Leptin Elevation as a Risk Factor for Slipped Capital Femoral Epiphysis Independent of Obesity Status

Abstract

Background:

Slipped capital femoral epiphysis (SCFE) is strongly associated with childhood obesity, yet the prevalence of obesity is orders of magnitude greater than the prevalence of SCFE. Therefore, it is hypothesized that obesity is not, by itself, a sufficient condition for SCFE, but rather one component of a multifactorial process requiring preexisting physeal pathology. Leptin elevation is seen to varying degrees in patients with obesity, and as leptin has been shown to cause physeal pathology similar to the changes seen in SCFE, we propose that leptin may be a factor distinguishing between patients with SCFE and equally obese children without hip abnormalities.

Methods:

Serum leptin levels were obtained from 40 patients with SCFE and 30 control patients with approximate body mass index (BMI) matching. BMI percentiles were calculated according to Centers for Disease Control and Prevention population data by patient age and sex. Patients were compared by demographic characteristics, leptin levels, odds of leptin elevation, and odds of SCFE.

Results:

The odds of developing SCFE was increased by an odds ratio of 4.9 (95% confidence interval [CI], 1.31 to 18.48; p < 0.02) in patients with elevated leptin levels, regardless of obesity status, sex, and race. When grouping patients by their obesity status, non-obese patients with SCFE showed elevated median leptin levels at 5.8 ng/mL compared with non-obese controls at 1.7 ng/mL (p = 0.006). Similarly, obese patients with SCFE showed elevated median leptin levels at 17.9 ng/mL compared with equally obese controls at 10.5 ng/mL (p = 0.039). Serum leptin levels increased in association with obesity (p < 0.001), with an increase in leptin of 0.17 ng/mL (95% CI, 0.07 to 0.27 ng/mL) per BMI percentile point.

Conclusions:

To our knowledge, this study is the first to clinically demonstrate an association between elevated serum leptin levels and SCFE, regardless of BMI. This adds to existing literature suggesting that SCFE is a multifactorial process and that leptin levels may have profound physiological effects on the development of various disease states. Despite a strong association with adiposity, leptin levels vary between patients of equal BMI and may be a vital resource in prognostication of future obesity-related comorbidities.

Level of Evidence:

Prognostic Level III. See Instructions for Authors for a complete description of levels of evidence.

Slipped capital femoral epiphysis (SCFE) is classified as either atypical or idiopathic¹. Atypical cases are associated with endocrinopathies such as hypothyroidism or hypogonadism, metabolic abnormalities such as renal osteodystrophy, or external causes such as radiation or chemotherapy². A majority of SCFE cases are idiopathic, which present in the absence of another preexisting diagnosis, and have been shown to be strongly associated with childhood obesity^3-7. The accepted pathogenesis presumes that increased biomechanical forces exceed the structural capacity of the physis, causing the epiphysis to translate, or slip, from its initial position, typically posteroinferiorly. As of 2012, 31.8% of American youth were overweight or obese⁸, yet the prevalence of SCFE is far less, ranging between 0.33 and 24.58 per 100,000 adolescents^9,10. This discrepancy between the prevalence of obesity and the prevalence of SCFE suggests that obesity is not, by itself, a sufficient condition for the development of idiopathic SCFE, but rather it is one component of a multifactorial process. The increased forces from obesity can only lead to a slip in the setting of preexisting physeal pathology.

This multifactorial theory is not new^11-13, yet in the absence of the above-mentioned atypical causes, clinical exploration for other factors that may have weakened the physis are rarely pursued. Histologic evaluation of physes in patients with SCFE show substantial distortion and widening in the columnar architecture of the zone of hypertrophy and zone of proliferation, with decreased cellularity, clustered degenerated chondrocytes, and irregular orientation and degradation of collagen fibers^13-19. The hormone leptin, which is secreted by adipose cells and primarily provides negative feedback at the hypothalamus to signal satiety, has been shown to cause similar physeal pathology when present in elevated levels in animal models^20-22. Obesity has been linked to leptin resistance, in which serum leptin levels must exceed the normal levels in non-obese peers to provide adequate satiety feedback, secondarily resulting in greater signaling at other sites of action^21,23. Leptin receptors have been identified on articular and physeal chondrocytes in animal and human studies^22,24-26. Animal studies have shown that, through these physeal receptors, leptin stimulates chondrocyte hypertrophy and proliferation, with alterations in leptin levels leading to an elongated physis, disturbed columnar structure, cellular apoptosis, and decreased expression and organization of the type-II and type-X collagen typically found throughout the hypertrophic zone^22,27-30.

