Copyright 2005 American Medical Association. All Rights Reserved. Applicable FARS/DFARS Restrictions Apply to Government Use.2005
To examine the relationship between childhood obesity as measured by body mass index (BMI) and long-term morbidity after an acute ankle sprain.
Six-month prospective cohort study with follow-up telephone questionnaires at 6 weeks and 6 months.
Cincinnati Children’s Hospital Medical Center emergency department.
Children between the ages of 8 and 18 years who presented with a chief complaint of an acute ankle injury were enrolled in the study. Children with ankle fractures were excluded. Exposed children were defined as those with a BMI in the 85th or greater percentile for age. Nonexposed children were defined as those with a BMI in the less than 85th percentile for age.
Main Outcome Measures
Persistent symptoms of pain, swelling, or weakness; pain during or after exercise; and recurrent ankle injury.
A total of 199 children were enrolled. The exposed and nonexposed groups were similar in terms of sex, age, and ethnicity. A total of 164 (93%) had complete follow-up at 6 months. Six months after an ankle injury, children with a BMI in the 85th or greater percentile for age were more likely to sustain persistent symptoms (relative risk, 1.70; 95% confidence interval, 1.10-2.61).
Overweight children are more likely to have persistent symptoms 6 months after an acute ankle sprain.
Childhood obesity is an epidemic in the United States.1 Today, 15% of children are obese, which is defined as a body mass index (BMI) of greater than the 95th percentile for age and sex, and nearly 25% are overweight (BMI between the 85th and 95th percentiles).2 The prevalence of children with a BMI in the 85th or greater percentile has doubled during the past 2 decades, with even higher rates among subpopulations of varying ethnicity and economically disadvantaged children.3
Obese children experience multiple chronic medical problems, including hypertension, type 2 diabetes mellitus, hypercholesterolemia, asthma, heart disease, and mental health concerns.4 Obesity in childhood is a strong predictor of adult obesity, and its comorbid conditions persist into adulthood.5,6 Relatively common comorbidities in obese children are orthopedic disorders such as Blount disease and slipped capital femoral epiphysis.7 In adults, obesity is a risk factor for spontaneous knee dislocation, hip fractures, and ankle fractures.8- 11 Furthermore, obesity is a risk factor for worse outcome following ankle fractures in adults.11,12
Previous work13- 15 has shown that regardless of an individual’s weight, an ankle injury carries significant morbidity. In adults, chronic ankle dysfunction persists in 40% of patients nearly 6 months after injury. In children, permanent morbidity was found 3 years later in 50% of children who sustained an ankle fracture and 23% of children who sustained an ankle sprain.16 However, to our knowledge, no studies have examined the relationship between childhood obesity and long-term morbidity following an acute ankle sprain. In this study, we hypothesized that 6 months after an acute ankle injury, overweight children would demonstrate greater ankle morbidity, defined as persistent pain, swelling, weakness, or recurrent ankle injury, than nonoverweight children.
This study was a prospective cohort study that compared long-term ankle morbidity (LTAM) between obese and nonobese children following an acute ankle sprain. The study was conducted in the emergency department (ED) at Cincinnati Children’s Hospital Medical Center, an urban tertiary care center with an annual patient census of close to 90 000 visits. The study was approved by the institutional review board.
Patients between the ages of 8 and 18 years who presented to the ED between March 2002 and September 2002 with a chief complaint of an ankle injury were eligible for inclusion in the study. During the study period, dedicated research assistants were present in the ED from noon until 11 PM 5 days a week to identify all potential study patients. Resident and fellow physicians familiar with the study also enrolled study patients in the absence of research assistants.
Patients were excluded if they were unable or unwilling to provide informed consent, were non–English speaking, were previously enrolled in the study, or did not have a telephone. Children who had sustained multiple trauma, children with neuromuscular disorders, and children with radiographically confirmed ankle fractures were excluded. Children diagnosed as having possible Salter-Harris I fractures of the tibia or fibula were enrolled in the study from the ED and were referred to the orthopedics department or their primary care physician in 1 week for reexamination. However, children were excluded from the study at the 6-week follow-up if they were subsequently diagnosed as having a Salter-Harris I fracture and placed in a cast.
