Patterns and influencing factors of dental implant anxiety in China Authors: An Lou, Yaping Zhou, Hongtai Pei, Zhongmin Wang, Ming Li Published: 2026 ABSTRACT To investigate the anxiety pattern and its association with perceived pain at different perioperative stages of dental implant surgeries in China, and identify key influencing factors of anxiety. A prospective observational perioperative study was conducted on 506 implant patients in Changsha Stomatological Hospital from January 2025 to June 2025. Patients assessed their current level of dental anxiety at each of five time points: pre-surgery night (T1), hospital waiting period (T2), local anesthesia administration (T3), implant placement procedure (T4), and post-surgery (T5). Pain perception was evaluated by patients using the Numerical Rating Scale (NRS) on the night before surgery, during surgery, and after surgery. Repeated-measures ANOVA, independent sample t-tests, binary logistic regression, and Kruskal–Wallis test were used for analysis. Anxiety levels peaked at T3 (2.61 ± 0.98) and T4 (2.51 ± 1.05), then decreased significantly at T1 (2.38 ± 1.06), T2 (2.21 ± 0.97), and T5 (1.43 ± 0.72). Predicted pain, intraoperative pain, and postoperative pain were significantly correlated with anxiety. Preoperative pain was not significantly associated with anxiety. Female (OR = 2.60), elderly patients (36–50 years old, OR = 2.40 51–72 years old, OR = 2.19), mandibular implant (OR = 1.81), and bimaxillary implant (OR = 9.63) are factors associated with increased odds of anxiety. The model showed acceptable discrimination (AUC = 0.704, 95% CI 0.658–0.750) and calibration (Hosmer–Lemeshow p = 0.42). Dental anxiety peaks during invasive procedural phases. Gender, age, implant location, and perceived pain degree are key influencing factors. Targeted interventions for high-risk groups and effective pain management can help mitigate anxiety and improve perioperative care quality. Introduction Dental implantation has become the preferred treatment for tooth loss due to its advantages of good masticatory function restoration and natural aesthetic effect1,2. For the past few years, dental implant surgeries in China have become increasingly popular. However, as a complex and invasive oral surgical procedure, dental implantation causes tissue trauma, lasts a long time, and can lead to potential risks of complications like implant loosening and nerve injury. These risks often induce anxiety in patients and affect patient satisfaction3–10. Dental anxiety not only affects the patient’s treatment experience but also may lead to physiological reactions such as increased blood pressure and heart rate. It may prolong the operation time, and even increase the risk of postoperative complications such as swelling and infection11,12. In recent years, domestic and foreign studies have shown that the incidence of dental anxiety ranges from 20 to 50%. Factors such as age, gender, surgical complexity, education level, and previous dental treatment experience are closely related to the occurrence of anxiety13–15. Existing studies on dental anxiety have primarily focused on common procedures like tooth extraction, where pain and bleeding are identified as primary anxiety triggers16,17. However, studies on anxiety regarding dental implant surgery, a more complex and higher-investment treatment, are still limited among the Chinese population. Most existing studies are conducted in the Western medical system or high-income countries, and their conclusions may not be directly applicable to medical scenarios and patient psychology in China18–20. This study aims to fill this gap by conducting a survey in a representative dental hospital in China, depicting the unique pattern of perioperative anxiety among Chinese dental implant patients, and deeply analyzing its influencing factors. This study aimed to systematically assess anxiety levels at five key perioperative time points in dental implant patients, analyze the correlation between anxiety and pain, and determine the independent influencing factors of anxiety. We supposed that dental implant anxiety in Chinese patients would be significantly influenced by pain perception, specific patient-level and procedure-level factors. The study results will enable clinicians to identify high-risk patients and implement early intervention. More broadly, these findings are expected to provide local evidence and a scientific basis for targeted anxiety management strategies, improving the overall quality of dental implant care in China, and ultimately enhancing patient satisfaction, treatment compliance and long-term treatment outcomes. This is important for the standardized and humanized development of oral implantology in China. Materials and methods Study design and participants Changsha Stomatological Hospital is a specialized dental institution with over 10 years of experience in dental implants. It performs about 1,500 implant procedures annually. This prospective observational perioperative study was conducted in Changsha Stomatological Hospital from January 2025 to June 2025. All patients participated in the study underwent immediate implant placement without any bone augmentation surgery, meaning implants were inserted immediately after tooth extraction. All procedures were conducted in accordance with the Declaration of Helsinki. All participants signed informed consent forms. Participants were explicitly informed that their participation was voluntary. They could withdraw at any time without affecting their dental care, and all data would be anonymized for analysis to ensure confidentiality and privacy. Exclusion criteria were listed as follows: 1. age ≤ 18 years old or ≥ 73 years old; 