Authors: Roberta Carminati Shimano, Ariane Zamarioli, Ana Paula Macedo, Viviane de Cassia Oliveira, Felipe Manuel Chiocca Nascimento, Patrícia Madalena San Gregório Guedes, Ricardo Tadeu Lopes, José Batista Volpon
Categories: Article, Metabolic disorders, Fracture repair
Source: Scientific Reports
Authors: Roberta Carminati Shimano, Ariane Zamarioli, Ana Paula Macedo, Viviane de Cassia Oliveira, Felipe Manuel Chiocca Nascimento, Patrícia Madalena San Gregório Guedes, Ricardo Tadeu Lopes, José Batista Volpon
Escitalopram oxalate is a widely used drug for treating depressive conditions in various age groups. However, it exerts systemic side effects beyond its primary action in the brain, including impacts on bone, with negative repercussions on bone quality. Yet, the influence of this drug on fracture healing has not been investigated, and this is the focus of the present study. Fractures were created in the right femur diaphysis in healthy adult rats and those with osteopenia induced by prior ovariectomy. All animals underwent right femur fracture, and the subsequent healing process was examined. Four groups were Healthy Adult Sham, Osteoporosis Sham, Healthy Adult Escitalopram, and Osteoporosis Escitalopram. Oral administration of escitalopram was conducted daily via gavage for 35 days, while sham treatment consisted of distilled water administration via gavage. The bone callus analysis included the determination of bone mineral density, Computed microtomography images, mechanical resistance testing, and histomorphometry. Exposure to escitalopram resulted in the disruption of the bone callus, characterized by a decrease in trabecular thickness, an increase in trabecular separation, and greater deposition of type I collagen. Bone mineral density and strength, however, remained unaffected. In conclusion, escitalopram oxalate negatively interfered with fracture repair in both healthy and osteopenic rat bones.
Antidepressants are commonly prescribed across different age groups. Recent research has established a positive correlation between the chronic use of selective serotonin reuptake inhibitors (SSRIs) and a decrease in bone mineral density^1–3^. This association raises concerns about an elevated risk of osteoporotic fractures among SSRI users. However, the precise influence of these drugs on the process of fracture healing remains uncertain.
Selective Serotonin Reuptake Inhibitors (SSRIs) represent a widely prescribed class of antidepressants that function by antagonizing the serotonin transporter. Certain medications have been linked to reduced bone density and an increased risk of osteoporosis, mainly when administered to postmenopausal women. However, the exact mechanism by which bone mass decreases is not fully understood and requires further research^2^. Previous research has elucidated the functional role of the serotonin signaling pathway in bone cells^4^. A recent study states that future analyses need to focus on the mechanistic pathway by which osteoclasts and the bone cell feedback cycle are altered with SSRI treatment. Additionally, the authors propose that osteoclasts are the most vulnerable cells in this relationship^5^.
For adults with a low risk of bone loss, there are questions about whether SSRI users should be categorized as a medically vulnerable population for osteopenia. It is important to note that SSRIs are not currently listed among the drugs known to cause or contribute to an increased risk of fractures^6^. In the case of menopausal women, heightened attention is warranted due to the metabolic and hormonal changes characteristic of this age group, which are recognized risk factors for bone loss and fractures^7^. Consequently, the administration of antidepressants may potentially accelerate or intensify the pre-existing bone loss^8^.
The precise impact of SSRIs on bone metabolism and healing has not been definitively established^9–11^, but there is a growing body of evidence suggesting that these drugs might disrupt osteoblast function, resulting in compromised bone strength and an increased risk of fractures^1^.
Furthermore, despite scientific evidence regarding the influence of antidepressants on reducing bone density^2,3,12^, the impact of antidepressants on fracture healing in adults and menopausal individuals is a subject that warrants further investigation.
To fill this gap in the existing literature, our study aims to assess the effects of escitalopram oxalate, an SSRI, on fracture healing in two distinct skeletal healthy bone and the osteoporotic bone. Examining these distinct scenarios will contribute to an enhanced understanding of the connection between antidepressants and their impact on bone healing, potentially having significant implications for patient care.
