Geographical incidence of testosterone supplementation in total hip and knee arthroplasty
Article History
-
Received: ,
Accepted: ,
Abstract
Abstract
Prescription testosterone usage in the United States declined after 2013 following reports of its association with myocardial infarction and stroke. However, more recently there has been a documented increase in testosterone prescriptions. Recently, testosterone levels have also been hypothesized to increase infection risk in patients undergoing elective shoulder arthroplasty. Furthermore, testosterone may increase the risk of venous thromboembolism. These complications are perioperative concerns for total hip and knee arthroplasties (THA/TKA). Therefore, the purpose of our study is to identify trends in the incidence of testosterone prescriptions in patients who underwent THA/TKA with respect to geographical population data.
We retrospectively reviewed 40,711 primary THAs and 50,893 primary TKAs performed in males between 1/1/2016 and 12/31/2021 using a commercial claims database. Records were reviewed for demographics, geographical location, and supplemental testosterone prescriptions within 1 year prior to surgery. Patient Metropolitan Statistical Area (MSA) was assessed with respect to United States Census Population Data.
We identified 91,604 males who underwent primary THA (n = 40,711) or TKA (n = 50,893). For THA/TKA, patients who were younger had a higher likelihood of having a supplemental testosterone prescription (OR = 0.99, 95 % CI [0.99–1.00], p < 0.001). TKA patients (2,507, 4.9 %) had a higher rate of testosterone prescriptions than THA patients overall (1,413, 3.4 %), (OR = 1.44 95 % CI [1.35, 1.54], p < 0.001) as well as within each year.
For THA and TKA patients, patients in the Midwest (OR = 1.61, p < 0.001), South (OR = 3.04, p < 0.001), and West (OR = 2.28, p < 0.001) regions all had higher testosterone prescription rates than the Northeast. Patients living in a city (MSA population ≥200,000) were more likely to be prescribed testosterone (OR = 1.20, p = 0.002).
Surgeons conducting TKA/THA should be aware that younger patients, those in higher population areas, and those in the Midwest, South, and West regions are more likely to be prescribed testosterone than those in the Northeast.
Keywords
This article was originally published by Reed Elsevier India Pvt. Ltd. and was migrated to Scientific Scholar after the change of Publisher.
1 Introduction
Testosterone deficiency (TD) is defined as two measurements of a total testosterone level below 300 ng/dL with clinical signs and symptoms including reduced energy, endurance, performance, motivation, mood, memory, infertility, and/or libido.1 Testosterone deficiency is typically treated with testosterone replacement therapy (TRT); however, it has also been demonstrated that there are many patients who were prescribed supplemental testosterone without this indication.2 Prescription testosterone usage in men in the United States peaked in 2013 but declined after reports of its association with myocardial infarction and stroke.3 However, a recent study found that from 2016 to 2019, there was a 54.6 % increase in testosterone supplementation claims and the incidence of testosterone prescriptions per prescriber also increased among several specialties.4 Of note, the incidence of testosterone supplementation is highest among patients who were age 40–64 and was as high as ∼0.75 % among these patients as of 2016.3 This patient age range is relevant to the total joint arthroplasty (TJA) population.
The cardiovascular risks of supplemental testosterone have been well-documented including the increased risk of venous thromboembolism.5–7 Despite this, there has been minimal investigation of the prevalence and impact of testosterone supplementation in the TJA population. One recent study found that increased testosterone levels were associated with increased levels of Cutibacterium on the skin in patients undergoing elective shoulder arthroplasty, which in turn is associated with an increased risk of prosthetic joint infection (PJI).8 PJI is a devastating complication following total knee arthroplasty and total hip arthroplasty (TKA/THA) resulting in significant morbidity and mortality.9–12 The treatment for PJI involves multiple operations, increasing the risk of temporary or permanent disability with an estimated 5-year mortality for TKA PJI necessitating 2-stage revision of 21.6 %.9–13 Therefore, any unidentified risk factors for PJI should be thoroughly investigated, and demographic information that may correlate with these risk factors should be consolidated.
