In the United States, the prevalence of obesity has been increasing at an alarming rate, and recent data show the prevalence of obesity in children is as high as18.5%. Rates of obesity were higher in adolescents and school-aged children versus preschool-aged children. In addition, Hispanic and non-Hispanic black children have a significantly higher prevalence of obesity when compared to non-Hispanic white and non-Hispanic Asian youth.Reference Hales, Carroll, Fryar and Ogden1 Concerning trends are emerging in the paediatric population where the rate of rise in worldwide childhood obesity is exceeding that among adults.Reference Collaborators, Afshin and Forouzanfar2 Childhood obesity is associated with multiple adverse health consequences including atherosclerosis, hypertension, obstructive sleep apnea, asthma, type 2 diabetes, fatty liver disease, and the metabolic syndrome.Reference Daniels3 Ray et al. reported low rates of physical activity in children born with congenital heart disease (CHD) compared to healthy children (10–14 years old), and an alarming 26% were obese, higher than the general adolescent population.Reference Ray, Green and Henry4 However, Tamayo et al, in a longitudinal cohort of 725 patients with CHD, showed a similar prevalence of obesity to the general population.Reference Tamayo, Manlhiot, Patterson, Lalani and McCrindle5 A few studies have reported the impact of obesity on length of stay, morbidity, and mortality in hospitalised critically ill children and adolescents, and the results have been conflicting.Reference Goh, Wakeham, Brazauskas, Mikhailov and Goday6-Reference Ross, Newth, Leung, Wetzel and Khemani8 In adults with CHD, obesity is associated with increased morbidity after pulmonary valve replacement, including longer hospitalisation, and higher risk for post-operative arrhythmias.Reference Buelow, Earing and Hill9 Furthermore, the society of thoracic surgeons database was queried between 2010 and 2015, and for all ages (10–35 years), being obese and underweight by body mass index were associated with increased risk of composite adverse outcome independently.Reference O’Byrne, Kim and Hornik10 However, data on the effect of obesity on outcomes in adolescents undergoing surgery for CHD are lacking. This study was performed with an objective to evaluate the short-term impact of obesity in the population of adolescents undergoing surgery for various CHD.
Materials and methods
This was a single center, retrospective study involving patients with CHD between 10 and 18 years of age who underwent cardiac surgery using cardiopulmonary bypass at Children’s Hospital of Michigan from January 2012 to December 2017. Patients with a known genetic or metabolic syndrome, patients undergoing orthotropic heart transplantation, and those with incomplete medical records were excluded. Approval for the study and waiver of parental consent and patient assent were obtained from the Institutional Review Board at Wayne State University. For patients requiring multiple surgeries during the study period, only the first surgery was considered for analysis. Demographic data including age, weight, height, gender, race, primary cardiac diagnosis, and details of surgery performed were collected. Body mass index (BMI) was used to classify overweight, underweight, obesity, and severe obesity in our cohort. Overweight was defined as a BMI between the 85th and 94th percentile for age and sex, and obesity was defined as a BMI ≥ 95th percentile for age and sex as plotted on Centers for Disease Control and Prevention growth curves.Reference Barlow and Expert11 Underweight was defined as a BMI less than 5th percentile for age and sex. In addition, severe obesity was defined as having a BMI ≥ 120% of the 95th percentile or an absolute BMI ≥ 35 kg/m2, whichever was lower based on age and sex.Reference Kelly, Barlow and Rao12
Perioperative variables such as aortic cross-clamp time, cardiopulmonary bypass time, duration of hospitalisation, time to first negative fluid balance (defined as time from ICU admission to the end of the first 8-hour shift in which total output exceeded input), length of initial intubation (from the time of ICU admission to first time extubation), and presence of acute kidney injury (defined by the pRIFLE criteria) were collected.Reference Lex, Toth and Cserep13 The primary outcome was composite adverse outcome, defined as any of the following before discharge: death, need for cardiopulmonary resuscitation, mechanical circulatory support, arrhythmias (defined as the presence of haemodynamically significant supraventricular tachycardia, junctional ectopic tachycardia, or ventricular tachycardia/fibrillation that required medical intervention), culture positive surgical wound infection requiring chest exploration or antibiotics, acute neurologic injury (defined as the presence of stroke demonstrated on imaging or clinical seizures that required medical intervention), and acute kidney injury. The complexity of their cardiac diagnoses and operations was categorised using the Society of Thoracic Surgery and European Association for Cardiothoracic risk stratification system with category 1 being the lowest mortality risk and 5 being the highest mortality risk.Reference O’Brien, Clarke and Jacobs14
Statistical analysis
Data were analysed using SPSS software for PC (IBM SPSS Statistics, Version 22.0. Armonk, NY: IBM Corp). Results were expressed as mean ± standard deviation (SD) or median (interquartile range) for continuous variables and numbers (percentage) for nominal/categorical variables. Comparison of variables among groups was performed by chi square test, student t-test, and one-way analysis of variance (ANOVA) as appropriate. Logistic regression was used for multivariate analysis since our dependent variable, composite adverse outcome, was dichotomous. Significance was taken as a p value of <0.05.
