Body Composition, Not Body Mass Index, Is Associated with Clinical Outcomes Following ACL Reconstruction

Abstract

Purpose:

Determine the associations of body mass index (BMI) and body composition (lean and fat mass) with i) patient-reported outcomes, ii) quadriceps strength, and iii) physical performance in anterior cruciate ligament reconstruction (ACLR) patients.

Methods:

Thirty-four individuals participated (56% female; 22±4 years; % body fat: 22.9±6.7%). Dual energy x-ray absorptiometry was used to measure percent lean (%LM) and fat (%FM) mass for the whole-body, and the whole-limb and thigh region of the ACLR limb. Knee Injury and Osteoarthritis Outcomes Score (KOOS) subscales were used to evaluate patient-reported outcomes. Isometric quadriceps strength (MVIC) was measured using an isokinetic dynamometer with the knee flexed at 90 degrees. Single-leg hop (SLH) distance was used to evaluate physical performance. Spearman rank correlations were used to separately evaluate the associations between BMI and body composition with KOOS, MVIC, and SLH.

Results:

BMI was not significantly associated with KOOS, MVIC, or SLH (p>0.05). Greater whole-body %LM associated with greater MVIC and SLH, while greater whole-body %FM associated with worse KOOS ADL, MVIC, and SLH (p<0.05). Greater whole-limb %LM associated with greater MVIC, while greater whole-limb %FM associated with worse KOOS Symptoms, Pain, QoL, and MVIC (p<0.05). Greater thigh %LM associated with better KOOS Symptoms and MVIC, and greater thigh %FM associated with worse KOOS Symptoms, Pain, QoL, and MVIC (p<0.05).

Conclusions:

Body composition characteristics of higher %LM and lower %FM associated with better clinical outcomes in ACLR patients of normal BMI status, thereby elucidating a potential modifiable target to mitigate poor ACLR-related outcomes. Future research should further evaluate mechanistic links between body composition and ACLR-related outcomes to inform clinical practice and rehabilitation frameworks.

INTRODUCTION

Individuals who have experienced an anterior cruciate ligament (ACL) injury often demonstrate persistent clinically relevant knee-related symptoms, impaired muscle function, and decreased physical performance for years following ACL reconstruction (ACLR) and completion of formal rehabilitation (1–5). Furthermore, half of these individuals develop radiographic knee osteoarthritis (KOA) within two decades of ACLR (6). Identifying modifiable factors linked to better patient-reported outcomes, muscle function, and physical performance following ACL injury and ACLR is paramount for improving long-term outcomes following injury and surgery.

High body mass index (BMI) is a modifiable risk factor that is associated with development and progression of symptomatic and radiographic KOA in the general population (7, 8). However, the link between BMI and poor knee-related outcomes following ACLR is limited and conflicting as individuals who are most likely to sustain an ACL injury are typically young (17–35 years old), physically active, and not classified as having obesity (i.e., BMI <30 kg·m⁻²) (6). Yet, there are strong associations between body composition characteristics of lower lean mass (LM) and higher fat mass (FM) and worse knee-related outcomes in individuals with and at-risk for KOA (9–12). However, limited studies have comprehensively evaluated the links between BMI and body composition with knee-related outcomes post-ACLR. Determining the associations between BMI and body composition with patient-reported outcomes, muscle strength, and physical performance is an important step in establishing more sophisticated screening metrics of future chronic disease onset and developing therapeutic targets for future interventions that can optimize joint health following ACL injury and ACLR.

BMI reportedly increases over the first 6 years following ACLR (13), but body weight changes are inconsistent across patients wherein some individuals demonstrate body weight increases while others demonstrate relatively stable body weight 1–5 years post-ACLR (14). Nonetheless, high BMI is a key indicator of poor outcomes after ACLR including KOA development (15), biomarkers linked to cartilage breakdown (16), persistent knee-related symptoms (13), altered strength recovery trajectories (17), decreased physical activity (18), and worse physical performance (19) within the first decade post-ACLR. Further, compared with uninjured controls, individuals with a history of traumatic knee joint injury (e.g., ACL injury and ACLR) demonstrate significantly higher BMI, greater percent body fat, and lower physical performance up to 10 years following injury (20). Greater FM is linked to worse knee pain, lesser strength, and decreased physical performance in young adults of normal BMI status both with and without joint injury (20–24). These data suggest that both BMI and underlying body composition characteristics such as body fat percentage may influence ACLR-related outcomes.

