- Email: [email protected]
Associations of body mass index with meniscal tears
Associations of Body Mass Index with Meniscal Tears Gregory M. Ford, MSPH, Kurt T. Hegmann, MD, MPH, George L. White, Jr, PhD, MSPH, Edward B. Holmes, MD, MPH Background: Meniscal tears are common knee injuries, with limited reported data on associated factors, let alone risk factors. The objective of this study was to determine whether associations exist between increasing obesity and meniscal tears leading to surgery. Methods:
We performed frequency-matched case– control studies using surgical case data for years 1996 to 2000 from administrative databases of two large Utah hospitals; each case was matched with three controls from a large cancer screening trial. Meniscal tear cases (262 male and 282 female) were determined by surgical procedures. Inclusion criteria were age (50 to 79) and body mass index (BMI) (17.00 to 54.99 kg/m2). Gender-specific, age-adjusted odds ratios with 95% confidence intervals (CIs) were calculated for BMI categories from ⬍20.00 to ⱖ40.00. The referent BMI category was 20.00 to 22.49.
Results:
Age-adjusted odds ratios for likelihood of meniscal surgery among those with a BMI of ⱖ40.00 were 15.0 (95% CI⫽3.8 –59.0) for men, and 25.1 (95% CI⫽10.3– 60.8) for women. All odds ratios for men and women with BMIs of ⱖ27.50 and ⱖ25.00, respectively, were statistically significantly elevated.
Conclusions: Significant associations were demonstrated between increasing BMI and meniscal surgeries in both genders, including obese and overweight adults. (Am J Prev Med 2005;28(4):364 –368) © 2005 American Journal of Preventive Medicine
Introduction
K
nee meniscal injuries are common with the prevalence of degenerative meniscal changes in reported asymptomatic individuals ranging between 11.1% and 31.5%.1–3 In autopsy studies the prevalence of horizontal cleavage lesions of the menisci has been reported between 18.6% and 60%.4,5 While common injuries, little epidemiologic data have been reported. Meniscal tears are commonly associated with athletic injuries.6,7 Athletes with greater mean heights and weights are noted to have significantly greater incidence of knee joint injuries.8 Height and weight have also been correlated with degenerative meniscal magnetic resonance imaging (MRI) findings.9 Degenerative meniscal changes are thought to predispose the meniscus to the development of symptomatic meniscal tears.9 Obesity is thought to increase subchondral bony stiffness, transmitting more force to overlying cartilage,10,11 possibly suggesting an injury mechanism involving obesity. While a relatively high incidence of knee injuries in From Public Health Programs (Ford, White) and The Rocky Mountain Center for Occupational and Environmental Health (Hegmann, Holmes), University of Utah School of Medicine, Department of Family and Preventive Medicine, Salt Lake City, Utah Address correspondence and reprint requests to: Kurt T. Hegmann, MD, MPH, The Rocky Mountain Center for Occupational and Environmental Health, 391 Chipeta Way, Suite C, Salt Lake City UT 84108. E-mail: [email protected].
364
obese people has been suggested,12 no epidemiologic studies assessing the risks for meniscal tears from obesity were found in a comprehensive review of the literature. The purpose of this study was to determine whether relationships exist between body mass index (kg/m2 [BMI]) and meniscal tears leading to surgery.
Methods This is a retrospective case– control study. Institutional Review Board approval was obtained.
Cases All cases were obtained for the years 1996 to 2000, and were limited to Utah residents to simultaneously limit the potential for referral biases, and to have access to a large populationbased data set from which to draw controls. The coding system that was relied on in each hospital setting to invoice services was the system used to define cases for this study. Cases from LDS Hospital in Salt Lake City, Utah, were defined by International Classification of Diseases, 9th Revision (ICD-9) procedural codes for partial and total meniscectomy (80.60). Cases from Utah Valley Regional Medial Center (UVRMC) in Provo, Utah, were identified by current procedural terminology (CPT) codes for meniscectomy (29880, 29881) and meniscal repair (29882, 29883). These comprised ⬎90% of the meniscal cases in the study, and those data were also hand-extracted from the medical records. CPT codes were accompanied by descriptions of the surgeries and were
Am J Prev Med 2005;28(4) © 2005 American Journal of Preventive Medicine • Published by Elsevier Inc.
0749-3797/05/$–see front matter doi:10.1016/j.amepre.2005.01.013
Table 1. Description of meniscal surgery cases and controlsa Male Age (years) BMI (kg/m2)
Cases
Controls
n⫽262 59.9 ⫾ 7.8 29.9 ⫾ 5.3
n⫽786 61.8* ⫾ 5.5 27.4* ⫾ 3.9
Female Age (years) BMI (kg/m2)
Cases
Controls
n⫽282 61.4 ⫾ 7.8 31.2 ⫾ 6.8
n⫽846 62.2 ⫾ 5.9 27.0* ⫾ 5.0
BMI, body mass index. a Values are mean ⫾ standard deviation. *p ⬍ 0.05, t-test comparison of means between cases and controls.
accurately coded in all cases. Heights and weights (used to calculate BMI) were measured prior to surgery and were used in this study. A computerized medical record for the LDS Hospital was used to obtain age, gender, and BMI data for cases from that facility. Coding accuracy rates of 94% for ICD-9 and 93% for CPT codes are reported by this facility. All cases from LDS Hospital were inpatients. At UVRMC, medical records were used to obtain age, gender, height, weight, and inpatient or outpatient status. Of the 515 meniscal surgery patients in this study, 29 (5.6%) had two meniscal surgeries, one on each knee, and both were included. All cases ⬍50 years were excluded due to the limited age range of the controls. These exclusions may also help to limit possible confounding due to athletic injuries from adolescent sports. All cases aged ⬎79 years were excluded because of the limited age range of the controls (as described below). Cases with BMI ⬍17 or ⬎55 were excluded due to possible information error.