2. cognitive impairment, language barriers, or inability to complete assessments even with assistance; 3. medical history of mental illness or severe systemic diseases, such as severe heart disease, or severe hypertension; 4. used sedative drugs within 30 days before enrollment; 5. pregnancy or lactation. In order to achieve 90% power with an alpha level of 0.05, we recruited patients visiting our hospital from January 2025 to June 2025, ensuring sufficient statistical power for the analysis. Sample size was calculated using R 4.4.0 for logistic regression with an anticipated OR of 2.0, baseline anxiety proportion of 50%, α = 0.05, and power = 0.90, yielding a required sample of 372. A total of 550 patients were initially screened. We excluded 44 patients who met the exclusion criteria or withdrew from the study. Finally, 506 patients finished the study, including 218 males and 288 females, aged 19–72 years, with an average age of 56.42 ± 15.23 years. Research tools Anxiety assessment In recent years, the use of MDAS to measure dental anxiety has been applied among the Chinese population21–24. We referred to the concept of the Modified Dental Anxiety Scale (MDAS) to design our anxiety assessment tool. We asked patients about their current level of dental anxiety at each of the five time points:T1: Pre-surgery night (assessed via online questionnaire sent to patients);T2: Hospital waiting period (assessed in the waiting room 30 min before surgeries);T3: Local anesthesia administration (assessed immediately after anesthesia injection);T4: Implant placement procedure (assessed midway during implant placement);T5: Post-surgery (assessed in the recovery room 30 min after surgeries). T1: Pre-surgery night (assessed via online questionnaire sent to patients); T1: Pre-surgery night (assessed via online questionnaire sent to patients); T2: Hospital waiting period (assessed in the waiting room 30 min before surgeries); T2: Hospital waiting period (assessed in the waiting room 30 min before surgeries); T3: Local anesthesia administration (assessed immediately after anesthesia injection); T3: Local anesthesia administration (assessed immediately after anesthesia injection); T4: Implant placement procedure (assessed midway during implant placement); T4: Implant placement procedure (assessed midway during implant placement); T5: Post-surgery (assessed in the recovery room 30 min after surgeries). T5: Post-surgery (assessed in the recovery room 30 min after surgeries). The anxiety scores range from 1 (no anxiety) to 5 (extreme anxiety), indicating that the level of anxiety gradually increases, with a total score ranging from 5 to 25 points. Cronbach’s α in this sample was 0.76 (95% CI 0.73–0.79), indicating acceptable internal consistency. Pain assessment The Numerical Rating Scale (NRS) is a widely adopted and linguistically simple tool. NRS was used to record pain perception at three time points in the study. At the night before surgeries, patients evaluated their preoperative pain and predicted intraoperative pain. During dental implant surgeries, patients evaluated their intraoperative pain. After completing dental implant surgeries, patients evaluated their postoperative pain. The NRS ranges from 0 (painless) to 10 (severe pain), with scores categorized as follows: 0 = painless, 1–3 = mild pain, 4–6 = moderate pain, and 7–10 = severe pain. Data collection and data analysis Anxiety and pain assessments were conducted by two trained researchers who were not involved in the clinical treatment of the patients to ensure the objectivity of the data. The anxiety assessment surveys were completed by patients at the night before surgeries (T1), 30 min before surgeries (T2), during dental implant surgeries (T3 and T4), and 30 min after surgeries (T5). The NRS surveys were completed by patients at the night before surgeries (preoperative pain and predicted intraoperative pain), during dental implant surgeries (intraoperative pain), and 30 min after surgeries (postoperative pain). During implant surgeries (time points T3 and T4), a trained research assistant not involved in the surgery asked patients aloud about their current anxiety levels and pain levels during surgical pauses. Each assessment took 5–10 s and did not interrupt the procedure. In order to elucidate the demographic features and clinical backgrounds of patients undergoing dental implant surgeries, we encouraged patients to complete a survey questionnaire at their first visit, and collected a series of variables through the completed questionnaires. Meanwhile, clinical data related to dental implantation were extracted from electronic medical records, specifically covering the anatomical positions of implants, and the cumulative number of teeth that underwent implant procedures. All data were entered into Excel for sorting, and statistical analysis was performed using R 4.4.0 software. Descriptive statistics were used to summarize participants’ demographic and clinical characteristics, with continuous variables reported as mean ± standard deviation (SD) and categorical variables as frequencies and percentages. Comparison between groups in anxiety and pain scores was assessed using independent sample t-tests or the Kruskal–Wallis test. Repeated-measures ANOVA was employed to examine changes in anxiety levels across the five perioperative time points. Mauchly’s test of sphericity was performed, and the Greenhouse–Geisser correction was applied when the assumption was violated. Post-hoc pairwise comparisons used Bonferroni correction. Univariate and multivariable binary logistic regression analyses were performed to identify factors associated with dental anxiety, with results reported as p values and odds ratios (ORs) with 95% confidence intervals (CIs) separately. The Kruskal–Wallis test was used to compare anxiety scores across different pain severity groups at different time points, with post‑hoc pairwise comparisons performed using Dunn’s test with Bonferroni correction. P < 0.05 or OR with 95% CIs excluding one was considered statistically significant. We have complied with the STROBE protocol. Results Baseline characteristics of participants A total of 550 questionnaires were distributed. However, 17 patients met the exclusion criteria, and 27 patients withdrew from the study because they were lost to follow-up or had no time. Ultimately, 506 valid questionnaires were collected (response rate: 92.0%). Based on previous literature, we have set the total anxiety scores 10 and 19 as the two cut-offs for dental anxiety25. Among the 506 patients, 291 had moderate level of dental anxiety (10 ≤ total anxiety scores ≤ 19 points), with an incidence rate of 57.5%. 14 had a severe level of dental anxiety (total anxiety scores ≥ 19 points), with an incidence rate of 2.8%. The baseline characteristics of the participants are shown in Table 1.Table 1Baseline characteristics of 506 participants.CharacteristicsNumber (n = 506)Gender Male218 (43.08%) Female288 (56.92%)Age (years, mean age 56.42 ± 15.23) 19–3563 (12.45%) 36–50136 (26.88%) 51–72307 (60.67%)Implant location Maxillary207 (40.91%) Mandibular278 (54.94%) Bimaxillary21 (4.15%)Number of implants 1402 (79.45%) ≥ 2104 (20.55%)Operation time (min) ≤ 60371 (73.32%) > 60135 (26.68%)Previous dental implant experience Yes40 (7.91%) No466 (92.09%) Table 1Baseline characteristics of 506 participants.CharacteristicsNumber (n = 506)Gender Male218 (43.08%) Female288 (56.92%)Age (years, mean age 56.42 ± 15.23) 19–3563 (12.45%) 36–50136 (26.88%) 51–72307 (60.67%)Implant location Maxillary207 (40.91%) Mandibular278 (54.94%) Bimaxillary21 (4.15%)Number of implants 1402 (79.45%) ≥ 2104 (20.55%)Operation time (min) ≤ 60371 (73.32%) > 60135 (26.68%)Previous dental implant experience Yes40 (7.91%) No466 (92.09%) Baseline characteristics of 506 participants. Patterns and influencing factors of anxiety Repeated-measures ANOVA showed that anxiety levels varied significantly across the five time points (p < 0.001). Mauchly’s W was 0.95 (p = 1.58e-03), indicating the data violated the sphericity assumption. Therefore, the Greenhouse–Geisser correction was applied (ε = 0.98). Subsequently, pairwise comparisons were conducted using the paired t-test, and Bonferroni correction was applied. Results showed that the anxiety levels of T3 and T4 were significantly higher than those of T1, T2, and T5 (p < 0.001). The difference between T1 and T2 was also statistically significant (p = 0.0033). There was no significant difference between T3 and T4 (p = 0.34), indicating that both the local anesthesia and implant placement stages induced similar anxiety peaks. The anxiety scores from highest to lowest were: T3 (2.61 ± 0.98), T4 (2.51 ± 1.05), T1 (2.38 ± 1.06), T2 (2.21 ± 0.97), and T5 (1.43 ± 0.72). Anxiety peaked during invasive stages (T3 and T4), which correspond to local anesthesia and implant placement. The scores dropped significantly post-surgery (T5), suggesting relief of patients after completing their operation. After an initial screening, we identified variables that might be influencing factors of dental anxiety. Because the number of people with severe anxiety (total anxiety scores ≥ 19 points) is relatively small, we combined patients with severe and moderate anxiety (total anxiety scores ≥ 10 points) and conducted a multivariable binary logistic regression analysis. Based on clinical relevance and prior researches, the following variables were entered into the multivariable binary logistic regression model: gender, age group (19–35, 36–50, 51–72 years), implant location, number of implants, operation time, and previous dental implant experience13,26–28. Univariate binary logistic regression analyses were also performed to identify these influencing factors of overall anxiety levels. The detailed results were listed in Table 2. Results show that the significant univariate predictors were gender, elder age, mandibular implant, bimaxillary implant, more than one implant, and more than one hour operation time. After adjusting the confounder “previous dental implant experience”, these variables remained independently associated with higher odds of moderate to severe level of anxiety except variables “more than one implant” and “more than one hour operation time” (p < 0.05). Therefore, gender, elder age, mandibular implant, and bimaxillary implant were identified as independent factors associated with increased odds of anxiety.Table 2Univariate and multivariable binary logistic regression analyses of factors associated with significant dental anxiety (total anxiety scores ≥ 10).Influencing factorsCategoriesAnxiety incidence rate(total anxiety scores ≥ 10)Unadjusted OR (95% CI)p valueAdjusted OR (95% CI)p valueGenderMale47.25%1 (reference)1 (reference)Female70.14%2.62 (1.82, 3.79)2.54e-072.60 (1.78, 3.84)1.16e-06Age (years)19–3546.03%1 (reference)1 (reference)36–5063.24%2.02 (1.10, 3.72)2.32e-022.40 (1.26, 4.65)8.52e-0351–7261.89%1.90 (1.10, 3.30)2.09e-022.19 (1.22, 3.99)9.31e-03Implant locationMaxillary50.24%1 (reference)1 (reference)Mandibular65.47%1.88 (1.30, 2.72)7.90e-041.81 (1.23, 2.67)2.63e-03Bimaxillary90.48%9.41 (2.64, 60.01)3.03e-039.63 (2.61, 62.56)3.29e-03Number of Implants156.97%1 (reference)1 (reference) ≥ 273.08%2.05 (1.29, 3.35)3.12e-031.59 (0.97, 2.67)7.19e-02Operation time ≤ 6057.68%1 (reference)1 (reference) > 6067.41%1.52 (1.01, 2.31)4.88e-021.39 (0.90, 2.18)0.15Previous dental implant experienceNo59.66%1 (reference)1 (reference)Yes67.50%1.41 (0.72, 2.88)0.331.46 (0.72, 3.11)0.31*Unadjusted ORs from separate simple logistic regressions. Adjusted ORs from the multivariable model including all variables. Table 2Univariate and multivariable binary logistic regression analyses of factors associated with significant dental anxiety (total anxiety scores ≥ 10).Influencing factorsCategoriesAnxiety incidence rate(total anxiety scores ≥ 10)Unadjusted OR (95% CI)p valueAdjusted OR (95% CI)p valueGenderMale47.25%1 (reference)1 (reference)Female70.14%2.62 (1.82, 3.79)2.54e-072.60 (1.78, 3.84)1.16e-06Age (years)19–3546.03%1 (reference)1 (reference)36–5063.24%2.02 (1.10, 3.72)2.32e-022.40 (1.26, 4.65)8.52e-0351–7261.89%1.90 (1.10, 3.30)2.09e-022.19 (1.22, 3.99)9.31e-03Implant locationMaxillary50.24%1 (reference)1 (reference)Mandibular65.47%1.88 (1.30, 2.72)7.90e-041.81 (1.23, 2.67)2.63e-03Bimaxillary90.48%9.41 (2.64, 60.01)3.03e-039.63 (2.61, 62.56)3.29e-03Number of Implants156.97%1 (reference)1 (reference) ≥ 273.08%2.05 (1.29, 3.35)3.12e-031.59 (0.97, 2.67)7.19e-02Operation time ≤ 6057.68%1 (reference)1 (reference) > 6067.41%1.52 (1.01, 2.31)4.88e-021.39 (0.90, 2.18)0.15Previous dental implant experienceNo59.66%1 (reference)1 (reference)Yes67.50%1.41 (0.72, 2.88)0.331.46 (0.72, 3.11)0.31*Unadjusted ORs from separate simple logistic regressions. Adjusted ORs from the multivariable model including all variables. Univariate and multivariable binary logistic regression analyses of factors associated with significant dental anxiety (total anxiety scores ≥ 10). *Unadjusted ORs from separate simple logistic regressions. Adjusted ORs from the multivariable model including all variables. Multicollinearity among the independent variables was assessed using variance inflation factors (VIFs). For categorical variables with more than two levels, the adjusted VIF was used. The results showed that the VIF and adjusted VIF values for all independent variables were less than 1.1, indicating no significant multicollinearity among the variables. The Hosmer–Lemeshow test yielded a non-significant result (χ2 = 8.14, df = 8, p = 0.42), indicating that the model fits the data well. The model demonstrated acceptable discrimination (AUC = 0.704, 95% CI 0.658–0.750). However, this model still needs improvement before it can be used for clinical decision-making. ROC curve for the multivariable logistic regression is listed below (Fig. 1).Fig. 1ROC curve for multivariable binary logistic regression. ROC curve for multivariable binary logistic regression. Multivariable binary logistic regression results in Table 2 showed that female patients were more likely to experience significant anxiety compared to males (OR = 2.60, 95% CI 1.78–3.84). Patients aged 36–50 years and 51–72 years were more than twice as likely to experience significant anxiety compared to the youngest group (OR = 2.40, 95% CI 1.26–4.65 and OR = 2.19, 95% CI 1.22–3.99). Implant location also played an important role, with mandibular implants being associated with a nearly 1.8 times anxiety risk as maxillary implants (OR = 1.81, 95% CI 1.23–2.67). Bimaxillary implants were associated with much more anxiety risk than maxillary implants (OR = 9.63, 95% CI 2.61–62.56). Similarly, patients who underwent more than one implant were much more likely to be anxious then those who underwent only one implant, but the influence was not statistically significant (OR = 1.59, 95% CI 0.97–2.67, p = 7.19e-02). The corresponding forest plot is shown in Fig. 2. The forest plot presents the adjusted ORs with 95% CIs for factors associated with moderate-to-severe dental anxiety from the multivariable binary logistic regression analysis. Each square represents the point estimate of the OR, and the horizontal line extends to the 95% CI. A vertical dashed line at the OR value 1 indicates no effect. Factors whose CI lies entirely to the right of this line are statistically significant (p < 0.05). ORs quantify the odds of anxiety in the exposed group relative to the reference group. An OR of more than 1 implies a higher risk. The forest plot thus enables rapid visual comparison of effect sizes and precision across different risk factors.Fig. 2Forest plot of factors associated with increased odds of dental anxiety. Forest plot of factors associated with increased odds of dental anxiety. To address the potential concern about interactions, we tested all two‑way interactions among the predictors in the multivariable logistic regression model. Likelihood ratio tests were performed, and p‑values were adjusted for multiple comparisons using the Bonferroni method. No interaction reached the adjusted significance level. Therefore, the reported effects for variables (gender, age, implant location, etc.) on anxiety levels are stable across different levels of the other covariates. The correlation between pain experiences and anxiety scores across perioperative stages was examined using Kruskal–Wallis test. Post‑hoc pairwise comparisons were performed using Dunn’s test with Bonferroni correction (Table 3). Kruskal–Wallis test showed that preoperative pain was not significantly correlated with anxiety scores at any time point (p > 0.05). Predicted intraoperative pain was strongly positively correlated with anxiety at the five time points (p < 0.05), underscoring the psychological impact of pain expectation. Intraoperative pain was significantly correlated with anxiety at T4 and T5 (p < 0.001). Postoperative pain also showed a strong correlation with anxiety scores at T5 (p < 0.01).Table 3Correlation between pain degree and anxiety scores at different time points.Grouping factorsNumberT1T2T3T4T5Preoperative painPainless3692.38 ± 1.062.20 ± 1.022.61 ± 1.042.51 ± 1.011.42 ± 0.73Mild pain782.31 ± 