Initially, the anesthesia method employed for ovariectomy and fracture surgeries involved the use of xylazine and ketamine. In the initial groups that underwent fracture procedures, 15% of the animals in the placebo group did not survive, with the majority succumbing either during surgery or on the following day, primarily due to complications associated with anesthesia. In contrast, among the groups that received escitalopram oxalate, a significant 90% did not survive beyond the day of surgery. Faced with these challenges, a decision was made to change the anesthetic approach. Subsequently, all groups underwent surgery using inhalational anesthesia with isoflurane, which effectively mitigated further losses. It is important to note that no losses could be attributable to surgical complications, such as excessive fragmentation of bone fragments, loosening of K-wires, or infection.
Throughout the observation period, there was a consistent increase in body mass across all animal groups. However, ovariectomized animals maintained a higher plateau (p < 0.001) compared to the others. Adult animals exhibited a significant increase in body mass from the second to the third week (p < 0.001), which was followed by a period of stabilization during the third and fifth weeks, and eventually, another increase in mass in the final week (p < 0.005), In contrast, ovariectomized animals showed a stable mass gain during the initial five weeks but experienced a decrease in the final week (p < 0.001). It is worth noting that the impact of escitalopram on body mass did not reach statistical significance (p = 0.347) (Fig. 1).Fig. 1Comparison of body mass (g) variation among AP (sham and placebo), AE (sham and escitalopram), OP (ovariectomy and placebo), OE (ovariectomy and escitalopram) Over a 6-week experimental period, the OP and OE groups initiated treatments (placebo or escitalopram) at an approximate mass of 300 g, while the AP and AE groups started with an average of 160 g. Fractures were performed in the fourth week. The intake of escitalopram did not influence the body mass.
Our study did not identify any statistically significant differences in bone mineral density or the biomechanical test results in relation to the fractured bones (Tables 1, 2).Table 1Parameters and statistical results of the analyzes conducted (Normal distribution).AP mean (dp)OP mean (dp)AE mean (dp)OE mean (dp)Surgery (S)Treatment (T)S × TppPTb.Th0.21 (0.04)0.16 (0.04)0.23 (0.05)0.15 (0.02)0.0010.5080.353Tb.NTb.SpP. AngleMasson3.41 (0.81)0.16 (0.03)66.97 (10.96)16.52 (3.26)4.05 (0.80)0.14 (0.01)70.58 (28.70)20.33 (6.70)3.32 (0.46)0.14 (0.01)71.95 (18.89)11.49 (5.17)3.99 (0.25)0.16 (0.01)73.93 (23.87)11.87 (3.76)0.0110.6610.7550.2710.7361.0000.6420.0010.9420.040.9270.364Statistically significant difference. For each analysis, 7 samples were used.dp standard deviation, Tb.Th thickness of the trabeculae, Th.N number of trabeculae, Tb.Sp spacing between trabeculae, P.Angle peak angle, Masson bone trabeculae, AP adult placebo, OP placebo ovariectomized, AE adult escitalopram, OE ovariectomized escitalopram.Table 2Parameters and statistical results of the analyzes conducted (non-normal distribution).AP median (ICM)OP median (ICM)AE median (ICM)OE median (ICM)Surgery (I)Treatment (T)S × TpppBMDBMCP. TorsionBV.BTConnCol ICol III0.162 (0.122;0.185)0.016 (0.012;0.019)0.002 (0.01;0.06)15.19 (10.14;30.73)0.49 (0.28;0.86)1.68 (− 0.38;5.64)16.48 (11.79;27.84)0.181 (0.173;0.204)0.018 (0.017;0.021)0.04 (0.03;0.07)17.70 (9.90;20.69)1.14 (0.82;1.30)1.20 (0.41;1.98)23.46 (20.65;25.61)0.164 (0.142;0,197)0.017 (0.015;0.020)0.04 (0.02;0.06)30.48 (19.51;42.30)0.70 (0.44;0.92)1.89 (0.65;3.47)11.66 (8.50;19.02)0.163 (0.149;0,189)0.016 (0.015;0.020)0.04 (0.02;0.05)20.47 (18.93;22.01)0.94 (0.89;0.99)13.51 (9.03;17.14)9.93 (5.99;15.37)0.0870.0730.4060.0200.0010.0010.9580.8470.8900.8320.0190.8320.0010.0010.0820.1300.1690.4270.2130.0010.153*Statistically significant difference. For each analysis, 7 samples were used.ICM 95% confidence interval for mean, BMD bone mineral density, BMC bone mineral content, P. Torsion peak torsion, BV.TV bone volume over total volume, Conn connection density between multiple trabeculae, Col I collagen type I, col III collagen type III, AP adult placebo, GP placebo growth, OP placebo ovariectomized, AE adult escitalopram, OE ovariectomized escitalopram.