Controls Controls were drawn from a group of 9944 Utah residents enrolled in the National Cancer Institute’s Prostate, Lung, Colorectal, and Ovarian Cancer Screening Trial from the years 1996 to 2000. Enrollees were healthy men and women aged 55 to 74, and are described elsewhere.13 Each control’s age and self-reported height and weight were provided. BMI was then calculated for each control. The information regarding these controls was optically scanned with an edit program that catches nonsense and out-of-range values.14
Statistical Methods A multistage frequency match was performed to age (within 5 years) and gender match three controls for each case. The cases were divided into ten categories (in 2.5 BMI increments), with 20.00 to 22.49 serving as the referent category. Crude gender-specific odds ratios and Mantel–Haenszel ageadjusted, sex-specific odds ratios were calculated. The age strata utilized for age-adjustment of male and female meniscal cases were 50 to 59, 60 to 69, and 70 to 79. Chi-square tests for linear trend were calculated for the odds ratios. Logistic regression models were analyzed with and without age as a continuous independent variable. All data analyses were performed in 2004 using SPSS 10.0 (SPSS, Inc., Chicago IL, 1999).
Results There were 262 male meniscal surgery cases (31 excluded for age ⬍50) and 282 female meniscal surgery
cases (21 excluded for age ⬍50 and 1 excluded for age ⬎79) included. No cases were out of the quality control range for BMI. Cases and controls were generally comparable regarding residence in Utah, age, and gender. However, male cases tended to be slightly younger than the controls (see Table 1). No significant differences in BMIs were found between inpatient and outpatient cases (male meniscus: p ⫽1.00, female meniscus: p ⫽0.11). There was no statistically significant difference between the BMIs of the 29 meniscal surgery patients who were included twice and the other meniscal surgery patients (p ⫽0.54). A relationship was identified between BMI and meniscal surgery. Chi-square tests of linear trend were highly significant for males and females (p ⬍0.0001). Elevated age-adjusted odds ratios were statistically significant in men with a BMI of ⱖ27.5, and in women with a BMI of ⱖ25.0. Mantel–Haenszel age-adjusted odds ratios for those with a BMI of ⱖ40.0 were 15.0 (95% confidence interval [CI], 3.8 –59.0) in men and 25.1 (95% CI, 10.3– 60.8) in women (Table 2). No appreciable differences existed between crude and Mantel–Haenszel age-adjusted odds ratios, indicating a negligible potential for residual confounding due to age. The odds ratios from the logistic regression models that adjusted for age were not appreciably different from the odds ratios of the models that did not adjust for age, again indicating a negligible potential for residual confounding due to age (unadjusted odds ratios not shown). The odds ratios from the logistic regression models were similar to the above-mentioned crude and Mantel–Haenszel adjusted odds ratios (see Table 2).
Discussion There is a dose–response relationship between BMI and meniscal surgery in middle-aged to older adults in both genders. A relationship had been previously suspected (T Sisk, The Orthopaedic Clinic, Memphis TN, personal communication, February 27, 2002), but never measured. Since 57.4% (164 million) of the U.S. adult population is either overweight (BMI⫽25 to 29.9) or obese (BMIⱖ30),15 this relationship has potentially large implications on a population basis for meniscal surgeries. Am J Prev Med 2005;28(4)
365
Table 2. ORs for meniscal surgeries by BMI in men and women Men
Women
BMI (kg/m2)
ORMH
95% CI
ORLR
95% CI
ORMH
95% CI
ORLR
95% CI
⬍20.00 20.00–22.49 22.50–24.99 25.00–27.49 27.50–29.99 30.00–32.49 32.50–34.99 35.00–37.49 37.50–39.99 ⱖ40.00
3.5 1.0 1.8 1.6 3.0 4.7 4.9 6.5 6.2 15.0
(0.7–16.2) — (0.7–4.6) (0.7–4.0) (1.2–7.4) (1.9–12.0) (1.9–12.6) (2.1–20.2) (1.7–23.0) (3.8–59.0)
4.5 1.0 1.9 1.6 3.0 4.8 4.7 5.6 7.5 14.6
(1.0–19.8) — (0.7–4.7) (0.7–4.0) (1.2–7.3) (1.9–12.1) (1.8–12.3) (1.9–16.5) (2.1–27.0) (4.0–53.7)
0.9 1.0 1.5 2.8 2.9 4.7 5.5 3.1 9.2 25.1
(0.3–3.1) — (0.7–3.0) (1.4–5.5) (1.5–5.8) (2.4–9.4) (2.5–11.9) (1.3–7.6) (3.4–24.5) (10.3–60.8)
1.0 1.0 1.5 2.7 2.9 4.7 5.1 2.9 9.6 24.3
(0.3–3.1) — (0.7–3.0) (1.3–5.3) (1.5–5.8) (2.4–9.3) (2.4–10.6) (1.2–7.1) (3.7–25.3) (10.3–57.2)
BMI, body mass index; CI, confidence interval; OR, odds ratio; ORMH, Mantel Haenszel age-adjusted odds ratio; ORLR, estimated age-adjusted odds ratio from logistic regression.
The menisci are thought to provide congruity and stabilization through load transfer across the knee joint, and act as shock absorbers, relieving stress transmitted through articular cartilage and subchondral bone.16 –19 Menisci are believed to transmit over half of the load to the knee joint.20 Thus, there is a biomechanical mechanism for the relationship between BMI and meniscal tears. As body mass index increases, the strain and torque in the knee joint during rotation likely increases, in theory leading to a higher risk for meniscal injuries. Whether other mechanisms contribute, such as reduced blood flow to the menisci17,21–24 or low-grade inflammation associated with obesity,25 is unclear. In those who are obese, blood supply to the menisci could be limited through either vascular compression effects of mechanical weight bearing or associations with cardiovascular risk factors. Meniscal blood supply is known to be limited to the periphery, with the presence of vessels ranging from the outer 10% to 25%.17,21–24 The blood supply is thought to be important for the healing of tears.17,22,26 A longitudinal tear in the avascular portion of the meniscus appears to require a vascular access channel at the midpoint to heal.22 Repairs of meniscal tears in the peripheral third zone also demonstrate superior healing rates compared with repairs in the central third zone (87% vs 59%).16 Because the focus of meniscal tears in this study was in older adults, healing of partial tears with respect to age may have a significant effect. However, while there is widespread belief of a strong effect of age on meniscal tear healing, this has not been well demonstrated in the literature.16,27–31 Still, the effects of aging on the menisci appear well established, and include biochemical changes,32 vascular changes,19,33,34 degenerative changes,1,9 cell surface changes,35 and prevalence of abnormalities on MRIs.36 Some studies report no distinct pattern with respect to age and healing,16,27,28 while others do report a distinct pattern with respect to age and healing.29 –31 366
The American Academy of Orthopaedic Surgeons Research Department estimates that surgical procedures of the meniscus are performed on approximately 850,000 patients each year. The meniscal surgeries in this study cost an average of $3000. Assuming that 57% of the U.S. adult population is overweight or obese, and conservatively assuming a threefold meniscal surgery risk among them, the excess surgery rate would be approximately 450,000 surgeries. This is a potential cost savings of $1.35 billion annually had they not been overweight or obese.