Results
The study cohort (n = 149) had a mean SD age of 14.9 (SD) and 65% were males. The primary cardiac diagnosis was Tetralogy of Fallot (n = 34; 23%), followed by atrial septal defects (n = 16; 11%). The cohort had a mean (SD) BMI of 22.6 ± 6.5 and included 24 overweight (16.1%, mean (SD) BMI = 24.6 ± 2.4), 27 obese (18.1%, mean (SD) BMI = 32.8 ± 7.7), and 98 normal weight (65.8%, mean (SD) BMI = 19.3 ± 2.7) patients. We had three underweight patients during the study period and were excluded from the statistical analysis. The three groups were similar in demographics in regard to age, gender, race, complexity of CHD, and operation complexity. Utilising the ANOVA statistic, there was no difference among the groups in time to first negative fluid balance, time of mechanical ventilation, post-operative arrhythmia, and acute renal injury. However, there were significant differences among the groups in regard to cardiopulmonary bypass time, aortic cross-clamp time, length of stay in cardiac intensive care unit, and rate of post-operative wound infection (Tables 1 and 2).
Table 1. Comparison of the baseline characteristics among the three groups
Data are presented as n (%) for categorical variables and median (25th–75th percentiles) for continuous variables. IQR: inter quartile range, STAT: Society of Thoracic Surgery and European Association for Cardiothoracic risk stratification system. Numbers in bold represent statistically significant results; p value <0.05
* For comparisons in between all three groups
Table 2. Comparison of perioperative variables among the three groups
Data are presented as n (%) for categorical variables and median (25th–75th percentiles) for continuous variables. IQR: inter quartile range. Numbers in bold represent statistically significant results; p value < 0.05
* Obese compared to normal weight group
The intraoperative variables, obese patients had significantly longer cardiopulmonary bypass time (mean (SD) = 166 ± 116 min) compared to the normal weight group (mean (SD) = 111 ± 60 min); p = 0.004. Similarly, the obese group had longer aortic cross-clamp time (mean (SD) = 87 ± 97 min) compared to the normal weight group (mean (SD) = 46 ± 57 min; p = 0.022). Post-operative variables, the obese group had statistically significant longer duration of mechanical ventilation (mean (SD) = 20 ± 16 hours versus mean (SD) = 11 ± 12 hours; p = 0.017) and prolonged length of stay in cardiac intensive care unit (mean (SD) = 112 ± 85 hours versus mean (SD) = 70 ± 67 hours; p = 0.034) compared to the normal weight group. There were no deaths recorded in our cohort, although two patients who died during the study period were excluded from the study due to incomplete medical records, specifically several physician and nurse perioperative notes were not scanned into the electronic medical records in those two patients.
There were 27 (18.1%) patients who had composite adverse outcome. Table 3 presents the overall relationship between the composite adverse outcome and BMI in our cohort. Overweight and obese patients were at higher risk to have composite adverse outcomes compared with normal weight patients (odds ratio (OR): 2.9; confidence interval (CI): 1–8.5, p = 0.046 and OR: 3; CI: 1–8.5, p = 0.03; respectively). However, there was no significant difference between the overweight and obese groups in regard to the composite adverse outcome (p = 0.79). Figure 1 illustrates the relationship between the adverse outcome, overweight, and obese variables.
Table 3. Relationship between BMI and composite adverse outcome
* The P value is for the chi-square test for categorical variables. Data are presented as n (%) for categorical variables
Figure 1. Composite adverse outcome distribution by BMI.
Multivariate analysis by logistic regression including the variables cardiopulmonary bypass time, aortic cross clamp time, time to extubation, arrhythmias, and obesity revealed that obesity remained an independent predictor of adverse outcome in our population (p = 0.04).
