The Longitudinal Relationship Between Body Composition and Patella Cartilage in Healthy Adults

Authors

  • Andrew J. Teichtahl,

    1. Department of Epidemiology and Preventive Medicine, Central and Eastern Clinical School, Alfred Hospital, Monash University, Melbourne, Victoria, Australia
    Search for more papers by this author
  • Yuanyuan Wang,

    1. Department of Epidemiology and Preventive Medicine, Central and Eastern Clinical School, Alfred Hospital, Monash University, Melbourne, Victoria, Australia
    2. Faculty of Life and Social Sciences, Swinburne University of Technology, Hawthorn, Victoria, Australia
    Search for more papers by this author
  • Anita E. Wluka,

    1. Department of Epidemiology and Preventive Medicine, Central and Eastern Clinical School, Alfred Hospital, Monash University, Melbourne, Victoria, Australia
    2. Baker Heart Research Institute, Melbourne, Victoria, Australia
    Search for more papers by this author
  • Maxine Szramka,

    1. Department of Epidemiology and Preventive Medicine, Central and Eastern Clinical School, Alfred Hospital, Monash University, Melbourne, Victoria, Australia
    Search for more papers by this author
  • Dallas R. English,

    1. Centre for MEGA Epidemiology, University of Melbourne, Carlton, Victoria, Australia
    2. Cancer Epidemiology Centre, The Cancer Council of Victoria, Carlton, Victoria, Australia
    Search for more papers by this author
  • Graham G. Giles,

    1. Department of Epidemiology and Preventive Medicine, Central and Eastern Clinical School, Alfred Hospital, Monash University, Melbourne, Victoria, Australia
    2. Centre for MEGA Epidemiology, University of Melbourne, Carlton, Victoria, Australia
    3. Cancer Epidemiology Centre, The Cancer Council of Victoria, Carlton, Victoria, Australia
    Search for more papers by this author
  • Richard O'Sullivan,

    1. MRI Unit, Mayne Health Diagnostic Imaging Group, Epworth Hospital, Richmond, Victoria, Australia.
    Search for more papers by this author
  • Flavia M. Cicuttini

    Corresponding author
    1. Department of Epidemiology and Preventive Medicine, Central and Eastern Clinical School, Alfred Hospital, Monash University, Melbourne, Victoria, Australia
    Search for more papers by this author

(flavia.cicuttini@med.monash.edu.au)

Abstract

Background: Although obesity is a risk factor for patellofemoral osteoarthritis (OA), it is unclear whether the components of body composition, such as muscle and fat mass, are major determinants of articular cartilage properties at the patella.

Objective: The aim of this study was to determine whether anthropometric and body composition measures, assessed over 10 years, were related to articular patella cartilage volume and defects in healthy adults with no clinical knee OA.

Methods and Procedures: Two hundred and ninety-seven healthy, community-based adults aged 50–79 years with no clinical history of knee OA were recruited. Anthropometric and body composition (fat-free mass and fat mass) data were measured at baseline (1990–1994) and follow-up (2003–2004). Patella cartilage volume and defects were assessed at follow-up (2003–2004) using magnetic resonance imaging (MRI).

Results: After adjustment for potential confounders, increased measures of obesity (weight, BMI, waist circumference, and fat mass) at baseline and follow-up were associated with an increased risk for the presence of patella cartilage defects at follow-up for both men and women (all P ≤ 0.03). Increased baseline values for these variables tended to be associated with reduced patella cartilage volume at follow-up for women (all P ≤ 0.11), but not men (all P ≤ 0.87).

Discussion: We have demonstrated that increased anthropometric measures of obesity, as well as fat mass, are associated with an increased risk for the presence of patella cartilage defects in both men and women. Women, but not men, with greater baseline body mass, particularly adipose-derived mass, appear to have an associated reduction in their patella cartilage volume. Interventions targeting a reduction in adipose tissue may help reduce the risk for the onset and progression of patellofemoral OA, particularly in women.

