Use of Psidium guajava fruit suspension may prevent treatment relapse following orthodontic tooth movement in Wistar rat model Authors: Arwa Aldeeri, Mohammed Alahmed, Yasser Alshawakir, Rita Khounganian, Abdullah Aldrees Published: 2026 ABSTRACT AbstractTo investigate the effect of Psidium guajava fruit suspension in preventing orthodontic treatment relapse in a Wistar rat model. Twenty-four male Wistar rats (aged 8–12 weeks) underwent orthodontic treatment on the left maxillary side for 48 days. After space creation, orthodontic appliances were removed, and a 35-day retention phase was initiated. Meanwhile, the rats were randomly allocated to two groups. The control group received normal saline, and the intervention group received a suspension of 250 mg/kg P. guajava via oral gavage. After retention, the resin retainers were removed to allow for a 7-day relapse phase. Polyvinyl siloxane impressions were obtained at the start (T0) and end (T1) of the relapse phase and scanned for analysis using OrthoCad software. Of the 24 rats initially recruited, 20 rats successfully completed the experiment and were included in the statistical analysis. The mean orthodontic tooth movement (T0) and relapse measurements (T1) in the intervention group were 0.570 ± 0.125 mm and 0.515 ± 0.094 mm, respectively. In the control group, the corresponding mean values were 0.630 ± 0.194 mm and 0.410 ± 0.204 mm, respectively. The intervention group demonstrated significantly lower mean orthodontic relapse (T1-T0) than that of the control group (-0.055 ± 0.089 mm and -0.220 ± 0.058 mm, respectively, P < 0.001). Orally administered P. guajava fruit suspension may reduce orthodontic relapse and improve treatment outcomes. Further studies are required to validate these findings and elucidate the underlying mechanisms before considering human trials.Keywords: Orthodontic retention, Orthodontic tooth movement, Psidium guajava, Relapse, Wistar rats Introduction The long-term stability of orthodontic treatment outcomes remains difficult to attain (Littlewood et al. 2017). Various strategies have been introduced to prevent relapse, including bonded lingual wires, removable retainers, or a combination of both. However, relapse is often inevitable, even with long-term retainer use and patient compliance, and the need for meticulous maintenance remains a key limitation (Bucur et al. 2022). Advances in understanding the biological mechanisms of orthodontic tooth movement have provided insights into post-treatment stability and relapse. Consequently, interest has grown in biological approaches that use pharmacological agents to modulate the chemical signals and biological pathways involved in this movement (Franzen et al. 2013, 2014; King et al. 1997). However, the therapeutic nature of these drugs and the need for their prolonged consumption have raised significant unaddressed health concerns. The Psidium guajava fruit is widely acknowledged as a source of β-cryptoxanthin, a carotenoid with demonstrated osteoanabolic activity that increases calcium content and alkaline phosphatase levels (Burri et al. 2016; Uchiyama et al. 2004; Yamaguchi 2008; Yamaguchi and Uchiyama 2003). Besides β-cryptoxanthin, P. guajava contains other osteoanabolic phytochemicals, including flavonoids, polyphenols, and vitamin C (Diyanat et al. 2018). In growing Wistar rats, oral administration of P. guajava fruit extract has been reported to promote bone regeneration at injury sites and improve bone mass parameters, such as bone mineral density (Porwal et al. 2017). In addition, ovariectomized Sprague–Dawley rats that ingested P. guajava fruit hydroalcoholic extract showed increased osteocyte and osteoblast numbers, as well as greater femoral bone weight, volume, and ash density (Diyanat et al. 2018). Collectively, these findings demonstrate the osteoanabolic properties of P. guajava. In addition, a recent systematic review of in vivo studies reported that P. guajava fruit exhibits minimal potential for drug interactions and shows no evidence of toxicity or mutagenicity (Takeda et al. 2023). To date, no studies have investigated the effects of P. guajava fruit or its derivatives on orthodontic tooth movement or post-treatment relapse. Therefore, we investigated the efficacy of P. guajava fruit suspension in preventing orthodontic relapse in a Wistar rat model via oral gavage. Materials and methods This study was approved by the Institutional Review Board (KSU-SE-23-55) and registered at King Saud University’s College of Dentistry Research Center (CDRC) (PR 0162). Ethical approval was granted by the Research Ethics Committee (REC) Office at King Saud University (KSU-SE-23-55). This study represents the third part of a series investigating the anti-relapse properties of P. guajava fruit suspensions. The molecular and histological findings from the same experimental model have been previously published (Aldeeri et al. 2025, 2026). Following the 3R principle of animal research (replacement, reduction, and