Risk factors for recurrence in benign paroxysmal positional vertigo patients and construction of a predictive nomogram

Abstract

This study aims to investigate potential factors associated with recurrence in patients with benign paroxysmal positional vertigo (BPPV) and establish and validate a predictive model. A retrospective study was conducted on BPPV patients who visited our hospital from January 2022 to December 2024, and their complete clinical information was collected. Univariate and multifactorial logistic regression were used to identify independent variables associated with recurrence, and a nomogram was constructed to visualize the model. Model performance was evaluated using receiver operating characteristic curves and area under the curve, and model fit was assessed using the Hosmer–Lemeshow test. To enhance the robustness of validation, we introduced the Bootstrap method combined with 10-fold cross-validation for internal validation, and assessed its clinical practicality through calibration plots and decision curves. The recurrence of BPPV is closely associated with multiple factors. The established predictive tool exhibits good predictive accuracy and can assist clinicians in identifying high-risk individuals for recurrence and implementing intervention measures in advance. A total of 276 BPPV patients were included, among whom 70 experienced recurrence within 6 months, with a recurrence rate of 25.36%. Of the 276 patients, 112 were male and 164 were female, with an age range of 21 to 76 years (mean age: 58.0 ± 12.3 years). Among the 70 patients who experienced recurrence, 22 were male and 48 were female, with a mean age of 61.2 ± 11.7 years. Statistical analysis identified 5 significant independent risk factors: insomnia (OR = 3.593, 95% CI: 1.541–8.379), depressive state (OR = 3.800, 95% CI: 1.082–13.348), hypertension (OR = 4.315, 95% CI: 1.014–18.365), diabetes (OR = 3.216, 95% CI: 1.392–7.425), and hyperlipidemia (OR = 3.912, 95% CI: 1.412–10.840). The results of receiver operating characteristic analysis, Hosmer–Lemeshow test, and internal cross-validation indicated that the model demonstrated good discriminative ability and fitting, with stable performance and clinical reference value.

1. Introduction

Benign paroxysmal positional vertigo (BPPV), also known as “ear stone syndrome,”^[1] is a transient vestibular disorder caused by the displacement of otoconia (ear stones) into the semicircular canals of the inner ear. Its primary clinical features include transient vertigo triggered by changes in head position relative to gravity (e.g., looking up, turning the head, lying down, or standing up), making it one of the most common peripheral vestibular disorders encountered in otolaryngology.^[2,3] In addition to vertigo, some patients may experience nystagmus, nausea, and vomiting,^[4,5] significantly impairing daily activities and quality of life.

Globally, BPPV affects approximately 2.4% of the general population, with higher prevalence among the elderly.^[6] In China, population-based studies have reported a prevalence of around 1.6%, and recurrence rates ranging from 15% to 30% within 1 year after treatment.^[7] Although BPPV is typically self-limiting and many patients experience resolution following repositioning maneuvers, a significant proportion of individuals suffer recurrent episodes.^[8] These recurrences not only increase the risk of falls and fractures, particularly in the elderly, but also exacerbate psychological conditions such as anxiety, depression, and panic, further reducing patient well-being.^[9] Therefore, early identification of recurrence risk factors is essential to enable timely interventions and personalized management strategies.

In recent years, several studies have explored potential risk factors for BPPV recurrence. For instance, Pan et al^[10] proposed a recurrence risk scoring system based on metabolic comorbidities, while Tang et al^[11] developed a prediction model using machine learning to identify high-risk patients. However, these approaches were limited by relatively small sample sizes, lack of external validation, or a narrow focus on specific BPPV subtypes. Thus, while preliminary predictive tools exist, their clinical applicability remains constrained.

Nomograms, as visually intuitive multivariable models, have been successfully applied in various medical fields to predict disease risk or prognosis. In otolaryngology, nomograms have been used to assess prognosis in idiopathic sudden sensorineural hearing loss^[12] and residual symptoms in BPPV patients,^[13] showing promising predictive performance. Building upon this, our study aims to construct and validate a nomogram for BPPV recurrence risk using a larger, clinically diverse patient cohort, with robust internal validation techniques (including bootstrapping and 10-fold cross-validation). This model is intended to serve as a reliable and practical tool for clinicians to identify high-risk patients early and implement targeted interventions to reduce recurrence and improve outcomes.

