Abstract
Objective
To investigate whether serum total testosterone level is associated with knee pain and function in men and women with severe knee osteoarthritis (OA).
Methods
We enrolled 272 adults age ≥60 years (mean ± SD age 70.4 ± 4.4 years, 53% women) who underwent unilateral total knee replacement (TKR) due to severe knee OA. Serum testosterone levels and Western Ontario and McMaster Universities Osteoarthritis Index (WOMAC) pain and function of the operated and contralateral knee were measured at 6–8 weeks after surgery. At the nonoperated knee, 56% of participants had radiographic knee OA with a Kellgren/Lawrence grade ≥2. Cross‐sectional analyses were performed by sex and body mass index (BMI) subgroups, using multivariable regression adjusted for age, physical activity, and BMI.
Results
At the operated knee, higher testosterone levels were associated with less WOMAC pain in men (B = –0.62, P = 0.046) and women (B = –3.79, P = 0.02), and less WOMAC disability scores in women (B = –3.62, P = 0.02) and obese men (B = –1.99, P = 0.02). At the nonoperated knee, testosterone levels were not associated with WOMAC pain in men or women, but higher testosterone levels were associated with less disability in women (B = –0.95, P = 0.02). Testosterone levels were inconsistently associated with pain and disability in BMI subgroups among men. Only among obese women, testosterone levels were inversely associated with radiographic knee OA (odds ratio = 0.10, P = 0.003).
Conclusion
Higher total testosterone levels were associated with less pain in the operated knee in men and women undergoing TKR and less disability in women. At the nonoperated knee, higher testosterone levels were inconsistently associated with less pain and disability.
Introduction
Knee osteoarthritis (OA) is the most common cause of difficulty walking in older adults (1). Thirty percent of adults will develop symptomatic knee OA by the age of 65 years and nearly 50 percent of adults will develop symptomatic knee OA by the age of 85 years, with the highest risk among those who were overweight during extended periods of their lifetime (2).
Significance & Innovations.
Among both men and women with severe knee osteoarthritis who underwent unilateral total knee replacement, higher serum testosterone levels were associated with less Western Ontario and McMaster Universities Osteoarthritis Index (WOMAC) pain in the operated knee, independent of age, body mass index (BMI), and physical activity.
Among women but not men, higher serum testosterone levels were also associated with less WOMAC disability in the operated and the nonoperated knee, independent of age, BMI, and physical activity.
The prevalence of symptomatic knee OA increases similarly with age in women and men until age 50 years. In the second half of life, women have a significantly higher prevalence of symptomatic knee OA (3) and greater disability from knee OA than men (4, 5). The sex difference in knee OA prevalence and severity in the second half of life has not been well understood (4, 6). However, given that cells in knee articular cartilage, underlying bone, and surrounding muscles in men and women express receptors for both estrogen and testosterone (T) (7, 8), hormonal factors are probably involved (4, 6) with a potential benefit of a greater physiologic testosterone exposure (7, 9, 10, 11).
With regard to changes in testosterone exposure with age, testosterone levels have been shown to decrease by approximately 1% per year in men starting at age 40 years (12). Clinical signs of testosterone deficiency in older men are a decrease in muscle mass and strength, a decrease in bone mass and an increase in central body fat (13). Women have 20‐fold lower testosterone levels compared to men (14), and their testosterone levels also decline with age, reaching a nadir after menopause, with a decline close to 15% of their premenopausal stage (15). Although the biologic role of testosterone in women remains unclear, the sharp and rapid decline after menopause may contribute to the age‐related decline in physical function among women (7, 9, 10, 11, 16).
With regard to studies that link testosterone levels to OA, lower serum testosterone levels have been associated with a higher prevalence of hand but not knee OA in 1 study (17). Further, a small cross‐sectional study among 45 healthy middle‐aged men suggested a positive association between higher serum testosterone levels and medial tibial cartilage thickness (9). In 1 larger study of 309 overweight adults age ≥60 years with knee OA, higher testosterone levels were found to be associated with less Western Ontario and McMaster Universities Osteoarthritis Index (WOMAC) stiffness among men and better WOMAC function among women (18).
With regard to pain sensitivity and potential relevance to OA, clinical and experimental investigations have consistently shown sex‐specific differences for both pain sensitivity and threshold (19). Although the underlying mechanisms for these differences are not well understood, an influence of testosterone on nociceptive processing is well established (19, 20). In fact, testosterone has been suggested as protecting men from chronic musculoskeletal pain conditions (21), either by a direct effect on nociceptive processing (19, 20), or indirectly by an increase in muscle mass, strength, and function, as suggested by 2 clinical trials with testosterone treatment among frail and hypogonadal older men (22, 23). Further potential benefits of higher physiologic testosterone levels on OA pain and function may be explained by prior data that linked higher testosterone exposure to a decrease in fat mass and inflammatory response among men (7, 9, 10, 11). Further, preoperative supraphysiologic testosterone administration has been suggested to confer some early functional benefit among older men undergoing knee replacement in a small study of 25 men (24).
