INTRODUCTION

Olfactory dysfunction possesses great potential as an early biomarker and diagnostic tool for Parkinson’s disease (PD), as it often appears years before motor symptoms do [1-3]. The Movement Disorder Society (MDS) has included olfactory loss as a supportive criterion for PD diagnosis due to its high sensitivity and specificity of over 80% [4]. Furthermore, olfaction can be a valuable biomarker for differential diagnosis, as it is differentially impaired in various Parkinsonian syndromes [1,2].

Drug-induced parkinsonism (DIP), brought by the use of drugs that can either block dopamine receptors or affect alternative pathways, ranks as the second most prevalent cause of parkinsonism [5]. However, differentiating “pure” DIP from PD without the aid of functional images measuring dopamine deficiency presents a considerable challenge. Earlier studies have suggested that olfactory impairment could aid in distinguishing between PD and DIP [6-13]. Nevertheless, insufficient evidence exists to thoroughly appraise the specific characteristics of olfactory dysfunction between these two conditions, particularly in the domains of olfactory threshold, discrimination, and identification (TDI), given the natural decline in olfactory function with age. In this respect, this study aims to assess the difference in olfactory profiles in PD and DIP specifically focusing on the olfactory TDI using various scents and calculating subscores.

MATERIALS & METHODS

Participants

We conducted a retrospective review of the Parkinson registry database at Hallym University Medical Center, which includes newly diagnosed patients with parkinsonism who consecutively visited the movement disorder outpatient clinic. We collected data, including demographics and clinical features, from 109 patients who were diagnosed with de novo PD or DIP and underwent an olfactory function test between August 2020 and December 2022. PD patients were in dopaminergic drug-naïve states and diagnosed based on the UK Parkinson’s Disease Society Brain Bank clinical diagnostic criteria [14]. DIP patients met specific criteria, including 1) the presence of at least two of four cardinal signs of parkinsonism (tremor, rigidity, bradykinesia, and impaired postural reflexes); 2) no history of parkinsonism or extrapyramidal disorders before use of the offending drug; and 3) onset of parkinsonism symptoms during treatment with an offending drug [5]. Additionally, all participants underwent ¹⁸F-N-(3-fluoropropyl)-2β-carboxymethoxy-3β-(4-iodophenyl) nortropane (¹⁸F-FP-CIT) positron emission tomography (PET) to confirm the diagnosis of PD or DIP and excluded DIP patients with abnormal dopamine transporter (DAT) imaging. We also excluded patients with 1) a history of rhinological disorders or surgery; 2) abnormal brain magnetic resonance imaging images with the presence of significant vascular lesions, brain contusion, and atrophy; and 3) dementia, as established by neuropsychological tests.

This study was approved by the institutional review board of Hallym University Medical Center (IRB 2023-05-018). Informed consent was waived due to the retrospective nature of the study, and all evaluations were conducted in accordance with the ethical standards of the institutional committee, as well as the 1964 Helsinki Declaration and its later amendments or comparable ethical standards.

Clinical assessment

All demographic data, including age, sex, underlying disease, drug history, education years, and duration of parkinsonism, were collected from all patients. Clinical evaluations were performed at the time of diagnosis. Motor and nonmotor symptoms were estimated by the Unified Parkinson’s Disease Rating Scale (UPDRS) and modified Hoehn and Yahr (H&Y) stage score. Cognitive functions were evaluated with the Korean version of the Mini-Mental State Examination-2 (K-MMSE-2) and Seoul Neuropsychological Battery (SNSB) [15], which contains tests for attention, language, visuospatial, memory, and frontal/executive function. Standardized z scores were derived using age- and education-matched norms for all tests with scores.

To assess the sense of smell in patients, we utilized the YSK olfactory function test (YOF test; Kimex Co., Suwon, Korea) to evaluate the threshold, discrimination, and identification aspects of olfactory function, known as TDI. The YOF test is a well-validated olfactory method that utilizes safe chemicals, specifically phenyl-ethyl alcohol (PEA), and uses culturally familiar odorants to Koreans [16,17]. In detail, the threshold test consists of 12 steps of PEA concentration, and the average of steps at which the subject recognizes PEA, in compliance with the internal test regulations, represents the olfactory threshold score. The discrimination and identification tests include 12 subtests each, encompassing major chemical functional groups such as ketone, terpene, aldehyde, aromatic, alcohol, ester, acid, and amine. The discrimination test assesses the ability to differentiate between different odors, while the identification test measures the ability to correctly name the odor, going beyond mere recognition of the odor stimulus. Each test is scored on a scale of 0 to 12, and the combined scores from the three domains can range from 0 to 36. Scores of 0–14.5 suggest anosmia or loss of smell, while scores between 14.5 and 21 indicate hyposmia or a reduced sense of smell. Scores above 21 are considered normal [16]. In this study, we defined anosmia or hyposmia using the YOF test score as olfactory dysfunction.