Introduction

Pathology involving the articular cartilage of the patella, such as patellofemoral osteoarthritis (OA) is common and has been implicated as the cause of significant symptoms and disability (1,2). McAlindon et al. demonstrated that although medial tibiofemoral and patellofemoral OA were both significantly associated with disability, higher disability scores were more common in people with patellofemoral OA (3). Despite this, few studies have examined factors associated with the health of patella cartilage. Moreover, these limited studies have predominantly been focused on biomechanical factors, such as the role of varus–valgus alignment (4,5,6), rather than systemic variables. Obesity, usually examined in the context of an increased BMI, is a well-recognized risk factor for OA at all compartments within the knee, including the patellofemoral joint (7,8,9,10). The mechanism by which obesity increases the risk of developing patellofemoral OA is unknown, and is perhaps related to the inability of the BMI to discriminate adipose from non-adipose body mass (11). It is unclear whether weight per se or the specific components of body composition such as fat mass or fat-free mass are associated with patella joint structure.

Measures of central adiposity, such as waist circumference and waist-to-hip ratio have proven to be better predictors of major public health problems such as diabetes and cardiovascular diseases than BMI (12). Although yet to be examined at the patellofemoral joint, most available data have suggested that fat distribution does not affect the risk of developing tibiofemoral OA (7,13,14). Nevertheless, these data may, in part, be limited by the use of conventional radiographic measurement, which provides only an approximation of articular cartilage. Using magnetic resonance imaging (MRI) to directly visualize cartilage three-dimensionally, it is possible to measure several properties of articular cartilage, including thickness, surface area and their product, cartilage volume, as well as surface lesions, such as cartilage defects. However, diurnal variations in cartilage thickness but not volume (15), as well as the difficulty in reselecting section locations at follow-up assessment to determine change in cartilage thickness (16) have been problematic factors associated with using cartilage thickness as an outcome measure. Subsequently, studies have tended to focus on the determinants of articular cartilage volume and defects, and we recently demonstrated a positive association between muscle mass and tibial cartilage volume in healthy subjects (17). No study has examined whether similar relationships between adipose and non-adipose tissue are apparent at the patella cartilage.

The aim of this study was to determine whether anthropometric measures, such as weight and the BMI, as well as body composition measures, such as fat and fat-free mass, measured over 10 years, were related to articular patella cartilage volume and defects, in a community-based population of healthy men and women with no clinical knee OA.

Methods and Procedures

Subjects. The study was conducted within the Melbourne Collaborative Cohort Study, a prospective cohort study of community-based people, aged 40–69 years at recruitment (1990–1994); with the aim of examining the role of lifestyle and genetic factors in the risk for cancer and chronic diseases from middle age and beyond (18). As our intent was to investigate subjects with no significant current or past knee disease, individuals were excluded if they had a clinical diagnosis of knee OA as defined by the American College of Rheumatology criteria (19), or if in the last 5 years they had had knee pain lasting for >24h; a previous knee injury requiring non-weightbearing treatment for >24h or surgery (including arthroscopy); or a history of any arthritis diagnosed by a medical practitioner. A further exclusion criterion was any contraindication to MRI. We invited subjects who fulfilled our inclusion criteria and who attended the first year of round 3 follow-up of the Melbourne Collaborative Cohort Study, which commenced in 2003. Furthermore, we used quota sampling whereby we stopped recruitment when we reached our target sample of ∼300 subjects.

There were no significant differences between this population and the original Melbourne Collaborative Cohort population which has the following profile: 61% females, aged 57.8 ± 3.0 years and BMI 25.7 ± 3.8kg/m2. There were no significant differences in dietary intake or other health-related behavior such as smoking: 60% subjects never smoked in this population vs. 57% in the Melbourne Collaborative Cohort population.

The study was approved by the Cancer Council Victoria Human Research Ethics Committee and the Standing Committee on Ethics in Research Involving Humans of Monash University. All participants gave written informed consent.