Given the histologic similarities between physes affected by SCFE and hyperleptinemic physes, as well as the link between elevated leptin and body mass index (BMI)^31-33, we propose that leptin may be the source of physeal pathology necessary for the development of SCFE. This link has previously been suggested by in vitro work by Kishida et al., but, to our knowledge, no clinical studies have investigated the association between leptin and SCFE²⁷. The purpose of this study was to evaluate the role of leptin as a distinguishing factor between obese children who develop SCFE and equally obese children without hip abnormalities.

Materials and Methods

Study Population

All patients presenting to the orthopaedic service at Monroe Carell Jr. Children’s Hospital at Vanderbilt Medical Center, Nashville, Tennessee, between January 26, 2015, and January 26, 2016, with a new or previous diagnosis of SCFE were consecutively approached for study enrollment. Control patients were prospectively identified over the same time period on presentation to our service for elective non-obesity-related issues or from a pediatric weight management clinic after being identified as matching the age and approximate BMI percentile of an enrolled case patient. Parental consent and prospective enrollment were obtained for 40 patients with SCFE (age, 5 to 18 years) and 30 control patients (age, 8 to 18 years) in this institutional review board-approved, case-control study. Exclusion criteria were non-English speaking, preexisting systemic illness other than obesity, preexisting syndromic diagnosis, known atypical cause for SCFE, and inability to obtain serum leptin levels or demographic information.

Data Collection

Serum leptin levels were obtained between 8:00 a.m. and 4:00 p.m. for all patients. Demographic information was recorded, with BMI percentile defined using age and sex-adjusted values³⁴. Per the Centers for Disease Control and Prevention (CDC) definition, obesity was defined as a BMI of ≥95th percentile for children of the same age and sex³⁵. Using this definition, patients were placed into 4 groups: the non-obese control group, the non-obese SCFE group, the obese control group, and the obese SCFE group.

Statistical Analysis

Univariate linear regression was used to identify how patient demographic characteristics were associated with serum leptin. The association of demographic characteristics with the presence of elevated serum leptin was further evaluated through multivariate logistic regression. Assessment of how the presence of demographic factors affects the odds of developing SCFE was performed via logistic regression. As variation exists in published normal leptin values, we used our institutional laboratory definition of leptin elevation as values of >12.7 ng/mL³⁶. Demographic characteristics and leptin values were assessed for normality by the Shapiro-Wilk test, with only leptin shown to be nonparametric. To assess the appropriateness of matching, comparison of demographic characteristics between groups was performed by the Mann-Whitney test. Leptin levels were compared within subgroups via the Mann-Whitney test and across all subgroups via the Dunn multiple comparison test for nonparametric data. Significance was set at p < 0.05. All statistics were calculated using R (version 3.2.5; R Foundation for Statistical Computing).

Results

Demographic Characteristics

Comparisons of patient age, sex, race, and BMI percentile between groups are shown in Table I, with subgroup comparisons by obesity status shown in Table II. Although the racial breakdown among all cases was similar to that among all controls, there was a lower percentage of black patients in the non-obese SCFE group compared with the non-obese control group (p = 0.04), and a higher percentage in the obese SCFE group compared with the obese control group (p = 0.01). The remaining demographic categories were similar between groups, suggesting that matching was adequate between cases and controls.

TABLE I.

Comparison of Demographic Characteristics Between the Control Group and the SCFE Group

	Control Group (N = 30)	SCFE Group (N = 40)	P Value
Age* (yr)	13.4 ± 2.6	12.8 ± 2.5	0.225
Sex†			0.668
Male	18	26
Female	12	14
Race†			0.457
Black	8	14
White	22	26
BMI percentile*‡ (%)	76.5 ± 28.6	90.7 ± 14.5	0.018

^*The values are given as the mean and the standard deviation.

^†The values are given as the number of patients.

^‡The SCFE group had a significantly higher BMI percentile than the control group at p < 0.05; no other demographic comparisons showed significant differences.

TABLE II.

