. Kriska and colleagues had already demonstrated the impact of season/weather on physical activity levels in a previous clinical trial in older women residing in a region with changing seasons (Newman et al., 2009). Reported baseline past-week physical activity levels of these women (as determined by the MAQ) varied dramatically throughout the 18-month recruitment period, with recruits being more active in summer and less active in winter even before the intervention started. Kriska and her team saw the same influence of season on activity levels prior to and during intervention in the community study. In addition to its impact on weight loss and physical activity, the lifestyle intervention positively improved several diabetes and cardiovascular risk factors—fasting glucose, hemoglobin A1c (%) (not measured at the military site), systolic blood pressure, diastolic blood pressure, LDL (low-density lipoprotein) cholesterol, HDL (high-density lipoprotein) cholesterol, and triglycerides—in the 286 participants from all of the sites combined (12-month data not yet available for military base participants). All factors except HDL cholesterol showed positive changes at 6 months, and all but LDL cholesterol showed positive changes at 12 months. These analyses were presented for all participants combined, as is often done in intervention efforts.
Kriska also ran the above analyses separately for participants with above-normal values at baseline for each risk factor—that is, those who needed improvement in the specific outcome variable of interest. When the data were analyzed using this “high-risk approach,” Kriska said, in every case and at every site, not only did the effect size improve for that variable, but it became statistically significant if not significant before (Kriska et al., 2014). This approach is most striking when the results for only one site are considered, representing a sample size more in line with current community translation efforts. For example, an analysis of 6-month changes among all worksite participants showed statistically significant positive changes in hemoglobin A1c (%), systolic blood pressure, and triglycerides. When only high-risk participants were included in the analysis, fasting glucose, LDL cholesterol, and HDL cholesterol showed statistically significant positive changes as well. This information is critical to understanding the effectiveness of lifestyle intervention translation efforts, considering that such studies typically report their results for all participants combined, likely underestimating the impact of the intervention for those at highest risk.
Summary of Lessons Learned
Kriska identified several key lessons learned from the DPP and the DPP-based community studies. First, results from both types of study (efficacy and effectiveness) suggest that physical activity levels can be increased in high-risk adults. The DPP provided further evidence that improvement in physical activity can have a significant impact on health outcomes, including the prevention of diabetes. Second, when evaluating the effects of community translation efforts, Kriska suggested considering use of a high-risk approach to understand the full extent of the impact of lifestyle intervention on participants’ health. Additionally, she suggested paying attention to the potential influence of seasonal variation on physical activity levels.
Finally, Kriska commented on newer research focusing on low intensities of physical activity and, conversely, sedentary behavior such as sitting, and their impact on health outcomes. As part of the DPP, she and colleagues assessed television (TV) watching at baseline (as a surrogate for sitting time) and then again at the 3-year follow-up. They found that, among lifestyle intervention participants, TV viewing declined more than it did in either the metformin or placebo group, despite the fact that this was not a primary goal of the intervention (Rockette-Wagner et al., 2015). Additionally, each hour per day of TV watching was associated with a 3.4 percent increased risk of developing diabetes over the follow-up period in this cohort of individuals with prediabetes at baseline (although the risk reduction became nonsignificant and was attenuated to 2.1 percent after controlling for weight).
Gaining a better understanding of the health impact of decreasing time spent sitting is the focus of Kriska’s next NIH-funded study. She is currently examining whether the significant beneficial changes in weight and in diabetes and cardiovascular risk factors that she and her colleagues observed in the NIH-funded community translation study will be seen if the goal is not to increase levels of moderate-intensity physical activity but to sit less.
Building the Conceptual Case
Gabriel Shaibi presented a three-part conceptual case for physical activity and exercise in youth who are already overweight or obese. First, long-term successful weight loss is possible but challenging, he said. Defining successful long-term weight loss as intentionally losing at least 10 percent of initial body weight and keeping it off for at least 1 year, Wing and Hill (2001) suggest that at least 20 percent of overweight or obese individuals who attempt to lose weight can achieve long-term success. Physical activity is a primary driver of that success, Shaibi said. The other 80 percent who are unable to lose weight or maintain their weight loss over the course of 1 year enter what Shaibi called a “vicious cycle.” Newer data have shown that weight cycling has a deleterious effect on health outcomes, including end organ damage.3 In Shaibi’s opinion, losing weight is the right thing to do only if the lost weight will not be regained. “If you are going to gain it back,” he said, “you are probably better off not losing weight in the first place.”