Anthropometric measures. At the baseline interview for the Melbourne Collaborative Cohort Study (1990–1994) (18), height, weight, and waist and hip circumferences were measured using standardized written protocols (20). BMI (weight/height2, kg/m2) and waist-to-hip ratio were calculated from these data, which were again collected at follow-up (2003–2004).

Bioelectrical impedance analysis was performed with a single frequency (50 kHz) electric current produced by a BIA-101A RJL system analyzer (RJL Systems, Detroit, MI) at baseline (1990–1994) in all 297 participants and again, in all 237 consecutive subjects, after this became available at follow-up (2003–2004). Resistance and reactance were measured with subjects in the supine position. We used bioimpedance analysis to estimate non-adipose mass, hereafter termed fat-free mass, as 9.1536 + (0.4273 × height2/resistance) + (0.1926 × weight) + (0.0667 × reactance) for men, and 7.7435 + (0.4542 × height2/resistance) + (0.119 × weight) + (0.0455 × reactance) for women (21). Adipose mass, hereafter termed fat mass (weight – fat-free mass) was subsequently calculated.

MRI and the measurement of patella cartilage volume and cartilage defects. Between 2003 and 2004, each subject had MRI performed on their dominant knee (defined as the lower limb the subject used to step-off from when initiating walking). Knees were imaged in the sagittal plane on a 1.5-T whole-body magnetic resonance unit (Phillips) using a commercial transmit–receive extremity coil. The following sequence parameters were used: a T1-weighted fat suppressed 3D gradient recall acquisition in the steady state; flip angle 55°; repetition time 58ms; echo time 12ms; field of view 16cm; 60 partitions; 513 × 196 matrix; one acquisition time 11min 56s. Sagittal images were obtained at a partition thickness of 1.5mm and an in-plate resolution of 0.31 × 0.83mm (512 × 196 pixels).

Patella cartilage and bone volumes were determined by image processing on an independent workstation using the Osiris software (University of Geneva). Contours were drawn around the patella in images 1.5mm apart on sagittal views. The coefficients of variation were 2.1% for patella cartilage volume and 2.2% for patella bone volume.

Cartilage defects were graded on the magnetic resonance images with a modification of a previous classification system at the patellofemoral site by a trained observer (22,23,24). Cartilage defects were said to be present when on at least two consecutive slices there were at least irregularities on the surface or base and loss of thickness <50%;. The same trained observer, as well as an independent observer, regraded the cartilage defects 1 month later blinded to the previous assessment. Intraobserver and interobserver reliability for the patellar compartment (expressed as an intraclass correlation) were 0.94 and 0.93, respectively (25).

Statistical analyses. Outcomes were patella cartilage volume and the presence of patella cartilage defects collected at follow-up. Patella cartilage volume was assessed for normality prior to linear regression. The presence/absence of patella cartilage defects was a dichotomous outcome, thus logistic regression was used. Multiple regression models were constructed to describe the relationship between anthropometric and body composition measures (at baseline and follow-up) and patella cartilage variables, adjusting for potential confounders including age and bone volume (for patella cartilage volume, as a measure of joint size). In terms of anthropometric and body composition measures as predictors of knee structure, we analyzed baseline parameters, follow-up parameters, and the change in these parameters (follow-up values – baseline values) separately. To further examine body mass distribution as a predictor of the different patella structural features, we fitted fat-free mass and fat mass simultaneously as continuous variables in the same model. Given the significant gender differences in body composition, all analyses were performed separately for men and women. P values < 0.05 were considered to be statistically significant. All analyses were performed using the SPSS statistical package (standard version 14.1.0, SPSS, Chicago, IL).

Results

Subject characteristics

The characteristics of the 297 participants are presented in Table 1, according to gender.

Table 1. . Characteristics of study subjects
 Total (n = 297)Males (111)Females (186)P*
  • Values are reported as mean (s.d.) except for percentages.

  • MRI, magnetic resonance imaging.

  • *

    Difference between males and females.