2. Data and methods

2.1. Data sources and collection

This study retrospectively organized and analyzed clinical data from patients diagnosed with BPPV and treated between January 2022 and December 2024 based on our hospital’s previous medical records. As this study is a non-interventional retrospective study, all data were obtained from routine clinical practice and no additional interventions were performed on the patients. The study strictly adhered to the relevant ethical guidelines and approval procedures of the hospital’s ethics review committee and complied with medical research ethics requirements.

All patient data were collected from Ya’an Hospital of Traditional Chinese Medicine, located in Sichuan Province, China. This institution serves as a regional tertiary care center covering urban and rural populations.

2.2. Inclusion and exclusion criteria

Inclusion criteria: participants in this study had to meet the following criteria: a definitive diagnosis of BPPV based on relevant guidelines or literature standards^[14]; clear consciousness and ability to fully cooperate with clinical examinations, including completing positional changes such as standing up, lying down, and turning over; the patient or their family members can truthfully provide basic medical history information and complete the required questionnaire assessments and follow-up content according to the study requirements.

Exclusion criteria: the following cases were excluded from this study: vertigo symptoms caused by central nervous system disorders (such as cerebellar or brainstem lesions); severe cervical spine disease that prevents safe performance of positional provocation tests or repositioning maneuvers; incomplete clinical data or missing key information that cannot meet the requirements for analysis; concurrent mental or cognitive impairments that prevent cooperation with the study process.

BPPV diagnosis was based on the criteria outlined in the 2017 Clinical Practice Guideline, and corresponding ICD-10 codes were used to identify cases in the electronic medical record system.

2.3. Collection of relevant variables

This study systematically collected clinical baseline information and relevant medical history data for all enrolled BPPV patients. The variables collected include: age, gender, body mass index, smoking and drinking habits, presence of significant fatigue, sleep disorders (such as insomnia), emotional status (including depression and anxiety), previous diagnosis of hypertension, diabetes, hyperlipidemia, coronary heart disease, osteoporosis, and history of migraine or head trauma.

All patients received standardized canalith repositioning maneuvers following initial diagnosis, with efficacy confirmed through clinical symptom improvement and eye movement examinations. For patients with successful repositioning, regular follow-up was scheduled for 6 months to observe whether recurrence occurred. If the patient reported recurrence of vertigo symptoms during follow-up and the results of positional provocation tests (such as the Dix–Hallpike or Roll test) were positive, the positional vertigo symptoms consistent with the initial presentation were confirmed as a “recurrence event.”

Depressive state was evaluated using the Chinese version of the Patient Health Questionnaire-9, with a cutoff score ≥ 10 indicating clinically relevant depression. Anxiety was assessed using the Generalized Anxiety Disorder-7 scale.

2.4. Statistical analysis

To systematically construct and validate a predictive model for the risk of recurrence in BPPV patients, this study used stratified random sampling to group the complete dataset. Using R software (version 4.2.1, R Foundation for Statistical Computing, Vienna, Austria), the collected clinical data were randomly divided into a modeling group (70%) and a validation group (30%) in a 7:3 ratio. The modeling group was used for model establishment and internal preliminary evaluation, while the validation group served as an independent dataset for external validation of the model’s generalization performance.

During the model construction phase, SPSS 27.0 statistical analysis software (IBM Corp., Armonk) was first used to conduct univariate analysis on the modeling group samples to explore the association between clinical variables and recurrence events. Categorical variables were tested using chi-square tests to examine the statistical significance of intergroup differences, and factors potentially associated with recurrence were preliminarily screened. The significance test criterion was set at P < .05, and the screened variables were included as candidates for further analysis.

Next, the variables with statistical significance from the univariate analysis were further introduced into a multivariate logistic regression model to identify independent risk factors for BPPV recurrence. The same P value threshold of < .05 was used to ensure that the predictive factors included in the model possessed independent predictive ability and statistical significance.