Comparison of Subgroup Demographic Characteristics Between the Control Group and the SCFE Group by Obesity Status^*

	Control Group	SCFE Group	P Value
Age† (yr)
Non-obese	13.3 ± 2.6	13.6 ± 2.7	0.701
Obese	14.1 ± 1.9	12.1 ± 2.5	0.133
Sex‡
Non-obese			0.358
Male	10	10
Female	8	4
Obese			0.761
Male	8	16
Female	4	10
Race‡§
Non-obese			0.040#
Black	7	1
White	11	13
Obese			0.013#
Black	1	13
White	11	13
BMI percentile† (%)
Non-obese	62.1 ± 29.1	76.8 ± 17.5	0.110
Obese	98.0 ± 1.4	98.2 ± 1.1	0.890

^*There were 18 patients in the non-obese control group, 14 patients in the non-obese SCFE group, 12 patients in the obese control group, and 26 patients in the obese SCFE group.

^†The values are given as the mean and the standard deviation.

^‡The values are given as the number of patients.

^§The non-obese control group and the obese SCFE group had significantly higher proportions of black patients than their comparison groups; no other demographic comparisons showed significant differences.

^#Significant.

Role of Demographic Characteristics in Leptin Elevation

Across all patients, linear regression showed that serum leptin levels increased in the setting of obesity, as expected. Patients with a BMI in the ≥95th percentile were shown to have an expected serum leptin level 9.86 ng/mL (95% confidence interval [CI], 5.69 to 14.03 ng/mL; p < 0.001) higher than their non-obese counterparts, with expected increases per BMI percentile point of 0.17 ng/mL (95% CI, 0.07 to 0.27 ng/mL; p < 0.001) when controlling for race and sex. The association between demographic characteristics and serum leptin elevation was assessed via a multivariate logistic odds model, showing that the odds of having elevated leptin was 10.32 times greater in obese patients than non-obese patients, and 3.56 times greater in female patients than in male patients, but race showed no significant association (p = 0.868) (Table III). Bilateral slipped capital femoral epiphysis was present in 10 patients, with 5 in both the obese and non-obese groups. There were no significant differences (p > 0.05) between patients with bilateral SCFE and those with unilateral SCFE with regard to BMI or leptin levels, although results of this analysis may be limited by the small number of patients with bilateral SCFE and the variable duration of patient follow-up, potentially missing the development of bilateral disease in patients previously treated unilaterally.

TABLE III.

Likelihood of Having Elevated Leptin

Factor	OR*	P Value
Sex (female vs. male)	3.56 (1.02 to 12.40)	0.046†
Race (white vs. black)	0.90 (0.26 to 3.13)	0.868‡
Obesity	10.32 (2.75 to 38.77)	0.001†

^*The values are given as the OR, with the 95% CI in parentheses.

^†Increased odds of elevated leptin (p < 0.05) were shown with female sex and obesity.

^‡Race showed no significant association.

Factors Predictive of SCFE

Univariate logistic regression comparing SCFE diagnosis with leptin elevation, obesity status, race, and sex showed a significant increase in the odds of SCFE diagnosis for obese patients (odds ratio [OR], 2.79 [95% CI, 1.05 to 7.40]; p = 0.04) and those with elevated leptin (OR, 5.00 [95% CI, 1.60 to 15.68]; p < 0.001). Race and sex showed no significant association (p > 0.05). All factors were then combined in a multivariate logistic regression, in which leptin elevation remained the only patient factor associated with a significantly increased odds of developing SCFE (OR, 4.93 [95% CI, 1.31 to 18.48]; p < 0.02), with obesity, sex, and race not contributing to the likelihood of developing SCFE (Table IV).

TABLE IV.

Likelihood of Developing SCFE

Factor	OR*	P Value
Sex (female vs. male)	0.49 (0.15 to 1.59)	0.238†
Race (white vs. black)	0.71 (0.22 to 2.26)	0.556†
Obesity	1.50 (0.49 to 4.57)	0.472†
Elevated leptin	4.93 (1.31 to 18.48)	0.018‡

^*The values are given as the OR, with the 95% CI in parentheses.

^†There was no significant association between the odds of SCFE and obesity, sex, or race.

^‡The increased odds of SCFE diagnosis were shown to be significant only with leptin elevation when incorporating all variables into the model.