Age when MRI performed (years)58.0 (5.5)59 (6.2)57 (4.7)0.01
At baseline (1990–1994)    
    Height (cm)168.1 (9.0)176 (6.7)163 (6.4)<0.0001
    Weight (kg)71.3 (13.3)80.7 (11.0)65.7 (11.2)<0.0001
    Body mass index (kg/m2)25.2 (3.8)26.0 (3.1)24.7 (4.1)0.002
    Waist circumference(cm)80.2 (12.6)89.7 (9.3)74.5 (10.8)<0.0001
    Hip circumference(cm)98.9 (7.899.8 (6.2)98.3 (8.6)0.09
    Waist-to-hip ratio0.81 (0.09)0.90 (0.1)0.76 (0.1)<0.0001
    Fat-free mass (kg)47.0 (9.7)58.1 (5.3)40.3 (3.8)<0.0001
    Fat mass (kg)24.3 (8.4)22.5 (7.6)25.4 (8.7)0.004
At follow-up (2003–2004)    
    Patella cartilage volume (mm3)2,656 (886)3,412 (837)2,204 (540)<0.0001
    Patella bonevolume (mm3)20,276 (4,667)24,836 (3,385)17,550 (2,849)<0.0001
    Patella cartilage defects (%;)2821320.03
    Weight (kg)73.2 (14.0)82.4 (12.5)68.1 (12.5)<0.0001
    Body mass index (kg/m2)25.9 (4.2)26.5 (3.4)25.6 (4.7)0.05
    Fat-free mass (kg)46.3 (10.1)58.7 (6.2)39.8 (3.8)<0.0001
    Fat mass (kg)26.4 (8.9)25.3 (8.3)27.3 (9.1)0.10
Change data (follow-up-baseline)    
    Weight (kg)2.2 (5.3)1.7 (5.2)2.5 (5.3)0.23
    Body mass index (kg/m2)0.8 (1.9)0.5 (1.7)0.9 (2.0)0.07
    Fat-free mass (kg)−0.1 (2.8)−0.2 (3.7)−0.2 (2.2)0.46
    Fat mass (kg)2.5 (4.3)2.2 (4.5)2.7 (4.2)0.45

Men had significantly greater anthropometric (height, weight, BMI, waist and hip circumference, and waist-to-hip ratio) and fat-free mass than women, although women had significantly greater fat mass than men at baseline. At the patella, men had greater patella cartilage and bone volume than women, although the women had a greater prevalence of patella cartilage defects than men (Table 1).

Relationship between anthropometric and body composition measures and the presence of patella cartilage defects

The associations between anthropometric and body composition measures and the presence of patella cartilage defects measured at follow-up are shown in Table 2 for men and Table 3 for women.

Table 2. . Relationship between anthropometric and body composition measures and presence of patella cartilage defects at follow-up for males only
inline image
Table 3. . Relationship between anthropometric and body composition measures and presence of patella cartilage defects at follow-up for females only
inline image

At baseline, weight, BMI, waist circumference, and fat mass were associated with an increased risk for the presence of patella cartilage defects for both men and women in univariate analyses. In multivariate analyses after adjusting for age and patella cartilage volume, these variables all remained significantly associated with the risk for the presence of patella cartilage defects for both men and women, with odds ratios (ORs) ranging from 1.07 to 1.31 (all P ≤ 0.03). However, after adjustment for fat-free mass and fat mass, i.e., including age, patella cartilage volume, baseline fat-free mass, and baseline fat mass in the regression model, fat mass was the only body composition measure to remain significantly associated with the risk for the presence of patella cartilage defects for both men (OR = 1.18 (95% confidence interval 1.07–1.30), P = 0.001) and women (OR = 1.10 (95% confidence interval 1.04–1.15), P = 0.001).

Similar trends to baseline were demonstrated when follow-up anthropometric and body composition measures were examined in relation to patella cartilage defects, where follow-up weight and BMI were both associated with an increased risk for the presence of patella cartilage defects for both men and women, before and after adjustment for potential confounders. Follow-up fat mass was associated with an increased risk for patella cartilage defects at follow-up for men (P = 0.037) and tended toward significance for women (P = 0.09) in multivariate analyses when adjusted for age, patella cartilage volume, fat-free mass, and fat-mass. Fat-free mass at follow-up was associated with the risk for patella cartilage defects at follow-up in women (P = 0.02) but not men (P = 0.34).