The aim of this cross‐sectional study was to investigate a possible association between total serum testosterone levels and symptoms of knee OA with regard to pain and disability in the operated and nonoperated knee among men and women age ≥60 years, who underwent unilateral total knee replacement (TKR) due to severe knee OA 6–8 weeks earlier. We chose this target population given their high risk of OA at the contralateral nonoperated knee (25, 26) as well as the potential shared benefits of higher physiologic testosterone levels on pain and disability at the operated and nonoperated knee.
Patients and Methods
Study design and participants
The current study is a cross‐sectional analysis of the baseline data from the Zurich Multiple Endpoint Vitamin D Trial in Knee OA Patients (27). The original study was a 2‐year, double blind, randomized controlled trial that investigated the effect of vitamin D (2,000 versus 800 IU/day cholecalciferol) on pain and disability related to the rehabilitation of the operated knee and contralateral knee among 273 seniors age ≥60 years (mean age 70.3 years, 53% women) who underwent elective surgery for unilateral TKR due to severe knee OA. The baseline assessment took place 6–8 weeks after surgery at the Centre on Aging and Mobility at the University of Zurich, Switzerland, from October 2007 to February 2013. In the original trial, participants were not selected based on their vitamin D status, and 31.4% of participants were vitamin D deficient at baseline (27). Of 273 participants enrolled, 1 male participant was excluded due to missing data on serum testosterone concentration, reducing the analytical sample size for this study to 272 (for the radiologic assessment data were only available for 270 participants). Exclusion criteria of the original trial were a history of inflammatory arthritis, chronic corticosteroid use, a history of malabsorption disorder, kidney disease (estimated creatinine clearance <30 ml/minute), current cancer, treatment with bisphosphonate, parathyroid hormone therapy, calcitonin therapy in the 6 months prior to enrollment, severe cognitive/visual/hearing impairments, and inability to walk at least 3 meters with or without a walking aid (27). All participants gave their written informed consent, and the study was approved by the Cantonal Ethical Commission of Zurich (protocol identifier STZ 20/07), Switzerland.
Measurement of serum testosterone concentration
Fasting blood samples were taken between 8:00 and 9:30 am. Serum concentration of total testosterone was measured by an electrochemiluminescence immunoassay (Roche Diagnostics) with an interassay coefficient of variation of 3.9% at a level of 7.3 nmoles/liter and 3.5% at a level of 18.8 nmoles/liter.
Assessment of covariates
Body mass index (BMI) was calculated as weight divided by height squared (kg/m2). BMI categories were defined according to World Health Organization guidelines as underweight (<18.5), normal weight (≥18.5 to <25), overweight (≥25 to <30), and obese (≥30). Physical activity levels were measured by an ankle‐worn ambulatory activity monitor (StepWatch Step Activity Monitor), which records the number of steps taken every minute. The StepWatch monitor has been validated for use in older adults (28) and has been used to monitor physical activity in several patient groups, including patients with knee OA (29). In the current study, participants were instructed to wear the monitor during the day for 7 consecutive days. A measurement was considered as valid if at least 3 days with ≥10 hours of recording were available, omitting blocks of >180 minutes of consecutive zeros, which was interpreted as device not worn. Minutes spent on moderate‐to‐vigorous physical activity were defined as the average minutes per day with ≥30 steps/minute according to the manufacturer’s software manual (StepWatch 3.1 software manual).
Outcome measurements
WOMAC pain and function score
Knee pain and function (disability) of the operated and the nonoperated knee were assessed by the respective subscales of the WOMAC questionnaire. The WOMAC is a broadly validated and commonly used self‐reported outcome tool for knee OA (30), including patients undergoing knee replacement (31). It consists of 24 items divided into 3 subscales: pain (5 items), stiffness (2 items), and physical function (17 items), with each item scoring on a 5‐point Likert scale (none, mild, moderate, severe, and extreme). For the pain and physical function subscale, these scores were transformed to a 0 to 100 score (where 0 = no symptoms and 100 = extreme symptoms) with higher scores indicating more pain and more disability. We used a knee‐specific paper‐version of the German WOMAC 3.1 (32).
Radiologic assessment of knee OA at the nonoperated knee
To rate knee OA at the nonoperated knee, we performed a plain radiograph of the contralateral knee in a semiflexed weight‐bearing position (we used the Multicentre Osteoarthritis Study center standardized radiograph assessment procedures [33]). Kellgren/Lawrence (K/L) grades were classified by a blinded knee OA radiology expert (0 = no radiographic features of OA; 1 = possible joint space narrowing and osteophyte formation; 2 = definite osteophyte formation with possible joint space narrowing; 3 = multiple osteophytes, definite joint space narrowing, sclerosis, and possible bony deformity; and 4 = large osteophytes, marked joint space narrowing, severe sclerosis, and definite bone deformity) (34).
Statistical analysis
Statistical analysis was performed using SAS software, version 9.4. Distributions of continuous variables were examined for normality. Differences in baseline characteristics between men and women were analyzed using a chi‐square test for categorical variables and a Student’s t‐test for continuous variables.