Children who met study inclusion criteria were administered a survey to ascertain demographic information and information on previous ankle injuries. Height and weight were measured in the ED. The BMI was calculated for each child as a measure of weight in kilograms divided by the square of height in meters. Values were plotted on standardized BMI charts developed by the National Center for Health Statistics.17 Standard-of-care ankle injury treatment and discharge instructions were provided to children before leaving the ED. All participants were contacted 6 weeks and 6 months after the initial ED visit. A questionnaire was administered over the telephone by a single interviewer (N.L.T.) blinded to the patient’s BMI.
Our primary outcome was the presence of LTAM at 6 weeks and 6 months as reported by child or parent. These symptoms included persistent pain, swelling, or weakness; pain during or after exercise; or recurrent ankle injury.
Our explanatory variable of interest was BMI. Children were categorized into 3 percentile groups of BMI weight based on age and sex using cutoffs established in the literature18 : (1) normal weight (<85th percentile), (2) overweight (≥85th percentile), and (3) obese (≥95th percentile). Other independent variables included those with a possible association with ankle injuries. These included ethnicity, sex, age, and history of previous ankle injury.19- 21
Logistic regression (SAS statistical software, version 8.2; SAS Institute Inc, Cary, NC) was used to perform bivariate analysis between the outcome of LTAM and independent variables of ethnicity, sex, age, previous injury, or BMI. Independent variables with a significance level of P≤.10 were candidates for inclusion in a multivariable logistic model. Backward selection process was used, and P≤.05 was considered statistically significant in the final model.
Sample size calculation was based on literature that showed 40% morbidity in adults following an ankle injury.13 We estimated that approximately 60% of obese children would have continued problems at 6 months. Based on this difference, we calculated a sample size of 80 patients per group. Statistical significance was defined as P<.05.
During the study period, 239 patients were evaluated in the ED. One patient refused enrollment and 39 were improperly enrolled with ankle fractures. Therefore, 199 patients entered the study protocol (85 exposed and 114 nonexposed). Telephone follow-up at 6 weeks revealed 23 children who subsequently had casts placed at orthopedic follow-up for Salter-Harris I fractures, and 5 children were lost to follow-up. These children were excluded from further analysis. A total of 171 (97%) of 176 patients were successfully contacted at the 6-week follow-up. Of the 171 children who completed the protocol at 6 weeks, 76 were in the exposed group and 95 were in the nonexposed group (Figure).
Patient enrollment in the study. BMI indicates body mass index.
Baseline demographic characteristics were similar between the 2 groups with respect to age, ethnicity, sex, prior treatment, and physical activity at the time of injury (Table 1). Children with a BMI in the 85th or greater percentile were more likely to report a previous ankle injury compared with those children with BMIs in the less than 85th percentile (47% of the children in the exposed group reporting a previous ankle injury compared with 30% in the nonexposed group).
Six weeks after an ankle injury, there was a trend toward a positive association between a BMI in the 85th or greater percentile for age and increased risk for each of the outcome measures, with 39% of children complaining of pain with activity. At 6 weeks, children with a BMI in the 95th or greater percentile showed a statistically significant relationship between BMI and persistent pain, swelling, or weakness and a positive correlation with reinjury (although not statistically significant [P=.35]). A total of 52% of children with a BMI greater than the 95th percentile complained of pain, swelling, weakness, or reinjury 6 weeks following an ankle injury (relative risk [RR], 1.74; 95% confidence interval [CI], 1.09-2.41).
A total of 164 (93%) of 176 study patients were contacted 6 months after their ankle injury. The 12 children lost to follow-up at 6 months were similar to the 164 with verifiable follow-up data with respect to demographic characteristics and BMI percentiles. Table 2 shows the effect of each of our independent variables of interest alone on LTAM. Age was dichotomized into children younger than and 15 years or older, the average age of skeletal maturity.22 Despite being associated with both the exposure obesity and outcome of LTAM, sex, ethnicity, and previous injury were not found to confound the relationship between obesity and LTAM.