After adjustment for age, patella cartilage volume and the respective baseline parameter, change in weight (P = 0.01), BMI (P = 0.02), fat mass (P = 0.02) and fat-free mass (P = 007) all tended to be significantly associated with the risk for patella cartilage defects in women only (OR ranging from 1.10 to 1.36). In men, only change in fat-free mass was associated with the risk for the presence of patella cartilage defects (P = 0.037). After fat mass and fat-free mass, as well as change in each variable were added to the regression equation, the association between change in fat-free mass and the risk for the presence of patella cartilage defects remained significant for both men (P = 0.04) and women (P = 0.003). Change in fat-mass tended toward significance for women (P = 0.08), but not men (P = 071).

Relationship between anthropometric and body composition measures and patella cartilage volume

The associations between anthropometric and body composition measures and the patella cartilage volume measured at follow-up are shown in Table 4 for men and Table 5 for women.

Table 4. . Relationship between anthropometric and body composition measures and patella cartilage volume at follow-up for males only
inline image
Table 5. . Relationship between anthropometric and body composition measures and patella cartilage volume at follow-up for females only
inline image

At baseline, there was a tendency for increased weight (P = 0.007), BMI (P = 0.02), waist circumference (P = 0.11), fat-free mass (P = 0.03) and fat mass (P = 0.01) to be associated with a reduction in follow-up patella cartilage volume after adjustment for age and patella bone volume for women only. However, after adjustment for fat-free mass and fat mass, i.e. including age, gender, patella bone volume, baseline fat-free mass and baseline fat mass in the regression model, fat mass was the only body composition measure that tended to remain significantly associated with a reduction in the follow-up measure of patella cartilage volume (P = 0.11) in women only.

Similar trends to baselinewere demonstrated when follow-up anthropometric and body composition measures were examined in relation to patella cartilage volume, where follow-up weight (P = 0.01), BMI (P = 0.03), fat-free mass (P = 0.02) and fat mass (P = 0.03) were all associated with a reduction in follow-up patella cartilage volume when adjusted for age and patella volume, for women only. When fat mass and fat-free mass were added to the regression equation, no body composition measures remained significantly associated with the patella cartilage volume at follow-up for either men or women.

Change in anthropometric or body composition variables was not associated with follow-up patella cartilage volume for either men or women.

Discussion

In this population of healthy adults with no clinical knee OA, we have demonstrated that increased anthropometric measures of obesity, as well as fat mass, are associated with deleterious changes in patella cartilage health (as determined by the presence of patella cartilage defects) for both men and women. Women, but not men, with greater baseline body mass, particularly adipose-derived mass, had reduced patella cartilage volume 10 years later. Moreover, women appear to be at specific risk for the presence of patella cartilage defects if they gain body mass over the period of 10 years.

Previously, Ding et al. demonstrated that BMI was negatively associated with patella cartilage thickness and volume in a larger (n = 372) cross-sectional study of people without clinical knee OA (26). Similarly, in this smaller (n = 297) study of healthy adults, our follow-up data demonstrated that women (but not men) with greater weight or BMI have an associated reduction in patella cartilage volume. We also demonstrated that women with greater weight or BMI at baseline have an associated reduction in patella cartilage volume 10 years later. Furthermore, we have shown that increased BMI and weight at baseline are associated with an increased risk for the presence of patella cartilage defects 10 years later in both men and women. To our knowledge, this is the first study to have examined and reported an association between either the BMI or weight and patella cartilage defects. Nevertheless, although the BMI is an adequate measure of body mass relative to height, it does not distinguish between adipose and non-adipose body mass (11).