Limitations Temporality issues cannot be resolved with a case– control study. However, the dose–response relationships reported in this study suggest a true association between increasing BMI and the need for meniscal surgery. The results of this study may not be generalized to conservatively managed meniscal tears since cases were selected using surgical codes. If conservatively managed meniscal tears were included, the odds ratios calculated could potentially have been greater. In ascertaining the appropriateness of the case surgeries, we relied on the accuracy of the hospital coding system and doctors’ reporting in medical records. Excluding cases aged ⬍50 years, due to the limited age range of the controls, was also thought to help limit possible confounding due to athletic injuries from adolescent sports. However, it is possible that those who suffer knee or joint injury in their youth may be at greater risk for having a similar injury later in life. Therefore, adolescent sports injuries may have played a role in the meniscal tear surgery cases used in this study despite our efforts to limit them with a suspicion that there may be interaction between BMI and those injuries. This study could be subject to information bias regarding BMI. Weight for each control was self-re-
American Journal of Preventive Medicine, Volume 28, Number 4
ported, whereas each case was weighed before surgery. Self-reported weight has been noted to be underestimated, especially for females.37–39 The magnitude of under-estimation38 regarding BMI has been estimated to be ⱕ1.14; thus, the results would not be appreciably different (data not shown). However, if a difference did exist, it would tend to reduce the reported odds ratios. The control set’s BMIs were compared to BMIs in both the Utah Behavioral Risk Factor Surveillance Survey data and the Utah portion of the National Health and Nutrition Examination Survey III data. (Both sets of BMI were from self-reported data, which makes them comparable to our controls, although this is a limitation with respect to the measured data of the cases.) In both comparisons, our controls actually had slightly higher BMIs than those of the two other data sets. The BMI of this study’s control group was approximately 1.96% greater than other Utah populationbased controls. Therefore, these results may be slightly biased toward a less strong association than actually exists. The study was limited to Utah residents to reduce possible bias from out-of-state referrals among cases, and to allow access to a large control population. Utah’s population has been reported to have a slightly lower proportion of overweight people than the national average (53.3% vs 57.4% [BMI⬎25.0], respectively).15 This may also lead to a slight under-estimation of true odds ratios for the U.S. population. However, as BMIs are not substantially different in other U.S. states, the results are considered generalizable. Data on the duration of obesity and other potential risk factors for meniscal tears (e.g., highest BMI ever, occupation, sports, tobacco use, genetics, ethnicity, and reproductive variables in women) were not available. Arthritis has been associated with obesity,10,11 and therefore could cause some confounding in this study, as well as potential interaction. We suspect that previous injury history also may be both an effect modifier and confounder, although data necessary for this analysis have similarly never been reported. If vascular models turn out to be important for this relationship due to a meniscal vascular supply impairment mechanism, smoking may be a potential confounder, in contrast with arthritis, where if anything, smoking is slightly protective for knee arthritis. Therefore, future research including these factors is needed.
Conclusion The strength of the association and dose–response characteristic demonstrated in this study suggest a significant relationship between increasing BMI and the need for meniscal surgery in both males and females. It also may imply that a population-based weight management program could decrease the future burden on orthopedic and medical care systems due to meniscal surgeries and treatment of other
What This Study Adds . . . There is a belief that increasing body mass index (BMI) is associated with meniscal tears. However, no data have been published testing this association. This case– control study demonstrates that there is a strong association between increasing BMI and meniscal tears. These results may serve as a basis for patient education and preventive measures. Future research should evaluate for potential interaction among age, obesity, and osteoarthritis.
obesity-related conditions. This study also suggests that not only obese adults, but also overweight adults, have an increased likelihood of needing meniscal surgery. We were unable to find another study testing this association. We are grateful for the help of Brent James, MD, in obtaining data from Intermountain Health Care. We also thank Saundra S. Buys, MD, for assisting with access to a large control population, the Prostate Lung Colorectal and Ovarian Cancer Screening Trial, University of Utah (contract N01-CN25524). This study was partially supported by the Centers for Disease Control and Prevention, National Institute for Occupational Safety and Health (training grant T42/CCT810426-06-1). No financial conflict of interest was reported by the authors of this paper.