1.092.16 ± 0.972.54 ± 0.792.47 ± 1.131.44 ± 0.76Moderate pain592.47 ± 1.052.35 ± 0.532.69 ± 0.782.57 ± 1.201.47 ± 0.61Kruskal–Wallis H (df)0.62(2)4.51(2)0.76(2)0.33(2)4.47(2)p value0.730.110.680.850.11Predicted intraoperative painPainless412.07 ± 1.212.16 ± 1.532.14 ± 1.212.11 ± 0.811.21 ± 0.58Mild pain1482.21 ± 0.832.03 ± 1.102.35 ± 0.882.31 ± 0.941.36 ± 0.57Moderate pain1962.45 ± 1.142.24 ± 0.502.71 ± 0.972.49 ± 1.161.45 ± 0.88Severe pain1212.57 ± 1.092.41 ± 1.102.93 ± 0.892.92 ± 0.941.56 ± 0.62Kruskal–Wallis H (df)11.69(3)20.10(3)32.72(3)31.40(3)30.80(3)p value8.53e-031.62e-043.69e-077.01e-079.38e-07Intraoperative painPainless253––-2.35 ± 1.131.31 ± 0.67Mild pain161–––2.58 ± 0.891.49 ± 0.71Moderate pain73–––2.69 ± 1.031.56 ± 0.73Severe pain19–––3.37 ± 0.722.07 ± 0.96Kruskal–Wallis H (df)–––25.39(3)33.68(3)p value–––1.28e-052.31e-07Postoperative painPainless367––––1.36 ± 0.65Mild pain124––––1.59 ± 0.87Moderate pain15––––1.76 ± 0.84Kruskal–Wallis H (df)––––12.67(2)p value––––1.78e-3*Post‑hoc pairwise comparisons (Dunn’s test with Bonferroni correction) for predicted intraoperative pain at T1 yielded the following adjusted p values: Painless vs. Mild = 1.00, Painless vs. Moderate = 0.12, Painless vs. Severe = 2.26e-02, Mild vs. Moderate = 0.57, Mild vs. Severe = 7.83e-02, Moderate vs. Severe = 1.00, at T2 yielded the following adjusted p values: Painless vs. Mild = 1.00, Painless vs. Moderate = 5.41e-02, Painless vs. Severe = 3.35e-02, Mild vs. Moderate = 3.93e-03, Mild vs. Severe = 3.30e-03, Moderate vs. Severe = 1.00, at T3 yielded the following adjusted p values: Painless vs. Mild = 1.00, Painless vs. Moderate = 6.62e-03, Painless vs. Severe = 6.65e-05, Mild vs. Moderate = 0.010, Mild vs. Severe = 1.60e-05, Moderate vs. Severe = 0.26, at T4 yielded the following adjusted p values: Painless vs. Mild = 1.00, Painless vs. Moderate = 0.28, Painless vs. Severe = 7.80e-05, Mild vs. Moderate = 1.00, Mild vs. Severe = 7.82e-06, Moderate vs. Severe = 7.10e-04, at T5 yielded the following adjusted p values: Painless vs. Mild = 0.22, Painless vs. Moderate = 0.58, Painless vs. Severe = 3.19e-05, Mild vs. Moderate = 1.00, Mild vs. Severe = 1.29e-03, Moderate vs. Severe = 1.98e-05.*Post‑hoc pairwise comparisons (Dunn’s test with Bonferroni correction) for intraoperative pain at T4 yielded the following adjusted p values: Painless vs. Mild = 3.02e-02, Painless vs. Moderate = 4.85e-02, Painless vs. Severe = 1.13e-04, Mild vs. Moderate = 1.00, Mild vs. Severe = 1.46e-02, Moderate vs. Severe = 5.82e-02, at T5 yielded the following adjusted p values: Painless vs. Mild = 1.64e-03, Painless vs. Moderate = 2.49e-03, Painless vs. Severe = 4.88e-05, Mild vs. Moderate = 1.00, Mild vs. Severe = 2.52e-02, Moderate vs. Severe = 0.13.*Post‑hoc pairwise comparisons (Dunn’s test with Bonferroni correction) for postoperative pain at T5 yielded the following adjusted p values: Painless vs. Mild = 1.92e-02, Painless vs. Moderate = 3.25e-02, Mild vs. Moderate = 0.47.*Adjusted p values indicate significant differences while p < 0.05. Table 3Correlation between pain degree and anxiety scores at different time points.Grouping factorsNumberT1T2T3T4T5Preoperative painPainless3692.38 ± 1.062.20 ± 1.022.61 ± 1.042.51 ± 1.011.42 ± 0.73Mild pain782.31 ± 1.092.16 ± 0.972.54 ± 0.792.47 ± 1.131.44 ± 0.76Moderate pain592.47 ± 1.052.35 ± 0.532.69 ± 0.782.57 ± 1.201.47 ± 0.61Kruskal–Wallis H (df)0.62(2)4.51(2)0.76(2)0.33(2)4.47(2)p value0.730.110.680.850.11Predicted intraoperative painPainless412.07 ± 1.212.16 ± 1.532.14 ± 1.212.11 ± 0.811.21 ± 0.58Mild pain1482.21 ± 0.832.03 ± 1.102.35 ± 0.882.31 ± 0.941.36 ± 0.57Moderate pain1962.45 ± 1.142.24 ± 0.502.71 ± 0.972.49 ± 1.161.45 ± 0.88Severe pain1212.57 ± 1.092.41 ± 1.102.93 ± 0.892.92 ± 0.941.56 ± 0.62Kruskal–Wallis H (df)11.69(3)20.10(3)32.72(3)31.40(3)30.80(3)p value8.53e-031.62e-043.69e-077.01e-079.38e-07Intraoperative painPainless253––-2.35 ± 1.131.31 ± 0.67Mild pain161–––2.58 ± 0.891.49 ± 0.71Moderate pain73–––2.69 ± 1.031.56 ± 0.73Severe pain19–––3.37 ± 0.722.07 ± 0.96Kruskal–Wallis H (df)–––25.39(3)33.68(3)p value–––1.28e-052.31e-07Postoperative painPainless367––––1.36 ± 0.65Mild pain124––––1.59 ± 0.87Moderate pain15––––1.76 ± 0.84Kruskal–Wallis H (df)––––12.67(2)p value––––1.78e-3*Post‑hoc pairwise comparisons (Dunn’s test with Bonferroni correction) for predicted intraoperative pain at T1 yielded the following adjusted p values: Painless vs. Mild = 1.00, Painless vs. Moderate = 0.12, Painless vs. Severe = 2.26e-02, Mild vs. Moderate = 0.57, Mild vs. Severe = 7.83e-02, Moderate vs. Severe = 1.00, at T2 yielded the following adjusted p values: Painless vs. Mild = 1.00, Painless vs. Moderate = 5.41e-02, Painless vs. Severe = 3.35e-02, Mild vs. Moderate = 3.93e-03, Mild vs. Severe = 3.30e-03, Moderate vs. Severe = 1.00, at T3 yielded the following adjusted p values: Painless vs. Mild = 1.00, Painless vs. Moderate = 6.62e-03, Painless vs. Severe = 6.65e-05, Mild vs. Moderate = 0.010, Mild vs. Severe = 1.60e-05, Moderate vs. Severe = 0.26, at T4 yielded the following adjusted p values: Painless vs. Mild = 1.00, Painless vs. Moderate = 0.28, Painless vs. Severe = 7.80e-05, Mild vs. Moderate = 1.00, Mild vs. Severe = 7.82e-06, Moderate vs. Severe = 7.10e-04, at T5 yielded the following adjusted p values: Painless vs. Mild = 0.22, Painless vs. Moderate = 0.58, Painless vs. Severe = 3.19e-05, Mild vs. Moderate = 1.00, Mild vs. Severe = 1.29e-03, Moderate vs. Severe = 1.98e-05.*Post‑hoc pairwise comparisons (Dunn’s test with Bonferroni correction) for intraoperative pain at T4 yielded the following adjusted p values: Painless vs. Mild = 3.02e-02, Painless vs. Moderate = 4.85e-02, Painless vs. Severe = 1.13e-04, Mild vs. Moderate = 1.00, Mild vs. Severe = 1.46e-02, Moderate vs. Severe = 5.82e-02, at T5 yielded the following adjusted p values: Painless vs. Mild = 1.64e-03, Painless vs. Moderate = 2.49e-03, Painless vs. Severe = 4.88e-05, Mild vs. Moderate = 1.00, Mild vs. Severe = 2.52e-02, Moderate vs. Severe = 0.13.