LM is a primary contributor to physical performance and demonstrates protective effects on joint tissue health (25), while FM contributes to a pro-inflammatory state that may promote deleterious changes in joint tissues following ACL injury and ACLR (22, 26). In young adults without injury and with normal BMI, greater whole-body LM and better physical performance are linked to greater tibial cartilage volume and thickness (27), and greater LM is a mediator of the relationship between better physical performance and greater tibial cartilage thickness (22). In individuals with incident KOA, lesser LM and greater FM is also associated with worse knee-related symptoms, lower quadriceps strength, and decreased physical performance (10). Similarly, in young adults of normal BMI without a knee injury, lower LM and greater FM are associated with greater knee pain incidence, lower muscular strength, and decreased physical performance (23)(24). Further, in a cohort of young women without a knee injury, those who reported knee pain demonstrated slightly higher BMI (25.6 vs 24.1 kg·m⁻²), greater whole-body %FM, lower %LM, and worse quadriceps strength, compared to those who did not report knee pain (23). A recent study also found that greater BMI, waist circumference, and thigh FM are linked to worse patient-reported outcomes and physical performance at 1 and 5 years post-ACLR in a cohort of adults primarily of normal BMI status (21). However, body composition outcomes associated with knee-related clinical outcomes including patient-reported outcomes, quadriceps strength, and physical performance remain unknown in ACLR patients of normal BMI status. Given robust previous evidence linking body composition to knee-related outcomes in individuals with KOA (9, 10, 12, 28, 29), along with recent evidence highlighting similar associations among both ACLR patients and uninjured young adults of normal BMI (20–24), there is a critical need to further evaluate the body composition outcomes (i.e., LM and FM) across various sites (i.e., whole-body, whole-limb, and thigh region) which may be important indicators of poor knee-related outcomes after ACLR. Moreover, as minimal BMI percentile changes may not be entirely reflective of critical body composition shifts that drive clinical ACLR outcomes, body composition assessments may be a critical, yet understudied outcome driving clinical outcomes after ACLR.

Patient-reported outcomes, quadriceps strength, and physical performance tests (e.g., single leg hop [SLH]) are integral to informing clinical decision making and determining the overall outcome of surgery and rehabilitation post-ACLR (30–32). Persistent poor patient-reported outcomes, muscle weakness, and decreased physical performance are linked to greater risk for sustaining a secondary ACL injury and developing KOA (30–33), yet the modifiable factors associated with these clinical ACLR outcomes remain unknown. Therefore, the primary purpose of this study was to separately determine the associations for BMI and body composition (percent LM [%LM] and FM [%FM] of the whole-body, whole-limb, and thigh region of the ACLR limb) with i) patient-reported outcomes, ii) quadriceps strength, and iii) physical performance (i.e., SLH test) in ACLR patients. We hypothesized that greater BMI would demonstrate weak associations, and lesser %LM and greater %FM would demonstrate moderate-to-strong associations with worse patient-reported outcomes, greater quadriceps weakness, and impaired physical performance in ACLR patients.

METHODS

Study Design

A cross-sectional analysis was used to determine the associations between continuous variables of BMI and body composition with patient-reported outcomes, quadriceps strength, and physical performance in individuals with a primary, unilateral ACLR. All data were collected during a single testing session. Demographic (e.g., age, height, weight, sex, time since ACLR) and patient-reported outcome data were collected first, followed by body composition, quadriceps strength, and physical performance outcomes. In order to describe the cohort, we also collected data regarding objective physical activity measures (i.e., average time spent in moderate-to-vigorous activity [MVPA], average steps per day), patient-reported physical activity/ sport participation, time spent in formal physical therapy following ACLR, time since completion of formal physical therapy, and descriptive self-reported data on exercises completed in physical therapy. Objective physical activity data were collected using a waist-worn accelerometer (ActiGraph wGT3X-BT) for 7 days, as previously described (34). All protocols were approved by the University’s Institutional Review Board. All participants provided written informed consent prior to study participation.

Participants

Participants were recruited from the university health clinics and surrounding community via flyers, email, and word of mouth. We included individuals with a primary unilateral ACLR between 0.5–7.0 years (6.0–84.0 months) prior to study enrollment who were between 18–35 years old and cleared by a physician for full return to activity. We excluded individuals from the study with: i) multiple ACL injuries, ii) a neurologic disorder, iii) a lower extremity fracture or any orthopaedic surgery except for ACLR, iv) a previous diagnosis of KOA or inflammatory arthritis, and v) lower extremity injury within 6 months prior to testing (35). We also excluded individuals who reported being pregnant at the time of testing (35) as DXA scans are not recommended for women who are pregnant unless the test is otherwise prescribed by their healthcare provider (36). Previous research has reported associations as strong as r: 0.45 between BMI, thigh LM, quadriceps strength, and patient-reported outcomes in individuals at high-risk for KOA (37). We estimated that we would need a total of 34 participants to be adequately powered (1-β=0.8) to detect statistically significant moderate association (Pearson’s r≥0.4) between BMI and body composition with patient-reported outcomes, quadriceps strength, and physical performance (37), using a two-tailed alpha level of 0.05 (G*Power, v3.1.9.2).

BMI and Body Composition: Dual Energy X-Ray Absorptiometry (DXA)

Height (m) and mass (kg) were recorded for each participant at the start of the data collection visit, and were used to calculate BMI (kg·m⁻²). A total body DXA scan was performed and analyzed as previously reported (10, 35) for LM and FM outcomes. DXA scans were performed in accordance with manufacturer recommendations and software (version 16; Encore Software, Madison, WI). Body composition outcomes of %LM and %FM were calculated for the whole-body, whole-limb, and thigh region of the ACLR and contralateral (uninvolved) limbs using a whole-body DXA scan (GE Lunar iDXA, GE Medical Systems Ultrasound and Primary Care Diagnostics, Madison, WI; Equations 1 and 2)(10, 35). The whole-limb region was defined as the region inferior to the iliac crest of the limb of interest (i.e., ACLR or uninvolved limb). A custom region-of-interest technique was utilized to manually segment the region from the most superior-lateral aspect of greater trochanter of the femur and pubic symphysis to the tibiofemoral joint line on each limb individually. This manually segmented region was defined as the thigh region (Figure 1).

Figure 1.