Based on the regression results, a nomogram was constructed using R software to visually present the relative contributions and weighted scores of each risk factor in predicting recurrence risk, thereby enhancing the clinical interpretability and practicality of the model. To systematically evaluate the model’s discriminative ability, receiver operating characteristic (ROC) curves were plotted in the modeling group and validation group, and the area under the curve (AUC) was calculated. The closer the AUC value is to 1.0, the stronger the model’s ability to distinguish between recurrent and non-recurrent patients, and the more ideal the prediction effect.

To evaluate the model’s goodness of fit, calibration curves were plotted to compare the consistency between predicted probabilities and actual observed incidence rates. If the calibration curve approaches the ideal diagonal line, it indicates that the model has good calibration performance. In addition, to further explore the clinical applicability of the model, decision curve analysis (DCA) is used to compare the net benefit of 3 strategies (using the prediction model, full intervention, or no intervention) at different probability thresholds. If the model shows higher net benefit over a wide range of probabilities, it suggests that it has strong clinical guidance potential.

To enhance the robustness and reliability of model validation, we further introduced the Bootstrap method (repeated sampling 1000 times) combined with 10-fold cross-validation for internal resampling validation to assess the model’s stability under different data distributions. Finally, by comprehensively analyzing the ROC curves, calibration plots, and DCA results, the discriminative performance, fit, and potential clinical utility of the predictive model were comprehensively evaluated to provide scientific basis and feasible tools for risk stratification management of BPPV patients.

Cases with missing key variables were excluded using listwise deletion. The proportion of missing data was below 5%, and imputation was deemed unnecessary due to the limited extent and random distribution of missingness.

Before conducting multivariate logistic regression, key assumptions were assessed: linearity of the logit was evaluated using Box–Tidwell transformation for continuous variables; multicollinearity was assessed via variance inflation factors, with all VIF values < 2.0 indicating no significant multicollinearity; and model specification was examined using residual plots and the link test. All tests supported the appropriateness of the logistic regression framework used.

3. Results

3.1. General characteristics

A total of 276 BPPV patients were included, among whom 70 experienced recurrence within 6 months, with a recurrence rate of 25.36%. Of the 276 patients, 112 were male and 164 were female, with an age range of 21 to 76 years (mean age: 58.0 ± 12.3 years). Among the 70 patients who experienced recurrence, 22 were male and 48 were female, with a mean age of 61.2 ± 11.7 years. To construct and validate a recurrence risk prediction model, all enrolled cases were stratified and randomly sampled in a 7:3 ratio using computer-generated random numbers. Among these, 193 patients were assigned to the modeling group for variable screening and model construction, while the remaining 83 patients served as the validation group to independently assess the model’s predictive performance and stability.

The overall baseline characteristics of the full study cohort (n = 276) are summarized in Table 1.

Table 1.

Baseline demographic and clinical characteristics of all BPPV patients (N = 276).

Variable	n (%)
Age ≤ 50 yr	103 (37.3%)
Male	104 (37.7%)
BMI ≤ 24	152 (55.1%)
Smoking	77 (27.9%)
Drinking	113 (41.0%)
Weariness	163 (59.1%)
Sleeplessness	140 (50.7%)
Depression	57 (20.7%)
Anxieties	140 (50.7%)
Hypertension	159 (57.6%)
Diabetes	138 (50.0%)
Hyperlipidemia	97 (35.1%)
Coronary heart disease	91 (33.0%)
Osteoporosis	120 (43.5%)
Migraineur	59 (21.4%)
History of head trauma	37 (13.4%)

BPPV = benign paroxysmal positional vertigo, BMI = body mass index.

Baseline comparability between the modeling and validation subsets was assessed, with no significant differences observed across key variables (Table 2).

Table 2.

Comparison of baseline characteristics between the modeling and validation groups.

Risk factors		Modeling group n = 193	Validation group n = 83	P
Age	≤50 yr	72	31	.979
	>50 yr	121	52
Sex	Male	73	31	.979
	Female	120	52
BMI	≤24	107	46	.927
	>24	86	37
Smoking	Yes	54	23	.887
	No	139	60
Drinking	Yes	79	34	.948
	No	114	49
Weariness	Yes	114	49	.927
	No	79	34
Sleeplessness	Yes	98	42	.988
	No	95	41
Depression	Yes	40	17	.991
	No	153	66
Anxiety	Yes	98	42	.988
	No	95	41
Hypertension	Yes	111	48	.975
	No	82	35
Diabetes	Yes	96	41	.981
	No	97	42
Hyperlipidemia	Yes	67	29	.995
	No	126	54
Coronary heart disease	Yes	63	27	.982
	No	130	56
Osteoporosis	Yes	82	35	.975
	No	111	48
Migraineur	Yes	41	18	.991
	No	152	65
History of head trauma	Yes	26	11	.985
	No	167	72

BMI = body mass index.