*Post‑hoc pairwise comparisons (Dunn’s test with Bonferroni correction) for postoperative pain at T5 yielded the following adjusted p values: Painless vs. Mild = 1.92e-02, Painless vs. Moderate = 3.25e-02, Mild vs. Moderate = 0.47.*Adjusted p values indicate significant differences while p < 0.05. Correlation between pain degree and anxiety scores at different time points. *Post‑hoc pairwise comparisons (Dunn’s test with Bonferroni correction) for predicted intraoperative pain at T1 yielded the following adjusted p values: Painless vs. Mild = 1.00, Painless vs. Moderate = 0.12, Painless vs. Severe = 2.26e-02, Mild vs. Moderate = 0.57, Mild vs. Severe = 7.83e-02, Moderate vs. Severe = 1.00, at T2 yielded the following adjusted p values: Painless vs. Mild = 1.00, Painless vs. Moderate = 5.41e-02, Painless vs. Severe = 3.35e-02, Mild vs. Moderate = 3.93e-03, Mild vs. Severe = 3.30e-03, Moderate vs. Severe = 1.00, at T3 yielded the following adjusted p values: Painless vs. Mild = 1.00, Painless vs. Moderate = 6.62e-03, Painless vs. Severe = 6.65e-05, Mild vs. Moderate = 0.010, Mild vs. Severe = 1.60e-05, Moderate vs. Severe = 0.26, at T4 yielded the following adjusted p values: Painless vs. Mild = 1.00, Painless vs. Moderate = 0.28, Painless vs. Severe = 7.80e-05, Mild vs. Moderate = 1.00, Mild vs. Severe = 7.82e-06, Moderate vs. Severe = 7.10e-04, at T5 yielded the following adjusted p values: Painless vs. Mild = 0.22, Painless vs. Moderate = 0.58, Painless vs. Severe = 3.19e-05, Mild vs. Moderate = 1.00, Mild vs. Severe = 1.29e-03, Moderate vs. Severe = 1.98e-05. *Post‑hoc pairwise comparisons (Dunn’s test with Bonferroni correction) for intraoperative pain at T4 yielded the following adjusted p values: Painless vs. Mild = 3.02e-02, Painless vs. Moderate = 4.85e-02, Painless vs. Severe = 1.13e-04, Mild vs. Moderate = 1.00, Mild vs. Severe = 1.46e-02, Moderate vs. Severe = 5.82e-02, at T5 yielded the following adjusted p values: Painless vs. Mild = 1.64e-03, Painless vs. Moderate = 2.49e-03, Painless vs. Severe = 4.88e-05, Mild vs. Moderate = 1.00, Mild vs. Severe = 2.52e-02, Moderate vs. Severe = 0.13. *Post‑hoc pairwise comparisons (Dunn’s test with Bonferroni correction) for postoperative pain at T5 yielded the following adjusted p values: Painless vs. Mild = 1.92e-02, Painless vs. Moderate = 3.25e-02, Mild vs. Moderate = 0.47. *Adjusted p values indicate significant differences while p < 0.05. For predicted intraoperative pain, the severe pain group consistently had significantly higher anxiety scores than the painless group at all time points, and significantly higher anxiety scores than the mild pain group at T2-T5 (all adjusted p < 0.05). At T2 and T3, the moderate pain group showed significantly higher anxiety than the mild pain group. The moderate pain group showed significantly higher anxiety than the painless group at T3. At T4 and T5, the severe pain group showed significantly higher anxiety than the moderate pain group. For intraoperative pain, patients with severe pain had significantly higher anxiety at T4 and T5 compared to those with no pain or mild pain. Patients with moderate pain or mild pain had significantly higher anxiety at T5 compared to those with no pain. For postoperative pain, both mild and moderate pain groups had significantly higher anxiety than the no pain group at T5. The corresponding heatmap is shown in Fig. 3. The heatmap displays mean anxiety scores (ranging from 1 = no anxiety to 5 = extreme anxiety) at five time points stratified by pain severity (painless, mild, moderate, severe) for four pain types: preoperative pain, predicted intraoperative pain, actual intraoperative pain, and postoperative pain. Colour intensity reflects the magnitude of the mean anxiety score, with darker red indicating higher anxiety. Darker red blocks in the heatmap facilitate rapid visual identification of high-risk groups throughout the perioperative period.Fig. 3Heatmap of correlations between pain levels and anxiety scores across perioperative stages of dental implant surgery. *Darker red indicates higher anxiety. Heatmap of correlations between pain levels and anxiety scores across perioperative stages of dental implant surgery. *Darker red indicates higher anxiety. To further elucidate the relationship among anxiety scores, predicted intraoperative pain, actual intraoperative pain, and postoperative pain levels, a three-dimensional visualization was generated (Fig. 4). The plot illustrates how total anxiety scores vary with patients’ expectations of pain and their experienced pain during and after the implant surgeries, providing an integrated perspective on the interplay between psychological anticipation and sensory experience. A smoothed regression surface is fitted to the data. The colour gradient (light blue to dark blue) represents increasing anxiety scores. The surface shows that higher anticipated pain is consistently associated with elevated anxiety scores across all levels of experienced pain. The steep slope along the anticipated‑pain axis indicates that pre‑operative pain expectation exerts a stronger influence on anxiety levels than the actual perceived pain. Anticipated and experienced pain were rated on the NRS from 0 (no pain) to 10 (worst imaginable pain). Total anxiety scores were derived from the sum of anxiety scores of T1 to T5 (each 1–5), giving a possible range of 5 (minimal anxiety) to 25 (extreme anxiety).Fig. 4Three-dimensional visualization of the