Custom regions-of-interest dual energy x-ray absorptiometry (DXA) thigh region segmentation defined as the region between the iliac crest and the pubic symphysis to the tibiofemoral joint line of the ACLR limb (highlighted region).

$$\text{Percent}\text{lean}\text{mass}(\text{\%LM})=(\text{lean}\text{mass}[\text{kg}]/\text{total}\text{body}\text{mass}[\text{kg}])\mathrm{*}100$$ Equation 1:

$$\text{Percent}\text{fat}\text{mass}(\text{\%FM})=(\text{fat}\text{mass}[\text{kg}]/\text{total}\text{body}\text{mass}[\text{kg}])\mathrm{*}100$$ Equation 2:

Test-retest reliability for LM and FM were as follows: ICCs of 0.998 and 0.998, respectively (12).

Patient-Reported Outcomes: Knee Injury and Osteoarthritis Outcomes Score (KOOS)

The KOOS is a reliable and valid 42-item questionnaire split into 5 subscales: Pain, Symptoms, Activities of Daily Life (ADL), Sports, and Quality of Life (QoL) (38). Participants completed the KOOS electronically, and scores for each subscale were independently normalized to a scale of 0–100 such that higher scores indicate better patient-reported outcomes (i.e., a score of 100 indicates no ACLR knee-related problems and a score of 0 reflects severe ACLR knee-related problems).

Quadriceps Strength: Maximal Voluntary Isometric Contraction (MVIC)

Maximal voluntary isometric contractions were performed to quantify quadriceps strength for both limbs using an isokinetic dynamometer (HUMAC NORM, CSMi, Stoughton, MA) with the knee positioned at 90 degrees of flexion and the hip at 85 degrees of flexion (39). Participants were seated upright, and straps were positioned across the torso, thigh, and shank to isolate quadriceps muscle action during the testing. A minimum of 3 warm-up trials were performed at 25, 50, and 75% of maximal effort. Following the warm-up trials, participants completed 2 practice MVIC trials wherein the average peak torque was calculated and used as biofeedback during the subsequent MVIC trials (39). The average of 2 practice MVIC trials were used to calculate a torque target that was 120% of that minimal threshold was displayed on a computer monitor. All participants were instructed to aim for the torque target to ensure participants were exerting maximal effort during each trial (LabVIEW; National Instruments Corp., Austin, TX) (39). A trial was considered to be successful if the peak torque was ≥90% of the average of the practice trials (39). Participants were allotted at least 1 minute of rest between MVIC trials to account for fatigue. Torque data were sampled at 2000 Hz, lowpass filtered at 50 Hz, and reduced to identify peak torque during each MVIC trial. MVIC values were normalized to body mass (kg) and are presented as Nm·kg⁻¹ (39).

Physical Performance: Single Leg Hop (SLH)

SLH tests were utilized to evaluate physical performance for the ACLR and contralateral limbs. SLH distance was defined as the maximal distance (m) travelled for a successful SLH trial. A trial was deemed successful if the participant jumped and landed on the ACLR limb without the contralateral foot encountering the ground at any point during the trial. Once the participant successfully landed and maintained balance for at least 3 seconds, they were allowed to place their contralateral limb on the ground to remain in position until the researcher recorded the hop distance for that trial. Participants were allowed 4 practice trials followed by 2 maximal SLH trials wherein the average distance of the 2 trials was calculated. Adequate rest (minimum 1 minute) was allotted between all trials. The average distance from 2 successful maximal SLH trials was calculated and utilized in analyses (40).

Primary Statistical Analyses

All statistical analyses were conducted in RStudio (Version 2023.06.1+523; PBC, Boston, MA). Shapiro-Wilk tests and Q-Q plots were used to assess normality of the data. As the assumption of normality was not met for BMI and body composition outcomes (%LM and %FM of the whole-body, whole-limb, and thigh region), we conducted separate non-parametric correlation analyses (i.e., Spearman rank correlation coefficients) to determine the associations of BMI and body composition with i) patient-reported outcomes (KOOS subscale scores); ii) quadriceps strength (MVIC), and iii) physical performance (SLH). As this study sought to be a descriptive and foundational analysis of the associations between body composition and ACLR-related outcomes, we did not adjust significance level to account for multiplicity.

Secondary Analyses

As previous data suggest that unilateral ACLR can influence bilateral outcomes (41), we conducted secondary analyses evaluating the associations described above in the primary analysis on the uninvolved, contralateral limb.

Post-hoc Analyses

As time-since ACLR has been found to associate with KOOS, MVIC, and SLH, we conducted post-hoc analyses evaluating the associations described in the primary and secondary analyses while controlling for time since ACLR. We also conducted between limb comparisons for %LM and %FM of the whole-limb and thigh region between the ACLR and contralateral limbs.

RESULTS

Data from 34 individuals with a history of primary, unilateral ACLR (n=19 female and 15 male; 22.2±4.2 years; BMI: 23.8±2.7 kg·m⁻²; % body fat: 22.9±6.7) were analyzed in the present study (Tables 1 and 2). Mean ± standard deviation (SD) and frequency (n, %) are reported for all continuous (age, time since ACLR, BMI, body composition, physical activity, time in formal physical therapy, and time since completion of formal physical therapy) and dichotomous (sex, returned to any level of sport, returned to same level of sport) variables in Tables 1, 2, and Supplemental Table 3 (Supplemental Digital Content). Physical activity patterns for ACLR individuals (i.e., average steps per day; average time spent in MVPA per day) in the current cohort (9201±2607 steps/day; 75.0±27.1 MVPA min/day) were similar to physical activity patterns reported in literature (42, 43) for other ACLR cohorts (BMI: 23.7±3.2 kg/m²; 0.5–7 years post-ACLR; 8652±2467 steps/day; 7982±3020 steps/day; 78.34±26.62 MVPA min/day), and uninjured controls of normal BMI (9945±2885 steps/day) (42, 43). All Spearman correlation coefficients are reported in Tables 3 and 4 for the ACLR limb and contralateral limb, respectively, with corresponding p-values and 95% confidence intervals (CI).