References 1. Jerosch J, Castro WHM, Assheuer J. Age-related magnetic resonance imaging morphology of the menisci in asymptomatic individuals. Arch Orthop Trauma Surg 1996;115:199 –202. 2. Kornick J, Trefelner E, McCarthy S, Lange R, Lynch K, Jokl P. Meniscal abnormalities in the asymptomatic population at MR imaging. Radiology 1990;177:463–5. 3. LaPrade RF, Burnett QM, Veenstra MA, Hodgman CG. The prevalence of abnormal magnetic resonance imaging findings in asymptomatic knees. Am J Sports Med 1994;22:739 – 45. 4. Noble J. Lesions of the menisci: autopsy incidence in adults younger than fifty-five years old. J Bone Joint Surg 1977;59A:480 –3. 5. Noble J, Hamblen DL. The pathology of the degenerate meniscus lesion. J Bone Joint Surg 1975;57B:180 – 6. 6. Chen E, Maffulli N, Chan KM. Knee injuries produced by recreational sports follows a different pattern than casual injuries. Bull Hosp Jt Dis 1998;57:74 –9. 7. Strickland JW. Philosophy of the treatment of athletes. Clin Sports Med 1995;14:285– 8. 8. Grace TG, Sweetser ER, Nelson MA, Ydens LR, Skipper BJ. Isokinetic muscle imbalance and knee-joint injuries. J Bone Joint Surg 1984;66:734 – 40. 9. Negendank WG, Fernandez-Madrid FR, Heilbrun LK, Teitge RA. Magnetic resonance imaging of meniscal degeneration in asymptomatic knees. J Orthop Res 1990;8:311–20. 10. Dequeker J, Goris P, Uytterhoeven R. Osteoporosis and osteoarthritis (osteoarthrosis). JAMA 1983;249:1448 –51. 11. Felson DT, Anderson JJ, Naimark A, Walker AM, Meenan RF. Obesity and knee osteoarthritis. Ann Intern Med 1988;109:18 –24. 12. Sisk TD. Traumatic disorders of the joints. In: Canale ST, Daugherty K, Jones L, eds. Campbell’s operative orthopaedics, vol. 3. 9th ed. St. Louis: Mosby, 1998:1465–766.
Am J Prev Med 2005;28(4)
367
13. Prorok PC, Andriole GL, Bresalier RS, et al. Design of the prostate, lung, colorectal, and ovarian (PLCO) cancer screening trial. Control Clin Trials 2000;21:273S–309S. 14. Hasson MA, Fagerstrom RM, Kahane DC, et al. Design and evolution of the data management systems in the prostate, lung, colorectal and ovarian (PLCO) cancer screening trial. Control Clin Trials 2000;21:329S– 48S. 15. Office of Public Health Assessment (Utah). Utah behavioral risk factor surveillance system local health district report, 1999 –2001. Salt Lake City UT: Utah Department of Health; 2003. 16. Asahina S, Muneta T, Yamamoto H. Arthroscopic meniscal repair in conjunction with anterior cruciate ligament reconstruction: factors affecting the healing rate. Arthroscopy 1996;12:541–5. 17. Hauger O, Frank LR, Boutin RD, et al. Characterization of the “red zone” of the knee meniscus: MR imaging and histologic correlation. Radiology 2000;217:193–200. 18. O’Meara PM. The basic science of meniscus repair. Orthop Rev 1993;22:681– 6. 19. Renström P, Johnson RJ. Anatomy and biomechanics of the menisci. Clin Sports Med 1990;9:523–38. 20. Kurosawa H, Fukubayashi T, Nakajima H. Load-bearing mode of the knee joint: physical behavior of the knee joint with or without menisci. Clin Orthop 1980;149:283–90. 21. Arnoczky SP, Warren RF. Microvasculature of the human meniscus. Am J Sports Med 1982;10:90 –5. 22. Arnoczky SP, Warren RF. The microvasculature of the meniscus and its response to injury: an experimental study in the dog. Am J Sports Med 1983;11:131– 41. 23. Chan PSH, Kneeland JB, Gannon FH, Luchetti WT, Herzog RJ. Identification of the vascular and avascular zones of the human meniscus using magnetic resonance imaging: correlation with histology. Arthroscopy 1998;14:820 –3. 24. Swiontkowski MF, Schlehr F, Sanders R, Limbird TA, Pou A, Collins JC. Direct, real time measurement of meniscal blood flow. An experimental investigation in sheep. Am J Sports Med 1988;16:429 –33. 25. Messier SP, Loeser RF, Mitchell MN, et al. Exercise and weight loss in obese older adults with knee osteoarthritis: a preliminary study. J Am Geriatr Soc 2000;48:1062–72.
368
26. King D. The healing of semilunar cartilages. J Bone Joint Surg 1936;18: 333– 42. 27. Henning CE, Lynch MA, Yearout KM, Vequist SW, Stallbaumer RJ, Decker KA. Arthroscopic meniscal repair using an exogenous fibrin clot. Clin Orthop 1990;252:64 –72. 28. Scott GA, Jolly BL, Henning CE. Combined posterior incision and arthroscopic intra-articular repair of the meniscus. An examination of factors affecting healing. J Bone Joint Surg 1986;68:847– 61. 29. Cannon WD, Vittori JM. The incidence of healing in arthroscopic meniscal repairs in anterior cruciate ligament-reconstructed knees versus stable knees. Am J Sports Med 1992;20:176 – 81. 30. Tenuta JJ, Arciero RA. Arthroscopic evaluation of meniscal repairs. Am J Sports Med 1994;22:797– 802. 31. Englund M, Roos EM, Roos HP, Lohmander LS. Patient-relevant outcomes fourteen years after meniscectomy: influence of type of meniscal tear and size of resection. Rheumatology 2001;40:631–9. 32. Takahashi M, Suzuki M, Kushida K, Hoshino H, Inoue T. The effect of aging and osteoarthritis on the mature and senescent cross-links of collagen in human menisci. Arthroscopy 1998;14:366 –72. 33. Cipolla M, Cerullo G, Puddu G. Microvasculature of the human medial meniscus: operative findings. Arthroscopy 1992;8:522–5. 34. Petersen W, Tillmann B. Age-related blood and lymph supply of the knee menisci: a cadaver study. Acta Orthop Scand 1995;66:308 –12. 35. Merkel KH. The surface of human menisci and its aging alterations during age: a combined scanning and transmission electron microscopic examination (SEM, TEM). Arch Orthop Trauma Surg 1980;97:185–91. 36. Boden SD, Davis Do, Dina TS, et al. A prospective and blinded investigation of magnetic resonance imaging of the knee: abnormal findings in asymptomatic subjects. Clin Orthop 1992;282:177– 85. 37. Jacobson BH, DeBock DH. Comparison of body mass index by self-reported versus measured height and weight. Percept Motor Skills 2001;92:128 –32. 38. Nawaz H, Chan W, Abdulrahman M, Larson D, Katz DL. Self-reported weight and height: implications for obesity research. Am J Prev Med 2001;20:294 – 8. 39. Rossouw K, Senekal M, Stander I. The accuracy of self-reported weight by overweight and obese women in an outpatient setting. Public Health Nutr 2001;4:19 –26.
