Promoting healthy lifestyle behaviours in youth: Findings from a novel intervention for children at risk of cardiovascular disease

Abstract

Objectives

Obesity is the most prevalent risk factor for cardiovascular disease (CVD) in children. We developed a 2-year lifestyle intervention for youth at risk of CVD. We assessed changes in body mass index z-scores (zBMI) and key cardiometabolic risk factors, physical fitness, and capacity among those who completed the program.

Methods

The CIRCUIT program is a multidisciplinary lifestyle intervention for children aged 4 to 18 years at risk of CVD, based on a personalized plan to improve cardiometabolic outcomes by increasing physical activity and reducing sedentary behaviours. Both at baseline and 2-year follow-up, we measured zBMI, blood pressure z-scores (zBP), adiposity (%body and %trunk fat), fasting blood glucose and lipid profile, aerobic (VO₂max) and anaerobic (5×5 m shuttle run test) fitness, and physical capacity indicators. Differences between baseline and follow-up were examined using paired t-tests (for age-sex standardized outcomes) and multivariable mixed effect models, adjusted for age and sex (for other outcomes).

Results

Among the 106 participants (53 males) who completed the 2-year program, mean age at baseline was 10.9 years (SD=3.2). After 2 years, zBMI and diastolic zBP decreased by 0.30SD (95% CI: −0.44; −0.16) and 0.43SD (95% CI: −0.65; −0.23), respectively. Participants improved %body and %trunk fat, lipid profile, aerobic and anaerobic fitness levels, and physical capacity (p<0.02). No changes in systolic zBP nor in fasting plasma glucose were observed.

Conclusion

Our findings showed improved zBMI, cardiometabolic outcomes, physical fitness, and capacity among children at risk of CVD, suggesting that CIRCUIT is a promising intervention.

Graphical Abstract

Graphical Abstract.

Childhood overweight and obesity is an international pandemic (1,2) currently affecting 27% of Canadian youth (3). It is the leading cause of cardiovascular disease (CVD) risk in children (4). While the optimal approach to treatment elicits intense debate (5–7), the US Preventative Task Force recommends referring children with obesity to intensive, comprehensive interventions that target lifestyle behaviours (8).

Although obesity is multifactorial, lifestyle factors including physical inactivity and excess sedentary behaviour are key contributors (9). Unhealthy lifestyle factors tend to originate in childhood and track into adulthood, accelerating CVD risk (10,11). Consequently, early-life intervention may be more effective than later attempts to alter behaviours (12).

Physical fitness, especially cardiorespiratory fitness, is associated with favourable physical health indicators among Canadian children (13). Additionally, muscular fitness developed during childhood and adolescence promotes healthier adiposity measures (e.g., body mass index [BMI] and skinfold thickness) and cardiometabolic health outcomes (e.g., triglycerides and cardiovascular risk scores) in later life (14). A recent study suggested a reciprocal relationship between physical fitness, capacity tests, and physical activity (15). Therefore, targeting physical fitness and capacity may enhance physical activity levels in lifestyle interventions.

The Centre d’Intervention en prévention et Réadataptation CardiovascUlaire pour Toute la famille (CIRCUIT) is a one-arm personalized 2-year intervention for children aiming to improve cardiometabolic outcomes by increasing physical activity and reducing sedentary behaviours. Respecting family preferences for personalized obesity treatment (16), CIRCUIT is tailored to each participant. We report the impact of CIRCUIT on adiposity and key cardiometabolic risk factors (main outcomes) and changes in physical fitness and capacity (additional outcomes), among participants who completed the 2-year program before July 2017.

METHODS

Setting

CIRCUIT has delivered paediatric lifestyle behaviour change interventions at the Sainte-Justine University Hospital Centre (CHU Sainte-Justine, Montreal, Canada) since 2011. Children aged 4 to 18 years with at least one CVD risk factor (e.g., obesity, metabolic syndrome, diabetes, dyslipidemia, high blood pressure) are referred by health care providers (55% by primary care physicians, 35% by specialists, 10% by allied healthcare professionals).