relationship between anxiety scores and pain levels in dental implant patients. Three-dimensional visualization of the relationship between anxiety scores and pain levels in dental implant patients. We further establish a prediction model incorporating the five identified independent factors associated with increased odds of dental anxiety (gender, aged 36–50, aged 51–72, mandibular implant, bimaxillary implant) and predicted pain level to predict the possibility of anxiety during dental implant surgeries (Fig. 5). AUC of the model is 0.743 (95%Cl: 0.700, 0.785), demonstrating acceptable discrimination. The Hosmer–Lemeshow test yielded a non-significant result (χ2 = 7.28, df = 8, p = 0.51), indicating that the model fits the data well. The internal validation was performed using bootstrapping, and the corrected AUC is 0.729 (95% Cl: 0.691, 0.770).Fig. 5Nomogram for predicting the risk of dental anxiety among patients undergoing dental implant surgery. Nomogram for predicting the risk of dental anxiety among patients undergoing dental implant surgery. The nomogram assigns a point value to each variable, and the total points correspond to the predicted probability of experiencing dental anxiety. This model would integrate these variables into a scoring system, allowing clinicians to stratify patients into low-risk, moderate-risk, and high-risk anxiety groups preoperatively. We can locate each variable level by drawing a vertical line to the “Points” axis to obtain points. Then we sum points and locate the total points on the “Total Points” axis to read the predicted probability. For example, a female patient aged 55 years, receiving a mandibular implant with the predicted pain score 6 would have 167 total points, which means her predicted probability of anxiety is about 80%. This predictive model could facilitate early targeted interventions for patients at potential risk of dental anxiety, such as preoperative counseling, enhanced pain management, and relaxation techniques, thereby potentially mitigating patients’ anxiety and improving overall surgical experience. However, this nomogram has not been externally validated. Therefore, it requires caution in clinical applications. Discussion Dental anxiety is a common psychological problem in patients experienced oral surgeries, and its occurrence is affected by multiple factors such as physiology, psychology, and society29–33. This study systematically explored the patterns and influencing factors of anxiety in patients undergoing dental implant surgeries in China, providing a detailed theoretical basis for the research and management in this field. It was found that the incidence of moderate or severe anxiety (total anxiety scores ≥ 10 points) was 60.3%, which was consistant with the level (20%-80%) reported in previous studies13–15,21,34. The results showed that anxiety levels peaked during local anesthesia administration (T3) and implant placement (T4). These two stages involve direct invasive operations (e.g., needle insertion for anesthesia, bone drilling for implant placement), and patients experience strong sensory stimulation (e.g., pressure, vibration) and fear of the unknown. The significant decrease in anxiety at T5 indicates that the relief of uncertainty effectively reduces the anxiety of patients. From the perspective of influencing factors, this study found that female patients had a higher incidence of anxiety than male patients, which is consistent with the previous research results26,35,36. The reason may be that women are more sensitive to pain and have stronger emotional vulnerability, making them more likely to feel anxious about the surgical process. In terms of age, older patients had a higher anxiety incidence, which may be due to their greater concern about the impact of implant surgeries on their health and daily life. The desire to maintain a youthful appearance and social confidence could exacerbate anxiety about unsatisfactory operation outcomes37,38. In addition, patients experienced more than one implant at one time, suffered mandibular teeth implantation, or underwent a long operation time were more likely to have higher anxiety scores. The relatively long operation duration may prolong the period of uncertainty, thereby making anxiety accumulate across perioperative phases. The notably elevated anxiety associated with mandibular implant placement can be attributed to anatomical and procedural factors of the lower jaw. Anatomically, the close proximity of implant sites to the inferior alveolar nerve raises the perceived risk of complications such as persistent pain and paresthesia within the lip or chin region, and patients often regard these complications as a severe outcome39,40. Procedurally, the denser cortical bone of the mandible frequently necessitates more extensive osteotomy, resulting in longer operation times, greater vibration, and more perception of invasiveness41,42. Furthermore, achieving profound anesthesia in the mandible via inferior alveolar nerve blocks is technically much more difficult than anesthesia in maxillary bone, aggravating the anxiety level associated with uncertainty43–45. To sum up, the above findings remind us to pay closer attention to these high-risk populations in clinical work and implement personalized intervention measures. The correlation analysis between pain degrees and anxiety levels showed that predicted intraoperative pain was strongly correlated with preoperative, intraoperative, and postoperative anxiety (p < 0.05), while actual pain was associated with intraoperative and postoperative anxiety (p < 0.005). This phenomenon indicates that pain anticipation plays a critical role in anxiety development, which aligns with the psychological