Associations between serum testosterone concentration and the WOMAC pain and function (disability) score were analyzed by using multivariable robust linear regression models. For these analyses, among women only, testosterone levels were natural logarithmically (ln) transformed to approach normality. At the nonoperated knee only, we also assessed the association between testosterone levels and K/L grades based on ordinal logistic regression models. All analyses were performed in an unadjusted way (model 1) and adjusted for age, BMI status (normal weight, overweight, obese, except in BMI strata), and physical activity (model 2), and for men and women separately. For all analyses, model 2 was considered the main model. Moreover, we performed a subgroup analysis by BMI (normal weight, overweight, and obese) to investigate whether specific BMI subgroups of seniors are more sensitive with regard to the relationship between testosterone and WOMAC scores or K/L grades. These analyses were performed because obesity is a very well documented risk factor for knee OA (35, 36) and several studies have shown that testosterone levels are inversely associated with BMI (37, 38). Statistical significance was set at a P value less than or equal to 0.05; reported P values are 2‐sided.
Results
The characteristics of the 272 study participants (127 men and 145 women) are shown in Table 1. There were no significant differences between men and women with regard to mean age, BMI, and K/L grades. Women were significantly less physically active and had lower serum testosterone levels than men. Moreover, women had significantly worse (higher) WOMAC pain and more disability (higher WOMAC function scores) at the operated knee and worse disability (higher WOMAC function scores) at the nonoperated knee compared with men. At the nonoperated knee, 48.4% of men (61 of 126) and 61.6% of women (90 of 144) had K/L grade ≥2 consistent with radiographic knee OA.
Table 1.
Characteristics of participants of the Zurich Knee Osteoarthritis Trial by sex*
| Characteristic | Men | Women | Sex difference P | Total |
|---|---|---|---|---|
| Subjects, no. (%) | 126 (47) | 146 (53) | 0.25 | 272 |
| Age, years | 70.3 ± 6.9 | 70.4 ± 6.0 | 0.83 | 70.4 ± 6.4 |
| Body mass index, kg/m2 | 27.6 ± 3.8 | 26.9 ± 4.1 | 0.11 | 27.2 ± 3.9 |
| MVPA, minutes/day | 45.3 ± 23.4 | 37.9 ± 21.6 | 0.009† | 42.3 ± 22.6 |
| Total testosterone, nmoles/liter | 13.0 ± 4.6 | 0.4 ± 0.4 | <0.0001† | 6.3 ± 7.0 |
| Kellgren/Lawrence grade, no. (%)‡ | ||||
| 0 | 32 (56) | 25 (44) | 0.11 | 57 (21) |
| 1 | 33 (53) | 29 (47) | – | 62 (23) |
| 2 | 18 (33) | 36 (67) | – | 54 (20) |
| 3 | 31 (42) | 42 (58) | – | 73 (27) |
| 4 | 12 (50) | 12 (50) | – | 24 (9) |
| WOMAC pain score (0–100) | ||||
| Operated knee | 23.9 ± 14.4 | 32.6 ± 14.4 | <0.0001† | 28.5 ± 15.0 |
| Nonoperated knee | 4.2 ± 8.1 | 5.2 ± 8.2 | 0.32 | 4.7 ± 8.1 |
| WOMAC functional score (0–100) | ||||
| Operated knee | 21.8 ± 12.6 | 29.2 ± 13.9 | <0.0001† | 25.8 ± 13.8 |
| Nonoperated knee | 2.9 ± 6.6 | 5.4 ± 8.9 | 0.01† | 4.3 ± 8.0 |
Data are crude means ± SD unless indicated otherwise. Differences between men and women were assessed by using Student’s t‐test for continuous variables and a chi‐square test for categorical variables. P values are 2‐sided; statistical significance is set at P < 0.05. MVPA = moderate‐to‐vigorous physical activity; WOMAC = Western Ontario and McMaster Universities Arthritis Index.
Statistically significant.
Values refer to the nonoperated knee (n = 270).
In multivariable‐adjusted analyses, at the operated knee (Table 2, model 2), higher testosterone levels were associated with less WOMAC pain in both men (B = –0.62 [95% confidence interval (95% CI) –1.23, –0.01]; P = 0.05) and women (B = –3.79 [95% CI –6.90, –0.69]; P = 0.02). This association was most pronounced in the subgroup of normal‐weight men (B = –1.16 [95% CI –2.17, –0.15]; P = 0.02) as well as in the subgroup of normal‐weight women (B = –9.27 [95% CI –15.56, –2.98]; P = 0.004). At the operated knee, higher testosterone levels were also associated with less disability among obese men (B = –1.99 [95% CI –3.62, –0.37]; P = 0.02) and less disability among all women (B = –3.62 [95% CI –6.65, –0.58]; P = 0.02), and most pronounced among normal‐weight women (B = –6.54 [95% CI –12.81, –0.28]; P = 0.04).
Table 2.
