The Effect of Body Mass Index on Outcomes of Isolated Medial Patellofemoral Ligament Reconstruction

Abstract

Background:

The effect of body mass index (BMI) on complication risk, recurrent instability risk, and patient-reported outcomes (PROs) after surgical intervention for recurrent patellar instability is unclear.

Purpose/Hypothesis:

The purpose was to evaluate the differences in complications, recurrence, and PROs in obese and nonobese patients undergoing isolated medial patellofemoral ligament reconstruction (MPFLR). It was hypothesized that obesity is associated with increased complication risk, increased risk of recurrent patellar instability, and poorer PROs after MPFLR.

Study Design:

Cohort study; Level of evidence, 3.

Methods:

A retrospective review identified all patients who underwent isolated MPFLR over an 8.5-year period at an academic medical center. Records were reviewed for demographic, physical examination, radiographic, surgical, and clinical outcome data. Patients were contacted to collect PROs, including the Norwich Patellar Instability score, Marx activity scale score, and Knee injury and Osteoarthritis Outcome Score (KOOS). Patients were stratified by BMI (<30 and ≥30 kg/m² for primary analysis and then by ≥35 kg/m² for secondary analysis) and complications and outcomes were compared. Regression analysis was then performed to evaluate the effects of increased BMI on PROs.

Results:

The records of 107 patients were analyzed in this investigation. Complication rates were similar across groups. Patients with a BMI ≥35 kg/m² demonstrated a lower Marx activity scale score compared with those with a BMI <30 kg/m² (P = .039). Regression analysis demonstrated no association between BMI and PROs adjusting for age, sex, and articular cartilage damage.

Conclusion:

No significant differences in complications or repeat dislocation risk after isolated MPFLR were noted based on BMI ≥30 or <30 kg/m². Patients with a BMI ≥35 kg/m² demonstrated lower activity level, but no other differences in PROs compared with patients with a BMI <30 kg/m².

Patellar dislocations are a common orthopaedic injury frequently affecting young, athletic individuals. The reported incidence of first-time dislocations ranges from 2.3 to 23.2 per 100,000 individuals annually.^15,22 While nonoperative management has often been the standard of care for first-time patellar dislocations, growing evidence has demonstrated high rates of recurrent instability after nonoperative management.^4,9 Recurrent instability is generally treated surgically.

Surgical intervention for patellar instability often involves medial patellofemoral ligament reconstruction (MPFLR). ¹⁴ Little is currently known regarding the outcomes of MPFLR in obese patients. Historically, surgeons may have approached this population with hesitancy or concern for poor outcomes, given that obesity has been associated with increased complications after other orthopaedic procedures.^5,7,8,21,24

The purpose of this study was to evaluate the effect of obesity on the risk of postoperative complications (deep vein thrombosis [DVT], infection, pain, repeat dislocation/instability, and stiffness), recurrent instability, and poorer patient-reported outcomes (PROs). We hypothesized that obesity would be associated with increased complication risk, increased risk of recurrent instability, and poorer PROs.

Methods

Study Design and Evaluation Methods

After receiving institutional review board approval, a retrospective review was performed to identify patients who underwent MPFLR by 1 of 4 sports medicine fellowship-trained orthopaedic surgeons (including D.C.F., C.C.K., R.A.M.) at a tertiary academic medical center between 2008 and 2016. Patients undergoing concomitant bony procedures such as tibial tubercle osteotomy were excluded. No age-based exclusion criteria were used, but all patients treated surgically at this institution were at least 13 years of age. Chart review identified 200 knees in 185 patients who underwent MPFLR between 2008 and 2016. Of these 200 knees, 35 (33 patients) underwent additional bony procedures and were excluded from the study. A total of 165 knees in 152 patients were eligible for study inclusion.