Table 1.

Participant Demographics, KOOS Scores, and Physical Activity Outcomes. Data presented as mean ± SD (range) or n (%), as appropriate.

Sex, M:F	M: 15 (44%) F: 19 (56%)
BMI, kg/m 2	23.4 ± 2.8 (19.8–33.6)
Age, years	22.18 ± 4.20 (18.0–33.0)
Months Since ACLR	51.9 ± 37.0 (9.0–146.0)
KOOS Symptoms	81.18 ± 15.55 (35.71–100.00)
KOOS Pain	91.83 ± 7.73 (69.44–100.00)
KOOS ADL	97.92 ± 4.57 (75.00–100.00)
KOOS Sport	85.59 ± 13.36 (45.00–100.00)
KOOS QOL	77.02 ± 16.69 (37.50–100.00)
Average Steps per Day, n/day	9201 ± 2607 (4374–14050)
Average Daily MVPA, min/day	75.0 ± 27.1 (23.6–136.7)
Time Spent in Formal PT, months	6.6 ± 2.0 (3.0–10.0)
Time Since Formal PT Completion, months	39.3 ± 35.1 (1.0–144.0)
Returned to Some Level of Sport, Y:N	Y: 32 (94%) N: 2 (6%)
Returned to Same Level of Sport, Y:N	Y: 20 (59%) N: 14 (41%)

Note: BMI, body mass index; ACLR, anterior cruciate ligament reconstruction; KOOS, Knee Injury and Osteoarthritis Outcomes Score; ADL, Activities of Daily Life; QOL, Quality of Life; MVPA, moderate-to-vigorous physical activity; PT, physical therapy

Table 2.

Body composition (%LM and %FM) for the whole-body, whole-limb, and thigh region of the ACLR and contralateral (uninvolved) limbs, and functional outcomes (MVIC, SLH) for the ACLR and contralateral limbs.

Body Composition And Functional Outcomes		Mean ± SD	P-valuea	Effect Sizeb
Whole-Body Composition	%LM	53.60 ± 12.40	-	-
	%FM	22.82 ± 6.72	-	-
Whole-Limb Composition	ACLR Limb %LM	29.84 ± 2.66	0.49	−0.12
	Uninv. Limb %LM	29.98 ± 2.75
	ACLR Limb %FM	10.05 ± 3.35	0.28	−0.19
	Uninv. Limb %FM	10.50 ± 3.84
Thigh Region Composition	ACLR Thigh %LM	19.80 ± 1.96	0.33	−0.17
	Uninv. Thigh %LM	19.90 ± 1.94
	ACLR Thigh %FM	6.50 ± 2.33	0.83	−0.04
	Uninv. Thigh %FM	6.51 ± 2.35
MVIC, Nm/kg	ACLR Limb	3.24 ± 0.77	0.03	−0.39
	Uninv. Limb	3.42 ± 0.81
SLH, m	ACLR Limb	1.24 ± 0.30	0.01	−0.52
	Uninv. Limb	1.32 ± 0.28

Between-limb comparisons (paired t-test) were conducted for whole-limb composition, thigh region composition, MVIC, and SLH. Data are presented as mean ± SD, paired t-test p-value^a, and Cohen’s d effect size^b, as appropriate.

Note: Bolded values represent statistically significant (p≤0.05) between limb difference. LM, lean mass; FM, fat mass; ACLR, anterior cruciate ligament reconstruction; uninv. limb, uninvolved limb; MVIC, maximal voluntary isometric contraction; Nm/kg, Newton meters per kilogram of bodyweight; SLH, single leg hop.

Table 3.

ACLR limb Spearman’s rank correlation coefficients (r), (p-value), and [95% CI] for associations between BMI, whole-body, whole-limb, and thigh-region composition (%LM and %FM) with patient-reported outcomes (KOOS subscale scores), quadriceps strength (MVIC), and functional performance (SLH).