3.2. Independent risk factors for recurrence in BPPV patients

In the modeling group, 16 clinical and medical history-related variables were included in the study. A univariate logistic regression analysis was performed to screen for potential risk factors associated with BPPV recurrence. The analysis results showed that 10 factors were statistically significantly associated with recurrence (P < .05), including increasing age, alcohol consumption, sleep disorders (insomnia), depressive mood, anxiety, previous hypertension, diabetes, hyperlipidemia, osteoporosis, and history of head trauma (see Table 3). These variables may play a role in the recurrence of symptoms and are therefore important for further analysis.

Table 3.

Univariate analysis of recurrence in patients with benign paroxysmal positional vertigo.

Risk factors	Recur (n = 49)	No recur (n = 144)	P
Age			.032
≤50 yr	12	60
>50 yr	37	84
Sex			.856
Male	18	55
Female	31	89
BMI			.292
≤24	24	83
>24	25	61
Smoking			.794
Yes	13	41
No	36	103
Drinking			.046
Yes	26	53
No	23	91
Weariness			.304
Yes	32	82
No	17	62
Sleeplessness			.007
Yes	33	65
No	16	79
Depression			.017
Yes	16	24
No	33	120
Anxieties			.043
Yes	31	67
No	18	77
Hypertension			.023
Yes	35	76
No	14	68
Diabetes			.012
Yes	32	64
No	17	80
Hyperlipidemia			.015
Yes	24	43
No	25	101
Coronary heart disease			.723
Yes	17	46
No	32	98
Osteoporosis			.039
Yes	27	55
No	22	89
Migraineur			.063
Yes	15	26
No	34	118
History of head trauma			.033
Yes	11	15
No	38	129

BMI = body mass index.

Subsequently, the significant variables identified in the univariate analysis were further included in a multivariate logistic regression model to identify key factors with independent predictive value for BPPV recurrence. The results indicated that insomnia (OR = 3.593, 95% CI: 1.541–8.379), depression (OR = 3.800, 95% CI: 1.082–13.348), hypertension (OR = 4.315, 95% CI: 1.014–18.365), diabetes (OR = 3.216, 95% CI: 1.392–7.425), and hyperlipidemia (OR = 3.912, 95% CI: 1.412–10.840) were identified as independent risk factors for recurrence (see Table 4). These variables remained significant in the multivariate analysis, indicating that they still have an important impact on BPPV recurrence after controlling for confounding factors and should be given clinical attention.

Table 4.

Multifactorial analysis of recurrence in patients with benign paroxysmal positional vertigo.

Risk factors	β	SE	Wald	P	OR	95% CI
Sleeplessness	1.279	0.432	8.765	.032	3.593	1.541–8.379
Depression	1.335	0.641	4.338	.016	3.800	1.082–13.348
Hypertension	1.462	0.739	3.914	<.001	4.315	1.014–18.365
Diabetes	1.168	0.427	7.482	.012	3.216	1.392–7.425
Hyperlipidemia	1.364	0.520	6.881	.004	3.912	1.412–10.840

It is worth noting that the confidence intervals for certain variables (e.g., depression, hypertension, and hyperlipidemia) were relatively wide. This may reflect variability within the sample or the influence of confounding variables not fully captured in the model. However, the associations remained statistically significant and consistent with prior literature, supporting their clinical relevance despite statistical uncertainty.

3.3. Nomogram development and validation

Based on the independent risk factors screened out in the multivariate logistic regression analysis, a nomogram model was further constructed to visually predict the probability of recurrence in BPPV patients during the follow-up period (see Fig. 1). This predictive tool assigns a corresponding risk score to each significant risk factor and maps the total score to the probability of recurrence, providing clinicians with a concise and intuitive method for individualized risk assessment. This facilitates rapid identification of high-risk individuals and the development of early intervention strategies in clinical practice.