Second, most youth who are obese will remain so for the rest of their lives. Adolescents who are obese by the age of 14 to 16 tend to remain obese into adulthood. Some data indicate that 8 to 10 percent of children between 0 and 2 years of age are already obese, Shaibi said, which to him “screams that there are strong biological drivers in this population.” Pediatric obesity is highly heritable, he noted, and there may be a genetic predisposition to obesity in childhood, especially early in life (Bouchard, 2009).
Finally, physical activity has been shown to be protective against morbidity and mortality, independent of obesity (Ekelund et al., 2015).
2 This section summarizes information and opinions presented by Gabriel Q. Shaibi, Ph.D., Arizona State University, Tempe.
3 End organ damage usually refers to damage occurring in major organs fed by the circulatory system (such as the heart, kidneys, brain, and eyes), which can sustain damage as a result of uncontrolled hypertension, hypotension, or hypovolemia.
Experimental Data Supporting the Role of Exercise in Health Promotion and Disease Prevention Among Youth Who Are Obese
In a review of four exercise-only interventions in youth who were already obese, Watts and colleagues (2005) found that on average, the youth actually gained weight after 2-5 months of exercise. Averaged over the four studies, however, observed weight gain was accompanied by an increase in fat-free mass and a slight but significant decrease in fat mass. In a meta-analysis of more than 300 studies, Kelley and Kelley (2013) identified similar reductions in percent body fat in youth who were already obese and who participated in an exercise program. They found no significant changes in any other measure of adiposity (i.e., BMI-related measures, body weight, and central adiposity).
Shaibi suggested that because youth are still growing, weight itself is probably not a good marker of health improvement following exercise in youth who are obese. Rather, changes in body composition may be better outcome measures.
Another alternative to weight for evaluating the impact of exercise in youth may be cardiometabolic health (i.e., combined cardiovascular disease and type 2 diabetes risk factors). At the population level, according to Shaibi, 30 percent of children or adolescents who are obese have the constellation of risk factors that puts them at risk for cardiometabolic disease later in life and a 25-fold increased odds ratio of having a cardiovascular event in adulthood. The Centers for Disease Control and Prevention (CDC) published data suggesting that as many as 50 percent of youth born in 2000 will develop type 2 diabetes in their lifetime. “This is a compelling argument for us to think about not necessarily weight loss in this population,” Shaibi said, “but targeted health promotion and disease prevention programs.”
Data from recent meta-analyses on the effects of exercise in children who are already overweight or obese indicate only small effects on blood pressure (Garcia-Hermoso et al., 2013), lipid profile (Escalante et al., 2012), and insulin resistance (Fedewa et al., 2014). However, Shaibi pointed out, while those results do not suggest that exercise has a robust or a very large effect on cardiometabolic outcomes, the outcomes analyzed in those three studies were based on clinical outcomes in adult populations. He questioned how relevant those outcomes are in younger populations. He suggested that exercise in youth can protect the cardiovascular system in ways that cannot be detected by changes in traditional clinical outcome measures. He referred to Joyner and Green’s (2009) notion of a risk factor gap, where exercise appears to be far more protective in terms of morbidity and mortality than it should be based on changes in traditional risk factors alone (e.g., blood pressure, lipid profile, insulin resistance). Relating that
notion to youth, he asked, what does a lowering of blood pressure by 5 or 10 percent mean in terms of long-term health outcomes 40, 50, 60, or 70 years down the road? What does a one- or two-point change in LDL cholesterol mean? The long subclinical period between elevated risk (in youth) and eventual disease outcome (in adulthood) suggests that the risk factor gap may be particularly wide in pediatric populations, in Shaibi’s opinion.