Most of the evidence surrounding hormone levels and supplementation in the field of arthroplasty has been focused on trials of sex steroid hormones for recovery following TJA and the correlative factors of sex steroid concentrations and osteoarthritis/TJA.14–18 Of note, one recent study by Collins et al., 2023 found that testosterone replacement therapy had an increased risk of revision, PJI, and complications within 2 years following THA and an increased rate of septic and aseptic revision and aseptic loosening following TKA.19
These recent findings have sparked an interest in understanding the prevalence of testosterone usage in patients undergoing TJA. The purpose of this study is to identify trends in the incidence of testosterone prescriptions in patients who underwent Total Hip or Knee Arthroplasty (THA/TKA) with respect to geographical population data.
2 Methods
2.1 Study design
After approval from our institutional review board, a retrospective cohort study was conducted using a large commercial claims database. Patients were identified as having a prescription for preoperative supplemental testosterone within 1 year of THA or TKA. Patients were retrospectively reviewed for demographics and geographical information.
2.2 Setting
This study included patients enrolled in the MarketScan Commercial Claims Database (Merative, Ann Arbor, Michigan) from January 1, 2016 to December 31, 2021.
2.3 Participants
A total of 244,080 primary THAs and 419,095 primary TKA claims were identified based on the Current Procedural Terminology and International Classification of Diseases Tenth Edition (ICD-10) Procedural Coding System codes listed in Supplementary Table 1. Patients with a preoperative diagnosis of gender dysmorphia were excluded using the ICD-10 codes described in Supplementary Table 1. Additional claims were excluded based on the criteria described in Figs. 1 and 2. A total of 40,711 THAs and 50,893 TKAs were included in the final study population.


2.4 Variables
Data collected included procedure type, presence of a supplemental testosterone prescription within one year of surgery, year of surgery, age at time of surgery, geographical region, metropolitan statistical area (MSA), population per MSA.
2.5 Data sources/management
Data was collected from the MarketScan Commercial Claims Database (Merative, Ann Arbor, Michigan). We chose to define a metropolitan statistical area (MSA) as a city if it had a population of ≥200,000 people. This is consistent with the Organization for Economic Co-operation and Development (OECD) definition of a middle-sized city as having a population of at least 200,000 people. United States Census Data was used to quantify the population in each MSA as well as define the geographical region.20–22
2.6 Bias
This study uses a large national deidentified commercial claims database. Therefore, this reduces publication bias, observer/experimenter bias, sampling bias, recall bias, selection bias, and the Hawthorne effect.
2.7 Study size
Patients with a preoperative diagnosis of gender dysmorphia were excluded using the ICD-10 codes described in Supplementary Table 1. Additional claims were excluded based on the criteria described in Figs. 1 and 2. A total of 40,711 THAs and 50,893 TKAs were included in the final study population.
2.8 Quantitative variables
Year of surgery, age at time of surgery, and prescription for supplemental testosterone were collected for all patients. Geographical region population in each MSA was defined using United States Census Data. City was defined as a population ≥200,000 according to the OECD.22
2.9 Statistical method
Python (Python Software Foundation, Beaverton, OR) was used for statistical analysis. THA and TKA cohorts were compared using chi-square and independent samples t-tests. Risk factors for being prescribed testosterone supplementation were evaluated using univariate and multivariate logistic regression. Odds ratios (OR) and 95 % confidence intervals (CI) were calculated. Significance was set at p < 0.05.
3 Results
3.1 Participants
Our study identified 91,604 males who underwent primary THA (n = 40,711) or TKA (n = 50,893) (Figs. 1 and 2).
3.2 Descriptive data
Demographic information regarding average local population size, geographical region, preoperative testosterone prescriptions, and area population type (city vs non-city) is included in Table 1.
| THA n = 40,711 | TKA n = 50,893 | Total n = 91,604 | p-value | |
| Age (years), mean (range) | 57.0 (18–97) | 59.6 (18–96) | 58.4 (18–97) | p < 0.001 |
| Geographical Region, n (%) | N/A | N/A | N/A | p < 0.001 |
| Northeast | 7573 (18.6) | 6781 (13.3) | 14,354 (15.7) | N/A |
| Midwest | 11,117 (27.3) | 15,195 (29.9) | 26,312 (28.7) | N/A |
| South | 16,234 (39.9) | 21,955 (43.1) | 38,189 (41.7) | N/A |
| West | 5787 (14.2) | 6962 (13.7) | 12,749 (13.9) | N/A |
| Preoperative Testosterone, n (%) | N/A | N/A | N/A | p < 0.001 |
| No Testosterone | 39298 (96.5) | 48386 (95.1) | 87684 (95.7) | N/A |
| Testosterone | 1413 (3.5) | 2507 (4.9) | 3920 (4.3) | N/A |
| City versus Non-City, n (%) | N/A | N/A | N/A | p < 0.001 |
| Non-City | 3331 (8.2) | 5053 (9.9) | 8384 (9.2) | N/A |
| City | 37380 (91.8) | 45840 (90.1) | 83220 (90.8) | N/A |
3.3 Outcome data
Age.