refinement) and aiming to use the minimum number of animals needed to deliver meaningful biological observations, a pilot group of 24 Wistar rats (8–12-week-old; weighing an average of 248 g) was enrolled. This sample size is consistent with that of published studies investigating orthodontic tooth movement and relapse in rodents. A veterinarian examined each rat to confirm complete dental development and exclude systemic or dental diseases. The rats were maintained under standardized laboratory conditions, including a 12-hour light/dark cycle, a controlled temperature of 21 ± 2 °C, and relative humidity of 50 ± 15%. An acclimatization period of 1 week preceded the experiment to allow adaptation to the environment. Moist food to avoid appliance breakage and water were provided ad libitum, and the animals’ teeth were cleaned daily. Appliance installation began after anesthetizing the rats via intraperitoneal injection of a mixture of ketamine and xylazine at doses of 80–120 mg/kg and 5–10 mg/kg, respectively, using a 22–27-gauge needle (Sigma Aldrich, Germany). Undercuts were prepared in the gingival third of the maxillary incisors and left first molar using a rotatory handpiece with a small round bur, followed by bonding. According to the manufacturer’s instructions, a Transbond self-etching primer (3 M Unitek Corp., St. Paul, MN, USA) was applied, rubbed onto the tooth surface for 5 s, and gently air-dried for 1–2 s using an air syringe. A 0.2 mm stainless-steel ligature wire (Dentaurum, Pforzheim, Germany) secured a stretched nitinol closed-coil spring (Ortho Technology, Clearwater, FL, USA) around the prepared notches on the incisor and first molar, delivering a continuous 100 g force (Fig. 1). Schematic design of the orthodontic appliance mounted between the maxillary incisors and left first molar in Wistar rats. (a) Occlusal view, (b) Sagittal view The ligature wire was then twisted, securely fitted, and firmly anchored to the tooth. A flowable resin composite (Transbond Supreme LV; 3 M Unitek Corp.) was applied over the ligature wire and corresponding undercuts and then light-cured. The mandibular incisors were trimmed twice weekly to accommodate continuous maxillary incisor eruptions and to prevent appliance breakage or dislodgment. The rats were inspected daily for appliance dislodgment, and cementation was performed immediately when necessary. This phase lasted 48 days. Intraperitoneal anesthesia was administered at the end of the active tooth movement phase to remove the appliance, initiating a 35-day retention phase. The proximal surfaces of the molars were prepared using Transbond self-etching primer (3 M Unitek Corp.). A light-cured resin (Filtek One Bulk Fill; 3 M Unitek Corp.) was used to fill these gaps (Fig. 2). The composite resin retainer bonded between the maxillary left first and second molars in the Wistar rats (blue arrow) The rats were divided into two groups using a random number generator: a negative control group (CG) receiving normal saline (0.9% NaCl; Pharmaceutical Solution Industry Ltd., Jeddah, Saudi Arabia) by which no effect of the supplement is expected, and an intervention group (IG) receiving P. guajava fruit suspension at 250 mg/kg body weight (Terravita, Mumbai, India). The preparation of the P. guajava suspension has been previously described (Aldeeri et al. 2026). Normal saline and P. guajava suspensions were administered once daily for 35 consecutive days via oral gavage. This route of administration was selected based on established evidence from studies investigating the effects of systemically administered agents on orthodontic tooth movement and relapse (Kebsch et al. 2007; Crowther et al. 2017; Huang et al. 2025). Intraperitoneal anesthesia was administered at the end of the retention phase, and the light-cured resin stop (orthodontic retainer) was removed using a low-speed handpiece fitted with a non-cutting carbide bur. Dental impressions were obtained using a custom impression tray and polyvinyl siloxane impression material (3 M Unitek Corp.) to record the space created between the molars at the end of the retention phase (T0). Impressions were scanned with an iTero intraoral scanning system (Align Technology, San Jose, CA, USA) before commencing the 7-day relapse phase. Subsequently, rats were euthanized using a triple overdose of ketamine. Vinyl polysiloxane impressions were repeated and scanned to measure the post-relapse distance (T1). Distances were measured using the OrthoCad software (Align Technology, San Jose, CA, USA). The primary outcome, the anti-relapse effect of P. guajava suspension, was quantified by measuring the separation between the mesial and distal marginal ridges of the second and first molars. The T0 value was subtracted from the T1 value, and a negative result indicated the relapse magnitude (Fig. 3). A summary of the experimental design and outcomes is shown in Fig. 4. Photographs showing the relapse assessment protocol as described by