As an example of a study on the effect of exercise on a nontraditional cardiometabolic health outcome measure in youth, Shaibi cited Ingul and colleagues’ (2010) study on the effects of 13 weeks of twice-weekly high-intensity interval training on cardiac function in adolescents who were obese. The researchers examined stroke volume (an indication of how much blood the heart is able to pump in relation to body size) and global strain. This was a small study with only 10 participants, so almost a proof of concept, Shaibi said. The interval training entailed extremely high intensity (90 percent of VO2 max [the maximum rate of oxygen consumption as measured during incremental exercise]) for 4 minutes, followed by 4 minutes off, then 4 minutes on, and so on, for 40 minutes. The researchers compared the 10 obese adolescents with 10 lean adolescents. Before the training started, the obese youth had significantly reduced stroke volume compared with their lean counterparts. After the 13 weeks of interval training, stroke volume in the obese adolescents had increased significantly, so much so that there was no statistical difference in stroke volume between them and the lean controls by the end of the study. The researchers observed similar changes in global strain, but no change in weight, lipid profile, or blood pressure. “In essence,” he said, “this intervention normalized myocardial function in this population.”
Regarding the feasibility of a twice-weekly 40-minute high-intensity exercise regimen, Shaibi noted that about 95 percent of the sessions in the Ingul et al. (2010) study were attended. Adolescents who are obese tend to like that kind of high-intensity interval training, in his opinion, more than they do a continuous exercise program.
In a similar study, Watts and colleagues (2004) examined the effects of 8 weeks of thrice-weekly circuit training on vascular function in 19 obese adolescents. They measured endothelial function using brachial artery flow-mediated dilatation (FMD) and compared outcome measures in the obese individuals and a lean control group. At baseline, they observed significant differences between the obese and lean adolescents. After the intervention, endothelial function in the obese adolescents was found to be normalized. As in the Ingul et al. (2010) study, there was no change in weight, lipid profile, or blood pressure. Again, Shaibi said, the results suggest that exercise can improve, in this case, vascular function—one of the earliest signs of atherosclerosis—without changing traditional risk measures.
Results of Shaibi’s own research on the effects of 16 weeks of twice-
weekly resistance training on insulin sensitivity in Latino adolescents who were overweight or obese suggest that exercise also increases insulin sensitivity (Shaibi et al., 2006). In addition to measuring insulin sensitivity, using the frequently sampled intravenous glucose tolerance test (FSIVGTT), Shaibi and his team measured body composition using dual-energy X-ray absorptiometry (DXA). They observed about a 45 percent increase in insulin sensitivity in the resistance training group compared with a nonexercising control group, but with tremendous individual variation in change in insulin sensitivity in the intervention group. Shaibi noted that the control group experienced a small increase in insulin sensitivity as well, which may have been a transient change related to puberty. The change in insulin sensitivity in the resistance training group was independent of any change in body composition. Individuals in the resistance training group also showed a subtle, but not statistically significant, decrease in fat mass and a slight, but again not statistically significant, increase in body weight.
The benefits of resistance training for obese youth are not just metabolic, in Shaibi’s opinion. There is also a psychosocial benefit. If one asks a classroom of youth to run a mile, he said, those who end up last are almost always those with the higher weights. By contrast, if one puts that same classroom of youth in a resistance training room, those with the higher weights end up being the strongest. In Shaibi’s opinion, resistance training supports children and adolescents who are obese in being physically active because it is an exercise they can do well.
Translating Research to Practice
Translating the research results described above into real-world practices is a challenge, Shaibi said. Most of the studies conducted in this area have been very small proofs of concept that exercise can improve health outcomes not typically measured in the clinic, but meaningful with respect to cardiometabolic disease.
Additionally, not all individuals respond to interventions in the same way. In collaboration with a local YMCA, Shaibi and colleagues (2012) developed a diabetes prevention program for Latino adolescents who were obese. It was a family-based program, with youth working out with other youth, and with nutrition education provided by health educators from the adolescents’ community. Overall, Shaibi and his team observed statistically significant improvements in insulin sensitivity and glucose tolerance. But on an individual basis, not all the adolescents showed improvement: nine exhibited increases in insulin sensitivity in response to the intervention (the “responders”), and six showed either no change or a decrease (the “nonresponders”).
Subsequently, Shaibi’s research team analyzed blood samples to see
whether there were any biochemical differences between the responders and nonresponders. They found that the responders had more than 1,000 genes that were significantly up- or down-regulated, compared with about 120 such genes in the nonresponders. The two groups of adolescents shared very few of the same up- or down-regulated genes. Shaibi interpreted these findings to mean that there may be a biological basis for responses to lifestyle intervention in adolescents. He suggested that clinicians keep this possibility in mind when they tell youth to exercise and the youth return to the clinic not having lost any weight. Nonresponders may require a different approach.