Figure 1.

Nomogram prediction model for recurrence in patients with benign paroxysmal positional vertigo.

To evaluate the discriminative ability and stability of the model, ROC curves were plotted in the modeling and validation sets, and AUC was calculated as the evaluation index. The Hosmer–Lemeshow goodness-of-fit test results showed that the modeling group had a χ² value of 7.634 and a P value of .455, while the validation group had a χ² value of 7.964 and a P value of .442, indicating that the predictive model fits well and has high consistency with the actual observed data. ROC analysis results show that the AUC of the modeling group is 0.851, and that of the validation group is 0.835 (see Fig. 2A and B), indicating that the model has good discriminative performance and can effectively distinguish between recurrent and non-recurrent patients.

Figure 2.

ROC curves of the nomogram for recurrence in patients with benign paroxysmal positional vertigo in the training set (A) and the validation set (B). Calibration curves of the nomogram for recurrence in patients with benign paroxysmal positional vertigo in the training set (C) and the validation set (D). DCA of the nomogram for recurrence in patients with benign paroxysmal positional vertigo in the training set (E) and the validation set (F). DCA = decision curve analysis, ROC = receiver operating characteristic.

Additionally, to enhance the internal validation stability and reliability of the model, internal resampling evaluation was conducted using the Bootstrap method (repeated sampling 1000 times) combined with 10-fold cross-validation. The results showed that the average AUC value was 0.846, sensitivity reached 83.2%, and specificity was 82.8%, further confirming the model’s good generalization performance and robustness across different samples.

To validate the fit between the model’s predicted values and the actual incidence rates, calibration curves were plotted in the modeling group and validation group (see Fig. 2C and D). The calibration curve results showed that the predicted recurrence probabilities were highly consistent with the actual observed incidence rates, indicating that the model has good calibration in risk estimation.

Finally, the clinical application value of the model at different risk thresholds was further evaluated using DCA (Fig. 2E and F). The DCA results showed that the line chart model demonstrated high net benefit across multiple threshold ranges, outperforming the “all intervention” or “no intervention” strategies, indicating that this predictive tool has practical application value in clinical decision-making and can provide effective support for BPPV recurrence risk management.

4. Discussion

The pathogenesis of BPPV is currently believed to be primarily associated with degenerative changes in the inner ear labyrinth and the ellipsoid sac, leading to the detachment of otoliths from their original attachment sites and their migration into the semicircular canals, resulting in disturbances in endolymphatic fluid dynamics and triggering positional vertigo symptoms.^[15,16] In clinical treatment, canalith repositioning maneuvers are widely recognized as the first-line therapy.^[17] This technique involves controlling the sequence, angle, and speed of the patient’s head and body movements, utilizing gravity and inertia to guide dislodged otoliths along the semicircular canal pathways back to the utricle, thereby effectively alleviating vertigo symptoms.^[18,19] However, although most patients achieve significant therapeutic effects through repositioning maneuvers, some individuals may experience symptom recurrence or relapse. Therefore, early identification and intervention for the risk of recurrence in BPPV are crucial for improving outcomes and reducing the incidence of subsequent episodes. Effective prevention and treatment strategies should be based on the precise identification of high-risk populations, which in turn relies on a thorough understanding of relevant risk factors. This study utilized multiple indicators commonly observed and easily accessible in clinical practice, combined with retrospective data analysis, to identify 5 independent risk factors closely associated with BPPV recurrence. These factors were then used as core variables to construct a nomogram prediction model. The model is structurally clear, easy to use, and convenient for clinicians to conduct individualized risk assessments in clinical practice. Through systematic internal validation analysis (including ROC curves, calibration curves, and DCA), the model demonstrated good predictive performance and clinical applicability, providing a practical quantitative tool for early warning and precise management of BPPV recurrence in the future.