Study population

This pre–post intervention study was performed with a cohort of CIRCUIT participants who attended the baseline visit and completed the 2-year program by attending the final visit. Data stem from baseline (Visit 1), year 1 (Visit 4), and year 2 (Visit 7) evaluations (Supplementary Appendix 1).

Inclusion and exclusion criteria

All participants that completed Visit 1 between January 2011 and July 2015 and visit 7 by July 2017 were considered eligible. Participants <4 years or >18 years at baseline, and those who did not complete the program up to Visit 7 were excluded.

Intervention

The program uses a conceptual framework for behaviour change inspired by the COM-B approach, a multidimensional construct whereby motivation, capabilities, and opportunities interact together to influence health behaviours (17,18). This framework has previously been used to understand child behaviours for which family plays an important role (19). Upon referral to the program, an initial meeting with each family is held, allowing an assigned kinesiologist to fully explain the nature and requirements of the CIRCUIT program. This is followed by a comprehensive evaluation of the child’s lifestyle habits, including sedentary pursuits such as televiewing, computer, and video game use; current physical activity patterns and preferences; family structure; and parental availability to support their child’s engagement in physical activities, using a questionnaire derived from a compendium of validated questions (20–22). The kinesiologist also documents information regarding their built environment and local resources, including proximity to grocery stores, parks, services, school, physical activity installations, and walking or cycling infrastructure within the child’s home or school neighborhoods. Perceptions of neighborhood safety are also assessed. We have also developed a physical activity compendium including resources and infrastructures available in each participant’s environment, organized by postal code. A personalized intervention plan is then developed in conjunction with the family and the child, prioritizing individual preferences and opportunities available within the family’s built environment to improve the child’s lifestyle habits. For example, to increase physical activity and reduce sedentary behaviour, the kinesiologist may suggest going to the park, walking to school daily, or swimming at the local pool on weekends. Local resources for fresh foods may be proposed to improve dietary habits. At the end of the appointment, each family usually has three goals to work on. Participants are subsequently contacted (by phone or email) regularly by the kinesiologist to enquire about their wellbeing, discuss their progress, and address any challenges in attaining their goals. In-clinic follow-up assessments are scheduled every 6 months where the personalized intervention plan is updated as needed (Supplementary Appendix 1). An example of a CIRCUIT visit and action plan is provided in Supplementary Appendix 2. Participants are seen individually or with a family member by the same kinesiologist throughout the program. Community activities and events are also regularly proposed to all participants, such as Fun Runs, No Screen Challenges and group activities at centrally located parks. To lessen burden to the families, participants have only four in-person clinic visits per year, and the number of phone/internet contacts are tailored to each participant’s needs (Supplementary Appendix 1). Contacts are further enhanced for families who participate in community activities and events that occur on a monthly basis. No monetary or other incentives were offered to participants.

Outcomes

For each test assessed, validated norms in children by age and sex were used. The same tests were performed for all participants. Measures are described succinctly here, with further detail provided in Supplementary Appendix 3.

Anthropometric measures

Height and weight were measured. BMI z-scores (zBMI) were obtained using World Health Organization growth charts, which are sex- and age-specific. Obesity was defined as having zBMI ≥ 2 (23). Total and truncal body fat percentage were measured using the validated bioelectric impedance (24).

Blood pressure measures

The mean of three measurements of systolic and diastolic blood pressure (BP) was computed and used to calculate systolic and diastolic BP z-scores (zBP), according to the National High Blood Pressure Education Program Working Group reference values for height, age, and sex (25).

Biochemical measures

A baseline blood test was performed within 3 months upon program initiation, and again after 1 and 2 years of follow-up. Fasting blood glucose (FBG) (26) and lipid-profile were evaluated.

Fitness measures

Aerobic fitness was estimated using peak oxygen consumption (VO₂max) during a treadmill stress test to volitional exhaustion adapted from the shuttle test, with indirect calorimetry measurements throughout the test (27–30). We also used allometrically scaled VO₂max for lean body mass (31). Anaerobic fitness was estimated using the 5×5-m shuttle run test where the ability to change direction and speed was assessed (32,33).