Predictor Variable		KOOS Symptoms	KOOS Pain	KOOS ADL	KOOS Sport	KOOS QOL	ACLR MVIC	ACLR SLH
BMI	r	0.26	0.24	−0.12	0.15	0.31	−0.04	−0.24
	(P-value)	(0.14)	(0.17)	(0.50)	(0.39)	(0.08)	(0.83)	(0.17)
	[95% CI]	[−0.08, 0.55]	[−0.07, 0.58]	[−0.44, 0.23]	[−0.20, 0.47]	[−0.03, 0.58]	[−0.37, 0.30]	[−0.53, 0.11]
Whole-Body %LM	r	0.45	0.46	0.14	0.10	0.39	0.31	0.12
	(P-value)	(0.01)	(0.01)	(0.44)	(0.59)	(0.02)	(0.08)	(0.51)
	[95% CI]	[0.13, 0.68]	[0.15, 0.69]	[−0.21, 0.46]	[−0.25, 0.42]	[0.06, 0.64]	[−0.03, 0.59]	[−0.23, 0.44]
Whole-Body %FM	r	−0.28	−0.28	−0.34	−0.21	−0.27	−0.55	−0.37
	(P-value)	(0.11)	(0.11)	(0.05)	(0.23)	(0.13)	(<0.01)	(0.03)
	[95% CI]	[−0.57, 0.06]	[−0.56, 0.07]	[−0.61, −0.00]	[−0.51, 0.14]	[−0.56, 0.08]	[−0.75, −0.26]	[−0.63, −0.03]
ACLR Limb %LM	r	0.19	0.15	0.24	0.11	0.13	0.44	0.20
	(P-value)	(0.29)	(0.40)	(0.18)	(0.53)	(0.47)	(<0.01)	(0.25)
	[95% CI]	[−0.16, 0.49]	[−0.20, 0.46]	[−0.11, 0.53]	[−0.24, 0.43]	[−0.22, 0.45]	[0.12, 0.68]	[−0.15, 0.51]
ACLR Limb %FM	r	−0.42	−0.38	−0.17	−0.16	−0.37	−0.42	−0.31
	(P-value)	(0.01)	(0.03)	(0.33)	(0.38)	(0.03)	(0.02)	(0.07)
	[95% CI]	[−0.67, −0.10]	[−0.63, −0.05]	[−0.48, 0.18]	[−0.47, 0.19]	[−0.63, −0.04]	[−0.66, −0.09]	[−0.59, 0.03]
ACLR Thigh %LM	r	0.34	0.32	0.27	0.18	0.29	0.55	0.23
	(P-value)	(0.05)	(0.06)	(0.12)	(0.31)	(0.10)	(<0.01)	(0.19)
	[95% CI]	[0.00, 0.61]	[−0.02, 0.59]	[−0.07, 0.49]	[−0.17, 0.49]	[−0.06, 0.57]	[0.26, 0.75]	[−0.12, 0.53]
ACLR Thigh %FM	r	−0.38	−0.34	−0.21	−0.16	−0.36	−0.42	−0.32
	(P-value)	(0.03)	(0.05)	(0.23)	(0.36)	(0.04)	(0.01)	(0.07)
	[95% CI]	[−0.64, −0.05]	[−0.61, −0.01]	[−0.51, 0.14]	[−0.47, 0.19]	[−0.62, −0.02]	[−0.67, −0.10]	[−0.59, 0.03]

Note: Bold values indicate statistically significant (p≤0.05) correlation. CI, confidence interval; %LM, percent lean mass; %FM, percent fat mass; BMI, body mass index; KOOS, Knee Injury and Osteoarthritis Outcomes Score; ADL, activities of daily life; QoL, quality of life; MVIC, maximal voluntary isometric contraction; SLH, single leg hop.

Table 4.

Contralateral (uninvolved) limb Spearman’s rank correlation coefficients (r), (p-value), and [95% CI] for associations between BMI, whole-body, whole-limb, and thigh-region composition (%LM and %FM) with patient-reported outcomes (KOOS subscale scores), quadriceps strength (MVIC), and functional performance (SLH).

Predictor Variable		KOOS Symptoms	KOOS Pain	KOOS ADL	KOOS Sport	KOOS QOL	Uninv. MVIC	Uninv. SLH
BMI	r	0.26	0.24	−0.12	0.15	0.31	0.02	−0.28
	(P-value)	(0.14)	(0.17)	(0.50)	(0.39)	(0.08)	(0.91)	(0.11)
	[95% CI]	[−0.08, 0.55]	[−0.07, 0.58]	[−0.44, 0.23]	[−0.20, 0.47]	[−0.03, 0.58]	[−0.32, 0.35]	[−0.56, 0.07]
Whole-Body %LM	r	0.45	0.46	0.14	0.10	0.39	0.37	0.12
	(P-value)	(0.01)	(0.01)	(0.44)	(0.59)	(0.02)	(0.03)	(0.48)
	[95% CI]	[0.13, 0.68]	[0.15, 0.69]	[−0.21, 0.46]	[−0.25, 0.42]	[0.06, 0.64]	[0.03, 0.63]	[−0.22, 0.44]
Whole-Body %FM	r	−0.28	−0.28	−0.34	−0.21	−0.27	−0.46	−0.34
	(P-value)	(0.11)	(0.11)	(0.05)	(0.23)	(0.13)	(<0.01)	(0.05)
	[95% CI]	[−0.57, 0.06]	[−0.56, 0.07]	[−0.61, −0.00]	[−0.51, 0.14]	[−0.56, 0.08]	[−0.69, −0.14]	[−0.61, −0.00]
Uninv. Limb %LM	r	0.24	0.36	0.43	0.28	0.30	0.32	0.20
	(P-value)	(0.18)	(0.04)	(0.01)	(0.11)	(0.09)	(0.07)	(0.27)
	[95% CI]	[−0.11, 0.53]	[0.02, 0.62]	[0.11, 0.67]	[−0.06, 0.57]	[−0.04, 0.58]	[−0.03, 0.59]	[−0.15, 0.50]
Uninv. Limb %FM	r	−0.32	−0.24	−0.13	−0.12	−0.24	−0.26	−0.20
	(P-value)	(0.06)	(0.17)	(0.47)	(0.49)	(0.18)	(0.13)	(0.26)
	[95% CI]	[−0.59, 0.02]	[−0.53, 0.11]	[−0.45, 0.22]	[−0.44, 0.22]	[−0.53, 0.11]	[−0.55, 0.08]	[−0.50, 0.15]
Uninv. Thigh %LM	r	0.28	0.36	0.39	0.22	0.32	0.36	0.13
	(P-value)	(0.10)	(0.04)	(0.02)	(0.21)	(0.06)	(0.04)	(0.45)
	[95% CI]	[−0.06, 0.57]	[0.02, 0.62]	[0.06, 0.65]	[−0.13, 0.52]	[−0.02, 0.60]	[0.02, 0.62]	[−0.21, 0.45]
Uninv. Thigh %FM	r	−0.39	−0.35	−0.22	−0.15	−0.36	−0.33	−0.28
	(P-value)	(0.02)	(0.05)	(0.22)	(0.40)	(0.04)	(0.06)	(0.12)
	[95% CI]	[−0.64, −0.06]	[−0.61, −0.01]	[−0.52, 0.13]	[−0.46, 0.20]	[−0.62, −0.03]	[−0.60, 0.01]	[−0.56, 0.07]

