Association Between Ambroxol at the Usual Dose and the Risk of Parkinson’s Disease in Chronic Lung Disease

Dear Editor,

Mutation of the glucocerebrosidase (GBA) gene that causes the lysosomal storage disorder called Gaucher disease is known to be the most common genetic risk factor for Parkinson’s disease (PD), reportedly being present in 5%–30% of PD patients.¹ The GBA gene encodes the lysosomal enzyme glucocerebrosidase (GCase). GCase activity was negatively correlated with the α-synuclein level, and α-synuclein accumulation has been found to be increased in both GBA cell and animal models.² The GCase activity was significantly decreased in the PD brain with GBA gene mutation and in the idiopathic PD brain, suggesting that its activity is closely associated with the pathogenesis of PD.

Ambroxol is a commercially available expectorant that has been suggested to be a disease-modifying drug for PD. Several studies have demonstrated that ambroxol increases GCase activity and decreases α-synuclein levels, and these findings were also demonstrated in PD irrespective of the presence of GBA gene mutation.² High-dose ambroxol was effective in halting the progression of Gaucher disease in case studies.³,⁴,⁵

While other studies have found ambroxol to be safe when administered at high doses in humans, the safety of long-term high-dose administration still needs to be confirmed. Moreover, there has been no investigation of the effects of administering ambroxol within the approved dose range in the real world. A high dose is defined as a dose that exceeds the permitted dose, and usually higher-than-permitted doses have been administered in previous PD clinical studies (e.g., 1.26 g/day).² The usual dose is that permitted in clinical practice (e.g., 90 mg/day according to the Korea Pharmaceutical Information Center at https://www.health.kr/searchDrug/result_drug.asp?drug_cd=2009071505405). We thus aimed to investigate the association between ambroxol at the usual dose and the risk of Parkinson’s disease in chronic lung disease (CLD) based on the nationwide Korean National Health Insurance Database (NHID).

We analyzed data extracted from nationwide population-based cohort data from the National Health Insurance Service (NHIS) of the Republic of Korea. The NHIS is a mandatory social health insurance service covering about 97% of the entire Korean population. The characteristics of the data source have been described in detail elsewhere.⁶ The Institutional Review Board of Hallym University Dongtan Sacred Hospital approved the study protocol. The study was exempted from needing to obtain informed consents because the data were anonymous and deidentified. All procedures were performed in accordance with the Declaration of Helsinki and followed relevant guidelines and regulations.

The study population and analysis methods are described in detail as supplementary data (Supplementary Materials, Supplementary Fig. 1, and Supplementary Table 1 in the online-only Data Supplement). The demographics of the CLD patients who were and were not ambroxol users are summarized in Supplementary Tables 2 and 3 (in the online-only Data Supplement). The Charlson Comorbidity Index was significantly higher in ambroxol users (p<0.001).

The risk of PD was significantly higher in CLD patients treated with ambroxol than in ambroxol nonusers (adjusted hazard ratio [aHR]=1.383, 95% confidence interval [CI]=1.272–1.504, p<0.0001; Supplementary Table 4 in the online-only Data Supplement). Kaplan-Meier analysis showed that the cumulative incidence of PD was significantly higher in ambroxol users during a 16-year follow-up (p<0.001 in log-rank test; Fig. 1). An analysis of the dose–incidence relationship showed that the risk of PD was highest in ambroxol users with cumulative defined daily doses >15, Supplementary Materials (in the online-only Data Supplement) (aHR=1.624, 95% CI=1.431–1.843, p<0.0001; Supplementary Table 5 in the online-only Data Supplement).

Fig. 1. Kaplan-Meier curves of the freedom from PD for ambroxol users and nonusers. The number of patients at risk over time are shown (p<0.001 by log-rank test). PD, Parkinson’s disease.

This study found that ambroxol users had a significantly higher PD risk than did ambroxol nonusers. It is possible that the usual dose of ambroxol does not produce any significant effect. As mentioned above, it has been suggested that high-dose ambroxol is effective at modifying PD progression. A previous investigation of nonhuman primates found that GCase activity was increased for ambroxol treatment at 100 mg/day but not at 22.5 mg/day.⁷ In transgenic mice, GCase activity increased significantly in a high-dose administration group (ambroxol at a concentration of 4 mM).⁸ However, since GCase activity was not increased at the highest dose tested (ambroxol at a concentration of 5 mM), it is not yet known whether higher doses provide more-effective treatments.⁸

Our study found that the incidence of PD was higher in ambroxol users, which is contrary to previous findings. Although adjusted analyses were performed, differences in the number of comorbidities between ambroxol users and nonusers may have been a bias in this study since numerous comorbidities associated with PD have been reported to increase the PD risk.⁹ We conducted several sensitivity analyses. It is generally better to maintain the balance of baseline characteristics between two groups by applying propensity-score analysis rather than multivariate analysis (traditional covariate adjustment). However, we also conducted a propensity-score-matching analysis, which produced similar results (Supplementary Tables 6 and 7 in the online-only Data Supplement). Conventional methods for survival analysis ignoring competing events, such as the Kaplan-Meier method and standard Cox proportional-hazards regression, may be inappropriate in the presence of competing risks, when alternative methods specifically designed for analyzing competing-risks data should be applied. Since death is a common competing risk, we also used a subdistribution proportional-hazards model with a competing risk to calculate the risks of PD, which yielded the same results (Supplementary Table 8 in the online-only Data Supplement). We also performed a sensitivity analysis including patients prescribed ambroxol at 1 year after a CLD diagnosis, which also produced similar results (aHR=1.41, 95% CI=1.31–1.52, p<0.0001).

This study had some limitations. First, the NHID contains data on which drugs were prescribed, which might differ from what patients actually took. Second, confounders such as smoking, alcohol consumption, physical activity, and obesity were not adjusted in this study, and so these factors might have exerted confounding effects. Third, there may have been differences in disease severity, comorbidities, and symptoms between CLD patients who were and were not prescribed ambroxol, which would have influenced the results. Fourth, we did not check that all of the study subjects took ambroxol at the usual dose. However, in clinical practice, it is unlikely that this drug would be prescribed at a dose higher than the approved dose. Fifth, this study involved patients with CLD, and so the results might not apply to individuals without CLD.

In conclusion, the usual dose of ambroxol was associated with an increased risk of PD. Further studies are needed that investigate the risk of PD incidence while more properly controlling for comorbidity.

Availability of Data and Material

All data generated or analyzed during the study are included in this published article (and its supplementary information files).

Supplementary Materials

The online-only Data Supplement is available with this article at https://doi.org/10.3988/jcn.2023.0172.

Supplementary Table 1

ICD-10 codes of medical comorbidities

Supplementary Table 2

Characteristics of the study population based on the ambroxol treatment

Supplementary Table 3

Detailed information on ambroxol use (n=73,421)

Supplementary Table 4

Crude and adjusted HRs for PD among patients with newly diagnosed chronic lung disease with treatment of ambroxol

Supplementary Table 5

Crude and adjusted HRs for PD for subgroups treated with different cDDDs of ambroxol

Supplementary Table 6

Characteristics of the study population based on the ambroxol treatment (propensity score matching by Greedy method)

Supplementary Table 7

Crude HRs for PD among patients with newly diagnosed chronic lung disease with treatment of ambroxol (propensity score matching by Greedy method)

Supplementary Table 8

Proportional subdistribution hazards model by Fine and Grey* for PD among patients with newly diagnosed chronic lung disease with treatment of ambroxol (competing risk analysis)

Supplementary Fig. 1

Flow chart of the study population.