Chart review was performed to collect demographic, surgical, and radiographic details of the included patients. Demographic variables included age at surgery, sex, height, and weight. Preoperative imaging was reviewed. Patellar height was assessed using the Caton-Deschamps index (CDI) on lateral plain radiographs at 30° of knee flexion. ¹ Trochlear dysplasia was evaluated on magnetic resonance imaging (MRI) scans via sulcus angle measurement, ² and tibial tuberosity–trochlear groove distance was measured on MRI as previously described. ¹⁶ Operative reports were reviewed to collect perioperative and surgical details (graft type and additional procedures performed). Articular cartilage status of the patellofemoral compartment was assessed at arthroscopy and quantified using the Outerbridge classification system. ¹²

Surgical Indications and Technique

All patients who experienced recurrent patellar instability during this period were treated with MPFLR. Recurrent patellar instability was defined as at least 2 patellar dislocations or persistent symptoms of patellar subluxation and instability with a positive patellar apprehension test after nonoperative treatment of a first-time patellar dislocation. Only patients who underwent isolated MPFLR without bony procedures were included in this study. The decision to perform a concomitant bony procedure was made on a case-by-case basis using a shared decision-making model between the attending surgeon and patient. General indications for the addition of a bony procedure were the presence of a J-sign or persistent apprehension in high degrees of knee flexion in the setting of an elevated tibial tuberosity–trochlear groove distance (>20 mm), patella alta (CDI >1.20), or high-grade trochlear dysplasia.

All procedures were performed by 1 of 4 sports medicine fellowship-trained orthopaedic surgeons. Diagnostic arthroscopy was performed, and any additional intra-articular pathology was addressed. MPFLR was performed according to each surgeon’s preferred technique with the goal of anatomic graft placement without tension. Patellar fixation was achieved using 1 or 2 anchors, and femoral fixation was performed using an interference screw in all cases. Femoral tunnel position was determined using anatomic and radiographic (Schöttle point) landmarks, ¹⁷ and patellar attachment was performed on the proximal half of the patella. Both allograft and autograft tissue were used according to surgeon preference.

Rehabilitation Protocol

All patients were instructed to initiate formal physical therapy within 1 week of surgery. ¹⁰ Brace use and weightbearing restrictions were per surgeon discretion. Rehabilitation consisted of superior and medial patellar glides; quadriceps, hamstring, core, and hip abductor strengthening; active and passive range of motion exercises; and gait retraining.

Outcome Assessment

Chart review of all eligible patients was undertaken to identify their last clinic follow-up data and to identify any documented episodes of recurrent patellar dislocations. Patients were subsequently contacted by mail and/or telephone with a short series of questions regarding subsequent treatment and injuries to their knee as well as PRO assessments. Patellar instability was assessed according to the Norwich Patellar Instability (NPI) score, with higher NPI scores correlating with greater instability. ¹⁹ Pain, symptoms, function, and quality of life were assessed using the Knee injury and Osteoarthritis Outcome Score (KOOS), with lower scores indicating poorer outcomes. ¹³ Activity level in the past year was evaluated using the Marx activity scale score. ¹¹

Patients were categorized by body mass index (BMI) status, with ≥30 kg/m² representing obesity and ≥35 kg/m² representing class 2 obesity according to the World Health Organization (WHO) definitions. Complication risk, risk of recurrent instability, and PROs were compared between groups based on BMI ≥30 kg/m² versus <30 kg/m². Subgroup analysis was performed to evaluate the outcomes of patients with class 2 obesity (BMI ≥35 kg/m²). Finally, the effect of BMI as a continuous variable on PROs controlling for age, sex, and articular cartilage status was performed using regression models.

Statistical Analysis

Summary statistics (means, standard deviations, and proportions) were calculated for patient characteristics and outcomes for the 2 groups. Complications and PRO measures (PROMs) between groups were compared using t tests and Fisher exact tests when appropriate after confirmation that continuous variable distributions were normal using the Kolmogorov-Smirnov test. Multivariable linear regression modeling was then performed using BMI as a continuous variable to evaluate the relationship between BMI and PROs controlling for patient age, sex, and cartilage status. Parameter estimates (linear regression) with 95% confidence intervals, as well as P values, were reported from multivariable regression analyses.

Pilot data suggested that between 25% and 30% of patients in the study would be in the group with a BMI ≥30 kg/m². A power analysis demonstrated that 90 patients with a 2.5 to 1 distribution between groups (64 in one group and 26 in the other) would be required to achieve 80% power to detect a clinically meaningful 10-point difference in KOOS values between groups with an alpha of .05 assuming a standard deviation of 15 points in the score distributions.^20,23

Results

A total of 107 patients (114 knees) (69.1%) completed a minimum follow-up of 2 years and comprised the study population. The mean follow-up time was 5.2 years after surgery. PROMs were completed by 88 eligible patients (95 knees) (82.2%) (Figure 1).

Figure 1.

Flowchart of patient inclusion in the study. MPFLR, medial patellofemoral ligament reconstruction; PROM, patient-reported outcome measure.