construct of anticipatory anxiety46,47. Anticipatory anxiety arises from the perception of an impending threat, which may result in high levels of anxiety and vigilance. In dental implant surgeries, the uncertainty surrounding the intensity of forthcoming pain can amplify dental anxiety profoundly. Catastrophic thinking often causes patients to overestimate pain severity, thereby elevating anxiety preoperatively and during procedural phases48. In a word, the lack of control associated with future pain contributes significantly to psychological distress and anxiety. In contrast, experienced pain would probably become a concrete sensation that can be assessed and managed. Thus, dental professionals in China should fully communicate with patients and manage their expectations before surgeries, providing accurate information about pain. For example, they could explain the impending pain as “Most patients report mild pain during anesthesia, similar to a mosquito bite.” Additionally, dental professionals should adopt effective analgesic measures (e.g., preoperative anti-inflammatory drugs, local anesthesia with long-acting anesthetics) to reduce actual pain and provide postoperative pain management guidance. For example, they can remind patients to take analgesics as prescribed if pain is moderate or severe after surgeries49,50. Although we have made some discoveries, this study also has several limitations. First, it was a single-center study. The participants were all from Changsha Stomatological Hospital. The sample source was relatively single, and findings may not generalize to other regions or primary care settings in China. Second, the follow-up time was only 30 min, and longer-term anxiety trajectories and their impact on postoperative outcomes (eg, wound healing, implant survival) were not examined. Third, the evaluation of the anxiety level and pain degree only used self-reported scales, which may result in subjective bias. For example, patients may underreport anxiety scores to avoid being seen as cowardly. Objective parameters such as heart rate and cortisol levels should be included to improve the accuracy of anxiety measurement in future studies. In light of the integrated findings, personalized anxiety management strategies can be formulated and implemented within the Chinese clinical setting. Preoperative health education and psychological counseling can help patients understand the treatment process and reduce their fear of the unknown51. During preoperative consultations, dental professionals should actively listen to patients’ anxiety triggers and provide evidence-based reassurance, like “Do not worry. Our department has a 98% implant success rate over the past 15 years”. Intraoperative distraction and real-time feedback can alleviate patients’ tension during dental surgeries52. Dental professionals should use proper local anesthetics to reduce intraoperative pain, and prescribe postoperative analgesics with clear instructions. Postoperative follow-up can timely mitigate patients’ anxiety and enhance their confidence in recovery. In addition, the comfortable environment and humanized care in the operating room can also play a positive role in relieving pain, thus relieving the perception of perioperative anxiety53,54. The full-process psychological intervention can not only alleviate the patient’s anxiety and pain, but also improve the patient’s recognition of the hospital’s medical services. This is of great significance for promoting the development of the dental implant discipline. In conclusion, the incidence of anxiety in patients undergoing dental implantation in China is relatively high. The influencing factors include gender, age, implant location, number of implants, operation time, and perceived pain degree. These findings can promote the development of localized psychological support protocols and perioperative interventions to relieve dental anxiety for dental implant patients in the Chinese healthcare environment. Ultimately, implementing such targeted anxiety management approaches could enhance not only individual patient experiences but also the overall quality and sustainability of implant dentistry in China. In the future, multi-center studies across diverse geographic and socioeconomic regions in China with long follow-up time would help validate and generalize findings of the study. Conclusions Dental anxiety among Chinese patients undergoing implant surgeries demonstrates distinct differences in different perioperative stages. It peaks during invasive procedural phases, specifically local anesthesia administration and implant placement. Female gender, elderly age, mandibular implant placement, bimaxillary implant placement, higher anticipated pain levels, and higher actual pain levels were identified as independent factors associated with elevated anxiety scores. It is suggested that full-process targeted psychological interventions for high-risk populations, such as the preoperative health education and psychological counseling, intraoperative distraction and real-time feedback, postoperative follow-up and guidance, should be integrated into clinical practice to effectively mitigate patients’ anxiety, improve treatment satisfaction, and further improve the quality of dental implant services in China. Hospitals in China can further optimize their intervention plans, such as introducing virtual reality (VR) technology for preoperative simulation55,56, in order to provide more high-quality and humanized medical services for implant patients. Future studies should adopt multi-center longitudinal designs to probe anxiety patterns in different areas of China, and explore the long-term impact of anxiety management strategies.