Within the bone callus region (Tables 1, 2), the group treated with escitalopram displayed a significant difference in bone volume (p = 0.002), with the placebo groups showing higher values. Ovariectomy resulted in a decrease in bone thickness (p < 0.001), a reduction in the number of trabeculae (p = 0.011), and a decline in connectivity (p = 0.001) fracture callus.
When examining the interaction between the two variables (surgery vs. medication) (p = 0.004), a more detailed analysis revealed the following When comparing the use of medication within the sham-operated groups, the placebo group showed higher bone thickness values than the escitalopram-receiving group (p = 0.031).Among the ovariectomized groups, it was clear that the placebo group had lower bone volume values compared to those treated with escitalopram (p = 0.031).
However, when evaluating the types of surgeries in the placebo groups, the sham-operated groups demonstrated higher bone volume values than the ovariectomized groups (p = 0.060). Nevertheless, evaluating the types of surgeries among those receiving escitalopram, no statistically significant difference was observed (p = 0.060).
In Fig. 2, it is possible to observe, highlighted in red, the total volume of the bone callus.Fig. 2Computed microtomography images of the experimental groups, in red the total volume of the bone callus is highlighted.
Masson’s trichrome staining results indicated that, regardless of the surgical procedure (sham or ovariectomized) or treatment (placebo or escitalopram), the fracture line remained discernible within the callus. At the fourteen-day mark following the fracture, all groups exhibited undifferentiated tissue, cartilage, and a limited number of trabeculae surrounding the bone fragments (Fig. 3). Notably, it was observed that the OP (ovariectomized and placebo) had a larger bone callus in comparison with the OM (ovariectomized and medication).Fig. 3Masson’s trichrome stained histological sections of the fracture callus in the different groups. AP (sham and placebo), AE (sham and escitalopram), OP (ovariectomy and placebo), OE (ovariectomy and escitalopram). It is evident that regardless of hormonal integrity (adult or ovariectomized) and treatment (placebo or escitalopram), there is a gap between the two main bone fragments of the fracture, and the absence of bone continuity in the callus between them, indicating that the fracture was not completely consolidated. Fourteen days after the fracture, in the groups that ingested escitalopram, a reduction in the volume of the bone callus was observed, characterized by undifferentiated tissue, cartilage, and some trabeculae. Magnification: 5x.
In quantitative terms (Table 1), it was evident that escitalopram was associated with a reduction in trabecular formation (p = 0.001). However, no statistically significant differences were observed when considering sham or ovariectomy (p = 0.271), and the interaction between surgery and medication (p = 0.364).
When examined under polarized light, picrosirius red staining brings out the inherent birefringence of the collagen fibers. This resulted in colors yellow, such as red, and green against a dark background. Type I fibers displayed a denser appearance and were distinguishable in yellow and red hues, whereas type III fibers were noticeable in green. The collagen fibers exhibited diverse orientations and were arranged in layers that intertwined with one another.
Qualitatively, it was observed that the bone callus of ovariectomized animals treated with escitalopram presented a reduction of type III collagen at the fourteen-day post-fracture (Fig. 4).Fig. 4Picrosirus red stained histological images of the fracture callus in all groups under polarized light. AP (sham and placebo), AE (sham and escitalopran), OP (ovariectomy and placebo), OE (ovariectomy and escitalopram). It’s observed that the bone callus of ovariectomized animals treated with escitalopram (OE group), presented a reduction of type III collagen (green and yellow colors), at the fourteen-day post-fracture. Furthermore, it is possible to observe that the ovariectomized groups had a higher number of type I collagen compared to the sham operated groups (red and orange colors). Magnification: 5x.