American Journal of Preventive Medicine, Volume 28, Number 4
We performed frequency-matched case– control studies using surgical case data for years 1996 to 2000 from administrative databases of two large Utah hospitals; each case was matched with three controls from a large cancer screening trial. Meniscal tear cases (262 male and 282 female) were determined by surgical procedures. Inclusion criteria were age (50 to 79) and body mass index (BMI) (17.00 to 54.99 kg/m2). Gender-specific, age-adjusted odds ratios with 95% confidence intervals (CIs) were calculated for BMI categories from ⬍20.00 to ⱖ40.00. The referent BMI category was 20.00 to 22.49.
Results:
Age-adjusted odds ratios for likelihood of meniscal surgery among those with a BMI of ⱖ40.00 were 15.0 (95% CI⫽3.8 –59.0) for men, and 25.1 (95% CI⫽10.3– 60.8) for women. All odds ratios for men and women with BMIs of ⱖ27.50 and ⱖ25.00, respectively, were statistically significantly elevated.
Conclusions: Significant associations were demonstrated between increasing BMI and meniscal surgeries in both genders, including obese and overweight adults. (Am J Prev Med 2005;28(4):364 –368) © 2005 American Journal of Preventive Medicine
Introduction
K
nee meniscal injuries are common with the prevalence of degenerative meniscal changes in reported asymptomatic individuals ranging between 11.1% and 31.5%.1–3 In autopsy studies the prevalence of horizontal cleavage lesions of the menisci has been reported between 18.6% and 60%.4,5 While common injuries, little epidemiologic data have been reported. Meniscal tears are commonly associated with athletic injuries.6,7 Athletes with greater mean heights and weights are noted to have significantly greater incidence of knee joint injuries.8 Height and weight have also been correlated with degenerative meniscal magnetic resonance imaging (MRI) findings.9 Degenerative meniscal changes are thought to predispose the meniscus to the development of symptomatic meniscal tears.9 Obesity is thought to increase subchondral bony stiffness, transmitting more force to overlying cartilage,10,11 possibly suggesting an injury mechanism involving obesity. While a relatively high incidence of knee injuries in From Public Health Programs (Ford, White) and The Rocky Mountain Center for Occupational and Environmental Health (Hegmann, Holmes), University of Utah School of Medicine, Department of Family and Preventive Medicine, Salt Lake City, Utah Address correspondence and reprint requests to: Kurt T. Hegmann, MD, MPH, The Rocky Mountain Center for Occupational and Environmental Health, 391 Chipeta Way, Suite C, Salt Lake City UT 84108. E-mail: [email protected].
364
obese people has been suggested,12 no epidemiologic studies assessing the risks for meniscal tears from obesity were found in a comprehensive review of the literature. The purpose of this study was to determine whether relationships exist between body mass index (kg/m2 [BMI]) and meniscal tears leading to surgery.
Methods This is a retrospective case– control study. Institutional Review Board approval was obtained.
Cases All cases were obtained for the years 1996 to 2000, and were limited to Utah residents to simultaneously limit the potential for referral biases, and to have access to a large populationbased data set from which to draw controls. The coding system that was relied on in each hospital setting to invoice services was the system used to define cases for this study. Cases from LDS Hospital in Salt Lake City, Utah, were defined by International Classification of Diseases, 9th Revision (ICD-9) procedural codes for partial and total meniscectomy (80.60). Cases from Utah Valley Regional Medial Center (UVRMC) in Provo, Utah, were identified by current procedural terminology (CPT) codes for meniscectomy (29880, 29881) and meniscal repair (29882, 29883). These comprised ⬎90% of the meniscal cases in the study, and those data were also hand-extracted from the medical records. CPT codes were accompanied by descriptions of the surgeries and were
Am J Prev Med 2005;28(4) © 2005 American Journal of Preventive Medicine • Published by Elsevier Inc.
0749-3797/05/$–see front matter doi:10.1016/j.amepre.2005.01.013
Table 1. Description of meniscal surgery cases and controlsa Male Age (years) BMI (kg/m2)
Cases
Controls
n⫽262 59.9 ⫾ 7.8 29.9 ⫾ 5.3
n⫽786 61.8* ⫾ 5.5 27.4* ⫾ 3.9
Female Age (years) BMI (kg/m2)
Cases
Controls
n⫽282 61.4 ⫾ 7.8 31.2 ⫾ 6.8
n⫽846 62.2 ⫾ 5.9 27.0* ⫾ 5.0
BMI, body mass index. a Values are mean ⫾ standard deviation. *p ⬍ 0.05, t-test comparison of means between cases and controls.
accurately coded in all cases. Heights and weights (used to calculate BMI) were measured prior to surgery and were used in this study. A computerized medical record for the LDS Hospital was used to obtain age, gender, and BMI data for cases from that facility. Coding accuracy rates of 94% for ICD-9 and 93% for CPT codes are reported by this facility. All cases from LDS Hospital were inpatients. At UVRMC, medical records were used to obtain age, gender, height, weight, and inpatient or outpatient status. Of the 515 meniscal surgery patients in this study, 29 (5.6%) had two meniscal surgeries, one on each knee, and both were included. All cases ⬍50 years were excluded due to the limited age range of the controls. These exclusions may also help to limit possible confounding due to athletic injuries from adolescent sports. All cases aged ⬎79 years were excluded because of the limited age range of the controls (as described below). Cases with BMI ⬍17 or ⬎55 were excluded due to possible information error.