Leptin Elevations Among Patients with SCFE

A comparison of serum leptin levels between the SCFE group and the control group is shown in Figure 1-A and Table V, with the subgroup comparison by obesity status shown in Figure 1-B and Table VI. The median serum leptin levels were significantly higher (p < 0.001) in the SCFE group at 13.0 ng/mL (interquartile range [IQR], 6.3 to 20.4 ng/mL) than the control group at 2.4 ng/mL (IQR, 1.0 to 10.4 ng/mL). In a subgroup comparison, the non-obese SCFE group showed elevated median leptin levels at 5.8 ng/mL (IQR, 1.0 to 14.0 ng/mL) compared with the non-obese control group at 1.7 ng/mL (IQR, 0.4 to 2.9 ng/mL) (p = 0.006); the obese SCFE group also showed elevated levels at 17.9 ng/mL (IQR, 11.1 to 22.5 ng/mL) compared with the obese control group at 10.5 ng/mL (IQR, 3.2 to 19.7 ng/mL) (p = 0.039). The mean serum leptin of our non-obese control group is similar to previously published values, as shown in the Appendix, which serves to ensure that our control group is in concert with the existing literature^37-42. Similar data for obese controls were not included because of a lack of consistent obesity definitions preventing accurate comparison with the obese groups in our study^42-45. When comparing leptin levels between all subgroups via the Dunn multiple comparison test, all comparisons between the 4 groups were significant (p < 0.05), with the exception of the comparison of the non-obese SCFE with the obese control group (p = 0.16).

Fig. 1.

Serum leptin levels compared according to diagnosis (Fig. 1-A) and obesity status and diagnosis (Fig. 1-B). The median values with IQRs are shown via bar-and-whisker plots for all groups. The SCFE group showed a significantly higher median leptin level (p < 0.001) at 13.0 ng/mL (IQR, 6.3 to 20.4 ng/mL) when compared with the control group at 2.4 ng/mL (IQR, 1.0 to 10.4 ng/mL). When grouped by obesity status, the non-obese SCFE group showed significantly higher median leptin levels (p = 0.006) at 5.8 ng/mL (IQR, 1.0 to 14.0 ng/mL) than the non-obese control group at 1.7 ng/mL (IQR, 0.4 to 2.9 ng/mL). There were similar findings (p = 0.039) between the obese SCFE group at 10.5 ng/mL (IQR, 3.2 to 19.7 ng/mL) and the obese control group at 17.9 ng/mL (IQR, 11.1 to 22.5 ng/mL). A single asterisk indicates significance at p < 0.05, a double asterisk indicates significance at p < 0.01, and a triple asterisk indicates significance at p < 0.001.

TABLE V.

Serum Leptin Levels by Group^*

Serum Leptin Values	Control Group (N = 30)	SCFE Group (N = 40)	P Value
Mean†	5.9 ± 7.3 (0.4 to 24.4)	14.9 ± 11.1 (0.4 to 53.5)	<0.001
Median‡	2.4 (1.0 to 10.4)	13.0 (6.3 to 20.4)	<0.001

^*Patients with SCFE were shown to have higher mean and median serum leptin levels compared with control patients.

^†The values are given as the mean and the standard deviation, with the range in parentheses, in nanograms per milliliter.

^‡The values are given as the median, with the IQR in parentheses, in nanograms per milliliter. As leptin data were nonparametric, the median and the IQR are the preferred descriptors (the mean leptin level was included for completeness).

TABLE VI.

Serum Leptin Levels by Obesity Subgroup

	Control Group	SCFE Group	P Value
Non-obese subgroup*	18	14
Mean†	2.3 ± 2.7 (0.4 to 11.3)	8.1 ± 7.5 (0.4 to 22.5)	0.002
Median‡	1.7 (0.4 to 2.9)	5.8 (1.0 to 14.0)	0.006
Obese subgroup*	12	26
Mean†	11.4 ± 8.6 (1.0 to 24.4)	18.6 ± 11.1 (3.8 to 53.5)	0.028
Median‡	10.5 (3.2 to 19.7)	17.9 (11.1 to 22.5)	0.039

^*The values are given as the number of patients.

^†The values are given as the mean and the standard deviation, with the range in parentheses, in nanograms per milliliter. Patients with SCFE were shown to have higher mean and median serum leptin levels compared with control patients when grouped by obesity status.

^‡The values are given as the median, with the IQR in parentheses, in nanograms per milliliter. As leptin data were nonparametric, the median and the IQR are the preferred descriptors (the mean leptin was included for completeness).

Discussion

Our work highlights a previously unrecognized clinical link between elevated leptin levels and the development of SCFE, regardless of obesity status. When incorporating serum leptin into factors known to be associated with SCFE (race, sex, and BMI percentile), only serum leptin maintained a significant association with the development of SCFE, with hyperleptinemic patients showing an increased odds of SCFE at 4.93 times greater than patients with normal leptin levels.