Quantitatively (Table 2), the analysis of collagen type III in the bone callus revealed that the placebo group exhibited significantly higher values than the escitalopram group (p < 0.001). However, neither the factors of surgery (p = 0.958) nor the interaction between surgery and medication (p = 0.153) resulted in statistically significant differences.
Regarding type I collagen, multiple statistical findings were Medication (p < 0.001); Surgery (p < 0.001); Medication vs. Surgery Interaction (p < 0.001).
When comparing the use of medication within the sham-operated groups no statistically significant difference was detected (p = 1.000). However, among the ovariectomized groups, individuals receiving placebos exhibited lower values compared with those receiving escitalopram (p < 0.001).
When comparing types of surgeries within the placebo groups, no significant difference was observed between sham-operated and ovariectomized groups (p = 1.000). However, among those taking escitalopram, the sham-operated groups displayed lower levels of type I collagen than the ovariectomized groups (p < 0.001).
In summary, escitalopram decreased bone trabeculae and type III collagen. As for type I collagen, statistical differences were observed in the interaction of variables. The groups that received escitalopram showed lower numbers of collagen fibers in the sham-operated groups compared to the ovariectomized groups. When comparing ovariectomized groups, the placebo group exhibited lower levels than those receiving medication.
Epidemiological studies have suggested that elderly patients receiving long-term antidepressant treatments may face a higher risk of experiencing fractures than menopausal women without antidepressant medication. This condition is likely the result of the superposing effect of menopause and the drug. However, to our best understanding, there have been no established links between the use of antidepressants and the process of fracture healing.
In the scope of this study, we conducted a comparative analysis involving healthy adult rats and those that had undergone ovariectomy. This decision was guided by the understanding that menopausal women experience substantial metabolic and hormonal changes, making them susceptible to osteoporosis, and thus categorizing them as individuals at a higher risk for experiencing fractures.
Currently, one of the most frequently prescribed classes of antidepressants belongs to the selective serotonin inhibitors. These drugs not only affect the central nervous system, but also extend their effects to other tissues, including bone.
Serotonin regulates gastrointestinal functions. The main enzyme involved in serotonin synthesis is tyrosine hydroxylase (TPH), which exists in two different TPH1 in the intestine and TPH2 in the brain. Serotonin is considered a monoamine, and much of it is synthesized in the periphery of the intestine, regulating various gut functions. In bone metabolism, it is believed that intestinal and brain serotonin have different actions, acting through different pathways^13^. A comprehensive review carried out by Ducy in 2010, highlighted that the effects of serotonin derived from the intestine have free circulation, which decreases the proliferation of osteoblasts, while serotonin from the brain decreases sympathetic production, favoring bone formation^14^. However, in the literature, the precise impact of serotonin on bone tissue remains inconsistent; a comprehensive understanding of how much SSRIs affect bone repair, especially in the context of bone healing, is a topic that warrants further investigation.
SSRIs negatively affect bone healing repair by inhibiting proliferation, bone differentiation, and mineralization^1^.
In our experimental protocol, the rats entered the experiment with an average body mass of approximately 180 g, with no significant differences among the groups at time zero (p > 0.05). Following ovariectomy, there was a substantial increase in the body mass, as expected^7^. The administration of escitalopram to those hormonally deprived animals did not affect their body mass gain. These findings are in accordance with Lam et al., (2022)^11^, who found that fluoxetine did not affect the body mass gain in rodents.
A review investigating the impact of antidepressants on bone tissue concluded that the use of SSRIs increases the risk of fractures^15^. This danger applies not only to patients with comorbidities, such as the elderly and menopausal women, but also extended to individuals on long-term antidepressant therapy^15^. In line with our results, our study compared rats with osteopenia to those with healthy bone under escitalopram regimen. We found that both groups were equally affected.