With respect to other novel practice and policy approaches, Shaibi suggested encouraging clinicians to prescribe exercise the same way they prescribe medications. “Give the kid a prescription pad,” he said, “and tell them what you want them to do.” Exercise prescriptions will require better models for community intervention programs, he noted, so that community center providers or exercise specialists will know what to do when children and adolescents show up with such prescriptions from their physicians. Additionally, he suggested incorporating behavioral change strategies into the prescriptions.
Shaibi suggested further that, with respect to reimbursement and incentives, it is not just providers who need to be incentivized but children and adolescents as well. He mentioned the development of novel programs that incentivize children and adolescents to exercise and how these programs may be beneficial if the evidence is supportive.
With respect to how scientific research can continue to help inform practice and policy, Shaibi asserted that scientists need, first, to gain a better understanding of outcomes beyond BMI that are relevant in children and more proximal to the cardiometabolic disease process. Shaibi suggested considering not just metabolic health but also psychosocial or emotional health. Additionally, scientists need to optimize exercise parameters by conducting dose–response studies and better-designed randomized controlled trials involving children and adolescents who are already obese.
Asking whether the focus should be physical activity (and its relationship to health) or obesity is the wrong approach, John Jakicic began. It may even be self-defeating, because one pathway by which physical activity may improve health is through its effect on body weight. If obesity is not part of
4 This section summarizes information and opinions presented by John M. Jakicic, Ph.D., University of Pittsburgh, Pennsylvania.
the discussion, he said, “We are probably going to miss the combined benefit of both addressing excess body weight and increasing physical activity.”
To explain why obesity matters, Jakicic highlighted baseline cross-sectional data from his own work with the Look AHEAD trial. Among its 5,145 participants, all with type 2 diabetes, decreased levels of fitness were associated with increased odds of high hemoglobin A1c (HbA1c),5 regardless of BMI (Wing et al., 2007). If one were to consider just those HbA1c data, Jakicic explained, one might conclude that having a higher BMI does not necessarily influence health and that health is significantly influenced by level of fitness. In this example, the higher the fitness level, the better individuals were able to manage their diabetes as reflected by level of HbA1c. With hypertension, however, which the researchers defined as either taking medication to control blood pressure or meeting the criteria for high systolic or diastolic blood pressure, the odds ratio increased as fitness decreased, but the odds ratio also was higher for individuals with class II or III obesity compared with those who were overweight or had class I obesity. In other words, Jakicic explained, in contrast with HbA1c, both fitness and body weight appear to influence blood pressure and diagnosis of hypertension. Thus, when thinking about how to maximize the influence on patient-centered outcomes broadly, he said, “Both body weight and fitness or physical activity have to be part of the discussion.”
In other work, based on self-report data, Hergenroeder and colleagues (2011) demonstrated an increase in both disability limitation and physical function limitation with increasing weight (i.e., across five weight categories—normal, overweight, obese I, obese II, and obese III). In a study of obese older adults with knee osteoarthritis, Messier and colleagues (2000) showed that exercise over 6 months improved various knee osteoarthritis outcomes, but exercise combined with weight loss yielded even greater improvement. Subsequent work by Messier and colleagues (2004) revealed a similar larger effect when weight loss was added to a diet and exercise program for older adults with knee osteoarthritis. Again, Jakicic said, exercise provides benefits, but weight loss induced by dietary change adds to the benefits observed.