Controls Controls were drawn from a group of 9944 Utah residents enrolled in the National Cancer Institute’s Prostate, Lung, Colorectal, and Ovarian Cancer Screening Trial from the years 1996 to 2000. Enrollees were healthy men and women aged 55 to 74, and are described elsewhere.13 Each control’s age and self-reported height and weight were provided. BMI was then calculated for each control. The information regarding these controls was optically scanned with an edit program that catches nonsense and out-of-range values.14
Statistical Methods A multistage frequency match was performed to age (within 5 years) and gender match three controls for each case. The cases were divided into ten categories (in 2.5 BMI increments), with 20.00 to 22.49 serving as the referent category. Crude gender-specific odds ratios and Mantel–Haenszel ageadjusted, sex-specific odds ratios were calculated. The age strata utilized for age-adjustment of male and female meniscal cases were 50 to 59, 60 to 69, and 70 to 79. Chi-square tests for linear trend were calculated for the odds ratios. Logistic regression models were analyzed with and without age as a continuous independent variable. All data analyses were performed in 2004 using SPSS 10.0 (SPSS, Inc., Chicago IL, 1999).
Results There were 262 male meniscal surgery cases (31 excluded for age ⬍50) and 282 female meniscal surgery
cases (21 excluded for age ⬍50 and 1 excluded for age ⬎79) included. No cases were out of the quality control range for BMI. Cases and controls were generally comparable regarding residence in Utah, age, and gender. However, male cases tended to be slightly younger than the controls (see Table 1). No significant differences in BMIs were found between inpatient and outpatient cases (male meniscus: p ⫽1.00, female meniscus: p ⫽0.11). There was no statistically significant difference between the BMIs of the 29 meniscal surgery patients who were included twice and the other meniscal surgery patients (p ⫽0.54). A relationship was identified between BMI and meniscal surgery. Chi-square tests of linear trend were highly significant for males and females (p ⬍0.0001). Elevated age-adjusted odds ratios were statistically significant in men with a BMI of ⱖ27.5, and in women with a BMI of ⱖ25.0. Mantel–Haenszel age-adjusted odds ratios for those with a BMI of ⱖ40.0 were 15.0 (95% confidence interval [CI], 3.8 –59.0) in men and 25.1 (95% CI, 10.3– 60.8) in women (Table 2). No appreciable differences existed between crude and Mantel–Haenszel age-adjusted odds ratios, indicating a negligible potential for residual confounding due to age. The odds ratios from the logistic regression models that adjusted for age were not appreciably different from the odds ratios of the models that did not adjust for age, again indicating a negligible potential for residual confounding due to age (unadjusted odds ratios not shown). The odds ratios from the logistic regression models were similar to the above-mentioned crude and Mantel–Haenszel adjusted odds ratios (see Table 2).
Discussion There is a dose–response relationship between BMI and meniscal surgery in middle-aged to older adults in both genders. A relationship had been previously suspected (T Sisk, The Orthopaedic Clinic, Memphis TN, personal communication, February 27, 2002), but never measured. Since 57.4% (164 million) of the U.S. adult population is either overweight (BMI⫽25 to 29.9) or obese (BMIⱖ30),15 this relationship has potentially large implications on a population basis for meniscal surgeries. Am J Prev Med 2005;28(4)
365
Table 2. ORs for meniscal surgeries by BMI in men and women Men
Women
BMI (kg/m2)
ORMH
95% CI
ORLR
95% CI
ORMH
95% CI
ORLR
95% CI
⬍20.00 20.00–22.49 22.50–24.99 25.00–27.49 27.50–29.99 30.00–32.49 32.50–34.99 35.00–37.49 37.50–39.99 ⱖ40.00
3.5 1.0 1.8 1.6 3.0 4.7 4.9 6.5 6.2 15.0
(0.7–16.2) — (0.7–4.6) (0.7–4.0) (1.2–7.4) (1.9–12.0) (1.9–12.6) (2.1–20.2) (1.7–23.0) (3.8–59.0)
4.5 1.0 1.9 1.6 3.0 4.8 4.7 5.6 7.5 14.6
(1.0–19.8) — (0.7–4.7) (0.7–4.0) (1.2–7.3) (1.9–12.1) (1.8–12.3) (1.9–16.5) (2.1–27.0) (4.0–53.7)
0.9 1.0 1.5 2.8 2.9 4.7 5.5 3.1 9.2 25.1
(0.3–3.1) — (0.7–3.0) (1.4–5.5) (1.5–5.8) (2.4–9.4) (2.5–11.9) (1.3–7.6) (3.4–24.5) (10.3–60.8)
1.0 1.0 1.5 2.7 2.9 4.7 5.1 2.9 9.6 24.3
(0.3–3.1) — (0.7–3.0) (1.3–5.3) (1.5–5.8) (2.4–9.3) (2.4–10.6) (1.2–7.1) (3.7–25.3) (10.3–57.2)
BMI, body mass index; CI, confidence interval; OR, odds ratio; ORMH, Mantel Haenszel age-adjusted odds ratio; ORLR, estimated age-adjusted odds ratio from logistic regression.
The menisci are thought to provide congruity and stabilization through load transfer across the knee joint, and act as shock absorbers, relieving stress transmitted through articular cartilage and subchondral bone.16 –19 Menisci are believed to transmit over half of the load to the knee joint.20 Thus, there is a biomechanical mechanism for the relationship between BMI and meniscal tears. As body mass index increases, the strain and torque in the knee joint during rotation likely increases, in theory leading to a higher risk for meniscal injuries. Whether other mechanisms contribute, such as reduced blood flow to the menisci17,21–24 or low-grade inflammation associated with obesity,25 is unclear. In those who are obese, blood supply to the menisci could be limited through either vascular compression effects of mechanical weight bearing or associations with cardiovascular risk factors. Meniscal blood supply is known to be limited to the periphery, with the presence of vessels ranging from the outer 10% to 25%.17,21–24 The blood supply is thought to be important for the healing of tears.17,22,26 A longitudinal tear in the avascular portion of the meniscus appears to require a vascular access channel at the midpoint to heal.22 Repairs of meniscal tears in the peripheral third zone also demonstrate superior healing rates compared with repairs in the central third zone (87% vs 59%).16 Because the focus of meniscal tears in this study was in older adults, healing of partial tears with respect to age may have a significant effect. However, while there is widespread belief of a strong effect of age on meniscal tear healing, this has not been well demonstrated in the literature.16,27–31 Still, the effects of aging on the menisci appear well established, and include biochemical changes,32 vascular changes,19,33,34 degenerative changes,1,9 cell surface changes,35 and prevalence of abnormalities on MRIs.36 Some studies report no distinct pattern with respect to age and healing,16,27,28 while others do report a distinct pattern with respect to age and healing.29 –31 366
The American Academy of Orthopaedic Surgeons Research Department estimates that surgical procedures of the meniscus are performed on approximately 850,000 patients each year. The meniscal surgeries in this study cost an average of $3000. Assuming that 57% of the U.S. adult population is overweight or obese, and conservatively assuming a threefold meniscal surgery risk among them, the excess surgery rate would be approximately 450,000 surgeries. This is a potential cost savings of $1.35 billion annually had they not been overweight or obese.