Children with a BMI in the 85th or greater percentile were more likely to have pain with activity (RR, 2.25; 95% CI, 1.25-4.02) and persistent swelling and weakness (RR, 2.40; 95% CI, 1.28-4.52). Overall, 31 (44%) of children with a BMI in the 85th or greater percentile were found to have persistent ankle symptoms at 6 months compared with 24 (26%) children with a BMI in the less than 85th percentile (RR, 1.70; 95% CI, 1.10-2.61) (Table 3). Similarly, almost half of the children with a BMI in the greater than 95th percentile had persistent symptoms at 6 months (21 or 46%) compared with 34 (29%) children with a BMI in the less than 95th percentile (RR, 1.58; 95% CI, 1.04-2.42) (Table 4).
Logistic regression was performed to examine the relationship between our independent variables of interest and LTAM. Nonwhite individuals were less likely to have long-term symptoms compared with white individuals (odds ratio [OR], 0.52; 95% CI, 0.27-1.01; P = .054). Children with a BMI in the 85th or greater percentile had a greater risk of having LTAM compared with children with a BMI in the less than 85th percentile (OR, 2.23; 95% CI, 1.15-4.31; P = .02). Similarly, children with a BMI in the greater than 95th percentile had a greater risk of having LTAM compared with children with a BMI in the less than 95th percentile (OR, 2.08; 95% CI, 1.03-4.20; P = .04). For every unit increase of BMI, the risk of having LTAM increased 0.66% (OR, 1.07; 95% CI, 1.02-1.12; P = .01). Sex, age, and previous injury did not significantly affect the outcome of LTAM. Ethnicity was no longer a predictor of LTAM after adjusting for BMI. Body mass index remained the only predictor of LTAM in the final logistic model (Table 5). Because of the high prevalence of LTAM in this population, the ORs are higher than the RRs, as expected. We would only expect the ORs to be similar to the RRs in a low prevalence population.
Our study demonstrates that overweight children are more likely to have persistent ankle morbidity 6 months following an acute ankle injury. A total of 44% of overweight children (BMI ≥85th percentile) had persistent symptoms at 6 months, whereas nearly half of obese children (BMI ≥95th percentile) had persistent ankle morbidity. In addition, we found that 34% of children in our population, regardless of BMI, had significant symptoms 6 months after an ankle injury, compared with 23% cited in previous work among athletic children.16
The greater the BMI, the more likely children were to have LTAM. As seen in the logistic regression model, BMI as a continuous or categorical variable showed this relationship. Ethnicity, sex, and previous injury were not found to be related to the outcome of LTAM once adjusted for BMI. Body mass index remained the only important predictor of LTAM in children following an acute ankle injury.
Our study had a number of strengths. Follow-up completion in this prospective cohort was 95.9% (164/171) at 6 months after presentation. To study a homogeneous ankle sprain group, every attempt was made to exclude all fractures from the population studied. Furthermore, the interviewer was blinded to the BMI of the patient during the telephone follow-up, thereby limiting interview bias.
Our study also had a number of limitations. This study was a convenience sample, and although every effort was made to enroll all children with ankle injuries, there were children with ankle injuries who were not approached for enrollment. In addition, children were not reexamined by a physician at 6 weeks and 6 months; however, previous research among athletes has shown that perceived clinical improvement is as useful as physical measures of organic dysfunction following an acute ankle sprain.23
In our study, ankle morbidity following an ankle sprain is a significant problem in children; however, overweight and obese children report greater symptoms 6 months after an ankle injury. Indeed, as BMI increased so did LTAM of ankle injuries. For this reason, it may be most important to target overweight and obese children for close follow-up and rehabilitation after an acute ankle injury. The next step is to explore the intervention of aggressive rehabilitation on the outcomes of LTAM in overweight and obese children.
Correspondence: Nathan L. Timm, MD, Division of Emergency Medicine, Cincinnati Children’s Hospital Medical Center, 3333 Burnet Ave, MLC 2008, Cincinnati, OH 45229-3039 (Nathan.firstname.lastname@example.org).
Accepted for Publication: July 22, 2004.
Timm NL, Grupp-Phelan J, Ho ML. Chronic Ankle Morbidity in Obese Children Following an Acute Ankle Injury. Arch Pediatr Adolesc Med. 2005;159(1):33-36. doi:10.1001/archpedi.159.1.33