For THA and TKA combined, patients who were younger had a higher likelihood of having a supplemental testosterone prescription (OR = 0.99, p<0.001) (Tables 2 and 3).
| OR (95 % CI) | No Preoperative Testosterone n = 87,684 | Preoperative Testosterone n = 3920 | Total n = 91,604 | p-value | |
| Age (years), mean (range) | 0.99 (0.99–1.00) | 58.5 (18–97) | 57.8 (22–90) | 58.4 (18–97) | p < 0.001 |
| Geographical Region, n (%) | N/A | N/A | N/A | N/A | N/A |
| Northeast | Ref | 14067 (16.0) | 287 (7.3) | 14354 (15.7) | Ref |
| Midwest | 1.61(1.41–1.84) | 25475 (29.1) | 837 (21.4) | 26312 (28.7) | p < 0.001 |
| South | 3.04 (2.68–3.44) | 35959 (41.0) | 2230 (56.9) | 38189 (41.7) | p < 0.001 |
| West | 2.28 (1.97–2.63) | 12183 (13.9) | 566 (14.4) | 12749 (13.9) | p < 0.001 |
| Procedure, n (%) | N/A | N/A | N/A | N/A | N/A |
| THA | Ref | 39298 (44.8) | 1413 (36.0) | 40711 (44.4) | Ref |
| TKA | 1.44 (1.35–1.54) | 48386 (55.2) | 2507 (64.0) | 50893 (55.6) | p < 0.001 |
| City versus Non-City, n (%) | N/A | N/A | N/A | N/A | N/A |
| Non-city | Ref | 8079 (9.2) | 305 (7.8) | 8384 (9.2) | Ref |
| City | 1.20 (1.07–1.36) | 79605 (90.8) | 3615 (92.2) | 83220 (90.8) | p = 0.002 |
| OR (95 % CI) | p-value | |
| Procedure | N/A | N/A |
| THA | Ref | Ref |
| TKA | 1.44 (1.35–1.55) | p < 0.001 |
| Geographical Region | N/A | N/A |
| Northeast | Ref | Ref |
| Midwest | 1.56 (1.35–1.80) | p < 0.001 |
| South | 2.90 (2.54–3.31) | p < 0.001 |
| West | 2.21 (1.90–2.56) | p < 0.001 |
| Age (years) | 0.99 (0.98–0.99) | p < 0.001 |
| City versus Non-City | 1.19 (1.05–1.34) | p = 0.005 |
Time.
We observed no significant difference in the rate of testosterone prescriptions for THA (OR 1.00, p = 1.0) and TKA (OR = 1.00, p = 0.2) from year-to year. (Fig. 3).

THA versus TKA.
TKA patients (2,507, 4.9 %) had a higher rate of testosterone prescriptions than THA patients overall (1,413, 3.4 %), (OR = 1.44, p<0.001) as well as within each year (Fig. 3; Tables 1 and 3).
3.4 Main results
Geographical Region.
For THA and TKA patients combined, patients in the Midwest (OR = 1.61, p<0.001), South (OR = 3.04, p<0.001), and West (OR = 2.28, p<0.001) regions all had higher testosterone prescription rates than the Northeast (Table 3).
3.5 City versus non-city
For the 63,101 THA/TKA patients with available MSA and Census data, patients living in an area with a population ≥200,000 were more likely to be prescribed testosterone (OR = 1.20, p = 0.002).
Other Analyses.
THA.