Cross‐sectional association between blood total testosterone concentration and WOMAC postoperative pain and functional scores in seniors age ≥60 years with knee osteoarthritis by sex*
| Men: model 1 | Men: model 2 | Women: model 1† | Women: model 2† | |||||
|---|---|---|---|---|---|---|---|---|
| B (95% CI) | P | B (95% CI) | P | B (95% CI) | P | B (95% CI) | P | |
| WOMAC pain score (0–100) | ||||||||
| Operated knee | ||||||||
| Total subjects (M/F) | –0.51 (–1.11, 0.09) | 0.09 | –0.62 (–1.23, –0.01) | 0.05‡ | –2.76 (–5.75, 0.23) | 0.07 | –3.79 (–6.90, –0.69) | 0.02‡ |
| BMI, kg/m2 § | ||||||||
| Normal weight | –1.25 (–2.23, –0.27) | 0.013 | –1.16 (–2.17, –0.15) | 0.02‡ | –5.67 (–11.81, 0.48) | 0.07 | –9.27 (–15.56, –2.98) | 0.004‡ |
| Overweight | –0.08 (–0.87, 0.72) | 0.85 | –0.01 (–0.81, 0.80) | 0.99 | –3.54 (–7.75, 0.68) | 0.10 | –3.88 (–8.20, 0.43) | 0.08 |
| Obese | –1.12 (–3.06, 0.82) | 0.26 | –1.36 (–3.54, 0.82) | 0.22 | 0.28 (–5.07, 5.63) | 0.92 | 1.30 (–3.62, 6.22) | 0.60 |
| P interaction | – | 0.29 | – | 0.38 | – | 0.09 | – | 0.01‡ |
| Nonoperated knee | ||||||||
| Total subjects (M/F) | –0.01 (–0.11, 0.08) | 0.77 | –0.07 (–0.17, 0.03) | 0.20 | 0.52 (–0.23, 1.28) | 0.18 | 0.42 (–0.55, 1.38) | 0.39 |
| BMI, kg/m2 § | ||||||||
| Normal weight | –0.27 (–0.54, 0.00) | 0.050 | –0.39 (–0.68, –0.10) | 0.009‡ | –0.08 (–1.10, 0.94) | 0.88 | 0.05 (–1.43, 1.54) | 0.94 |
| Overweight | 0.0007 (–0.10, 0.10) | 0.99 | 0.003 (–0.12, 0.13) | 0.97 | 0.73 (–0.45, 1.91) | 0.22 | 0.73 (–0.70, 2.16) | 0.32 |
| Obese | –0.63 (–1.29, 0.03) | 0.06 | –0.33 (–0.90, 0.24) | 0.25 | –0.73 (–4.98, 3.53) | 0.74 | 0.69 (–3.51, 4.90) | 0.75 |
| P interaction | – | 0.18 | – | 0.14 | – | 0.51 | – | 0.88 |
| WOMAC functional score (0–100) | ||||||||
| Operated knee | ||||||||
| Total subjects (M/F) | –0.01 (–0.53, 0.51) | 0.97 | –0.11 (–0.66, 0.44) | 0.70 | –3.57 (–6.53, –0.61) | 0.02‡ | –3.62 (–6.65, –0.58) | 0.02‡ |
| BMI, kg/m2 § | ||||||||
| Normal weight | –0.62 (–1.67, 0.44) | 0.25 | –0.60 (–1.77, 0.56) | 0.31 | –6.51 (–12.34, –0.68) | 0.03‡ | –6.54 (–12.81, –0.28) | 0.04‡ |
| Overweight | 0.31 (–0.34, 0.95) | 0.35 | 0.31 (–0.39, 1.00) | 0.39 | –3.41 (–7.29, 0.47) | 0.09 | –4.04 (–8.12, 0.04) | 0.052 |
| Obese | ‐0.96 (–2.53, 0.61) | 0.23 | –1.99 (–3.62, –0.37) | 0.02‡ | –0.48 (–7.15, 6.20) | 0.89 | 1.26 (–4.57, 7.08) | 0.67 |
| P interaction | – | 0.95 | – | 0.71 | – | 0.13 | – | 0.11 |
| Nonoperated knee | ||||||||
| Total subjects (M/F) | –0.0005 (–0.04, 0.03) | 0.98 | –0.0006 (–0.03, 0.03) | 0.97 | –0.66 (–1.33, 0.01) | 0.05 | –0.95 (–1.73, ‐0.17) | 0.02‡ |
| BMI, kg/m2 § | ||||||||
| Normal weight | 0.04 (–0.32, 0.39) | 0.83 | 0.02 (0.25, 0.30) | 0.88 | 0.07 (–0.62, 0.76) | 0.85 | 0.05 (–0.82, 0.92) | 0.91 |
| Overweight | 0.004 (–0.01, 0.02) | 0.64 | –0.0006 (–0.20, 0.20) | 0.87 | –1.02 (–2.09, 0.04) | 0.06 | –1.24 (–2.49, 0.02) | 0.054 |
| Obese | –0.07 (–0.28, 0.15) | 0.55 | 0.001 (–0.02, 0.02) | 0.99 | –2.89 (–5.25, –0.53) | 0.02 | –2.50 (–5.14, 0.14) | 0.06 |
| P interaction | – | 0.35 | – | 0.16 | – | 0.17 | – | 0.13 |
Values (n = 272), unless indicated otherwise, are unstandardized regression coefficients (B) and 95% confidence intervals (95% CIs) for the association between blood total testosterone concentration (nmoles/liter) and Western Ontario and McMaster Universities Arthritis Index (WOMAC) pain and functional scores derived from multivariable linear robust regression models unadjusted (model 1) or adjusted for age, body mass index (BMI) status (normal weight, overweight, and obese, except BMI strata) and moderate‐to‐vigorous physical activity (model 2) for men and women separately. P values are 2‐sided and uncorrected; statistical significance was set at P value ≤ 0.05. M/F = male/female.