Physical capacity measures

Physical capacity included a combination of tests ranging from motor skills to physical strength detailed in Supplementary Appendix 3 (33–39).

Statistical analysis

Differences between baseline and year 2 for age-sex-standardized measures (e.g., zBMI) were tested using paired student t-tests. For the other outcomes, in order to account for time-varying changes (e.g., expected increase in strength as children grow), multivariable hierarchical linear models incorporating measures at three different time points (baseline, year 1, and year 2) and adjusting for age and sex were employed. Using baseline measures set as the reference category, binary variables were used to indicate the visit during which measures were taken. The beta estimates for the year 2 dummy variable indicate the average difference between baseline and postintervention measures, and are reported in the results section. Participants who took anti-hypertensives (n=4), statins (n=1), or hypoglycaemic medications (n=2) were excluded from the analyses involving blood pressure, lipid profile, and fasting glucose, respectively. We set statistical significance at α of 0.05 and present the corresponding 95% confidence intervals. All statistical analyses were performed using SAS 9.4 for Windows.

Ethics

The study was conducted in accordance with the Declaration of Helsinki (40) and approved by the Medical Ethics Committee of the CHU-Sainte-Justine (2011–333, 3257). Written informed parental consent and child assent were obtained for all participants. The study was registered at ClinicalTrials.GOV with the number NCT01736748. This manuscript was written following the TREND statement (41).

RESULTS

As of July 2015, a total of 330 participants had initiated CIRCUIT. Two hundred and seventeen had dropped out and an additional 7 were absent for Visit 7, leaving 106 participants who completed the 2-year intervention included in this analysis (Supplementary Appendix 4). About 47% of participants (46% males) who initiated the program dropped out prior to Visit 4 (42).

The average age of participants (50% males) was 10.9 years at baseline. While the majority were Caucasian, a broad range of other ethnic origins were represented. Of the 106 study participants, 5.7% were diagnosed with overweight, 36.8% with obesity, and 54.7% with severe obesity. Also, 28% of participants had hypertension, 37% were classified as having dyslipidemia, and 12% as having diabetes at baseline. Baseline descriptive data for all, males and females are shown in Supplementary Appendix 5.

Over the 2-year program, the mean zBMI of all participants improved (−0.30SD, 95%CI: −0.44; −0.16; Figure 1). Overall, 21% of participants classified as having severe obesity at baseline transitioned to being obese and 28% of the participants with obesity transitioned to having overweight or normal weight after program completion. In addition, 50% of participants demonstrated a decrease in zBMI of at least 0.25SD, previously identified as a clinically important indicator of success (43). Regarding BP measures, only diastolic zBP decreased statistically significantly from baseline to year 2 (0.43SD, 95% CI: −0.64; −0.23; Figure 1).

Figure 1.

Two-year changes in age-sex standardized measures of baseline body mass index and blood pressure among study participants of the CIRCUIT program. From participants initiating the program from 2011 to 2015, Montreal, Canada. BMI z-score: body mass index z-score; sBP z-score: systolic blood pressure z-score adjusted for height, age and sex; dBP z-score: diastolic blood pressure z-score adjusted for height, age and sex. *Statistically significantly different from baseline at α = 0.0001.

Over the 2-year program, there were modest, but statistically significant, improvements in total cholesterol and LDL-cholesterol levels, but no change in FBG was observed (Table 1). On average, participants decreased their total %body fat by 1.93 percentage points (95%CI: −3.48; −0.38) and their %trunk fat by 2.00 percentage points (95%CI: −3.64; −0.37; Table 1).

Table 1.

Baseline to year 2 changes in cardiometabolic outcomes of the CIRCUIT program study participants

Output (units)	Baseline Mean (SD)	Year 2 Mean (SD)	Baseline to Year 2	CI	p-value
Total cholesterol (mmol/L)	4.47 (0.92)	4.22 (0.93)	−0.25	−0.47 to −0.04	0.02
Triglycerides (mmol/L)	1.30 (0.71)	1.39 (1.05)	−0.18	−0.39 to 0.04	0.11
HDL cholesterol (mmol/L)	1.23 (0.48)	1.13 (0.29)	−0.05	−0.14 to 0.05	0.32
LDL cholesterol* (mmol/L)	2.70 (0.72)	2.48 (0.72)	−0.20	−0.36 to −0.03	0.02
FBG (mmol/L)	5.13 (1.59)	5.25 (1.40)	−0.17	−0.50 to 0.16	0.31
Percent body fat (%)	37.99 (7.53)	37.69 (9.02)	−1.93	−3.48 to −0.38	0.01
Percent trunk fat (%)	32.58 (8.02)	33.25 (9.28)	−2.00	−3.64 to −0.37	0.02

From participants initiating the program from 2011 to 2015, Montreal, Canada.