The overall mean BMI for the study population was 26.8 kg/m². The final patient cohort included 78 patients (83 knees) with a BMI <30 kg/m² and 29 patients (31 knees) with a BMI ≥30 kg/m². The final cohort included 74 knees (64.9%) in female patients and 40 knees (35.1%) in male patients. The mean patient age at surgery was 24.7 years (range, 13-57 years). No significant differences in patient characteristics were noted between the 2 BMI groups (Table 1).

Table 1.

Demographics of Patients With BMI <30 and ≥30 kg/m^{2 a}

	BMI, kg/m2		P Value
	<30	≥30
	(n = 83)	(n = 31)
Age, y	25.0 ± 10.1	24.0 ± 7.1	.92
Female, n (%)	53 (63.9)	21 (67.7)	.82
Dislocation history, n (%)			.81
1 or 2 previous dislocations	35 (42.2)	13 (41.9)
3-5 previous dislocations	14 (16.9)	5 (16.1)
>5 previous dislocations	30 (36.1)	10 (32.3)
Unknown/not specified	4 (4.8)	3 (9.7)
Age at first dislocation, y	16.0 ± 4.8	17.8 ± 5.8	.27
Interval from first dislocation until surgery, y	8.2 ± 9.9	6.0 ± 5.8	.28
BMI, kg/m2	23.7 ± 3.4	34.8 ± 3.8	NA

^aValues are reported as mean ± SD unless otherwise noted. BMI, body mass index; NA, not applicable.

Preoperative radiographs were available in 101 knees (88.6%). The mean CDI for all knees was 1.10. Preoperative MRI scans were available for 80 knees (70.2%). Articular cartilage damage of Outerbridge grade 3 or 4 was noted on the patella in 55 knees (48.2%) and on the trochlea in 5 knees (4.4%). Most lesions were noted to be medial on the patella and lateral on the trochlea. No statistically significant differences in imaging or perioperative factors were noted between the 2 groups (Table 2).

Table 2.

Imaging and Perioperative Findings of Patients With BMI <30 and ≥30 kg/m^{2 a}

	BMI, kg/m2		P Value
	<30	≥30
	(n = 83)	(n = 31)
CDI	1.12 ± 0.18	1.15 ± 0.15	.64
Crossing sign, n (%)	11 (13.3)	3 (9.7)	.60
ST spur, n (%)	8 (9.6)	4 (12.9)	.61
TT-TG distance, mm	17.2 ± 4.9	18.1 ± 5.2	.53
Sulcus angle, deg	146.2 ± 8.5	143.0 ± 10.5	.16
Sulcus depth, mm	4.1 ± 1.5	4.4 ± 1.6	.46
Patellar tilt, deg	11.6 ± 7.2	9.9 ± 6.3	.39
Patellotrochlear index, %	48 ± 10	50 ± 10	.58
Patellar cartilage grade, n (%)			.98
0-2	43 (51.8)	16 (51.6)
3 or 4	40 (48.2)	15 (48.4)
Patellar defect size, cm2	2.6 ± 1.7	3.5 ± 2.8	.57
Trochlea cartilage grade, n (%)			.61
0-2	80 (96.4)	29 (93.5)
3 or 4	3 (3.6)	2 (6.5)
Trochlea defect size, cm2	2.0 ± 1.4	2.3 ± 1.5	.79
Graft source, n (%)			.84
Autograft	23 (27.7)	8 (25.8)
Allograft	60 (72.3)	23 (74.2)

^aValues are reported as mean ± SD unless otherwise noted. BMI, body mass index; CDI, Caton-Deschamps index; ST, supra-trochlear; TT-TG, tibial tuberosity–trochlear groove.

The overall complication incidence for the group included: 3 postoperative DVTs (2.6%), 13 cases (11.4%) requiring reoperation, 22 cases (19.3%) with persistent pain, and 9 cases (7.9%) with postoperative stiffness. Subjective postoperative patellar instability was reported in 39 knees (34.2%), and recurrent dislocation was noted in 5 knees (4.4%). There were no statistically significant differences in complication incidences between groups (Table 3).

Table 3.