In women, the blood concentration of total testosterone was natural logarithmically (ln) transformed to approach normality.
Statistically significant.
Normal weight: 18.5–24.9; overweight: 25.0–29.9; obese: ≥30.0. One female senior was marginally underweight with BMI = 18.2.
At the nonoperated knee (Table 2), higher testosterone levels were associated with less WOMAC pain only in normal‐weight men (B = –0.39 [95% CI –0.68, –0.10]; P = 0.009). Further, at the nonoperated knee, higher testosterone levels were associated with less disability (better WOMAC function), among all women (B = –0.95 [95% CI –1.73, –0.17]; P = 0.02). With regard to the association between testosterone level and K/L grade, only among obese women (Table 3), per 1 nmole/liter increase in testosterone level, the odds of a higher K/L grade was reduced 10‐fold (odds ratio = 0.10 [95% CI 0.02, 0.45]; P = 0.003).
Table 3.
Cross‐sectional association between blood total testosterone concentration and K/L grade in participants of the Zurich Knee Osteoarthritis Trial stratified by sex*
| K/L grade (0–4) | Men: model 1 | Men: model 2 | Women: model 1 | Women: model 2 | ||||
|---|---|---|---|---|---|---|---|---|
| OR (95% CI) | P | OR (95% CI) | P | OR (95% CI) | P | OR (95% CI) | P | |
| Total subjects (M/F) | 1.01 (0.94, 1.08) | 0.84 | 1.00 (0.93, 1.08) | 0.90 | 0.54 (0.25, 1.19) | 0.13 | 0.66 (0.29, 1.50) | 0.32 |
| BMI, kg/m2 † | ||||||||
| Normal weight | 1.03 (0.90, 1.19) | 0.67 | 0.99 (0.85, 1.15) | 0.89 | 1.63 (0.23, 11.53) | 0.62 | 2.85 (0.31, 26.32) | 0.35 |
| Overweight | 0.98 (0.89, 1.07) | 0.59 | 0.99 (0.90, 1.08) | 0.75 | 1.01 (0.30, 3.39) | 0.99 | 0.94 (0.27, 3.26) | 0.93 |
| Obese | 1.10 (0.89, 1.38) | 0.38 | 1.11 (0.87, 1.41) | 0.39 | 0.21 (0.05, 0.87) | 0.03 | 0.10 (0.02, 0.45) | 0.003‡ |
| P interaction | – | 0.54 | – | 0.72 | – | 0.11 | – | 0.06 |
Values (n = 270), unless indicated otherwise, are odds ratios (ORs) and 95% confidence intervals (95% CIs) for the association between blood total testosterone concentration (nmoles/liter) and Kellgren/Lawrence (K/L) grade derived from ordinal logistic regression models unadjusted (model 1), or adjusted for age, body mass index (BMI) status (normal weight, overweight, obese, except BMI strata) and moderate‐to‐vigorous physical activity (model 2) for men and women separately. P values are 2‐sided and uncorrected; statistical significance is set at P value ≤ 0.05. M/F = male/female.
Normal weight: 18.5–24.9; overweight: 25.0–29.9; obese: ≥30.0. One female senior was marginally underweight with BMI = 18.2.
Statistically significant.
Discussion
In this cross‐sectional study among 272 men and women with severe knee OA who underwent unilateral TKR 6–8 weeks prior to enrollment, we found an association of testosterone level with symptoms of OA in the operated and the nonoperated knee. Specifically, among both men and women, higher serum testosterone levels were associated with less WOMAC pain in the operated knee. Further, among women but not men, higher serum testosterone levels were also associated with less WOMAC disability in the operated and the nonoperated knee. Notably, these associations were independent of age, BMI, and physical activity. With regard to subgroups by BMI, there were inconsistent signals for men in the nonoperated knee. Further, only among obese women, higher testosterone levels were associated with fewer radiographic changes due to knee OA. Consistent with the literature (25, 26), 48% of men and 42% of women had radiographic knee OA at the nonoperated knee according to standardized radiographs performed in all participants.
Our findings are consistent with an earlier study describing a nonsignificant inverse association between testosterone level and WOMAC disability among women and WOMAC stiffness among men with symptomatic knee OA age ≥60 years (18). As outlined in the introduction of this article, a potential benefit of testosterone level with regard to pain and disability among patients with severe knee OA could be explained by prior findings supporting a positive association of higher testosterone levels with better muscle strength (39, 40) and better muscle function (41) in patients with knee OA. Further, an influence of testosterone on nociceptive processing (19, 20) and inflammatory response has been suggested in several studies (7, 9, 10, 11), all of which concern important pathways in the development of OA (42, 43).
Except among obese women, we did not find an association between testosterone levels and radiographic changes of knee OA, despite the high prevalence of radiographic OA at the nonoperated knee in our study. On the one hand, this finding is in line with prior studies where symptoms and the extent of radiographic changes show discrepant findings (36, 44). On the other hand, we may have missed such an association due to the sample selection, where the most severely affected knee had undergone surgery and could not be assessed radiographically with respect to K/L grade. Notably, however, symptoms are considered the patient‐relevant feature of OA rather than the extent of radiographic changes (45).