Statistical analysis

Categorical variables are presented as absolute frequencies (proportions), while continuous variables are expressed as the mean ± standard deviation (SD). Group differences between patients with PD and DIP were assessed using the Mann‒Whitney U test for continuous variables and chi-squared or Fisher’s exact test for categorical variables. To adjust for heterogeneity between the two groups, propensity score matching (PSM) was conducted. Propensity scores were estimated as probabilities of individuals having propensity scores based on the given covariates, which were sex and age as common variables. The matching method involved performing a 1:1 matching between the PD group and DIP group using nearest neighbor matching, with the closest individuals from the control group identified based on the calculated propensity scores. The test for differences in sociodemographic characteristics between the two groups was conducted after PSM. Post hoc analysis was performed using the Holm method for multiple comparisons. In addition, to assess the diagnostic performance of the TDI test, we conducted receiver operating characteristic (ROC) curve analysis. The significance threshold was set at p < 0.05. Statistical analysis was performed with SPSS version 27.0 (IBM Corp., Armonk, NY, USA) and R version 4.3.0 (https://cran.r-project.org/bin/windows/base/old/4.3.0/).

RESULTS

Patient characteristics

A total of 78 patients with PD and 31 patients with DIP were enrolled in the study. Among the DIP patients, the most commonly implicated drug was levosulpiride (n = 15, 48.4%), and 5 patients had two or more offending drugs (all drugs are listed in Supplementary Table 1 in the online-only Data Supplement). Table 1 presents the demographic characteristics of both groups. The DIP group exhibited a significantly higher age than the PD group (p < 0.001) and a significant female predominance (p = 0.001). However, there was no difference in the duration between the onset of parkinsonism and diagnosis between the groups. Patients with DIP had a lower level of education than those with PD (p < 0.001) and showed poorer global cognition (K-MMSE-2), despite the exclusion of dementia. In the SNSB, across all cognitive domains except for memory and attention, individuals with DIP performed worse than those diagnosed with PD (refer to Supplementary Table 2 in the online-only Data Supplement). The DIP group had higher UPDRS scores than the PD group (p < 0.001), while the H&Y stage did not show significant differences. Additionally, diabetes was more prevalent among patients with DIP than among those with PD (p = 0.022).

Comparisons of each domain and item of the YOF test

Table 2 presents a comparison of the results from the olfactory function test between the groups. Olfactory dysfunction, characterized by a state of hyposmia or anosmia (YOF total score ≤ 21), was observed in 70.5% of the PD group and 77.4% of the DIP group, without statistical significance in prevalence (p = 0.466).

When examining each domain of the TDI triad, there were no significant differences in the YOF total score, the threshold test score, and the identification score between the two groups. However, the discrimination test scores were significantly lower in the PD group than in the DIP group.

Comparison between the two groups after PSM

Upon implementing age and sex matching through the PSM approach, we successfully paired a subset of 26 individuals from the DIP cohort with their counterparts in the PD group (Table 3). In addition to age and sex, other variables, including disease duration, underlying disease, H&Y, UPDRS parts II and III, years of education, and K-MMSE-2 scores, did not exhibit any significant differences between the cohorts, while UPDRS Part I and total score were different between the groups. As shown in the analysis of all patients (Table 2), the total scores for the YOF test, the threshold test scores, and identification test scores exhibited no significant differences between the two groups, while in terms of the discrimination tests, there was a notable and statistically significant underperformance among the PD patients when compared to their DIP counterparts (p = 0.002). Additionally, the area under the ROC curve obtained from the discrimination test scores was 0.738 (95% confidence interval: 0.602–0.875, p = 0.003) in distinguishing between PD and DIP.