The results of this study indicate that insomnia is an important risk factor for recurrence in BPPV patients. Numerous studies have confirmed that insomnia not only affects an individual’s sleep quality but also has negative impacts on overall health and quality of life.^[20] Notably, many BPPV patients had varying degrees of sleep disorders prior to their initial diagnosis.^[21] Related studies have shown that insomnia is positively correlated with the risk of developing BPPV, particularly in patients with chronic insomnia, who have a significantly higher probability of developing BPPV.^[22] Although the direct pathophysiological mechanism linking insomnia and BPPV remains unclear, recent studies suggest that sleep deprivation may lead to enhanced asymmetry in vestibular-ocular reflex function, thereby impairing the normal functioning of the vestibular system.^[23] This mechanism provides a potential physiological explanation for how insomnia influences BPPV recurrence, suggesting that improving sleep quality may be important for preventing recurrence. Therefore, clinicians should actively educate patients about good sleep hygiene habits, recommend maintaining regular sleep schedules, avoiding caffeine and alcohol, and creating a quiet, comfortable, and well-lit sleep environment. Additionally, for patients with significant insomnia symptoms, cognitive behavioral therapy is recommended to help patients identify and adjust negative cognitive and behavioral patterns that may contribute to sleep disturbances, thereby improving sleep quality from the root cause.^[24] These comprehensive intervention measures not only help improve patients’ overall quality of life but may also reduce the risk of BPPV recurrence.

This study indicates that hypertension is a significant risk factor for recurrence in BPPV patients. In elderly patients with hypertension, prolonged elevated blood pressure can cause vascular endothelial damage, promote the formation of atherosclerosis, and thereby impair microcirculatory blood flow to the inner ear.^[25] Insufficient inner ear blood flow may result in abnormal otolith metabolism and abnormal detachment of free otoliths, thereby increasing the risk of BPPV recurrence.^[26] Additionally, vascular lesions caused by hypertension may weaken the functional stability of the vestibular organs, further exacerbating disease susceptibility. Therefore, for elderly patients with hypertension, in addition to actively controlling blood pressure, comprehensive intervention measures should be adopted, including reasonable control of dietary intake, strict prohibition of smoking and drinking, and avoidance of excessive consumption of stimulating beverages such as strong tea. At the same time, psychological and behavioral adjustments are equally important, such as maintaining psychological balance and emotional stability through relaxation training and emotional management. Only through comprehensive management can the negative impact of hypertension on inner ear function be effectively reduced, thereby lowering the risk of BPPV recurrence. In clinical practice, health education for hypertensive patients should be strengthened, emphasizing the importance of blood pressure monitoring and proper medication use, combined with individualized lifestyle interventions to maximize patient outcomes and reduce the likelihood of recurrence.

The results of this study indicate that hyperlipidemia is an important risk factor influencing BPPV recurrence. Elderly patients with long-standing dyslipidemia are prone to the formation of atherosclerosis, which can lead to narrowing and thrombosis of inner ear microvessels, particularly small arteries, resulting in impaired inner ear microcirculation.^[27] This ischemic and hypoxic state of the vestibular organs can trigger metabolic disorders and abnormal detachment of otoliths in the utricle, significantly increasing the risk of disease recurrence.^[28,29] For patients with hyperlipidemia, especially the elderly, clinical management should prioritize effective control of lipid levels. Increasing intake of foods rich in vitamins and dietary fiber, such as fresh fruits and vegetables, whole grains, and nuts, can help improve lipid metabolism. Combined with moderate physical exercise, this can promote lipid metabolism balance, lower lipid levels, improve vascular elasticity, and thereby optimize the blood supply environment to the inner ear. Furthermore, health education should be incorporated into management plans to help patients understand the potential impact of hyperlipidemia on vestibular function and motivate them to adhere to a scientific diet and regular exercise regimen to reduce the risk of recurrence. Future studies should further explore the specific mechanisms underlying the association between lipid metabolism abnormalities and BPPV recurrence to provide theoretical basis for precise treatment.