Note: Bold values indicate statistically significant (p≤0.05) correlation. CI, confidence interval; %LM, percent lean mass; %FM, percent fat mass; BMI, body mass index; KOOS, Knee Injury and Osteoarthritis Outcomes Score; ADL, activities of daily life; QoL, quality of life; MVIC, maximal voluntary isometric contraction; SLH, single leg hop.

Primary Analysis: ACLR Limb (Table 3)

Body Mass Index (BMI):

BMI was not significantly associated (p>0.05) with any KOOS subscales (r range: −0.06 to 0.26; Table 2), MVIC (r: −0.10), or SLH (r: −0.16).

Whole-body Composition:

Whole-body composition outcomes (%LM and %FM) demonstrated weak to moderate associations (r range: −0.55 to 0.56) with KOOS, MVIC, and SLH. Specifically, greater whole-body %LM was associated with greater MVIC (r: 0.56) and SLH distance (r: 0.36). Greater whole-body %FM was associated with worse KOOS ADL (r : −0.34), lesser MVIC (r: −0.55), and shorter SLH distance (r: −0.36).

Whole-limb Composition:

Similarly, whole-limb %LM and %FM demonstrated weak to moderate associations (r range: −0.42 to 0.44) with ACLR outcomes (KOOS, MVIC, and SLH). Greater whole-limb %LM was associated with greater MVIC (r: 0.44). Greater whole-limb %FM associated with worse KOOS Symptoms (r: −0.42), Pain (r: −0.38), QoL (r: −0.37), and lesser MVIC (r: −0.42).

Thigh Region Composition:

Lastly, our custom thigh region of interest DXA analysis found thigh region composition characteristics of %LM and %FM associated with KOOS, MVIC, and SLH (r range: −0.42 to 0.55). Greater thigh %LM was associated with better KOOS Symptoms (r: 0.34) and greater MVIC (r: 0.55), and greater thigh %FM associated with worse KOOS Symptoms (r: −0.38), Pain (r: −0.34), QoL (r: −0.36), and lesser MVIC (r: −0.42).

Secondary Analysis: Contralateral Limb (Table 4)

Body Mass Index):

BMI did not associate with contralateral limb MVIC (r: 0.02) and SLH (r: −0.28).

Whole-body Composition:

Similar to the ACLR limb, greater whole-body %LM (r: 0.37) and lesser %FM (r: −0.46) associated with better contralateral limb MVIC. Unlike the ACLR limb, whole-body %LM and %FM did not associate with contralateral limb SLH performance (r: 0.12 and −0.34, respectively).

Whole-limb Composition:

Greater whole-limb %LM of the contralateral limb was associated with better KOOS Pain (r: 0.36) and KOOS ADL (r: 0.43) while contralateral whole-limb %FM was not associated with KOOS, MVIC, or SLH.

Thigh Region Composition:

Greater %LM of the contralateral thigh was associated with better KOOS Pain (r: 0.36), KOOS ADL (r: 0.39), and better contralateral limb MVIC (r: 0.36). Greater thigh region %FM of the contralateral limb was associated with worse KOOS Symptoms (r: −0.39), Pain (r: −0.35), KOOD QOL (r: −0.36), but not with MVIC (r: −0.36) or SLH (r: −0.28).

Post-Hoc Supplementary Analysis: Controlling for Time Since ACLR (Supplemental Tables 1 and 2, Supplemental Digital Content)

When controlling for time since ACLR, greater whole-body %LM was associated with better KOOS Pain scores (Supplementary Table 1, Supplemental Digital Content). Further, greater whole-body %FM was associated with lower ACLR and contralateral limb MVIC, along with lesser ACLR limb SLH performance. Greater ACLR limb and thigh region %LM was associated with greater ACLR limb MVIC. No additional or other significant associations were identified when controlling for time since ACLR in post-hoc analyses.