With regard to the association between demographic characteristics and our outcomes, female sex was associated with greater odds of leptin elevation compared with male sex by a factor of 3.6, which parallels but exceeds the previously published increased leptin levels in girls compared with boys^46,47, but did not translate into increased odds of SCFE. This is consistent with reported male predominance of SCFE⁴⁸, which has been linked to the anabolic effects of testosterone during puberty in males causing an elongated, weaker physis compared with female patients, where estrogen directly induces growth plate closure and therefore stability¹¹. Race did not alter leptin levels, consistent with the majority of previous research^49-51, although discrepant race associations with leptin have been reported^52,53. Race also did not affect SCFE prevalence in our cohort, which differs from reported relative SCFE frequencies ranging from 2.2 to 4.6 times greater in black patients compared with white patients^9,10,48,54.

Of particular interest to highlight are the 14 patients with idiopathic SCFE despite not meeting criteria for obesity. Although their mean BMI percentile of 76.8 is still elevated compared with a theoretical perfect non-obese control population with a BMI percentile of 50.0, it was not significantly different from the non-obese control group mean BMI percentile of 62.1. Additionally, despite being elevated, the mean BMI percentile for this group does not even meet CDC criteria for overweight (BMI of the ≥85th percentile for age and sex)³⁵. This non-obese SCFE group did not show a significant difference in leptin levels from obese patients without SCFE, which is in concert with our findings that although obesity was shown to be strongly associated with an increase in leptin, the diagnosis of SCFE shows that same association regardless of BMI. Of these non-obese patients with SCFE, the most interesting were the 2 with the highest leptin levels, a 12-year-old white female patient with a BMI percentile of 83 and a leptin level of 22.5 ng/mL, and a 13-year-old white male patient with a BMI percentile of 81 and a leptin level of 20.5 ng/mL. Both patients, although close to reaching overweight criteria, had elevated leptin levels far exceeding what is expected from increased adiposity alone.

To our knowledge, we are among the first to describe patients with abnormally elevated leptin levels in the absence of obesity or a known additional diagnosis. A comprehensive literature search showed that among non-obese patients, differences in leptin levels have been shown with multiple other endocrine conditions (decreased in hypothyroidism⁵⁵ and elevated in precocious puberty^56,57, hyperinsulinemia⁵⁷, elevation of thyroid autoantibodies⁵⁸, idiopathic intracranial hypertension⁵⁹, and hypothalamic dysfunction⁶⁰), pulmonary conditions (decreased in smokers^32,61 and elevated in asthmatics⁶², adults with impaired lung function⁶³, obesity hypoventilation syndrome⁶⁴, and obstructive sleep apnea⁶⁵), and multiple myeloma⁶⁶ and in the setting of elevated uric acid⁶⁷. Despite this extensive list, the magnitude of difference in leptin levels between cases and unaffected controls in these studies was usually only a few nanograms per milliliter, far smaller than the variations we have presented herein.

Although in our study the mean patient age at the time of SCFE diagnosis fits published norms, the age range of our patients with SCFE at the time of laboratory acquisition, from 5 to 18 years of age, exceeded the typically reported range of 12 to 15 years at idiopathic SCFE presentation⁴⁸ and deserves further discussion. Of note, limiting our analysis to only patients who were 12 to 15 years of age did not alter the significance of our results. The single patient with SCFE who was older than the standard range was 18 years of age at the time of enrollment and acquisition of serum leptin levels, but had been diagnosed with SCFE at the age of 15 years, fitting the expected age of idiopathic SCFE. The 2 patients with idiopathic SCFE who were younger than the expected range were particularly notable, as they presented at the surprisingly young ages of 5 and 9 years, both with BMI of >95th percentile by age. Given the recommendation of evaluation for endocrine and metabolic aberrations for patients with SCFE who were younger than 10 years (for girls) or 12 years (for boys)¹², thorough serology work-up was performed and both patients were shown to have no atypical sources for SCFE. Despite their expected serum leptin levels based on their age and sex being <5 ng/mL, both patients had leptin levels of >20 ng/mL. Notably, both patients eventually developed bilateral SCFE, with the 9-year-old patient initially presenting after the second side became symptomatic, and the 5-year-old patient developing a contralateral unstable slip 2 years after fixation of the initial side, which at that point had already progressed to osteonecrosis.