This study shows that diabetes is a significant risk factor for recurrence in BPPV patients. Previous studies have indicated that diabetic patients are more prone to BPPV recurrence than non-diabetic patients.^[30] Long-term hyperglycemia can lead to microvascular lesions and peripheral neuropathy, resulting in insufficient blood supply to the inner ear terminal vessels and impairing vestibular system function.^[31] Additionally, elevated blood glucose concentrations may penetrate the endolymph, altering its acid–base balance (pH), thereby affecting the solubility of otoliths in the endolymph. This can disrupt the metabolism of otoliths in the utricle and cause abnormal detachment, ultimately increasing the risk of BPPV recurrence.^[32,33] Given this, clinical management for elderly patients with diabetes should prioritize strict control of blood glucose levels. Patients are advised to avoid high-sugar diets, adopt a diet pattern of small, frequent meals, and take hypoglycemic medications regularly as prescribed to maintain stable blood glucose levels. Additionally, regular monitoring of blood glucose and glycated hemoglobin levels is essential, with timely adjustments to treatment regimens to prevent disease progression and recurrence. Psychological counseling and lifestyle guidance should also be emphasized to enhance patients’ self-management capabilities and reduce the risk of recurrence. In summary, proper blood glucose management plays an important role in the comprehensive treatment of BPPV patients and provides a key guarantee for reducing recurrence. Future studies should further explore the pathological mechanisms of diabetes and BPPV recurrence to develop more precise prevention and treatment strategies.

The results of this study indicate that depression is an important risk factor for recurrence in BPPV patients. Previous studies have suggested that prolonged depression may lead to vasospasm of the inner ear capillaries, resulting in reduced local blood supply and impaired normal function of the cochlea and vestibular system.^[34] Additionally, patients with depression exhibit significantly reduced secretion of mucus by local cells in the inner ear, with the mucus becoming thinner and less adhesive, making it difficult to effectively fix otoliths.^[35] This alteration increases the risk of otolith particle displacement, thereby triggering the onset and recurrence of BPPV.^[36] For patients with depression, psychological intervention is particularly critical, especially cognitive behavioral therapy, which has been proven effective in alleviating depressive symptoms, improving patients’ quality of life, and enhancing their self-regulation abilities. Additionally, depending on the severity of depression, appropriate antidepressant medications should be selected and administered to assist patients in restoring their mental health.^[37] The implementation of multidisciplinary collaboration and integrated psychological and medical treatment plans may help reduce the recurrence rate of BPPV patients. In summary, depression is not only a psychological disorder but may also indirectly exacerbate the risk of BPPV recurrence by affecting the microcirculation and mucus secretion of the inner ear. Therefore, clinical treatment should strengthen the assessment and management of patients’ psychological status to achieve comprehensive physical and mental rehabilitation.

However, this study has several limitations that should be noted. First, due to the retrospective study design, inevitable errors such as incomplete information collection or data bias may have affected the accuracy of the results. Second, the study was conducted in a single tertiary hospital, and most patients were hospitalized with severe or complex conditions, limiting the representativeness of the sample and potentially introducing selection bias, which may affect the generalizability and extrapolability of the results. Finally, the risk prediction model developed in this study was based solely on data from a single center and has not undergone independent external validation from different regions or multiple centers. Therefore, its general applicability and reliability require further confirmation through prospective studies involving multiple centers and large samples. Future studies should focus on addressing these limitations by incorporating multicenter collaboration and prospective design, expanding sample sources, and improving data quality to enhance the model’s robustness and clinical applicability, thereby providing more precise and reliable evidence for risk assessment and management of patients with BPPV.

5. Conclusions

The results of this study indicate that factors such as insomnia, depression, hypertension, diabetes, and hyperlipidemia significantly increase the risk of recurrence in BPPV patients. Based on these independent risk factors, we developed a nomogram model for predicting the risk of BPPV recurrence. This model not only demonstrates high predictive accuracy and stability but also provides clinicians with a convenient tool for early identification and intervention management of patients at high risk of recurrence, thereby effectively reducing recurrence rates and improving patients’ quality of life and treatment outcomes. This predictive model holds significant application value in guiding individualized clinical decision-making and provides scientific evidence and practical support for the prevention and treatment of BPPV in the future.

Acknowledgments

The author would like to thank Deng Guanghua for his guidance in analyzing the data and submitting this thesis.

Author contributions

Conceptualization: Ping Wang.

Data curation: Ping Wang.

Formal analysis: Ping Wang.

Investigation: Ping Wang.

Methodology: Ping Wang.

Software: Ping Wang.

Validation: Ping Wang.

Visualization: Ping Wang.

Writing – original draft: Ping Wang.

Writing – review & editing: Ping Wang.