A secondary analysis was performed with only two groups: 27 obese (18.1%, mean (SD) BMI = 32.8 ± 7.7) and 122 non-obese (81.9%, mean (SD) BMI = 20.3 ± 3.3) patients, which included our previous overweight and normal weight groups. Obese patients had longer bypass time, longer duration of mechanical ventilation, and longer hospital length of stay, compared to their non-obese counterparts, as shown in Table 4. In addition, obesity significantly increased the risk of post-operative wound infection (OR 7.14; 95% C.I.: 1.13–44.98, p = 0.037). Further subgroup analysis was performed for the patients who were severely obese (n = 11) within the obese group. Similarly, severely obese patients had higher risk for surgical wound infection (p = 0.005), longer cardiac intensive care stay (p < 0.001), longer hospital stay (p = 0.006), and longer need for mechanical ventilation (p = 0.007) compared to the rest of the cohort. There was no significant difference between the obese and severe obese groups in regard to composite adverse outcome.
Table 4. Obese patients compared to the rest of the cohort
* The P value is for the chi-square test for categorical variables and T test for continuous variables for between-group comparisons. Data are presented as n (%) for categorical variables and median (25th–75th percentiles) for continuous variables. IQR: inter quartile range, BMI: body mass index
Discussion
Previous studies have shown a prevalence of obesity in adolescent patients with CHD ranging at 17–26%.Reference Ray, Green and Henry4,Reference Tamayo, Manlhiot, Patterson, Lalani and McCrindle5,Reference Ogden, Carroll, Kit and Flegal15 In our cohort, 18.1% patients were obese, although there is selection bias as not all adolescents with CHD and obesity require surgical intervention. A few studies have investigated the short-term deleterious effect of obesity following cardiovascular surgery in the population with CHD. In the adult population, a cohort that included 71 post-operative patients after pulmonary valve replacement reported that obesity was independently associated with hospital length of stay >5 days (odds ratio (OR) = 5.2; 95% CI: 1.5–18.2, P = 0.01) and with increased post-operative arrhythmias (OR = 4.2; 95% CI: 1.7–40, P < 0 .01).Reference Buelow, Earing and Hill9 An analysis from the Society of Thoracic Surgeons Congenital Heart Surgery Database (STS), which included patients aged 10–35 years, revealed that obese (OR: 1.28 p = 0.008) and underweight subjects (OR: 1.39 p = 0.002) were independently associated with increased risk of adverse composite outcome.Reference O’Byrne, Kim and Hornik10
Obesity and short-term outcomes in General Paediatric Critical Care
The relationship between obesity and short-term outcome in the critically ill pediatric population has been investigated in recent years with conflicting results. Goh et al reviewed a cohort of 1030 patients aged 2- to 18-year-old admitted to a general paediatric intensive care unit. The risk-adjusted mortality rates in obese (OR, 0.68; 95% CI, 0.31–1.48) and severely obese patients (OR, 1.02; 95% CI, 0.45–2.34) were not significantly different compared to the normal weight group, and surprisingly obesity had a small protective effect. Interestingly, they defined obesity and severe obesity using only weight Z scores (1.66–2.33 and >2.33; respectively) for age and without taking into consideration the height of the patients.Reference Goh, Wakeham, Brazauskas, Mikhailov and Goday6 Brown et al. reported the effect of obesity in 316 paediatric trauma patients (6–19 years old). They reported no difference in mortality among obese and non-obese patients, but obese children had more complications (41% versus 22%, p = 0.04) and required longer intensive care unit stay (8 ± 9 versus 6 ± 6 days, p = 0.05).Reference Brown, Neville, Salim, Rhee, Cologne and Demetriades7 In a review of the virtual PICU systems database that included 127,607 patients, Ross et al. reported that risk-adjusted PICU mortality significantly increased as BMI increased into the overweight and obese ranges. Being overweight was independently associated with increased PICU mortality after controlling for severity of illness with the Pediatric Index of Mortality 2 score and pre-existing comorbidities.Reference Ross, Newth, Leung, Wetzel and Khemani8 We included height into the determination of obesity in our cohort by calculating BMI, and although, we did not have mortality in our study cohort, the short-term outcomes of obese and overweight patients were adversely affected.
