Characterizing severe obesity in children and youth referred for weight management

Obesity is one of the most common chronic disorders affecting the health and well-being of children and youth and its prevalence remains high [1, 2]. Recent data from the Canadian Health Measures Survey revealed that almost one-third of Canadian 5 – 17 year-olds are either overweight or obese [3]. Even more striking is recent evidence from the US indicating an increasing prevalence of severe obesity (SO). For example, between the National Health and Nutrition Examination Survey (NHANES) 1988-1994 and NHANES 1999-2004, the prevalence of SO in children and youth increased by 72% from 2.2% to 3.8% [4]. With evidence suggesting that the overall prevalence of pediatric overweight and obesity has remained stable [2], these data are cause for concern because SO is associated with even worse cardiometabolic risk factors profiles, including elevated total cholesterol, triglycerides, serum glucose, and systolic blood pressure and low HDL-cholesterol [2, 5, 6]. Compounding these health risks is the observation that SO tends to be more challenging for clinicians and families to manage successfully. For instance, a lower degree of obesity predicts successful weight loss during pediatric weight management [7, 8], suggesting that as the degree of obesity in children and youth increases, it becomes more entrenched and resistant to therapeutic interventions.

Recent reports have highlighted a growing research and clinical focus on SO in pediatrics, with data characterizing high prevalence levels [2, 4‐9] and cardiometabolic health risks [10‐14]. This information is highly relevant from a health services perspective because many children and youth referred to weight management clinics present with SO and because most clinical programs use the severity of obesity and/or presence of obesity-related co-morbidities as referral criteria [11]. Clinical practice guidelines [12] and expert recommendations [13] highlight the importance of making positive lifestyle (nutrition and physical activity) changes in order to manage obesity. With this in mind, the objectives of our research were to characterize and examine differences in demographic, anthropometric, cardiometabolic and lifestyle variables in children and youth with SO versus their less overweight and obese (OW/OB) peers among a cohort of individuals referred for weight management.

In total, 345 participants were included in this report, the majority of whom were girls (56.2%), Caucasian (78.7%), and had family incomes > $50,000 CDN/year (65.7%). The anthropometric characteristics of children and youth grouped according to OW/OB and SO categories are presented in Table 2. By design, the SO group was both heavier and more overweight than their OW/OB peers. We also completed exploratory analyses between OW (n = 26) and OB (n = 158) sub-groups. Among all of the demographic, cardiometabolic, and lifestyle variables, the only group differences related to systolic blood pressure (OW: 119 ± 13mmHg versus OB: 110 ± 11 mmHg; p < 0.05) and dietary energy from saturated fat (OW: 203 ± 78 kcal/d versus OB: 254 ± 119 kcal/d; p < 0.05). These findings highlighted the homogeneity within the OW/OB group and justified our decision to combine these sub-groups.

Comparisons of cardiometabolic risk factors (continuous data) between OW/OB and SO groups are shown in Table 3, with lower TC and HDL-C as well as higher ALT, fasting insulin, HOMA-IR, SBP, and DBP in the SO vs OW/OB group. After we dichotomized cardiometabolic risk factors (see Table 1 for thresholds), we found that the proportions of children and youth with cardiometabolic risk factors were higher in the SO group compared to the OW/OB group for most variables, including WC, TG, HDL-C, ALT, HOMA-IR, SBP and DBP (see Figure 1).

Comparisons of nutrition- and physical activity-related behaviors between the SO and OW/OB groups are shown in Table 4. Overall, the SO group consumed a greater amount of total energy, which was reflected in their higher intakes of all macronutrients (carbohydrate, protein, and fat) as well as servings/day of Grain Products and Meat and Alternatives. In addition, in relation to their OW/OB peers, the SO group reported accumulating fewer daily steps, less MVPA, and more LTST.

The proportions of participants in the OW/OB and SO groups that met each of the nutrition- and physical activity-related recommendations (defined in Table 1) are shown in Figure 2. While we failed to observe any significant group differences according to the food groups, in general, a large proportion (70 – 80%) of our sample met the recommendations for daily intakes of Grain Products and Meat and Alternatives whereas a small proportion (10 – 25%) satisfied the Vegetables and Fruit and Milk and Alternatives guidelines. Similar to the nutrition-related recommendations, we observed that most children and youth in our study did not satisfy the physical activity-related recommendations. However, almost twice as many OW/OB participants met the MVPA recommendation compared to their SO peers (χ² = 7.6; p < 0.01). Similar differences were revealed with a greater proportion of the OW/OB group meeting LTST recommendations (χ² = 6.7; p < 0.05) than the SO group.

Our stratified analyses revealed no significant (i) weight status (OW/OB vs SO) × gender (boys vs girls), (ii) weight status × age group (children vs youth), or (iii) weight status × ethnicity (Caucasian vs non-Caucasian) interactions across our cardiometabolic risk factors and nutrition and physical activity-related habits (all interaction effects >0.05).

In our study of children and youth referred for weight management, we showed that (i) those with SO were at increased cardiometabolic health risk compared to their OW/OB peers, and (ii) while the study sample as a whole had sub-optimal nutrition and physical activity habits, those with SO tended to have poorer lifestyle behaviors than their OW/OB counterparts.