In a subcohort analysis of THA patients, there was no significant trend with respect to age (OR 0.99, p = 0.7). However, univariate (p < 0.001) and multivariate (p < 0.001) analyses with respect to geographical region demonstrated significant variation (Tables 4 and 5). Additionally, univariate (OR = 1.46, p < 0.001) and multivariate (OR = 1.42, p < 0.002) analyses demonstrated significantly higher likelihood of testosterone prescriptions for patients in cities.
| OR (95 % CI) | No Preoperative Testosterone n = 39,298 | Preoperative Testosterone n = 1413 | Total n = 40,711 | p-value | |
| Age (years), mean (range) | 0.99 (0.99–1.00) | 57.0 (18–97) | 56.9 (22–90) | 57.0 (18–97) | p = 0.7 |
| Geographical Region, n (%) | N/A | N/A | N/A | N/A | N/A |
| Northeast | Ref | 7437 (18.9) | 136 (9.6) | 7573 (18.6) | Ref |
| Midwest | 1.43 (1.16–1.76) | 10834 (27.6) | 283 (20.0) | 11117 (27.3) | p = 0.001 |
| South | 2.74 (2.28–3.29) | 15460 (39.3) | 774 (54.8) | 16234 (39.9) | p < 0.001 |
| West | 2.16 (1.74–2.68) | 5567 (14.2) | 220 (15.6) | 5787 (14.2) | p < 0.001 |
| City versus Non-city, n (%) | N/A | N/A | N/A | N/A | N/A |
| Non-City | Ref | 3249 (8.3) | 82 (5.8) | 3331 (8.2) | Ref |
| City | 1.46 (1.17–1.83) | 36049 (91.7) | 1331 (94.2) | 37380 (91.8) | p = 0.001 |
| OR (95 % CI) | p-value | |
| Geographical Region | N/A | N/A |
| Northeast | Ref | Ref |
| Midwest | 1.48 (1.18–1.86) | p = 0.001 |
| South | 2.76 (2.29–3.32) | p < 0.001 |
| West | 2.18 (1.75–2.70) | p < 0.001 |
| Age (years) | 1.00 (0.99–1.01) | p = 0.471 |
| City versus Non-City | 1.42 (1.13–1.78) | p = 0.002 |
TKA.
In a subcohort analysis of TKA patients, younger patients were found to have an increased likelihood of testosterone supplementation (OR 0.97, p < 0.001) (Tables 6 and 7). Additionally, univariate (p < 0.001) and multivariate (p < 0.001) analyses with respect to geographical region demonstrated significant variation. There were no significant trends observed with respect to City versus Non-City (univariate, p = 0.076; multivariate, p = 0.196) (Tables 6 and 7).
| OR (95 % CI) | No Preoperative Testosterone n = 48,386 | Preoperative Testosterone n = 2507 | Total n = 50,893 | p-value | |
| Age (years), mean (range) | 0.97 (0.97–0.98) | 59.7 (18–96) | 58.3 (34–86) | 59.6 (18–96) | p < 0.001 |
| Geographical Region, n (%) | N/A | N/A | N/A | N/A | N/A |
| Northeast | Ref | 6630 (13.7) | 151 (6.0) | 6781 (13.3) | Ref |
| Midwest | 1.66 (1.38–1.99) | 14641 (30.3) | 554 (22.1) | 15195 (29.9) | p < 0.001 |
| South | 3.11 (2.63–3.70) | 20499 (42.4) | 1456 (58.1) | 21955 (43.1) | p < 0.001 |
| West | 2.30 (1.89–2.79) | 6616 (13.7) | 346 (13.8) | 6962 (13.7) | p < 0.001 |
| City versus Non-City, n (%) | N/A | N/A | N/A | N/A | N/A |
| Non-City | Ref | 4830 (10.0) | 223 (8.9) | 5053 (9.9) | Ref |
| City | 1.13 (0.99–1.31) | 43556 (90.0) | 2284 (91.1) | 45840 (90.1) | p = 0.076 |
| OR (95 % CI) | p-value | |
| Geographical Region | N/A | N/A |
| Northeast | Ref | Ref |
| Midwest | 1.70 (1.41–2.03) | p < 0.001 |
| South | 3.07 (2.59–3.64) | p < 0.001 |
| West | 2.28 (1.87–2.77) | p < 0.001 |
| Age (years) | 0.97 (0.97–0.98) | p < 0.001 |
| City versus Non-City | 1.10 (0.95–1.26) | p = 0.196 |
4 Discussion
Testosterone replacement therapy has been associated with an increased risk of cardiovascular complications.5–7 Given the overlap of these complications (e.g. myocardial infarction, venous thromboembolism) with those observed with major surgical procedures such as total joint arthroplasty, recent studies have sought to investigate the impact of testosterone supplementation on outcomes in total joint arthroplasty. Collins et al., 2023 found that testosterone replacement therapy had an increased risk of revision, PJI, and complications within 2 years following THA and TKA.19 Therefore, we sought to identify the demographic and geographic patient factors correlated with testosterone supplementation to inform surgeons to screen patients for testosterone usage.