DISCUSSION

Measures of body composition (%LM and %FM), but not BMI, were associated with patient-reported outcomes, quadriceps strength, and physical performance post-ACLR in the present descriptive analyses. The current study introduces novel findings that body composition – a modifiable factor linked to knee-related outcomes (patient-reported outcomes, strength, and physical performance) in young adults without obesity and individuals with incident KOA (11, 12, 20–24, 26, 44) – is also associated with important knee-related clinical outcomes in ACLR patients. Our findings are the first to establish a cross-sectional link between body composition and clinical ACLR outcomes, and future research should consider determining associations between body composition and development of chronic joint-related disease and disability following ACLR. Additionally, future research should evaluate the existence of a mechanistic link between early changes in post-ACLR body composition and longitudinal changes in clinical outcomes post-ACLR to determine the potential for targeted body composition intervention strategies to improve long-term joint health following ACLR.

Patient-reported outcome surveys (e.g., KOOS) are implemented to clinically track rehabilitation progress and success, yet few modifiable factors are known to associate with patient-reported outcomes after ACLR. We found that body composition is associated with persistent knee-related symptoms (lower KOOS scores) in individuals who are 4–7 years post-ACLR and of normal BMI status, suggesting that body composition may be an integral contributing factor to poor outcomes after ACLR. Greater quadriceps intramuscular and periarticular FM have been linked to worse knee-related pain and symptoms in individuals 3–70 months post-ACLR (1, 45, 46), along with radiographic evidence of KOA (47). Lower %LM and greater %FM were linked to worse patient-reported function which is similar to previous studies which have used ultrasonography and MRI to quantify FM in a cross-section of the thigh region (21, 46). Our findings add to this previous evidence linking soft tissue composition to knee-related outcomes as we found that lesser ACLR limb and thigh region %LM, and greater ACLR limb and thigh region %FM influences worse knee-related pain, symptoms, and quality of life. Further, whole-body %LM and %FM associated with broader assessments of patient-reported function during activities of daily life (KOOS ADL) rather than knee-specific outcomes (KOOS Pain, Symptoms, QoL) in the present study. Similar links have been reported between greater whole-body adiposity and worse patient-reported function in individuals with KOA (29), yet limited evidence has linked body composition characteristics to knee-related outcomes such as KOOS Pain, Symptoms, and QoL. Our findings suggest that similar to individuals with KOA, lesser %LM and greater %FM associate with worse KOOS scores after ACLR. The findings in the present study demonstrate associations of similar strength (i.e., weak to moderate associations) as recently published data which found that measures of whole-body, central, and peripheral adiposity are associated with worse KOOS and shorter SLH distances at 1 and 5 years post-ACLR (21). Future work should determine whether modifying body composition (e.g., improving LM:FM ratio) or maintaining LM in individuals with high LM leads to improved patient-reported outcomes post-ACLR.

Quadriceps strength is another primary clinical outcome used to assess rehabilitation success and inform clinical decisions such as return to sport (30–32). Persistent quadriceps strength deficits and muscle dysfunction are linked to joint tissue changes consistent with early KOA development after ACLR (48). Previous research (49) suggests that non-modifiable factors such as age, sex, graft type, and time since ACLR strongly influence quadriceps strength. However, understanding the modifiable factors that contribute to quadriceps strength deficits is important for mitigating quadriceps weakness following ACLR. Our results indicate that both lower global (i.e., whole-body) and region-specific (i.e., whole-limb, and thigh region) %LM, along with greater global and region-specific %FM are indicators of worse quadriceps muscle strength and function in the ACLR limb, but not the contralateral limb. These findings are similar to previous evidence in individuals with obesity (50), incident KOA (51), and post-ACLR (52). There is also evidence supporting that greater quadriceps intramuscular fat post-ACLR is linked to quadriceps muscle dysfunction, lesser strength, and worse knee-related pain in patients 50.5±29.4 months post-ACLR, which is a similar timeframe post-ACLR as the participants in the present study (46). As most factors that have previously been linked to quadriceps strength after ACLR are non-modifiable (e.g., age, sex, time since ACLR) (49), the present study introduces novel findings which highlight modifiable factors (i.e., body composition) linked to quadriceps strength in ACLR patients of normal BMI status. Conversely, whole-body %LM and %FM were the only body composition outcomes associated with physical performance (SLH) in the present study (i.e., lower whole-body %LM and greater whole-body %FM associated with worse SLH performance). Lean mass is critical for producing force during movement, and global lean mass deficits (i.e., lesser whole-body %LM) indicate lesser force-producing capabilities while greater whole-body %FM indicates greater non-force-producing soft tissue. Lesser %LM and greater %FM yield worse physical performance during dynamic, whole-body movements such as SLH, which may also be linked to the worse KOOS ADL scores, as observed in the present study. Therefore, suboptimal whole-body LM:FM ratios of (i.e., lower %LM and greater %FM) may not only result in greater perceived difficulty of functional activities of daily life (worse KOOS ADL) but also worse performance on physical function assessments (SLH). Data in ACLR cohorts 4–6 years post-ACLR indicate that overall soft tissue composition and quadriceps strength of the ACLR limb may remain altered for years post-ACLR compared with controls (21, 46, 53). Improving and maintaining quadriceps strength and physical function is integral to reducing the risk of future injury, mitigating KOA development, and maintaining overall function after ACLR.