Our study has several strengths. The results for the association of testosterone levels and pain are consistent for men and women at the operated knee, and disability for women at the operated and nonoperated knee. Also, our study has a moderately large sample size, with 272 men and women age ≥60 years. Further, we used the WOMAC questionnaire considered the gold standard for pain and disability measurement among patients with knee OA with and without TKR (31, 46). Finally, WOMAC scores of our study are representative of those reported in the literature among similar patient groups (47, 48).
Our study also has limitations. Its cross‐sectional design does not allow the exploration of cause and effect. Also, our study is a secondary analysis of a randomized controlled trial not powered for the association of testosterone level and symptoms of knee OA. Therefore, the study was possibly underpowered to show significant results for all subgroups. Further, assessments were obtained 6–8 weeks after unilateral TKR. Thus, analgesic regimens, and especially the use of opioid pain medications (49), may have influenced the association between testosterone levels and WOMAC pain and disability. Notably, in 1 large cross‐sectional study (National Health and Nutrition Examination Survey) (50), participants on opioids had a higher odds of having low testosterone levels than those unexposed to opioids, which is consistent with further literature suggesting that opioids can suppress gonadal hormone production, possibly reducing testosterone levels (49). Unfortunately, we were not able to explore the potential confounding by use of pain medication (and specifically opioid use) on pain and disability in our study. However, all participants were recruited in a stable health state, 6–8 weeks after surgery, with rehabilitation efforts largely completed, and thereby there was a reduced likelihood of exposure to opioid pain medications.
In summary, the current study suggests a possible advantage of higher sex‐specific physiologic testosterone levels among both men and women undergoing unilateral TKR due to severe knee OA. Additional studies with a prospective design are needed to further explore and clarify the role of higher physiologic testosterone levels in patients with symptomatic knee OA.
Author Contributions
All authors were involved in drafting the article or revising it critically for important intellectual content, and all authors approved the final version to be submitted for publication. Dr. Bischoff‐Ferrari had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.
Study conception and design
Orav, Egli, Theiler, Felson, Bischoff‐Ferrari.
Acquisition of data
Orav, Egli, Theiler, Felson, Bischoff‐Ferrari.
Analysis and interpretation of data
Freystaetter, Fischer, Orav, Egli, Theiler, Münzer, Felson, Bischoff‐Ferrari.
ClinicalTrials.gov identifier: NCT00599807.
Supported by the Vontobel Foundations. Dr. Bischoff‐Ferrari’s work was supported by a Swiss National Foundation Professorship (grant PP00B‐114864) and by the Baugarten Foundation Centre Grant to the Centre on Aging and Mobility at the University of Zurich and University Hospital Zurich.
Drs. Freystaetter and Fischer contributed equally to this article.
No potential conflicts of interest relevant to this article were reported.
References
- 1. Guccione AA, Felson DT, Anderson JJ, Anthony JM, Zhang Y, Wilson PW, et al. The effects of specific medical conditions on the functional limitations of elders in the Framingham Study. Am J Public Health 1994;84:351–8. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2. Murphy L, Schwartz TA, Helmick CG, Renner JB, Tudor G, Koch G, et al. Lifetime risk of symptomatic knee osteoarthritis. Arthritis Rheum 2008;59:1207–13. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3. Oliveria SA, Felson DT, Reed JI, Cirillo PA, Walker AM. Incidence of symptomatic hand, hip, and knee osteoarthritis among patients in a health maintenance organization. Arthritis Rheum 1995;38:1134–41. [DOI] [PubMed] [Google Scholar]
- 4. Hame SL, Alexander RA. Knee osteoarthritis in women. Curr Rev Musculoskelet Med 2013;6:182–7. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5. Srikanth VK, Fryer JL, Zhai G, Winzenberg TM, Hosmer D, Jones G. A meta‐analysis of sex differences prevalence, incidence and severity of osteoarthritis. Osteoarthritis Cartilage 2005;13:769–81. [DOI] [PubMed] [Google Scholar]
- 6. O'Connor MI. Osteoarthritis of the hip and knee: sex and gender differences. Orthop Clin North Am 2006;37:559–68. [DOI] [PubMed] [Google Scholar]
- 7. Koelling S, Miosge N. Sex differences of chondrogenic progenitor cells in late stages of osteoarthritis. Arthritis Rheum 2010;62:1077–87. [DOI] [PubMed] [Google Scholar]
- 8. Ushiyama T, Ueyama H, Inoue K, Ohkubo I, Hukuda S. Expression of genes for estrogen receptors alpha and beta in human articular chondrocytes. Osteoarthritis Cartilage 1999;7:560–6. [DOI] [PubMed] [Google Scholar]