CI 95% confidence interval; FBG Fasting blood sugar; HDL High-density lipoprotein; LDL Low-density lipoprotein.

*Calculated with Friedewald equation.

We also noted significant improvements in both aerobic and anaerobic fitness and several physical capacity indices, even after adjusting for age and sex (Table 2). Although underpowered, stratification by sex yielded similar results (Supplementary Appendix 6).

Table 2.

Baseline to year 2 changes in physical capacity, aerobic and anaerobic fitness of the CIRCUIT program study participants

Outcomes (units)	Baseline Mean (SD)	Year 2 Mean (SD)	Baseline to Year 2	CI	p-value
Relative VO2max [mL/(kg × min)]	27.91 (5.28)	29.83 (7.06)	3.55	2.08 to 5.02	<0.01
Allometrically scaled VO2max [mL/(kg0.75 × min)]	80.45 (13.69)	88.32 (18.32)	10.57	6.52 to 14.62	<0.01
5×5-m shuttle run test (sec)	13.25 (2.66)	11.72 (2.05)	−1.21	−1.66 to −0.77	<0.01
Sit and Reach test (cm)	24.74 (9.38)	26.15 (9.37)	1.37	−0.35 to 3.09	0.12
Right Grip (kg)	22.15 (9.13)	29.87 (11.09)	3.63	2.01 to 5.24	<0.01
Left Grip (kg)	21.05 (8.33)	28.32 (10.74)	2.98	1.49 to 4.48	<0.01
Hand Grip (Left + Right) (kg)	43.20 (17.18)	58.19 (21.40)	6.71	3.75 to 9.66	<0.01
Sit-ups test (reps)	14.34 (9.58)	20.59 (8.02)	4.06	2.00 to 6.13	<0.01
Push-ups test (reps)	7.60 (7.24)	12.33 (9.82)	3.50	1.66 to 5.31	<0.01
Long-Jump test (cm)	104.04 (29.39)	119.29 (34.99)	9.71	4.20 to 15.23	<0.01
Balance test (sec)	16.21 (16.60)	20.05 (17.83)	3.92	−0.22 to 8.99	0.06
Throwing test (points)	3.66 (2.85)	5.26 (2.92)	0.80	0.11 to 1.49	0.02

From participants initiating the program from 2011 to 2015, Montreal, Canada.

CI 95% confidence intervals; reps Repetitions; VO ₂ max Maximum oxygen volume. All models adjusted for age and sex of the participants.

DISCUSSION

In this study, we found that participants who completed the 2-year CIRCUIT intervention program decreased both their zBMI and diastolic zBP. They also showed improvements in their aerobic and anaerobic fitness tests, physical capacity, adiposity (%body and %trunk fat), and lipid profile during the intervention period. No changes were observed in systolic zBP and FBG.

As childhood obesity is associated with several negative physical health outcomes (44), any intervention lowering obesity and its associated cardiometabolic risks is important in preventing future poor health. Our findings are therefore encouraging, with a mean decrease in zBMI of 0.30SD (p<0.01), suggesting that CIRCUIT may be an effective strategy for children at risk of CVD (45). Emerging evidence point to a zBMI reduction of 0.25 as sufficient to lead to clinically significant improvement in CVD risk factors among overweight children (46). Half of this study’s subjects successfully achieved such a reduction after completing the program. In our program, personalization was achieved through simple queries that prioritized family preferences and involvement. This is in line with previous research suggesting that tailoring paediatric lifestyle interventions is more effective than standard interventions for treating obesity (47).