Qi et al. (2019). Left: baseline measurement (T0). Right: measurement after relapse (T1) Summary of the experimental design and the main outcome All measurements were performed twice by one examiner at a 1-week interval to determine intra-examiner consistency; a score of Cohen’s Kappa ≥ 0.80 indicated acceptable agreement. Data was recorded, and the mean distances in millimeters were analyzed using GraphPad Prism version 8. Statistical analysis Descriptive statistical analyses, including means and standard deviations, were performed to summarize the findings of the study groups. Box-and-whisker plots were used to visually depict the distribution of numerical data and identify potential outliers. The normal distribution of continuous variables was assessed using the Shapiro–Wilk test. Subsequently, within-group differences in mean measurements at T0, T1, and T1 − T0 were assessed via one-sample t-test, whereas between-group comparisons of the corresponding variables were performed using unpaired t-tests. Statistical significance was set at P < 0.05. Results Twenty Wistar rats aged 8–12 weeks with a mean weight of 248 g completed the experiment. Normality was confirmed for all continuous variables (P > 0.05). The summary statistics of the enrolled rats are presented in Table 1. Table 1Descriptive statistics of relapse data of the two study groupsVariableGroupNumberMean (mm)SDMaximum (mm)Minimum (mm)Distance T0 Control 100.6300.1941.0000.500 Intervention 100.5700.1250.7000.400Distance T1 Control 100.4100.2040.7500.200 Intervention 100.5150.0940.7000.400Distance T1-T0 Control 10−0.2200.058−0.300−0.100 Intervention 10−0.0550.089−0.2000.050 Descriptive statistics of relapse data of the two study groups Relapse values (T1-T0) ranged from − 0.200 mm to 0.050 mm in the intervention group, and from − 0.300 mm to − 0.100 mm in the control group. Median relapse values in the intervention and control groups were 0.000 mm and − 0.200 mm, respectively (Fig. 5). An unpaired t-test comparing relapse (T1–T0) between the groups revealed a significant difference (P < 0.001). The observed effect size corresponded to a large magnitude (Cohen’s d = 2.197; 95% confidence interval [CI] [1.09, 3.31]) (Table 2). Box-and-Whisker plot of the relapse extent (T0 − T1) in the two groups Table 2Unpaired t-test results of the two study groupsGroupDistance T0Distance T1Distance T1-T0Cohen’s dMean (mm)SDP-valueMean (mm)SDP-valueMean (mm)SDP-valueControl0.6300.1940.4230.4100.2040.156−0.2200.058< 0.001*2.197Intervention0.5700.1250.5150.094−0.0550.089* P < 0.001 Unpaired t-test results of the two study groups * P < 0.001 Discussion The entire experiment lasted 90 days and proceeded uneventfully, except that two rats developed respiratory symptoms after appliance installation, and two were prematurely sacrificed. The lost rats were evenly distributed among the study groups, resulting in 20 rats completing the study. Working with Wistar rats, although feasible, proved challenging due to their fragility, susceptibility to respiratory symptoms following oral gavage, limited visibility, and tendency to achieve tipping rather than bodily tooth movement. Nonetheless, they represent a well-established model for studies involving orthodontically induced tooth movements (Ren et al. 2004; Qi et al. 2019). In recent decades, multiple studies have explored the potential of both pharmaceutical and natural agents to improve orthodontic retention or reduce reliance on retention appliances (Arias and Marquez-Orozco 2006; Al-Hamdany et al. 2017; Asefi et al. 2018; Tokhtah and Alhadlaq 2022). Arias and Marquez-Orozco (2006) compared the effects of aspirin, acetaminophen, and ibuprofen on orthodontic tooth movement in Wistar rats. Rats treated with aspirin and ibuprofen had significantly reduced tooth movement than control rats. In a study conducted by Al-Hamdany et al. (2017), the effect of olive oil supplementation at doses of 15.4 mL/kg on orthodontic retention in the albino rabbit model significantly reduced treatment relapse in the intervention group. This finding was accompanied by elevated counts of osteoblasts and osteocytes and increased bone remineralization compared with the control. Another study investigated the effect of curcumin oil, the most active compound in turmeric, on orthodontic tooth movement in Wistar rats (Asefi et al. 2018). While the intervention did not inhibit orthodontic tooth movement, it significantly reduced root resorption, probably by suppressing the NF‑κB pathway and downregulating osteoclastogenesis, as reported by Bharti et al. (2004). Nonetheless, the body of evidence regarding these agents is growing, and definitive conclusions regarding their use in orthodontics need to be established through future investigations. Bone disease research has shown that P. guajava exerts osteoanabolic effects. A randomized controlled trial by Kakuo et al. (2018) reported that a 12-week regimen of daily intake of P. guajava leaf extract significantly reduced