- 9. Cicuttini FM, Wluka A, Bailey M, O'Sullivan R, Poon C, Yeung S, et al. Factors affecting knee cartilage volume in healthy men. Rheumatology (Oxford) 2003;42:258–62. [DOI] [PubMed] [Google Scholar]
- 10. Trigunaite A, Dimo J, Jørgensen TN. Suppressive effects of androgens on the immune system. Cell Immunol 2015;294:87–94. [DOI] [PubMed] [Google Scholar]
- 11. Bischoff‐Ferrari HA, Orav EJ, Dawson‐Hughes B. Additive benefit of higher testosterone levels and vitamin D plus calcium supplementation in regard to fall risk reduction among older men and women. Osteoporos Int 2008;19:1307–14. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12. Feldman HA, Longcope C, Derby CA, Johannes CB, Araujo AB, Coviello AD, et al. Age trends in the level of serum testosterone and other hormones in middle‐aged men: longitudinal results from the Massachusetts male aging study. J Clin Endocrinol Metab 2002;87:589–98. [DOI] [PubMed] [Google Scholar]
- 13. Bhasin S, Cunningham GR, Hayes FJ, Matsumoto AM, Snyder PJ, Swerdloff RS, et al. Testosterone therapy in men with androgen deficiency syndromes: an Endocrine Society clinical practice guideline. J Clin Endocrinol Metab 2010;95:2536–59. [DOI] [PubMed] [Google Scholar]
- 14. Goodman‐Gruen D, Barrett‐Connor E. Sex differences in the association of endogenous sex hormone levels and glucose tolerance status in older men and women. Diabetes Care 2000;23:912–8. [DOI] [PubMed] [Google Scholar]
- 15. Zumoff B, Strain GW, Miller LK, Rosner W. Twenty‐four‐hour mean plasma testosterone concentration declines with age in normal premenopausal women. J Clin Endocrinol Metab 1995;80:1429–30. [DOI] [PubMed] [Google Scholar]
- 16. Horstman AM, Dillon EL, Urban RJ, Sheffield‐Moore M. The role of androgens and estrogens on healthy aging and longevity. J Gerontol A Biol Sci Med Sci 2012;67:1140–52. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17. Sowers MF, Hochberg M, Crabbe JP, Muhich A, Crutchfield M, Updike S. Association of bone mineral density and sex hormone levels with osteoarthritis of the hand and knee in premenopausal women. Am J Epidemiol 1996;143:38–47. [DOI] [PubMed] [Google Scholar]
- 18. Miller GD, Nicklas BJ, Davis CC, Legault C, Messier SP. Basal growth hormone concentration increased following a weight loss focused dietary intervention in older overweight and obese women. J Nutr Health Aging 2012;16:169–74. [DOI] [PubMed] [Google Scholar]
- 19. Maurer AJ, Lissounov A, Knezevic I, Candido KD, Knezevic NN. Pain and sex hormones: a review of current understanding. Pain Manag 2016;6:285–96. [DOI] [PubMed] [Google Scholar]
- 20. Gupta S, McCarson KE, Welch KM, Berman NE. Mechanisms of pain modulation by sex hormones in migraine. Headache 2011;51:905–22. [DOI] [PubMed] [Google Scholar]
- 21. Cairns BE, Gazerani P. Sex‐related differences in pain. Maturitas 2009;63:292–6. [DOI] [PubMed] [Google Scholar]
- 22. Srinivas‐Shankar U, Roberts SA, Connolly MJ, O'Connell MD, Adams JE, Oldham JA, et al. Effects of testosterone on muscle strength, physical function, body composition, and quality of life in intermediate‐frail and frail elderly men: a randomized, double‐blind, placebo‐controlled study. J Clin Endocrinol Metab 2010;95:639–50. [DOI] [PubMed] [Google Scholar]
- 23. Storer TW, Basaria S, Traustadottir T, Harman SM, Pencina K, Li Z, et al. Effects of testosterone supplementation for 3 years on muscle performance and physical function in older men. J Clin Endocrinol Metab 2017;102:583–93. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24. Amory JK, Chansky HA, Chansky KL, Camuso MR, Hoey CT, Anawalt BD, et al. Preoperative supraphysiological testosterone in older men undergoing knee replacement surgery. J Am Geriatr Soc 2002;50:1698–701. [DOI] [PubMed] [Google Scholar]
- 25. Gunther KP, Sturmer T, Sauerland S, Zeissig I, Sun Y, Kessler S, et al. Prevalence of generalised osteoarthritis in patients with advanced hip and knee osteoarthritis: the Ulm Osteoarthritis Study. Ann Rheum Dis 1998;57:717–23. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26. McMahon M, Block JA. The risk of contralateral total knee arthroplasty after knee replacement for osteoarthritis. J Rheumatol 2003;30:1822–4. [PubMed] [Google Scholar]
- 27. Bischoff‐Ferrari HA, Orav EJ, Egli A, Dawson‐Hughes B, Fischer K, Staehelin HB, et al. Recovery after unilateral knee replacement due to severe osteoarthritis and progression in the contralateral knee: a randomised clinical trial comparing daily 2000 IU versus 800 IU vitamin D. RMD Open 2018;4:e000678. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28. Resnick B, Nahm ES, Orwig D, Zimmerman SS, Magaziner J. Measurement of activity in older adults: reliability and validity of the step activity monitor. J Nurs Meas 2001;9:275–90. [PubMed] [Google Scholar]