Evidence suggests that paediatric weight management interventions must be sufficiently long in duration (6 months for initial weight loss phase) and have a strong family component (48) in order to achieve optimal results. The CIRCUIT program is a personalized intervention that incorporates these components, while also minimizing the burden imposed on families by using telephone/e-mail contacts and requiring only four in-person visits per year.

We observed a decrease in diastolic zBP but not systolic zBP. This is in line with a meta-analysis of randomized controlled trials on the impact of exercise on blood pressure in children and adolescents reporting that exercise results in greater decreases in diastolic BP compared to systolic BP (3% versus 1%, respectively) (49). Physical activities proposed as part of the CIRCUIT program were varied and included exercises involving endurance, isometric and dynamic resistance. As suggested by Cornelissen and Smart, when a physical activity program includes a variety of exercise types, diastolic BP tends to decrease before an impact is seen on systolic BP (50).

CIRCUIT participants showed modest improvements in their lipid profile but no difference in FBG over the 2 years. Participants, on average, had FBG within the normal range at baseline, which is not unusual in paediatric obesity (51). A recent meta-analysis found that FBG remains unchanged even after aerobic, resistance, and combined exercise training among children with overweight and obesity (52). Other lifestyle modification interventions have similarly observed little or no effect on lipids (53,54). Despite significant improvements in BMI, Blüher et al. observed no changes in lipid profile and FBG among 115 participants with overweight/obesity, aged 7 to 18 years, following a 1-year physical activity intervention (55). It may be that paediatric lifestyle interventions studied typically do not include a long enough follow-up period to detect beneficial effects on lipid and glucose profiles, or that more significant changes in weight or lifestyle are required to affect these measures.

There is accumulating evidence that improvement in fitness levels can have beneficial effects on CVD risk factors in both children and adolescents (56–58). Low muscular strength has been found to be a risk factor for young adulthood mortality, including from cardiovascular diseases, with an effect comparable to that of being obese (59). Similarly, normalized strength was shown to be associated with lower cardiometabolic risk among healthy children. It has been shown that muscular fitness is highly influenced by body weight in children (60), however several individual studies have shown that muscular fitness measured both in absolute terms (61), and relative to body weight (62–64), is inversely associated with adiposity later in life. After the 2-year intervention, we observed clinically significant improvements in aerobic and anaerobic fitness tests, as well as in physical capacity tests, even after adjusting for age and sex. Hruby et al. showed, in a large sample of American school children, that achieving and maintaining adequate muscular fitness resulted in significantly greater odds of a healthy weight at follow-up (65). Although children’s aerobic capacity and consequently their ability to exercise for longer periods of time increase as they grow (66), it is clear that aerobic capacity in youth increases with physical activity as well (67,68). This highlights the importance of targeting fitness levels in lifestyle interventions for children at risk of CVD.

Strengths of this study include the use of a personalized intervention posing a low burden to participants, potentially transposable to other health care settings. This novel approach was based on socioecological models of behaviour that emphasize the importance of environments. In addition, outcomes were measured using rigorous and standardized measurement protocols. Our attrition rate (42), albeit elevated, was comparable to other clinical paediatric lifestyle interventions (69,70). Engagement and motivation in lifestyle modification are notoriously difficult in this population, and remain an important obstacle to intervention success (71). Given that the strongest evidence in favour of any intervention is gathered from randomized controlled trials, the main limitation of the present study is the absence of a control group. Moreover, although participants were evaluated at the end of the intervention, follow-up after completion of the program is required to evaluate the long-term effectiveness of CIRCUIT. Nonetheless, our results suggest that CIRCUIT is a promising approach to improving cardiometabolic health as well as fitness and physical capacity levels among children at-risk of CVD. The results of the CIRCUIT intervention suggest that a comprehensive approach, centred on healthy lifestyle choices such as enhancing physical activity, may impart the necessary confidence, skills, and knowledge to children and families for the promotion of healthy lifestyles and ultimately improve cardiometabolic health. Given its demonstrated effectiveness in a real-life clinical setting, future initiatives should explore scalability and adaptability for the implementation of similar programs in other settings.