Detailed characteristics of discrimination and identification

Regarding the discrimination test, patients with PD tend to have trouble distinguishing between scents of styrallyl acetate and para-cresol methyl ether, as indicated in Supplementary Table 3 in the online-only Data Supplement. In the identification test, PD patients had difficulty correctly identifying the scents of spearmint and scorched rice compared to DIP patients, despite these smells being familiar to Koreans, as indicated in Supplementary Table 4 in the online-only Data Supplement. However, the post hoc analysis did not yield statistically significant differences, as indicated in each supplementary table. This is noteworthy considering that in the standardization process using the YOF test, over 90% of the general population answered these two questions correctly (refer to Supplementary Table 4 in the online-only Data Supplement) [16,17].

DISCUSSION

This study demonstrated that impaired ability to discriminate odors is a key characteristic that distinguishes PD from DIP, whereas the olfactory threshold, identification and overall severity of olfactory dysfunction do not differ significantly. Despite both conditions showing olfactory dysfunction in more than 70% of patients, the PD group exhibited lower discrimination scores than the DIP group, even after accounting for variables such as age, sex, diabetes, disease duration, and cognitive function.

Distinguishing between PD and DIP poses a considerable challenge in clinical practice. Unlike other movement disorders, the diagnostic criteria for DIP are not well established, and there is ongoing debate regarding the time it takes for parkinsonism to recover after discontinuing the causative drug [5]. Inaccurate patient information, along with possible pathophysiological factors, can contribute to a misdiagnosis and subsequent reclassification of PD and DIP cases [5,12]. The intricate interaction between the effects of the causative drugs and preexisting neurodegenerative changes in the dopamine system may contribute to the diagnostic complexities encountered in differentiating DIP from PD in clinical practice.

Previous studies have evaluated clinical features, laboratory findings, and imaging to differentiate DIP and PD [6-13]. Some phenotypes, such as the presence or absence of motor and nonmotor symptoms, motor symmetricity, and response to levodopa, have been highlighted to aid in the differential diagnosis of PD and DIP [5-7]. However, phenotypic frequencies also differ among studies, with many patients deviating from these typical patterns.

Several studies have suggested that olfactory impairment can play a role in distinguishing between PD and DIP [8-13]. Regarding the detailed features of olfactory function, Bovi et al. [9] found that 13 PD patients exhibited lower olfactory functions across all TDI domains compared to 9 DIP patients in similar age brackets utilizing Sniffin’ Sticks. However, the majority of the studies primarily focused on the identification test within TDI domains.

Elderly individuals are particularly susceptible to developing DIP, which contributes to the higher average age of symptom onset in DIP patients compared to PD patients [5]. It needs to be considered that the normal aging process can result in declining olfactory acuity ranging from subtle changes to severe decrement, which can contribute to an olfactory disturbance in DIP [18,19]. Unlike previous research, our study revealed that the DIP group exhibited a notable occurrence of olfactory dysfunction, comparable to what was observed in the PD group, suggesting that relying solely on overall olfactory decline makes it challenging to distinguish between PD and DIP in a real-world clinical context.

This study showed that PD patients showed impaired olfactory discrimination compared to DIP patients. The differing patterns of olfactory dysfunction in PD and DIP observed in this study may arise from the underlying pathophysiology of the diseases and the anatomical substrate of the olfactory TDI domains. Although the neural processes involved in these olfactory TDI are not fully understood, the anatomical substrate for olfactory function comprises a complex network connecting various brain regions, including the olfactory epithelium, bulb, olfactory cortex, limbic system, and hippocampus [18,19].

The olfactory threshold, which refers to the minimum concentration of an odorant that can be detected, primarily involves the peripheral olfactory system, specifically the olfactory receptor neurons located in the olfactory epithelium [18-20]. These neurons send information to the olfactory bulb, which processes and transmits olfactory signals to other brain regions. On the other hand, higher-order processing of olfactory information, such as discrimination and identification, occurs more in the olfactory cortex, which includes regions such as the piriform cortex and orbitofrontal cortex, in addition to the peripheral olfactory system. The piriform cortex is involved in encoding the unique characteristics of different odors for discrimination, while the orbitofrontal cortex is responsible for associating perceived odors with specific memories or labels. The limbic system, which is involved in emotion and memory, and the hippocampus, which is critical for learning and memory, also play roles in olfactory processing. The amygdala is thought to be involved in emotional responses to odors influencing the perception of odors, while the hippocampus is implicated in the formation of olfactory memories, including odor identification.