4.1 Key results
Our study found that patients who were younger, underwent TKA, were in areas of the country outside of the Northeast, and who lived in an area with a population ≥200,000 were more likely to be prescribed supplemental testosterone. To the best of our knowledge, Collins et al., 2023 is the only other study that has investigated the prevalence of hormone replacement therapy in TJA. Their study found similar results with respect to geographical distribution of TRT, in that the Northeast had the lowest prevalence of TRT users and the South had the highest.19 These results are also similar to those found by Baillargeon et al., 2018 for all patients in the USA from 2002-2016.3 Collins et al., 2023 also found that there was a higher prevalence of depression, coronary artery disease, and osteoporosis for patients who were utilizing TRT.19 Of note, this study represented a different patient population as this study utilized the PearlDiver Mariner Database (PearlDiver Inc., Colorado Springs, CO, USA) database whereas the MarketScan database was used for this study.
The rest of the literature on hormone levels and supplementation in arthroplasty has been limited. These studies have included small trials of sex steroid hormones for recovery following TJA.14,15,23 One small study conducted by Hohmann et al., 2020 followed 5 control patients and 5 patients who were given 50 mg nandrolone decanoate intramuscularly bi-weekly for 6 months following TKA.14 Their results showed higher quadriceps strength and knee society scores (KSS) in the treatment group. Amory et al., 2002 conducted a slightly larger study of 25 patients using 600 mg testosterone enanthate intramuscularly once weekly for 4 weeks postoperatively and found that there was shorter length of stay and an improvement in the ability to stand at postoperative day 3.23 However, a 2012 scoping review of these two studies determined that these were the only trials to date investigating this question and asserted that these two studies did not amount to sufficient evidence to guide practice.15
To the best of our knowledge the only other studies to investigate sex hormone profiles in osteoarthritis and total joint arthroplasty are the three studies investigating circulating and salivary steroid hormone profiles in patients who had osteoarthritis/TJA.16–18 One study found that higher levels of dihydrotestosterone (DHT) and testosterone were correlated with higher rate of hip replacement.16 Another study found that higher levels of androstenedione were associated with a decreased rate of hip and knee arthroplasty. Perego et al., 2015 found improved pain scores following total hip and knee arthroplasty with higher testosterone to cortisol ratios.18 Overall, these findings are difficult to compare and synergize because they investigate different derivatives of testosterone and have varying and conflicting outcomes.
Though the literature is limited on this topic with respect to the impact of sex hormone supplementation on outcomes in arthroplasty, there have been at least two studies that have demonstrated adverse effects of supplementation of testosterone in TJA.8,19 One of these studies displayed similar findings to our study regarding geographic distribution, but they did not identify differences with age or geographical population. Therefore, our findings will add to the literature and inform surgeons who live in cities located in the West, Midwest, and Southern United States to be cognizant of the fact that there are higher rates of testosterone supplementation in their area. We advise that surgeons utilize this information as rationale to screen for this potential risk factor prior to TJA. Further investigation is needed on the impact of testosterone supplementation on PJI and other complications.
4.2 Limitations and strengths
This study was a large database study which is inherently prone to administrative and medical coding errors. Additionally, though we did utilize the MSA data to estimate population size, there are other epidemiological factors including physician practice factors that may influence testosterone prescribing patterns that were likely unaccounted for. It is difficult to determine the potential direction or magnitude of these factors’ influence. Another potential factor that may not be accounted for is illegal distribution or acquisition of prescription testosterone. In addition, our study is limited in scope to testosterone and does not capture other hormone therapies that might have provided additional context to demographic distribution data.