knee pain and stiffness in patients with symptomatic osteoarthritis compared to a placebo. Porwal et al. (2017) reported that a 12-day regimen of 250 mg/kg P. guajava extract promoted fracture healing in adult female Wistar rats. The extract stimulates osteoblastic differentiation via the Wnt/β-catenin pathway (Porwal et al. 2017). Although precise phytochemical characterization of the active constituents is lacking, flavonoids present in P. guajava, particularly quercetin and kaempferol, have been reported to increase osteoprotegerin (OPG) expression and decrease matrix metalloproteinase MMP-2, MMP-9, and receptor activator of nuclear kappa B ligand (RANKL) expression in bone-resorbing cells (Gürsoy et al. 2024; Yamaguchi 2008). This alteration in key regulators of bone remodeling is thought to preserve alveolar bone integrity. Additionally, ellagic acid, another polyphenol found in P. guajava fruit, has been reported to facilitate gingival wound healing by promoting collagen fiber stabilization and collagenase inhibition (Gutiérrez et al. 2008). In this study, orthodontic tooth movement of the maxillary left first molar yielded mean distances (T0) of 0.630 ± 0.194 mm in the control group and 0.570 ± 0.125 mm in the intervention groups (P = 0.423). This level of tooth movement is within the range reported in other studies using the same animal model (MirHashemi et al. 2013; Khalaf and Almudhi 2022). Following the relapse phase (T1), the mean distance between the first and second molars was 0.410 ± 0.204 mm in the control group, compared with 0.515 ± 0.094 mm in the intervention group. Intergroup differences were minimal and did not reach statistical significance (P = 0.156). Notably, the mean magnitude of treatment relapse, represented by the subtraction of the values of T1 from T0, was − 0.055 ± 0.089 mm in the intervention group, compared to − 0.220 ± 0.058 mm in the control group. An unpaired t-test revealed that the intervention group exhibited significantly fewer relapses than the control group (P < 0.001). Although an effect size of 0.35 was assumed, the observed effect size was six-fold larger (Cohen’s d = 2.197, 95% CI [1.09, 3.31]). This large effect size indicated the potential of the intervention, which may have been more potent than initially anticipated. Although such a large-scale intervention is not unprecedented in orthodontic research using animal models, it offers a valuable opportunity for future studies to make meaningful contributions. For instance, it highlights the importance of reporting multiple intervention doses to determine whether the observed effects scale proportionally with higher doses or plateau. In addition, extended follow-up is warranted to confirm the persistence of the observed effects. Meanwhile, the observed large effect size in the current investigation could be attributed to the presence of β-cryptoxanthin, as investigated by (Yamaguchi and Uchiyama 2003; Torre 2017; Martiniakova et al. 2022), or flavonoids, polyphenols, and vitamin C as reported by others (Yamaguchi et al. 2006; Ramesh et al. 2021). Considering these potential factors, the synergistic role of controlled experimental conditions in amplifying biological responses to interventions should not be overlooked. The exact roles and magnitudes of individual factors in preventing relapse remain to be elucidated. In a previous experiment using the same animal model, molecular assessment using an enzyme-linked immunosorbent assay measured alkaline phosphatase (ALP) and osteocalcin (OC) levels (Aldeeri et al. 2025). In the intervention group, the mean ALP and OC concentrations were 0.075 ± 0.005 pg/mL and 0.692 ± 0.011 pg/mL, respectively. However, in the control group, the mean ALP and OC concentrations were 0.074 ± 0.012 pg/mL and 0.696 ± 0.003 pg/mL, respectively. Histological assessment, on the other hand, reported more pronounced tissue changes (Aldeeri et al. 2026). While 70% of the specimens from the pressure and tension sides in the intervention group showed uniform PDL fibers, 90% of the control specimens exhibited significant irregularities with excessively wide and narrow PDL fibers (P < 0.001). Furthermore, 70% of the pressure-side specimens in both groups showed a moderate number of osteoclasts. The remaining 30% either lacked osteoclasts (intervention group) or had few osteoclasts (control group). The groups differed significantly from one another (P < 0.05). The collective evidence from the molecular, histological, and mechanical assessments indicates that, while the molecular assessment exhibited partial divergence, the observed effect of the intervention, represented by a significant reduction in treatment relapse metrics, is concordant with, and likely reflects, histologic changes. A plausible explanation for the divergence between the molecular findings and both the histological and mechanical outcomes may be the differential attrition rate between the groups comprising ten and four rats, respectively. Furthermore, ALP