Odors that stimulate the olfactory receptors on the neuroepithelium are composed of organic compounds with varying combinations of chemical structures [21]. Odors with closely related molecular structures can exhibit overlapping chemical and psychophysical properties, making it challenging to differentiate them from odors with significantly distinct structures. This similarity in configuration suggests that these odors are likely to be recognized by similar sets of odorant receptors. Additionally, different regions of the mammalian nose detect various odors [22]. The complexity in odorant structure and receptor distribution may contribute to the selective olfactory deficit observed in PD [23].

Selective olfactory dysfunction in PD, specifically for discriminating or identifying certain odorants, has been reported [23]. However, the nature of this selective olfactory dysfunction varies across studies due to the utilization of different odorants and diverse methodologies. Therefore, further efforts are needed to address the question of whether there is a selective olfactory deficit in PD and whether this abnormality is strongly related to the structural characteristics of odorants and olfactory receptors in PD.

In this study, the findings indicate that olfactory dysfunction in PD is closely related to both the central olfactory pathway and peripheral olfactory system. The precise cause of olfactory dysfunction in PD is not fully understood but could be linked to neuropathological alterations within the olfactory system, spanning from the olfactory bulb to olfaction-related cortical structures [1-3,24-28]. On the other hand, the observation that the peripheral olfactory system, specifically the olfactory threshold, is primarily affected in DIP in this study aligns with the understanding that the olfactory threshold is typically the first aspect of olfactory function to be affected, followed by discrimination and then identification in the normal aging process [18]. This can explain the results showing an impaired odor threshold in both groups without significant differences but significantly more diminished discrimination in the PD group than in the DIP group, regardless of age.

Moreover, several studies have revealed olfactory impairment in DIP patients when compared to age-matched healthy controls. One study identified olfactory impairment in DIP patients with depression [29]. Another study, even in cases of “pure DIP” distinguished by DAT imaging, observed a decrease in odor threshold in DIP patients compared to age-matched healthy controls, resulting in lower total TDI scores [9]. The findings suggest that in addition to the expected effects of aging, there may be other factors, such as diabetes, contributing to the development of olfactory dysfunction in DIP patients. Further research is necessary to investigate the underlying reasons for olfactory impairment in DIP.

In this study, we adopted the YOF test using culturally familiar odorants, which encompasses the assessment of all TDI components, to evaluate the various facets of olfactory function. In addition to subjective self-assessment (e.g., nonmotor symptoms scale for PD), there are various kinds of psychophysical olfactory tests widely used in a quite objective way to evaluate olfactory function by scoring a subject [16,17,21]. These tests typically encompass one or more domains of the TDI triad. Because the subject’s previous exposure to odor significantly affects the psychophysical olfactory tests, there have been many attempts to modify widely used olfactory function tests to localized tests reflecting the cultural experience of each country. For this reason, we used the YOF test, a newly developed olfactory function test to assess the olfactory TDI domain, using a safe chemical, PEA, and to reselect culturally familiar odorants to Koreans for upgrading to the correct answer rate [16,17].

This study has a limitation in that it lacks control groups for comparison, such as patients with PD combined with DIP patients or healthy individuals without parkinsonism, because of the retrospective nature of this study. As a result, it is difficult to determine whether PD with DIP exhibits distinct olfactory features compared to PD alone, which partly hinders the interpretation of the study’s findings. Additionally, we did not provide a clear explanation for the remarkably high occurrence of hyposmia in the DIP group. Notably, one of the patients in our study experienced a progressive decline in cognitive dysfunction and was subsequently diagnosed with Alzheimer’s disease. Hence, further research is required to observe whether the cause of hyposmia in the pure DIP population is linked to future neurodegenerative disease or other factors. Furthermore, a limitation of this study lies in the absence of a comparison involving agematched groups. Given the higher average age of onset in DIP, age-related olfactory dysfunction might have exerted a more pronounced influence. To overcome this limitation, we conducted PSM analysis between the two groups. In addition, the issue of multiple comparisons arises when examining various dependent variables in each TDI domain across two groups, which can diminish the discriminating power. Last, it is important to note that the absence of quantitative analyses on DAT images constitutes a limitation in this study. This omission could potentially result in the inclusion of patients with subtle nigrostriatal degeneration, hinting at a prodromal stage of synucleinopathy.