5 Conclusions
We found that younger males, those who live in areas with a population of ≥200,000, those who underwent TKA, and those in the West, Midwest, and Southern regions of the United States were more likely to have testosterone supplementation within the year leading up to TJA. We advise that surgeons in these areas screen for this potential risk factor prior to TJA.
Funding
This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.
Ethical approval and patient consent
The work described has been carried out in accordance with The Code of Ethics of the World Medical Association (Declaration of Helsinki) for experiments involving humans. All procedures were performed in compliance with relevant laws and institutional guidelines. This work was approved by our institutional review board. This study used a large national deidentified database.
CRediT authorship contribution statement
Gloria Coden: Methodology, Software, Formal analysis, Investigation, Resources, Data curation, Writing – original draft, Visualization. Andrew Grant: Methodology, Conceptualization, Writing – original draft, Writing – review & editing, Validation, Visualization, Resources, Investigation, Formal analysis. Mikhail Kuznetsov: Writing – review & editing, Validation. Eric L. Smith: Conceptualization, Project administration, Supervision.
References
- Evaluation and management of testosterone deficiency: AUA guideline. J Urol. 2018;200(2):423-432.
- [Google Scholar]
- Characteristics of men undergoing testosterone replacement therapy and adherence to follow-up recommendations in metropolitan multicenter health care System. Urology. 2015;85(6):1382-1388.
- [Google Scholar]
- Testosterone prescribing in the United States, 2002-2016. JAMA. 2018;320(2):200-202.
- [Google Scholar]
- Trends in testosterone prescription during the release of society guidelines. Int J Impot Res 2023:1-5.
- [Google Scholar]
- Association of testosterone therapy with risk of venous thromboembolism among men with and without hypogonadism. JAMA Intern Med. 2020;180(2):190-197.
- [Google Scholar]
- Testosterone therapy: what we have learned from trials. J Sex Med. 2020;17(3):447-460.
- [Google Scholar]
- Association between serum testosterone levels and Cutibacterium skin load in patients undergoing elective shoulder arthroplasty: a cohort study. JB JS Open Access. 2021;6(4)
- [Google Scholar]
- Periprosthetic joint infection: the incidence, timing, and predisposing factors. Clin Orthop Relat Res. 2008;466(7):1710-1715.
- [Google Scholar]
- Mortality during total knee periprosthetic joint infection. J Arthroplasty. 2018;33(12):3783-3788.
- [Google Scholar]
- Periprosthetic joint infection increases the risk of one-year mortality. J Bone Joint Surg Am. 2013;95(24):2177-2184.
- [Google Scholar]
- Are we winning or losing the battle with periprosthetic joint infection: trends in periprosthetic joint infection and mortality risk for the medicare population. J Arthroplasty. 2018;33(10):3238-3245.
- [Google Scholar]
- Socioeconomic disadvantage predicts decreased likelihood of maintaining a functional knee arthroplasty following treatment for prosthetic joint infection. J Arthroplasty. 2024;39(7):1828-1833.
- [Google Scholar]
- Anabolic steroids after total knee arthroplasty. A double blinded prospective pilot study. J Orthop Surg Res. 2010;5:93.
- [Google Scholar]
- Anabolic steroids in patients undergoing total knee arthroplasty. BMJ Open. 2012;2(5)
- [Google Scholar]
- Sex steroids and osteoarthritis: a mendelian randomization study. Front Endocrinol. 2021;12
- [Google Scholar]
- Relationship between circulating sex steroid hormone concentrations and incidence of total knee and hip arthroplasty due to osteoarthritis in men. Osteoarthritis Cartilage. 2016;24(8):1408-1412.
- [Google Scholar]
- Perisurgical and intra-rehabilitative salivary steroid hormone profiles in bicompartmental arthroplasty. J Biol Regul Homeost Agents. 2015;29(4):953-960.
- [Google Scholar]
- Hormone replacement therapy does not eliminate risk factors for joint complications following total joint arthroplasty: a matched cohort study. Pathophysiology. 2023;30(2):123-135.
- [Google Scholar]
- Urban population by city size (indicator) 2024
- [Google Scholar]
- Preoperative supraphysiological testosterone in older men undergoing knee replacement surgery. J Am Geriatr Soc. 2002;50(10):1698-1701.
- [Google Scholar]