Limitations Temporality issues cannot be resolved with a case– control study. However, the dose–response relationships reported in this study suggest a true association between increasing BMI and the need for meniscal surgery. The results of this study may not be generalized to conservatively managed meniscal tears since cases were selected using surgical codes. If conservatively managed meniscal tears were included, the odds ratios calculated could potentially have been greater. In ascertaining the appropriateness of the case surgeries, we relied on the accuracy of the hospital coding system and doctors’ reporting in medical records. Excluding cases aged ⬍50 years, due to the limited age range of the controls, was also thought to help limit possible confounding due to athletic injuries from adolescent sports. However, it is possible that those who suffer knee or joint injury in their youth may be at greater risk for having a similar injury later in life. Therefore, adolescent sports injuries may have played a role in the meniscal tear surgery cases used in this study despite our efforts to limit them with a suspicion that there may be interaction between BMI and those injuries. This study could be subject to information bias regarding BMI. Weight for each control was self-re-
American Journal of Preventive Medicine, Volume 28, Number 4
ported, whereas each case was weighed before surgery. Self-reported weight has been noted to be underestimated, especially for females.37–39 The magnitude of under-estimation38 regarding BMI has been estimated to be ⱕ1.14; thus, the results would not be appreciably different (data not shown). However, if a difference did exist, it would tend to reduce the reported odds ratios. The control set’s BMIs were compared to BMIs in both the Utah Behavioral Risk Factor Surveillance Survey data and the Utah portion of the National Health and Nutrition Examination Survey III data. (Both sets of BMI were from self-reported data, which makes them comparable to our controls, although this is a limitation with respect to the measured data of the cases.) In both comparisons, our controls actually had slightly higher BMIs than those of the two other data sets. The BMI of this study’s control group was approximately 1.96% greater than other Utah populationbased controls. Therefore, these results may be slightly biased toward a less strong association than actually exists. The study was limited to Utah residents to reduce possible bias from out-of-state referrals among cases, and to allow access to a large control population. Utah’s population has been reported to have a slightly lower proportion of overweight people than the national average (53.3% vs 57.4% [BMI⬎25.0], respectively).15 This may also lead to a slight under-estimation of true odds ratios for the U.S. population. However, as BMIs are not substantially different in other U.S. states, the results are considered generalizable. Data on the duration of obesity and other potential risk factors for meniscal tears (e.g., highest BMI ever, occupation, sports, tobacco use, genetics, ethnicity, and reproductive variables in women) were not available. Arthritis has been associated with obesity,10,11 and therefore could cause some confounding in this study, as well as potential interaction. We suspect that previous injury history also may be both an effect modifier and confounder, although data necessary for this analysis have similarly never been reported. If vascular models turn out to be important for this relationship due to a meniscal vascular supply impairment mechanism, smoking may be a potential confounder, in contrast with arthritis, where if anything, smoking is slightly protective for knee arthritis. Therefore, future research including these factors is needed.
Conclusion The strength of the association and dose–response characteristic demonstrated in this study suggest a significant relationship between increasing BMI and the need for meniscal surgery in both males and females. It also may imply that a population-based weight management program could decrease the future burden on orthopedic and medical care systems due to meniscal surgeries and treatment of other
What This Study Adds . . . There is a belief that increasing body mass index (BMI) is associated with meniscal tears. However, no data have been published testing this association. This case– control study demonstrates that there is a strong association between increasing BMI and meniscal tears. These results may serve as a basis for patient education and preventive measures. Future research should evaluate for potential interaction among age, obesity, and osteoarthritis.
obesity-related conditions. This study also suggests that not only obese adults, but also overweight adults, have an increased likelihood of needing meniscal surgery. We were unable to find another study testing this association. We are grateful for the help of Brent James, MD, in obtaining data from Intermountain Health Care. We also thank Saundra S. Buys, MD, for assisting with access to a large control population, the Prostate Lung Colorectal and Ovarian Cancer Screening Trial, University of Utah (contract N01-CN25524). This study was partially supported by the Centers for Disease Control and Prevention, National Institute for Occupational Safety and Health (training grant T42/CCT810426-06-1). No financial conflict of interest was reported by the authors of this paper.
References 1. Jerosch J, Castro WHM, Assheuer J. Age-related magnetic resonance imaging morphology of the menisci in asymptomatic individuals. Arch Orthop Trauma Surg 1996;115:199 –202. 2. Kornick J, Trefelner E, McCarthy S, Lange R, Lynch K, Jokl P. Meniscal abnormalities in the asymptomatic population at MR imaging. Radiology 1990;177:463–5. 3. LaPrade RF, Burnett QM, Veenstra MA, Hodgman CG. The prevalence of abnormal magnetic resonance imaging findings in asymptomatic knees. Am J Sports Med 1994;22:739 – 45. 4. Noble J. Lesions of the menisci: autopsy incidence in adults younger than fifty-five years old. J Bone Joint Surg 1977;59A:480 –3. 5. Noble J, Hamblen DL. The pathology of the degenerate meniscus lesion. J Bone Joint Surg 1975;57B:180 – 6. 6. Chen E, Maffulli N, Chan KM. Knee injuries produced by recreational sports follows a different pattern than casual injuries. Bull Hosp Jt Dis 1998;57:74 –9. 7. Strickland JW. Philosophy of the treatment of athletes. Clin Sports Med 1995;14:285– 8. 8. Grace TG, Sweetser ER, Nelson MA, Ydens LR, Skipper BJ. Isokinetic muscle imbalance and knee-joint injuries. J Bone Joint Surg 1984;66:734 – 40. 9. Negendank WG, Fernandez-Madrid FR, Heilbrun LK, Teitge RA. Magnetic resonance imaging of meniscal degeneration in asymptomatic knees. J Orthop Res 1990;8:311–20. 10. Dequeker J, Goris P, Uytterhoeven R. Osteoporosis and osteoarthritis (osteoarthrosis). JAMA 1983;249:1448 –51. 11. Felson DT, Anderson JJ, Naimark A, Walker AM, Meenan RF. Obesity and knee osteoarthritis. Ann Intern Med 1988;109:18 –24. 12. Sisk TD. Traumatic disorders of the joints. In: Canale ST, Daugherty K, Jones L, eds. Campbell’s operative orthopaedics, vol. 3. 9th ed. St. Louis: Mosby, 1998:1465–766.