Lam and colleagues (2022) investigated the effects of fluoxetine, a SSRI, on healthy adult rodents. They concluded that the group receiving fluoxetine experienced a significant increase in bone mass. They suggested that the detrimental effects on bone were primarily linked to psychological stress. These results contrast with our findings as we observed a negative impact of the drug administration on fracture callus. One possible explanation for the differing results could be the variation in the treatment duration and drug dosage. Those authors administered 10 mg/kg of fluoxetine for 16 weeks, whereas we used a higher dose for a shorter duration of seven weeks.
Conversely, a study involving mice revealed that SSRIs influence bone remodeling through various regulatory mechanisms, with the timing of administration being linked to their specific mode of action. According to the authors, peripherally administered fluoxetine interfered with osteoclast differentiation and functioned through an independent serotonin reuptake mechanism that relies on the Ca2 + levels and the Nfatc1 factor. In the brain, serotonin under sympathetic control, increases bone resorption, potentially harming to bone tissue^12^.
A recent study in mice^16^ revealed impaired craniofacial healing secondary to SSRI treatment. The study concluded that the drug affected bone healing by enhancing collagen formation, which subsequently led to problems in cartilage formation and the functioning of osteoclasts, an essential part of normal bone formation. These findings support our results, as our animals treated with SSRIs displayed a significantly higher amount of collagen type I compared with the placebo groups.
Our findings offer a new perspective on the utilization of some antidepressants. We have identified their detrimental impact on fracture healing in healthy and osteopenic bones. These outcomes should encourage more research, both in laboratory settings and clinically**,** and emphasize the need for caution when using these medications in osteoporotic patients.
The main clinical implication of the use of SSRIs is osteoporosis and a consequent increase in the risk of fractures. A recent longitudinal cohort study investigated the association between the use of SSRIs and bone tissue; the authors concluded that using SSRIs is related to an increased risk of bone loss and highlights the importance of carefully monitoring this population^2^. Another study^17^, aimed to describe the association between exposure to different antidepressant drugs and hip fractures in an elderly population, observed a substantial increase in the risk of hip fractures associated with chronic use of SRIs.
The limitations of the present study, as with all experiments involving animal models, pertain to the translation of findings to humans. Additionally, we did not investigate the long-term progression of the fracture, leaving questions about the duration required for full consolidation unanswered.
In summary, we found that escitalopram negatively affected fracture healing by reducing trabeculae, increasing trabecular spacing, and elevating type I collagen deposition. These effects were consistent in both healthy and osteopenic conditions. This highlights the importance of considering these potential consequences for regular users of antidepressants. Healthcare professionals treating osteoporotic fractures should also be mindful of these findings when providing treatment.
The experimental protocol received approval from the local Ethics Committee for Animal Experimentation (FMRP Ethics Committee on the Use of Animals (CEUA-FMRP), # 001/2019) and adhered ARRIVE guidelines for the use of experimental animals.
A total of 56 female SPF Wistar Hannover rats (7–8 week, when the animal reaches sexual maturity)^18^, were included in this study. Twenty-eight rats were adult healthy animals and twenty-eight were previously ovariectomized, resulting osteoporotic conditions (confirmed by the increase in weight of the animals)^19,20^. All rats received either oral escitalopram or placebo and underwent a closed mid-diaphyseal femoral fracture. Therefore, four experimental groups (n = 14) were 1—AE (healthy adults + escitalopram), 2—AP (healthy adults + placebo), 3—OE (osteopenia adults + escitalopram), and 4—OP (osteopenia adults + placebo).
The healthy adult groups underwent sham surgery, while in the osteopenic groups, the rats underwent ovariectomy 90 days prior to their inclusion in the experimental groups. During the ovariectomy, the ovaries were surgically removed under anesthesia, and the muscular wall was reflected to access the abdominal cavity, location and removal of the ovary. Then, the muscular and cutaneous planes were sutured, and the same procedure was performed on the other side. In the sham groups, the ovaries were only exposed and repositioned into the abdominal cavity^19^.
Escitalopram was administered daily via gavage at a dose of 20 mg/kg, once in the morning. The dose was adjusted weekly in response to the animal’s weight gain^10^, and its administration for 35 days qualifies as a chronic treatment^21^. Sham animals received an equivalent volume of distilled water via gavage.