Clinical Outcomes of Patients With BMI <30 and ≥30 kg/m^{2 a}

	BMI, kg/m2		P Value
	<30	≥30
	(n = 83)	(n = 31)
DVT	2 (2.4)	1 (3.2)	.81
Reoperation	7 (8.4)	6 (19.4	.13
Hardware removal, pain	1 (1.2)	1 (3.2)
Infection	1 (1.2)	0 (0)
MUA, lysis of adhesions	5 (6.0)	3 (9.7)
Other cause	0 (0)	2 (6.5)
Postoperative pain	14 (16.9)	8 (25.8)	.30
Postoperative stiffness	5 (6.0)	4 (12.9)	.25
Repeat instability	26 (31.3)	13 (41.9)	.38
Repeat dislocation	3 (3.6)	2 (6.5)	.61

^aValues are reported as n (%). BMI, body mass index; DVT, deep vein thrombosis; MUA, manipulation under anesthesia.

PROMs were completed for 95 knees (83.3%) at a mean follow-up of 5.2 years. Completion rates, follow-up intervals, and NPI score, KOOS, and Marx activity scale score were similar between groups (Table 4).

Table 4.

Patient-Reported Outcome Measures of Patients With BMI <30 and ≥30 kg/m^{2 a}

	BMI, kg/m2		P Value
	<30	≥30
	(n = 70)	(n = 25)
Length of follow-up, mo	67.3 ± 29.0	58.9 ± 16.9	.27
NPI score	15.4 ± 16.8	20.4 ± 17.4	.22
KOOS subscales
Symptoms	78.8 ± 19.5	80.1 ± 19.1	.75
Pain	80.1 ± 20.8	81.1 ± 16.5	.81
ADL	88.4 ± 17.0	87.9 ± 15.0	.81
Sport	72.2 ± 26.5	74.8 ± 23.0	.86
QoL	66.6 ± 26.1	65.8 ± 28.7	.97
Marx activity scale score	6.3 ± 5.4	5.6 ± 5.0	.62

^aValues are reported as mean ± SD. ADL, Activities of Daily Living; BMI, body mass index; KOOS, Knee injury and Osteoarthritis Outcome Score; NPI, Norwich Patellar Instability; QoL, Quality of Life.

Subgroup analysis among patients with WHO class 2 or greater obesity (BMI ≥35 kg/m²) demonstrated similar complication findings to those observed when comparing at a BMI cutoff of 30 kg/m². Notably, patients with WHO class 2 obesity (BMI ≥35 kg/m²) demonstrated a lower Marx activity scale score (P = .039) and a trend toward increased patellar instability (NPI score; P = .07) compared with those with a BMI <30 kg/m² (Table 5).

Table 5.

PROMs of Patients With BMI <30 and ≥35 kg/m^{2 a}

	BMI, kg/m2		P Value
	<30	≥35
	(n = 83)	(n = 15)
DVT	2 (2.4)	1 (6.7)	.40
Reoperation	7 (8.4)	2 (13.3)	.62
Postoperative pain	14 (16.9)	3 (20.0)	.72
Postoperative stiffness	5 (6.0)	1 (6.7)	.93
Repeat instability	26 (31.3)	7 (46.7)	.25
Repeat dislocation	3 (3.6)	0 (0)	.45
PROMs	<30	≥35
	(n = 70)	(n = 12)
NPI score	15.4 ± 16.8	23.7 ± 17.3	.07
KOOS subscales
Symptoms	78.8 ± 19.5	77.0 ± 21.9	.82
Pain	80.1 ± 20.8	77.3 ± 16.0	.30
ADL	88.4 ± 17.0	85.9 ± 16.0	.51
Sport	72.2 ± 26.5	75.0 ± 18.5	.95
QoL	66.6 ± 26.1	64.1 ± 31.0	.88
Marx activity scale score	6.3 ± 5.4	2.6 ± 2.6	.039

^aValues are reported as mean ± SD unless otherwise noted. ADL, Activities of Daily Living; BMI, body mass index; DVT, deep vein thrombosis; KOOS, Knee injury and Osteoarthritis Outcome Score; NPI, Norwich Patellar Instability; PROM, patient-reported outcome measure; QoL, Quality of Life.

Multivariable linear regression models analyzed the effect of BMI on PROMs, controlling for patient age, sex, and cartilage status. When accounting for cartilage status and patient demographics, BMI did not significantly correlate with any of the captured PROMs (Table 6).

Table 6.