In conclusion, to the best of our knowledge, this is the first study to compare olfactory function between patients with PD and DIP through the utilization of the YOF test. Our primary discovery indicated no significant variations in the YOF total scores, the threshold test scores, and the identification scores between the two groups, but patients with PD exhibited decreased scores in the discrimination domain compared to those with DIP. These results imply that the PD and DIP groups showed different olfactory profiles, even though typical aging or other factors can influence them. Further research is warranted to explore the underlying mechanisms and potential clinical implications of these olfactory differences.

Supplementary Material

Notes

Conflicts of Interest

The authors have no financial conflicts of interest.

Funding Statement

This research was supported by a grant of the Korea Health Technology R&D Project through the Korea Health Industry Development Institute (KHIDI), funded by the Ministry of Health & Welfare, Republic of Korea (grant number: RS-2023-00265159), and Basic Science Research Program through the National Research Foundation of Korea (NRF) funded by the Ministry of Education (RS-2023-00246655).

Author contributions

Conceptualization: In Hee Kwak, Jeongjae Lee. Data curation: In Hee Kwak, Suk Yun Kang. Formal analysis: In Hee Kwak. Funding acquisition: Young Eun Kim. Investigation: In Hee Kwak, Joong Seob Lee. Methodology: In Hee Kwak, Dong A Yea. Supervision: Hyeo-il Ma, Young Eun Kim, Joong Seob Lee. Writing—original draft: In Hee Kwak, Min Seung Kim. Writing—review & editing: Hyeo-il Ma, Young Eun Kim, Suk Yun Kang.

Acknowledgments

We extend our heartfelt thanks to researcher Sion Lee from Medical Statistics Office at Hallym University Medical Center. Her invaluable contribution has been instrumental in advancing our research.