The femoral fracture was produced on the 21st day the animals entered the experiment, under isoflurane inhalation anesthesia administered to the rats. First, the right thigh of each rat was shaved and positioned on two metallic supports. Subsequently, a lever-controlled blunt blade was directed at the mid-thigh area. Pressure was exerted until a sudden reduction in resistance indicated bone fracture. Following this, the pelvic limb was cleaned with a 0.5% alcohol-based chlorhexidine solution, and a 1.0 cm long straight incision was made on the lateral side of the mid-thigh, over the fracture site. The incision extended along the intermuscular septum until the fractured bone was reached, allowing for minimal exposure and examination of the bone ends. Rats with fractures not situated in the mid-diaphysis or with more than three fragments were excluded from the study. Next, a 1.0 mm thick orthopedic Kirschner wire (K-wire) was inserted into the medullary canal of the proximal fragment. With direct visualization, the fracture was aligned and secured by advancing the K-wire into the medullary canal of the distal fragment until it reached the femoral condyles. The excess length of the K-wire protruding from the trochanteric region was trimmed at the tip of the greater trochanter. It was then embedded within the muscles. The wound was closed in layers using absorbable sutures. Post-surgery radiographs confirmed whether the fragments were well-aligned and securely fixed (Fig. 5). For postoperative pain management, subcutaneous tramadol was administered^22^.Fig. 5Radiographic image of the animals to confirm suitability and fixation.
At the conclusion of the experiment (fourteen days post-fracture) and following 35 days of escitalopram administration, the animals were euthanized using an excessive intraperitoneal dose of Thiopental®. The right (fractured) femur of each animal was removed, dissected, and stored in accordance with specific protocols for subsequent analysis.
The experimental design is represented in Fig. 6.Fig. 6Experimental design of groups.
Bone densitometry was conducted with the femurs submerged in saline to a depth of 2.0 cm, using a Lunar® DPX-IQ model dual-energy X-ray absorptiometer (DXA), equipped with a software designed for small samples. To encompass the entire bone callus area, a range of geometric shapes and sizes were used as regions of interest (ROIs).
The samples were scanned using the SkyScan system, model 1273 (Kontich, AN, Belgium). The parameters were set to 60 kV and 250 μA, pixels of 20.50 μm, and a rotation step of 0.5°. The entire bone callus was chosen to determine bone volume fraction by total volume (BV/TV—%), trabecular thickness (Tb. Th), number of trabecular (Tb. N–1/mm), trabecular separation (Tb. Sp—mm), and connection density between multiple trabecular (Conn.D–1/mm3) (American Society of Bone and Mineral Research)^23^.
After the non-destructive DXA (Dual-energy X-ray absorptiometry) and μCT assessments, the entire femurs were positioned in a torsion testing machine fitted with a 2.0 Nm load cell (Instron 55MT, Norwood, USA), and subjected to counterclockwise rotation at a rate of 0.5°/s until failure occurred. With the help of a specific software script, the peak torsion and peak angle were obtained.
The femurs were fixed in 10% formaldehyde for 24 h, then decalcified in 10% EDTA solution, dehydrated in an ascending series of alcohols, cleared using xylene, and embedded in paraffin. From the paraffin blocks, 5 µm thick longitudinal sections were prepared and examined under the AxioImager® Z2 microscope (ZeissTM, Oberkochen, Germany) using the Axiovision software (ZeissTM) for data quantification. The software identified trabecular bone, type I and type III collagens, based on color and quantified the volume in bone callus following standard procedures^24^.
Statistical analysis was carried out using IBM SPSS Statistics 20 (IBM Corporation, Armonk, New York, USA) with a significance level set at 5%. Data distribution was assessed using the Shapiro–Wilk test. For datasets exhibiting a normal distribution, a two-way analysis of variance (ANOVA) was performed, followed by Tukey’s post-test with Bonferroni adjustment (p ≤ 0.05). The Wald Test was applied within a generalized linear model, to analyze data with a non-normal distribution, incorporating multiple comparisons with Bonferroni adjustment. Additionally, factorial ANOVA (age/treatment) and repeated measures (weeks) were performed to evaluate changes in body mass.