Am J Prev Med 2005;28(4)
367
13. Prorok PC, Andriole GL, Bresalier RS, et al. Design of the prostate, lung, colorectal, and ovarian (PLCO) cancer screening trial. Control Clin Trials 2000;21:273S–309S. 14. Hasson MA, Fagerstrom RM, Kahane DC, et al. Design and evolution of the data management systems in the prostate, lung, colorectal and ovarian (PLCO) cancer screening trial. Control Clin Trials 2000;21:329S– 48S. 15. Office of Public Health Assessment (Utah). Utah behavioral risk factor surveillance system local health district report, 1999 –2001. Salt Lake City UT: Utah Department of Health; 2003. 16. Asahina S, Muneta T, Yamamoto H. Arthroscopic meniscal repair in conjunction with anterior cruciate ligament reconstruction: factors affecting the healing rate. Arthroscopy 1996;12:541–5. 17. Hauger O, Frank LR, Boutin RD, et al. Characterization of the “red zone” of the knee meniscus: MR imaging and histologic correlation. Radiology 2000;217:193–200. 18. O’Meara PM. The basic science of meniscus repair. Orthop Rev 1993;22:681– 6. 19. Renström P, Johnson RJ. Anatomy and biomechanics of the menisci. Clin Sports Med 1990;9:523–38. 20. Kurosawa H, Fukubayashi T, Nakajima H. Load-bearing mode of the knee joint: physical behavior of the knee joint with or without menisci. Clin Orthop 1980;149:283–90. 21. Arnoczky SP, Warren RF. Microvasculature of the human meniscus. Am J Sports Med 1982;10:90 –5. 22. Arnoczky SP, Warren RF. The microvasculature of the meniscus and its response to injury: an experimental study in the dog. Am J Sports Med 1983;11:131– 41. 23. Chan PSH, Kneeland JB, Gannon FH, Luchetti WT, Herzog RJ. Identification of the vascular and avascular zones of the human meniscus using magnetic resonance imaging: correlation with histology. Arthroscopy 1998;14:820 –3. 24. Swiontkowski MF, Schlehr F, Sanders R, Limbird TA, Pou A, Collins JC. Direct, real time measurement of meniscal blood flow. An experimental investigation in sheep. Am J Sports Med 1988;16:429 –33. 25. Messier SP, Loeser RF, Mitchell MN, et al. Exercise and weight loss in obese older adults with knee osteoarthritis: a preliminary study. J Am Geriatr Soc 2000;48:1062–72.
368
26. King D. The healing of semilunar cartilages. J Bone Joint Surg 1936;18: 333– 42. 27. Henning CE, Lynch MA, Yearout KM, Vequist SW, Stallbaumer RJ, Decker KA. Arthroscopic meniscal repair using an exogenous fibrin clot. Clin Orthop 1990;252:64 –72. 28. Scott GA, Jolly BL, Henning CE. Combined posterior incision and arthroscopic intra-articular repair of the meniscus. An examination of factors affecting healing. J Bone Joint Surg 1986;68:847– 61. 29. Cannon WD, Vittori JM. The incidence of healing in arthroscopic meniscal repairs in anterior cruciate ligament-reconstructed knees versus stable knees. Am J Sports Med 1992;20:176 – 81. 30. Tenuta JJ, Arciero RA. Arthroscopic evaluation of meniscal repairs. Am J Sports Med 1994;22:797– 802. 31. Englund M, Roos EM, Roos HP, Lohmander LS. Patient-relevant outcomes fourteen years after meniscectomy: influence of type of meniscal tear and size of resection. Rheumatology 2001;40:631–9. 32. Takahashi M, Suzuki M, Kushida K, Hoshino H, Inoue T. The effect of aging and osteoarthritis on the mature and senescent cross-links of collagen in human menisci. Arthroscopy 1998;14:366 –72. 33. Cipolla M, Cerullo G, Puddu G. Microvasculature of the human medial meniscus: operative findings. Arthroscopy 1992;8:522–5. 34. Petersen W, Tillmann B. Age-related blood and lymph supply of the knee menisci: a cadaver study. Acta Orthop Scand 1995;66:308 –12. 35. Merkel KH. The surface of human menisci and its aging alterations during age: a combined scanning and transmission electron microscopic examination (SEM, TEM). Arch Orthop Trauma Surg 1980;97:185–91. 36. Boden SD, Davis Do, Dina TS, et al. A prospective and blinded investigation of magnetic resonance imaging of the knee: abnormal findings in asymptomatic subjects. Clin Orthop 1992;282:177– 85. 37. Jacobson BH, DeBock DH. Comparison of body mass index by self-reported versus measured height and weight. Percept Motor Skills 2001;92:128 –32. 38. Nawaz H, Chan W, Abdulrahman M, Larson D, Katz DL. Self-reported weight and height: implications for obesity research. Am J Prev Med 2001;20:294 – 8. 39. Rossouw K, Senekal M, Stander I. The accuracy of self-reported weight by overweight and obese women in an outpatient setting. Public Health Nutr 2001;4:19 –26.
American Journal of Preventive Medicine, Volume 28, Number 4