REFERENCES

• 1. Doty RL. Olfactory dysfunction in Parkinson disease. Nat Rev Neurol 2012;8:329–339.ArticlePubMedPDF
• 2. Haehner A, Hummel T, Reichmann H. Olfactory loss in Parkinson’s disease. Parkinsons Dis 2011;2011:450939.ArticlePubMedPMCPDF
• 3. Fullard ME, Morley JF, Duda JE. Olfactory dysfunction as an early biomarker in Parkinson’s disease. Neurosci Bull 2017;33:515–525.ArticlePubMedPMCPDF
• 4. Postuma RB, Berg D, Stern M, Poewe W, Olanow CW, Oertel W, et al. MDS clinical diagnostic criteria for Parkinson’s disease. Mov Disord 2015;30:1591–1601.ArticlePubMed
• 5. Shin HW, Chung SJ. Drug-induced parkinsonism. J Clin Neurol 2012;8:15–21.ArticlePubMedPMC
• 6. Brigo F, Erro R, Marangi A, Bhatia K, Tinazzi M. Differentiating drug-induced parkinsonism from Parkinson’s disease: an update on non-motor symptoms and investigations. Parkinsonism Relat Disord 2014;20:808–814.ArticlePubMed
• 7. Morley JF, Duda JE. Use of hyposmia and other non-motor symptoms to distinguish between drug-induced parkinsonism and Parkinson’s disease. J Parkinsons Dis 2014;4:169–173.ArticlePubMed
• 8. Lee PH, Yeo SH, Yong SW, Kim YJ. Odour identification test and its relation to cardiac 123I-metaiodobenzylguanidine in patients with drug induced parkinsonism. J Neurol Neurosurg Psychiatry 2007;78:1250–1252.ArticlePubMedPMC
• 9. Bovi T, Antonini A, Ottaviani S, Antonioli A, Cecchini MP, Di Francesco V, et al. The status of olfactory function and the striatal dopaminergic system in drug-induced parkinsonism. J Neurol 2010;257:1882–1889.ArticlePubMedPDF
• 10. Kim JS, Youn J, Shin H, Cho JW. Nonmotor symptoms in drug-induced parkinsonism and drug-naïve Parkinson disease. Can J Neurol Sci 2013;40:36–41.ArticlePubMed
• 11. Morley JF, Cheng G, Dubroff JG, Wood S, Wilkinson JR, Duda JE. Olfactory impairment predicts underlying dopaminergic deficit in presumed drug-induced parkinsonism. Mov Disord Clin Pract 2016;4:603–606.ArticlePubMedPMCPDF
• 12. Lee SH, Kim HK, Lee YG, Lyoo CH, Ahn SJ, Lee MS. Clinical features indicating nigrostriatal dopaminergic degeneration in drug-induced parkinsonism. J Mov Disord 2017;10:35–39.ArticlePubMedPMCPDF
• 13. Morley JF, Pawlowski SM, Kesari A, Maina I, Pantelyat A, Duda JE. Motor and non-motor features of Parkinson’s disease that predict persistent drug-induced parkinsonism. Parkinsonism Relat Disord 2014;20:738–742.ArticlePubMed
• 14. Gelb DJ, Oliver E, Gilman S. Diagnostic criteria for Parkinson disease. Arch Neurol 1999;56:33–39.ArticlePubMed
• 15. Kang Y, Jahng S, Na DL. Seoul neuropsychological screening battery, 2nd edition (SNSB-II). Seoul: Human Brain Research & Consulting Co; 2012.
• 16. Ha JG, Kim J, Nam JS, Park JJ, Cho HJ, Yoon JH, et al. Development of a Korean culture-friendly olfactory function test and optimization of a diagnostic cutoff value. Clin Exp Otorhinolaryngol 2020;13:274–284.ArticlePubMedPMCPDF
• 17. Cho HJ, Ha JG, Kim CH. The YSK olfactory function test: development of a new Korean olfactory test. J Rhinol 2022;29:61–68.ArticlePDF
• 18. Kondo K, Kikuta S, Ueha R, Suzukawa K, Yamasoba T. Age-related olfactory dysfunction: epidemiology, pathophysiology, and clinical management. Front Aging Neurosci 2020;12:208.ArticlePubMedPMC
• 19. Bhatia-Dey N, Heinbockel T. The olfactory system as marker of neurodegeneration in aging, neurological and neuropsychiatric disorders. Int J Environ Res Public Health 2021;18:6976.ArticlePubMedPMC
• 20. Su B, Bleier B, Wei Y, Wu D. Clinical implications of psychophysical olfactory testing: assessment, diagnosis, and treatment outcome. Front Neurosci 2021;15:646956.ArticlePubMedPMC
• 21. Hummel T, Whitcroft KL, Andrews P, Altundag A, Cinghi C, Costanzo RM, et al. Position paper on olfactory dysfunction. Rhinol Suppl 2017;54:1–30.ArticlePubMed
• 22. Tan L, Xie XS. A near-complete spatial map of olfactory receptors in the mouse main olfactory epithelium. Chem Senses 2018;43:427–432.ArticlePubMedPMC
• 23. Boesveldt S, Verbaan D, Knol DL, Visser M, van Rooden SM, van Hilten JJ, et al. A comparative study of odor identification and odor discrimination deficits in Parkinson’s disease. Mov Disord 2008;23:1984–1990.ArticlePubMed
• 24. Sui X, Zhou C, Li J, Chen L, Yang X, Li F. Hyposmia as a predictive marker of Parkinson’s disease: a systematic review and meta-analysis. Biomed Res Int 2019;2019:3753786.ArticlePubMedPMCPDF
• 25. Xiao Q, Chen S, Le W. Hyposmia: a possible biomarker of Parkinson’s disease. Neurosci Bull 2014;30:134–140.ArticlePubMedPMCPDF
• 26. Doty RL. Olfaction in Parkinson’s disease. Parkinsonism Relat Disord 2007;13(Suppl 3):S225–S228.ArticlePubMed
• 27. Kranick SM, Duda JE. Olfactory dysfunction in Parkinson’s disease. Neurosignals 2008;16:35–40.ArticlePubMedPDF
• 28. Hoyles K, Sharma JC. Olfactory loss as a supporting feature in the diagnosis of Parkinson’s disease: a pragmatic approach. J Neurol 2013;260:2951–2958.ArticlePubMedPDF
• 29. Krüger S, Haehner A, Thiem C, Hummel T. Neuroleptic-induced parkinsonism is associated with olfactory dysfunction. J Neurol 2008;255:1574–1579.ArticlePubMedPDF

Figure & Data

References

Citations

Citations to this article as recorded by  

Comments on this article