Obesity and short-term outcomes in Paediatric Cardiac Critical Care
Shamszad et al. reported the association of obesity and post-operative outcomes after cardiothoracic surgery in the adolescent population. Utilising the Pediatric Health Information System database, they performed a case–control study including 629 obese or diabetic patients and 629 controls. The median age was 17 years (interquartile range (IQR) 12–32). Cases had longer paediatric intensive care unit length of stay than controls (3 versus 2 days, P = 0.001) The mortality was low and similar between cases and controls (2% versus 1.9%, P = 0 .692), and surgical complications occurred similarly between cases and controls (13.5% versus 12.4%, P = 0 .535).Reference Shamszad, Rossano, Marino, Lowry and Knudson16 In the analysis by O’Byrne et al of the STS Database (2010–2015), they included 18,337 patients, 10–35 years of age and found that obese (OR: 1.28 p = 0.008), severely underweight (OR: 1.29 p < 0.0001), and underweight subjects (OR: 1.39 p = 0.002) subjects were associated with increased risk of adverse composite outcome.Reference O’Byrne, Kim and Hornik10 Although we had similar outcomes, our cohort differs from these two previous studies in that we did not include adult patients or patients with diabetes since it could lead to confounders in the determination and measurement of the short-term effect of obesity in the adolescent population undergoing cardiothoracic surgery. In addition, given the small number of underweight patients during the study period (n = 3), no statistical analysis was performed in this small group, and they were excluded from the study cohort.
How obesity increases perioperative morbidity?
Obesity-related complications may have a significant impact in the perioperative period in adolescents. Obese adolescents have a higher prevalence of obstructive sleep apnea, which can increase pulmonary vascular resistance due to persistent hypoxia.Reference Daniels3 This would lead to difficulty in weaning the ventilator support after surgery, leading to prolong intensive care time and increased morbidity after cardiothoracic surgery. In addition, the chronic inflammatory state associated with obesity in children is another factor affecting the post-operative outcomes in this population.Reference Singer, Eng, Lumeng, Gebremariam and M Lee17 Our data show that the obese patients had increased risk of surgical wound infection after cardiothoracic surgery; similarly, Train et al. query the National Surgical Quality Improvement Program database from 2012 to 2014, including patients aged 2–17 years who underwent a surgical procedure. They reported that obesity was associated with a greater risk of wound complications (OR 1.24, 95% CI 1.13–1.36).Reference Train, Cairo, Meyers, Harmon and Rothstein18 Obese patients have relative hypoxia in the subcutaneous tissue that impairs wound healing.Reference Kabon, Nagele and Reddy19,Reference Sen20 Additionally, antibiotics have different pharmacokinetics and clearance in obese patients that may require dosing adjustment for the desired prophylactic or therapeutic effect.Reference Falagas and Karageorgopoulos21,Reference Martin, Saleem and Looke22
Obesity as an opportunity for improvement
Obesity is a known modifiable risk factor for various morbidities in adults. A recent study from a group in Philadelphia had reported the positive effect of exercise on BMI in the adolescent population with CHD and found that increasing exercise duration was associated with lower BMI (P = 0.01).Reference O’Byrne, McBride, Paridon and Goldmuntz23 The possibility of intervention into a modifiable risk factor before non-emergent cardiothoracic surgery represents a great opportunity to improve outcomes in this population. Further studies should evaluate if intervention in the preoperative period with a structured program aimed to decrease BMI can improve short-term post-operative outcomes in this population.
Our study has limitations including its retrospective nature and the limitation in data collection by the extent of documentation in the electronic medical record. However, this study is one of few to evaluate the relationship of obesity to short-term post-operative outcomes exclusively in the adolescent population growing with CHD. In addition, even though we controlled for complexity of repair and other variables, there is the possibility that other factors may be affecting the outcomes in this population; therefore, a direct cause and effect relationship is difficult to establish.
Conclusions
Obesity is associated with short-term adverse outcome and increases health resource utilisation in adolescents following CHD surgery. Our findings strengthen the importance of routinely encouraging a multidisciplinary approach to a healthy lifestyle in adolescent patients with CHD in order to prevent obesity. Further studies should evaluate if intervention in the preoperative period with a structured program aimed to decrease BMI can improve short-term post-operative outcomes in this population.
Acknowledgements
None.
Financial Support
This research received no specific grant from any funding agency, commercial, or not-for-profit sectors.
Conflicts of Interest
None.