The limitations to our study included the lack of exact 1:1 matching by age, sex, race, and obesity status, although demographic information showed significant differences between the groups only when looking at race. The disproportionate racial distribution of our cohort may explain why our findings regarding the role of race on leptin levels and SCFE risk do not match previous reports. The only races represented in our study were white and black, leaving several racial groups unrepresented. Additionally, with the exception of patients whose age alone prompted laboratory evaluation for atypical causes of SCFE, no formal investigations beyond history and physical examination were pursued to evaluate for other medical conditions that could alter leptin levels or put patients at risk for SCFE. Of note, history and physical examination did adequately identify 5 patients with previously undiagnosed endocrinopathies, prompting exclusion from the study. Lastly, the study was from a single institution, with enrollment of cases after SCFE had already occurred. With regard to inter-laboratory measurement issues, although the use of leptin values from a single institution avoids the reported variability of leptin reference ranges between institutions, this variability is a recognized limitation itself when combining our results with those from an earlier study³⁶.

Despite the diurnal profile of serum leptin levels⁶⁸, we do not believe that the lack of standardization in the timing of blood draws is a limitation. All values were obtained between 8:00 a.m. and 4:00 p.m., and although serum leptin does spike between midnight and 8:00 a.m. to roughly 25% greater than the 8:00 a.m. value, it shows a relatively constant level from 8:00 a.m. to midnight, with <10% deviation, which is far less than the intergroup differences that we identified⁶⁸. Although non-obese patients have a higher proportional change in this diurnal profile, the reported differences of <2 ng/mL between morning and evening values do not explain our findings⁶⁹. Additionally, some patients had laboratory values obtained in clinic without a nil per os (NPO) period, but this should not alter our results as food intake yields no acute alteration in serum leptin levels⁷⁰. Lastly, the age at laboratory evaluation was as much as 3 years after the initial diagnosis of SCFE, as was the case of the 18-year-old patient discussed above. Although serum leptin does change somewhat throughout youth and adulthood⁷¹, the expected changes in youth are based on the progression of puberty, with negligible differences within a single pubertal stage. Of several studies showing normal serum leptin levels throughout puberty^33,40,46,47, the greatest reported difference between prepuberty and postpuberty is a drop of 2.9 ng/mL in boys⁴⁷ and a rise of 3.7 ng/mL in girls⁴⁶, with much smaller differences between adjacent puberty stages. Accordingly, we believe that the time between SCFE diagnosis and laboratory acquisition does not detract from our findings. To ensure that a change in obesity between these 2 time points did not alter our results, patients with delay between SCFE and laboratory acquisition underwent post hoc comparison of BMI at both time points, without any significant changes identified.

Our results support the multifactorial pathogenesis of SCFE, while acknowledging its clear connection with obesity. The development of SCFE requires preexisting physeal pathology to weaken the physis, allowing the increased biomechanical forces from obesity to cause a slip. For patients diagnosed with atypical SCFE, the cause of their physeal pathology has been recognized as an endocrinopathy or other existing diagnosis. We propose that idiopathic SCFE similarly requires another diagnosis to explain the physeal pathology and that diagnosis very well may be hyperleptinemia, regardless of obesity status or severity. Recognition of this abnormality in orthopaedic patients and referral for early management of associated comorbidities may be vital to prevent subsequent orthopaedic and non-orthopaedic complications. Elevated leptin has been shown to be a marker for the risk of future cardiovascular disease in non-obese healthy children and is associated with future obesity⁷². Additionally, leptin has been repeatedly shown to be a key mediator in the pathophysiology of hypertension^73-77, which our group has previously shown to cause abnormal physeal changes similar to those in SCFE, and patients with SCFE have a significantly increased prevalence of undiagnosed hypertension compared with equally obese control patients⁷⁸.

In conclusion, the findings of this study demonstrate an association between elevated serum leptin levels and SCFE, regardless of BMI. This adds to existing literature suggesting that SCFE is a multifactorial process, leptin elevation has variations between patients regardless of obesity status, and leptin levels may have profound physiological effects on the development of various disease states. Clinically, recognition of this link may lead to improved patient prognostication for detrimental orthopaedic and non-orthopaedic conditions in patients found to have elevated serum leptin levels.

Appendix

A table showing serum leptin levels among healthy non-obese controls from prior studies is available with the online version of this article as a data supplement at jbjs.org (http://links.lww.com/JBJS/C862).