Our observations are consistent with those of Skelton et al.[4] who applied the same classification system to define weight status categories in a sample (n ~ 12,000) of boys and girls from the National Health and Nutrition Examination Surveys II, III, and 1999-2004 and also reported a higher degree of cardiometabolic risk among the SO group in relation to their OW/OB peers. To our knowledge, we are the first to report differences in lifestyle behaviors between children and youth with SO versus their OW/OB peers. This is a noteworthy finding and emphasizes the point that a one-size-fits-all approach to weight management is inappropriate since those with SO appear to have greater potential to improve their lifestyle habits, which may have led to their higher degree of obesity. Viewed another way, the less favorable lifestyle habits of the SO group provides an opportunity for them to make positive improvements to nutrition and physical activity that can contribute to weight management success and improvements in other health outcomes, independent of substantial weight loss. For instance, a recent review of 22 studies of immersion programs examined treatments that combined a controlled diet, physical activity, nutrition education and therapy and/or education supportive of behavioral change. The participants in the reviewed treatments lost an average of 23.9% of their overweight during treatment and 20.6% during follow up [38]. Although there are exceptions [39], we recently reported [16] that it is challenging for most boys and girls with obesity to improve their weight status by making modest lifestyle changes; for many individuals, preventing further weight gain by making nutrition and physical activity changes that lead to weight stabilization is an important and achievable goal, albeit one that may or may not satisfy the expectations and desires of families [40].

Despite the differences we observed between SO and OW/OB groups, overall, most children and youth in our study did not achieve current lifestyle recommendations. In addition to supporting previous findings from our clinic [41], these results are aligned with other larger, population-based studies. For instance, Shields showed low levels of vegetable and fruit intakes as a reflection of poor diet quality that may increase the risk of pediatric obesity [42]. Similarly, Garriguet and colleagues found that a substantial number of children and youth do not meet the recommendations for vegetables and fruit, and milk and alternatives intake [21, 43]. He et al. suggested that increased consumption of fruits and vegetables might reduce long-term weight gain and obesity risk [44]. Lowering the amount of added fructose in diets may also benefit children and youth with obesity by improving their markers for cardiometabolic dysfunction, which may be particularly beneficial for those at risk for developing type 2 diabetes or nonalcoholic fatty liver dysfunction [45]. Although our data are cross-sectional, our observations suggest that interventions that emphasize improvements in diet quality and quantity are likely to be important strategies for managing pediatric obesity [12].

It is clear that a physically active lifestyle has numerous health promoting benefits (beyond weight management) [46], but it remains a challenge for most boys and girls to achieve current recommendations. The Canadian Physical Activity and Sedentary Behaviour Guidelines[34] suggest that children and youth should accumulate ≥60 minutes of daily MVPA and limit their sedentary time to <2 hours daily [47]. Following these guidelines can help improve body composition, cardiorespiratory and musculoskeletal health, social behaviors, self-esteem and academic success [48]. A relatively high proportion (30 – 40% according to the self-reported data) of our sample met the physical activity recommendations; however, based on objective physical activity data collected using accelerometers, <10% of Canadian children and youth accumulate ≥60 minutes of daily MVPA [49]. The difference in the proportion of individuals meeting MVPA recommendations in our sample versus national normative data suggests boys and girls in our study may have overestimated their physical activity levels. Our observation that a lower percentage of our sample achieved the daily steps recommendation versus self-reported MVPA provides indirect evidence in support of this point, but is validated by other data showing that young individuals may overestimate their physical activity using self-report tools [50].

Our study has several limitations to acknowledge. First, since all of the participants were referred by physicians to a weight management clinic, our findings may not apply to all children and youth with SO, especially those who do not enroll in health services for obesity management. Notably, our clinic is one of eight that is participating in a new Canadian multi-center study to characterize children, youth and families receiving health services for managing pediatric obesity [51]. Evidence generated from this registry of ~1,600 participants will show whether our findings are generalizable to other similar settings. Second, given our cross-sectional study design, we cannot comment on whether our observed differences between SO and OW/OB groups remain consistent over time or whether changes in response to therapy are similar across groups. Third, while our analyses included objective measurements of anthropometric variables and cardiometabolic risk factors, self-reported nutrition and physical activity data are subjective in nature, which can reduce measurement reliability and accuracy. The fact that we still observed group differences based on these self-reported data provides some confidence that real differences exist between SO and OW/OB groups. Finally, a recent report [52] highlighted the need for consistent terms and definitions of SO in pediatrics. While the current manuscript applied one of several different definitions of SO in the literature, applying the newer recommended criteria in future studies is needed to establish its superiority.

In conclusion, our study revealed that children and youth with SO were at increased cardiometabolic health risk and tended to have less healthy nutrition and physical activity habits compared to their OW/OB counterparts. In their discussions with families, clinicians can use this information to highlight the health risks associated with a higher degree of obesity. By applying family- and client-centered counseling strategies [53, 54], they can also emphasize the potential that exists for boys and girls with obesity to make healthy changes since most do not satisfy current lifestyle recommendations.