Jakicic emphasized that he was not arguing that a focus on obesity is more important than a focus on physical activity. Again, he used data from the Look AHEAD trial, in this case follow-up data, to illustrate how both need to be part of the discussion to maximize the effect of intervention. Trial participants were randomized into either a diabetes support and
5 High levels of HbA1c are associated with an increased risk of diabetes-related complications.
education (DSE) group or an intensive lifestyle intervention (ILI) group. The emphasis of the intervention was on weight loss, with participants in the ILI group receiving weekly intervention contact; individuals in the DSE group received standard care and little intervention contact. Four-year data from the trial showed, generally, a decline in HbA1c with increasing fitness (Jakicic et al., 2013). More specifically, in the DSE group and when DSE and ILI data were combined, the data revealed an almost dose–response relationship, Jakicic observed: the greater the increase in fitness, the greater the decline in HbA1c. In the ILI group, HbA1c levels declined in a dose–response way until the highest increase in fitness level was reached (>10 percent increase in fitness), at which point the levels increased slightly. All of these changes in HbA1c were observed after controlling for weight change and use of medication to treat diabetes. Based on these findings, in Jakicic’s opinion, if people with type 2 diabetes were told that all they needed to do was take their medication and lose weight, other potential benefits of fitness (e.g., lower HbA1c) would be missed.
Jakicic suspects that in addition to directly impacting health-related outcomes, physical activity may impact health indirectly through its effects on body weight and adiposity. Physical activity also may change dietary behavior, again indirectly impacting health-related outcomes. Jakicic encouraged a greater understanding of the different pathways that impact health-related outcomes.
The Important Role of Physical Activity in Maintaining Weight Loss
Not only is physical activity important for weight loss, but it is also important for the maintenance of weight loss, Jakicic continued. In a study of about 150 individuals who participated in an 18-month weight loss program, he and his research team classified participants into three categories based on their physical activity levels and examined their weight loss trajectories (Jakicic et al., 1999). The researchers showed that individuals in the highest activity category (average dose of 280 minutes/week) lost more weight during the first 6 months than individuals in the other categories and were able to maintain that lost weight over the next 12 months. Individuals in the other categories lost less weight during the first 6 months and regained more of their lost weight over the next 12 months.
In a separate trial involving about 170 individuals, Jakicic and colleagues (2008) analyzed the effect of physical activity on 24-month weight loss and found that, when participants were stratified by percent weight loss into four groups, every group showed an increase in physical activity during the first 6 months. However, the only group that showed sustained physical activity (1,500 kcal/week or 275 minutes/week above baseline)
over the course of the entire 24-month period was the group that lost the most weight (10 percent or greater weight loss).
The Jakicic et al. (1999, 2008) studies both were based on self-reported physical activity data. More recently, Jackicic and his research team used an objective measure of physical activity obtained from a wearable device to determine whether activity patterns differ for people who maintain a 10 percent weight loss over time compared with those who are unable to sustain that weight loss (Jakicic et al., 2014). The researchers separated participants into four groups based on whether they had achieved and then maintained a 10 percent weight loss at both 6 and 18 months and analyzed their activity patterns. The group that had lost weight at the 6-month mark and maintained that weight loss over the course of the entire 18 months lost and sustained, on average, an impressive 18 percent of body weight, Jakicic said. That is the same magnitude of weight loss typically observed following gastric band surgery, he noted, and it may even be greater. In terms of activity patterns, the “maintainers” compared with the other groups sustained much greater levels of moderate- to vigorous-intensity physical activity (MVPA), measured in bouts of 10 minutes’ or longer duration, over time (about 1,200 metabolic equivalent of task [MET]-minutes per week at both 12 months and 18 months). In contrast, there appeared to be no relationship between weight loss and bouts of MVPA shorter than 10 minutes in duration.
In the same study, Jakicic and colleagues also examined effects of light physical activity on maintenance of long-term weight loss. Again, they found that those who lost 10 percent or more and maintained that weight loss over time showed a sustained increase in activity over time. Jakicic interpreted this finding to mean that individuals who successfully lost 10 percent or more of their body weight and maintained that loss over time did not simply convert what had formerly been light activity into MVPA. Rather, the people who were most successful with their weight loss management were those who increased both types of activity. He said, “We should be targeting both of those behaviors in order to potentially improve long-term weight loss outcomes.”
Variability in Response to Physical Activity
Like Gabriel Shaibi before him, Jakicic expressed curiosity about variability in responses to physical activity (see the previous section for a summary of Shaibi’s presentation). Data from the Midwest Exercise Study, a 16-month study involving 4 to 5 days per week of supervised 45-minute exercise sessions at about 70 to 75 percent of VO2 max, showed that about half of the women in the exercise intervention gained weight, while the other half lost weight (Donnelly and Smith, 2005; Donnelly et al., 2003).