and OC showed transient and time-sensitive expression, indicating that their peak levels were not coincident with the sampling time point. Lastly, while treatment relapse is multifactorial in nature, it may be suppressed by upregulating osteoanabolic activity or downregulating resorptive activity. In that sense, it is plausible that the intervention mechanism of action was the downregulation of resorptive activity rather than upregulation of osteoanabolic activity, a process better captured by biomarkers, such as tartrate-resistant acid phosphatase (TRAP) and the RANKL/OPG ratio. However, caution should be exercised when interpreting the results of this study to achieve a balanced understanding. The strengths of this study include the implementation of a controlled and reproducible experimental model with consistent force application, utilization of a bioavailable and nutritional supplement, standardized experimental conditions throughout the study period, as indicated by the lack of significant differences between groups at T0 (P = 0.423, Table 1), predefined outcome variables, non-invasive route of supplement administration, and a multi-dimensional approach of assessment incorporating molecular, histological, and mechanistic analyses. Although the preliminary findings from this hypothesis-generating study show potential, key limitations should be acknowledged. First, the current sample size, while comparable to that of similar orthodontic relapse studies and justified by the exploratory nature of the experiment, undermines the generalizability of the findings. Second, the lack of a dose-dependent study design, although methodologically sound in the presence of a negative control group, limits the narrative of the intervention’s potential. Third, implementing a split-mouth design, in which the contralateral side serves as an internal control, would provide additional within-animal comparisons. Fourth, this series of experiments reported the molecular (Aldeeri et al. 2025), histological (Aldeeri et al. 2026), and mechanistic aspects. However, phytochemical characterization using liquid chromatography-mass spectrometry can provide a deeper understanding of the specific bioactive compounds responsible for these observed effects. Fifth, although this study investigated the effect of gastrointestinal administration of P. guajava fruit suspension, the potential role of local delivery systems warrants further consideration. Therefore, future studies should conduct priori sample size calculations, distinguish between different types of tooth movement, report dose-dependent responses across multiple intervention groups, explore alternative drug-delivery systems, and implement a multi-dimensional analytical approach. Identifying a bioavailable, cost-effective supplement that reduces the reliance on conventional orthodontic retainers while promoting systemic health would mark a major advancement in orthodontic research and clinical practice. Conclusion In this preclinical study, the intervention group experienced a significant reduction in treatment relapse compared to the negative control group, with an observed effect size exceeding the original assumption, suggesting a potent intervention effect. Acknowledgements The authors extend their appreciation to the College of Dentistry Research Centre (CDRC) at King Saud University (PR 0162) and the Saudi National Institute of Health [SNIH-RO-HRT01-2302-KSU- 3630310] for their support in conducting this study. Author contributions Conceptualization, Arwa Aldeeri; Funding acquisition, Arwa Aldeeri; Investigation, Arwa Aldeeri, Mohammed Alahmed, and Yasser Alshawakir; Methodology, Arwa Aldeeri, Mohammed Alahmed, Yasser Alshawakir, Rita Khounganian, and Abdullah Aldrees; Project administration, Abdullah Aldrees; Supervision, Rita Khounganian, and Abdullah Aldrees; Visualization, Arwa Aldeeri, Mohammed Alahmed, Yasser Alshawakir, Rita Khounganian, and Abdullah Aldrees; Writing – original draft, Arwa Aldeeri; Writing – review & editing, Mohammed Alahmed, Yasser Alshawakir, Rita Khounganian, and Abdullah Aldrees. Funding This project was partially funded by the Saudi National Institute of Health [SNIH-RO-HRT01-2302-KSU- 3630310]. Data availability The datasets used and/or analyzed in the current study are available from the corresponding author on reasonable request. Declarations Ethical approval All procedures were approved by the Institutional Review Board (KSU-SE-23-55) and registered at College of Dentistry Research Center (CDRC) (PR 0162) at King Saud University. Ethical approval was obtained from the Research Ethics Committee (REC) Office of King Saud University (KSU-SE-23-55). Competing interests The authors declare no competing interests. Footnotes Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. References Associated Data Data Availability Statement The datasets used and/or analyzed in the current study are available from the corresponding author on reasonable request.