No other study has examined the relationship between body composition measures and patella cartilage defects and volume. We have demonstrated that increased fat mass (at baseline and follow-up) tended to be significantly associated with an increased risk for the presence of patella cartilage defects in both men and women. There was also a tendency for increased fat mass at baseline to be associated with a reduction in patella cartilage volume 10 years later in women, but not men. Cartilage defects have previously been shown to be associated with subsequent loss in cartilage volume, leading to the inference that the defects may represent early cartilage abnormalities (27). Why women, but not men with increased baseline BMI, weight and fat mass are particularly susceptible to an associated reduction in their cartilage volume after 10 years is unclear. Several studies have previously documented a female predisposition toward a reduction in cartilage volume that is associated with increased weight (28,29), although the reasons for this gender disparity remain elusive. It is also unclear why gain in weight, BMI, fat-free mass, and fat mass are associated with an increased risk for patella cartilage defects in women, but not men. It appears that the effects of longitudinal gain in body mass, particularly fat mass, is more detrimental to the patella cartilage of women rather than men. The greater magnitude of adipose mass in women, relative to men, may provide useful clues toward the potential mechanisms that might account for this gender disparity.

Mechanistically, it may be that increased joint loading and the potential changes in retropatellar forces imparted by any increase in body mass mediates patella cartilage defects. However, if this were the case then it would be expected that increased fat-free mass would consistently be associated with patella cartilage defects. Given that this was not the case, it may be that the association between BMI, weight, and fat mass and the presence of patella cartilage defects is reliant upon adipose/systemic-derived factors, rather than biomechanical ones. Adipose tissue was previously thought to be a passive store of energy but is now considered an endocrine organ, releasing a multitude of factors, including cytokines such as tumour necrosis factor and interleukin 1, as well as adipokines, such as leptin, adiponectin, and resistin (30). Dysregulation of lipid homeostasis may therefore be crucial in mediating the obesity–OA relationship. For instance, leptin receptors have been found at articular cartilage (31) and significant levels of leptin were observed in the cartilage and osteophytes of people with OA (32). It may also be that obesity has an indirect effect via elevations in cytokines such as interleukin 1 and tumor necrosis factor-α, both of which have recently been shown to play a key role in cartilage destruction in OA (33,34). Nevertheless, the potential contribution to joint destruction in OA from obesity-related increases in these adipokines and cytokines is unclear. Interventions targeting a reduction in adipose tissue may potentially help to reduce the risk for the onset and or progression of patellofemoral OA, particularly in women.

In this study, we examined asymptomatic people and used novel and very sensitive measures to assess knee joint structural features that have been shown to form part of a continuum from the normal to the diseased joint, namely cartilage defects and cartilage volume (25,27). Nevertheless, this study was limited by the absence of baseline MRI data, so we were unable to examine longitudinal change in cartilage volume and defects. However, this study was strengthened by the availability of prospectively collected anthropometric and body composition data over ∼10 years. Moreover, this study was strengthened by the separate analyses performed for men and women, given that body composition is markedly different between the two sexes.

This study has demonstrated that increased anthropometric measures of obesity, as well as fat mass, are associated with deleterious changes in patella cartilage health (as determined by the presence of patella cartilage defects) for both men and women. Women with greater baseline body mass, particularly adipose-derived mass, appear more prone than men to having associated reductions in their patella cartilage volume with the passage of time. Moreover, women appear to be at specific risk for the presence of patella cartilage defects if they gain body mass over 10 years. Interventions targeting a reduction in adipose tissue may help reduce the risk for the onset and/or progression of patellofemoral OA, particularly in women.

Disclosure

The authors declared no conflict of interest.

Acknowledgments

The Melbourne Collaborative Cohort Study recruitment was funded by VicHealth and The Cancer Council of Victoria. This study was funded by a program grant from the National Health and Medical Research Council (NHMRC; 209057) and was further supported by infrastructure provided by The Cancer Council of Victoria. We acknowledge the NHMRC (project grant 334150), Colonial Foundation, and Shepherd Foundation for support. Dr Wang is the recipient of an NHMRC PhD Scholarship. Dr Wluka is the recipient of an NHMRC Public Health Fellowship. We especially thank the study participants who made this study possible.

Ancillary