In other words, there was great variability in terms of gaining versus losing weight with the same exercise intervention. The question for Jakicic was, why? He said he was exploring the possibilities and was especially interested in eating behavior.
Jakicic described a study conducted in his laboratory that served as the basis for a doctoral student’s dissertation. Women were brought into the lab on two separate occasions—once to sit for 45 minutes and once to exercise at about 70-75 percent of their age-predicted maximum heart rate for 45 minutes—and then provided free access to food (Unick et al., 2010). Half the women ate more on the day they exercised, and the other half ate more on the day they rested. Jakicic suggested that this finding may indicate a differential response to physical activity whereby physical activity may stimulate hunger in one person and produce a satiety effect in another. He called for a greater understanding of the hunger versus satiety response to exercise.
Finally, Jakicic expressed curiosity about expenditure patterns of physical activity. For example, the expenditure patterns of three 10-minute bouts of MVPA within a 60-minute period and one 30-minute bout within that same 60-minute period are very different. Jakicic suggested that different patterns may produce different responses.
Following Jakicic’s presentation, he and the other speakers participated in a panel discussion with the audience.
An audience member asked about weight cycling and what it bodes for weight loss and maintenance in the future. Jakicic observed that weight cycling is not discussed now as much it was 20 years ago, yet it is still, in his opinion, a “huge deal.” It is difficult to conduct randomized controlled trials of weight cycling. He and his team have done a small amount of work on the psychological response to weight regain and the vicious cycle it starts, beginning with people becoming less adherent. In an 18-month study, they proactively targeted some participants in places where they would tend to slip and observed a 10 percent sustained weight loss among those participants (Jakicic et al., 2015). Participants who were not proactively targeted regained their lost weight. Kriska added that, in both the DPP and its follow-up study, help was provided to participants who started slipping with either their weight or their physical activity in determining what they needed to do to get back on track.
Diabetes Prevention Program (DPP) Results
Kriska was asked why the lifestyle intervention in the community study did not work as well at the workplace site as at the other sites. Kriska replied that the program worked quite well at the worksite, but not as well as at the other two sites, probably because of the younger and employed nature of the worksite participants. In the DPP, she said, older individuals did better, and she speculated that the reason may be that they were retired and could devote more time to their healthy lifestyle program. She suspects that the same may have been true of the community intervention program.
A question was raised as to whether the lifestyle intervention arm of the DPP was more expensive than the metformin arm and if so, whether it is realistic for communities to consider an intense lifestyle intervention. Kriska replied that in the DPP, while the metformin arm was somewhat more economical than the lifestyle intervention, the one-on-one delivery of the lifestyle intervention contributed to its cost. As these effective lifestyle intervention programs move into communities, the intervention generally is being offered in a group setting, with the bulk of the cost related to the provision of lifestyle coaches to lead these groups. Because the programs are offered in groups, they can be more feasible to run in the community setting, as well as more cost-efficient. Kriska and others are also working on developing other cost-effective lifestyle intervention approaches, such as offering the intervention online or via DVD.
Obesity as a Physical Disability
A comment was made about obesity being a physical disability in the sense that it keeps people from being active. Jakicic recalled his work in a physical therapy setting where patients presented with back pain. It was clear to Jakicic that many of these people were carrying significant amounts of body weight, yet nothing was done in that setting to address their weight issues. He said, “I think we need to be thinking cross-disciplinarily . . . in order to have the biggest effect.” Another audience member observed that, at least with respect to osteoarthritis, physical therapists today are much more aware of weight than they have been in the past.
Physical Activity Versus Obesity
An audience member asked Jakicic what would be lost by dissociating physical activity from weight loss. If someone engages in physical activity, in the questioner’s opinion, eventually their obesity will be managed. The real tragedy, he believes, is when people discontinue their weight loss behaviors because they have not lost any weight. Jakicic stated that very
few people with a BMI of more than about 35 are physically active at a level that would help them achieve the health benefits of physical activity. In his opinion, for people with BMIs greater than around 35, one of the best ways to start may be with weight loss. “There are a lot of people out there,” he said, “who need to lose weight just to be functional.” In his opinion, dissociating physical activity from weight loss would be doing those people an injustice.
This page intentionally left blank.