- 29. White DK, Tudor‐Locke C, Zhang Y, Fielding R, LaValley M, Felson DT, et al. Daily walking and the risk of incident functional limitation in knee osteoarthritis: an observational study. Arthritis Care Res (Hoboken) 2014;66:1328–36. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30. Brazier JE, Harper R, Munro J, Walters SJ, Snaith ML. Generic and condition‐specific outcome measures for people with osteoarthritis of the knee. Rheumatology (Oxford) 1999;38:870–7. [DOI] [PubMed] [Google Scholar]
- 31. Bombardier C, Melfi CA, Paul J, Green R, Hawker G, Wright J, et al. Comparison of a generic and a disease‐specific measure of pain and physical function after knee replacement surgery. Med Care 1995;33:AS131–44. [PubMed] [Google Scholar]
- 32. Stucki G, Meier D, Stucki S, Michel BA, Tyndall AG, Dick W, et al. Evaluation of a German version of WOMAC (Western Ontario and McMaster Universities) arthrosis index. Z Rheumatol 1996;55:40–9. In German. [PubMed] [Google Scholar]
- 33. Thorlund JB, Felson DT, Segal NA, Nevitt MC, Niu J, Neogi T, et al. Effect of knee extensor strength on incident radiographic and symptomatic knee osteoarthritis in individuals with meniscal pathology: data from the Multicenter Osteoarthritis Study. Arthritis Care Res (Hoboken) 2016;68:1640–6. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 34. Kellgren JH, Lawrence JS. Radiological assessment of osteo‐arthrosis. Ann Rheum Dis 1957;16:494–502. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 35. Zheng H, Chen C. Body mass index and risk of knee osteoarthritis: systematic review and meta‐analysis of prospective studies. BMJ Open 2015;5:e007568. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 36. Felson DT. The epidemiology of knee osteoarthritis: results from the Framingham Osteoarthritis Study. Semin Arthritis Rheum 1990;20:42–50. [DOI] [PubMed] [Google Scholar]
- 37. DeFina LF, Radford NB, Leonard D, Wilson RK, Cooper TC, Clark SM, et al. The association of cardiorespiratory fitness, body mass index, and age with testosterone levels at screening of healthy men undergoing preventive medical examinations: the Cooper Center Longitudinal Study. Maturitas 2018;118:1–6. [DOI] [PubMed] [Google Scholar]
- 38. Deutschbein T, Mann K, Petersenn S. Total testosterone and calculated estimates for free and bioavailable testosterone: influence of age and body mass index and establishment of sex‐specific reference ranges. Horm Metab Res 2015;47:846–54. [DOI] [PubMed] [Google Scholar]
- 39. Nam YS, Lee G, Yun JM, Cho B. Testosterone replacement, muscle strength, and physical function. World J Men's Health 2018;36:110–22. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 40. Magnussen LV, Hvid LG, Hermann AP, Hougaard DM, Gram B, Caserotti P, et al. Testosterone therapy preserves muscle strength and power in aging men with type 2 diabetes: a randomized controlled trial. Andrology 2017;5:946–53. [DOI] [PubMed] [Google Scholar]
- 41. Metcalfe D, Watts E, Masters JP, Smith N. Anabolic steroids in patients undergoing total knee arthroplasty. BMJ Open 2012;2:e001435. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 42. Felson DT, Zhang Y, Hannan MT, Naimark A, Weissman B, Aliabadi P, et al. Risk factors for incident radiographic knee osteoarthritis in the elderly: the Framingham Study. Arthritis Rheum 1997;40:728–33. [DOI] [PubMed] [Google Scholar]
- 43. Felson DT, Lawrence RC, Dieppe PA, Hirsch R, Helmick CG, Jordan JM, et al. Osteoarthritis: new insights. Part 1: the disease and its risk factors. Ann Intern Med 2000;133:635–46. [DOI] [PubMed] [Google Scholar]
- 44. Hannan MT, Felson DT, Pincus T. Analysis of the discordance between radiographic changes and knee pain in osteoarthritis of the knee. J Rheumatol 2000;27:1513–7. [PubMed] [Google Scholar]
- 45. Zhang W, Doherty M, Peat G, Bierma‐Zeinstra MA, Arden NK, Bresnihan B, et al. EULAR evidence‐based recommendations for the diagnosis of knee osteoarthritis. Ann Rheum Dis 2010;69:483–9. [DOI] [PubMed] [Google Scholar]
- 46. Escobar A, Quintana JM, Bilbao A, Aróstegui I, Lafuente I, Vidaurreta I. Responsiveness and clinically important differences for the WOMAC and SF‐36 after total knee replacement. Osteoarthritis Cartilage 2007;15:273–80. [DOI] [PubMed] [Google Scholar]
- 47. Waimann CA, Fernandez‐Mazarambroz RJ, Cantor SB, Lopez‐Olivo MA, Zhang H, Landon GC, et al. Cost‐effectiveness of total knee replacement: a prospective cohort study. Arthritis Care Res (Hoboken) 2014;66:592–9. [DOI] [PubMed] [Google Scholar]
- 48. Gstoettner M, Raschner C, Dirnberger E, Leimser H, Krismer M. Preoperative proprioceptive training in patients with total knee arthroplasty. Knee 2011;18:265–70. [DOI] [PubMed] [Google Scholar]
- 49. Katz N, Mazer NA. The impact of opioids on the endocrine system. Clin J Pain 2009;25:170–5. [DOI] [PubMed] [Google Scholar]
- 50. Cepeda MS, Zhu V, Vorsanger G, Eichenbaum G. Effect of opioids on testosterone levels: cross‐sectional study using NHANES. Pain Med 2015;16:2235–42. [DOI] [PubMed] [Google Scholar]