Most previous research and clinical applications have largely sought to evaluate the link between BMI and ACLR-related outcomes, as BMI can be assessed at early time-points post-ACLR and is an easily measured outcome linked to future knee-related declines in individuals with high BMI (15). While BMI may be a clinically feasible measure, it cannot discern the contributions of LM and FM to total body mass. Therefore, BMI cannot provide a comprehensive evaluation of key soft tissue characteristics (%LM and %FM) that are associated with determinants of long-term joint tissue health and disease (e.g., KOA). Similar to previous literature in cohorts of similar BMI as the participants in the present study (23.8±2.7 kg·m⁻²) (17, 52, 54, 55), we found weak and non-significant associations between BMI and patient-reported outcomes, quadriceps strength, and physical performance in individuals post-ACLR who were of normal BMI status. BMI is a non-specific measure for quantifying soft tissue characteristics and some previous research has found BMI to be associated with KOOS and SLH 1–5 years post-ACLR (21). However, participants in previous studies have slightly higher BMI (26.1±3.1 kg·m⁻²) compared with the average BMI in the present study (23.8±2.7 kg·m⁻²).Thus, while high BMI (≥25 kg·m⁻²) is a significant predictor for radiographic and symptomatic KOA development in ACLR patients (56) and the general population (57), most patients with ACLR are of normal BMI status (17, 52, 54, 55) thus highlighting the limited potential for BMI to be used as a clinical predictor and intervention target for ACLR outcomes. It is possible that in individuals with normal BMI, but higher whole-body and region-specific %FM may still be indicative of worse outcomes. Previous research has shown that improvements in %LM and decreased %FM after an exercise intervention in individuals with KOA associates with better physical performance, strength, and patient-reported knee function (11, 12). Thus, factors beyond BMI such as body composition represent a critical and modifiable target to consider when evaluating clinical knee-related outcomes in populations at risk for declines in knee joint health. Our findings further support previous evidence in individuals with KOA (9–12) and young adults of similar BMI status with and without knee joint injury (20–22, 46) indicating that body composition is linked to knee-related outcomes in patients after ACLR. Importantly, body composition is a modifiable factor which can be assessed and modified pre- and early post-operatively (58) – time periods at which strength and function rehabilitation interventions are limited for ACLR patients. Future research should seek to determine whether assessing and modifying body composition at these critical timepoints post-ACLR can subsequently yield improved short- and long-term outcomes for ACLR patients.

While the present study introduced novel associations between body composition and important clinical outcomes following ACLR, there are some limitations that can inform future research. The present study determined cross-sectional associations between ACLR outcomes and body composition and thus causality cannot be inferred from these data. Future research should aim to identify longitudinal and mechanistic associations between body composition and ACLR outcomes to determine causality between these factors. Further, we assessed individuals who were 52 (±37) months post-ACLR, and future research should seek to evaluate body compositional changes within the first 6 months post-ACLR – a time period in which body composition changes (e.g., decreased LM and increased FM) occur (59). When accounting for time since ACLR in post-hoc analyses, the associations between body composition and KOOS, MVIC, and SLH were influenced by time since surgery. Therefore, future research may seek to evaluate early post-ACLR body composition shifts and associations with poor long-term outcomes as this study is one of the first to identify key cross-sectional links between body composition and clinical ACLR outcomes. BMI did not associate with any clinical ACLR-related outcomes collected in the present study, thereby highlighting the need for future research to incorporate measures of body composition when seeking to understand factors linked to poor ACLR-related outcomes. It should be noted that the average BMI within our cohort was 23.8±2.7 kg·m⁻², which is similar to other ACLR cohorts and uninjured matched controls (22, 56), but future research should seek to determine whether BMI is a better predictor for clinical outcomes in ACLR patients with high BMI. The demographic characteristics of the cohort in the present analysis are reflective of previous ACLR cohorts reported in literature (i.e., young adults, normal BMI, 9000+ steps/day, 75+ min MVPA/day) (42). Future research may seek to expand upon these findings and the findings of previous studies conducted in older adults and individuals with obesity as a comprehensive assessment evaluating the influence of body composition on knee-related outcomes. Such data can ultimately be used to identify individuals at-risk for worse long-term knee outcomes, along with aid in developing precision medicine interventions that target specific factors liked to poor outcomes (e.g., thigh region LM and FM). Lastly, body composition characteristics commonly differ between males and females, yet we were not adequately powered to conduct sex-specific subgroup analyses. Research should seek to identify whether differences in the results would be found after stratifying by sex. Despite these limitations, this is the first study to evaluate links between body composition and ACLR outcomes which may aid in explaining the conflicting evidence linking BMI to ACLR outcomes in those of normal BMI status as most ACLR individuals and uninjured controls are of normal BMI status (22, 56). We highlight that body composition is an important, modifiable factor linked to clinical outcomes in ACLR patients and might represent a factor to be considered in future research and clinical assessments.

CONCLUSIONS

Measures of body composition, but not BMI, were associated with clinically significant ACLR outcomes (e.g., patient-reported outcomes, strength, and physical performance). Quadriceps strength (MVIC) associated with all body composition outcomes (%LM and %FM of whole-body, whole-limb, and thigh region). Soft tissue composition of the ACLR limb may be a key factor when evaluating knee joint-specific outcomes (KOOS Symptoms, Pain, QoL) as whole-limb and thigh region composition associated with these knee joint-specific outcomes. Yet, when evaluating whole-body, dynamic movements (SLH) and patient-reported function (KOOS ADL), whole-body %LM and %FM seem to be more representative of these ACLR outcomes. Nonetheless, across all associations, lower %LM and greater %FM were associated with worse ACLR outcomes. Altogether, these data show novel associations between body composition and key outcomes following ACLR in patients who are at-risk for KOA development and persistent knee-related disability.