Multivariable Linear Regression Analysis: PROMs ^a

	Beta	95% CI	P Value
NPI score	1.29	−0.54 to 3.1	.16
KOOS subscales
Symptoms	−0.28	−1.0 to 0.44	.43
Pain	−0.35	−1.1 to 0.38	.94
ADL	−0.45	−1.1 to 0.16	.15
Sport	−0.45	−1.4 to 0.57	.38
QoL	−0.17	−1.1 to 0.78	.72
Marx activity scale score	−0.17	−0.36 to 0.01	.07

^aADL, Activities of Daily Living; KOOS, Knee injury and Osteoarthritis Outcome Score; NPI, Norwich Patellar Instability; PROM, patient-reported outcome measure; QoL, Quality of Life.

Discussion

The most important findings of this study are that elevated BMI is not associated with increased complications after isolated MPFLR, and the effects of increased BMI on PROs after isolated MPFLR are minimal. Recent attention has been paid to BMI and its effect on outcomes after surgery. While imperfect, BMI is a known independent risk factor for complications and poorer outcomes after numerous orthopaedic surgeries.^21,24

To our knowledge, this study is one of the largest to date to evaluate the relationship between BMI and complications and PROs after isolated MPFLR. Among 114 knees with >2 years of follow-up, we observed comparable rates of DVT, infection, persistent pain, and stiffness across all groups. While PROMs were mostly similar across groups, we did observe a trend toward higher NPI and lower KOOS values and Marx activity scale scores in the higher-BMI groups. Additional work with larger numbers is required to further evaluate this potential effect.

Other authors have evaluated the relationship between BMI and outcomes of MPFLR. Sherman et al ¹⁸ reported the safety and efficacy of MPFLR in patients with a BMI ≥30 kg/m², demonstrating improvement in the majority of PROMs and low complication rates when compared with a cohort of patients with a BMI <30 kg/m². We attempted to further study obesity by stratifying our patients into 2 large groups in line with similar investigations, which have shown an increase in complications with an increase in obesity class, ³ and to further quantify these effects through detailed multivariable analysis and class 2 obesity subgroup analysis. In the current study, complication risk was not increased in obese patients.

In 2014, Enderlein et al ⁵ prospectively assessed clinical outcomes after MPFLR among 240 knees and reported an association between female sex, BMI ≥30 kg/m², age >30 years, and grade 3 or 4 cartilage injury with poorer Kujala Anterior Knee Pain scores. In our investigation, we observed a statistically insignificant association with regard to some but not all of the functional outcome measures. In regard to the KOOS, when adjusting for age and sex alone, only the Activities of Daily Living (ADL) subscale demonstrated a statistically significant association with BMI. When including cartilage status in this multivariable analysis, significance was no longer observed. It is important to note that elevated BMI was also strongly correlated with decreased activity as assessed through the Marx activity scale. Decreased activity level has been shown to mitigate negative pressure on other PROs and may allow patients to compensate for poorer knee function.^4,6 The elevated NPI scores in this population suggest that decreased activity level and poorer KOOS-ADL function may be related to ongoing patellar instability symptoms.

Limitations

This study has several limitations. The MPFLR procedures were performed by multiple surgeons with heterogeneous techniques and graft choices and postoperative protocols. While these factors increase study heterogeneity, they also reflect the realities of clinical practice, making the findings broadly generalizable. Furthermore, with a mean follow-up of 4 years and minimum follow-up of 2 years, the study is unable to provide insights into potential medium- or long-term effects of BMI on MPFLR outcome. In addition, the study is subject to selection bias as only 69% of eligible patients completed 2-year follow-up, and only 57% of patients completed PROs. While we had relatively few patients with a BMI ≥35 kg/m², there were extremely few patients with a BMI >40 kg/m², precluding meaningful analysis of outcomes in this population. This subgroup analysis among a relatively rare cohort of patients was subsequently underpowered. Furthermore, the presented study outcomes only pertain to isolated MPFLR. The effects of elevated BMI on outcomes of MPFLR with tibial tubercle osteotomy or trochleoplasty procedures remain less clear. Because cartilage damage has been shown to be associated with poorer PROMs, ⁷ this variable was included in the multivariable analysis to control for its potentially confounding effects. A larger series of patients is needed to confirm the results of this investigation and to evaluate additional related populations.

Conclusion

No significant differences in complications or repeat dislocation risk after isolated MPFLR were noted based on BMI ≥30 or <30 kg/m². Patients with a BMI ≥35 kg/m² demonstrated a lower activity level, but no other differences in PROs compared with patients with a BMI <30 kg/m².