Stimulated by Hwang et al 2025.[1]
PD – Parkinson’s disease
IPD – idiopathic Parkinson’s disease
NHIS – national health insurance service
ICD – international classification of diseases
IRB – institutional review board
c – circa (from Latin and written before a number to mean it is an approximation)
PSM – propensity score matching
aHR – adjusted hazard ratio
CI – confidence interval
This is a large retrospective cohort study from Korea that examines the association between receiving acupuncture and subsequent mortality in patients newly diagnosed with IPD.
IPD has been increasing in Korea from just under 100k in 2016 to over 120k in 2022.[2] The population of Korea is just over 51 million and they are all on the NHIS database. This research used a ‘customised cohort’ provided by NHIS and judging by the numbers it looks to me as if it was taken from the whole database rather than from a smaller random sample as we have seen in the many of the large retrospective cohort studies from Taiwan.
New patients with IPD (ICD code G20) were identified over a 5-year period from the start of 2012 to the end of 2016. Interestingly, the study was exempted from IRB review because the data was anonymised. There were over 300k patients identified over the 3-year period, but just over 100k were excluded because their first diagnosis was prior to the inclusion dates of this study. A further c50k were excluded due to existing disability, leaving just over 150k in the cohort. The next group to be excluded were c60k who had received acupuncture within the previous 6 months prior to the index date (ie when G20 appeared on their record in the database).
To determine the cohort exposed to acupuncture following their G20 diagnosis, the patients needed to have at least 6 treatments within a year of diagnosis. Excluding those who had 5 or fewer sessions of acupuncture and those who developed dementia or died within the first year (total c50k) left 6400 exposed to 6 or more sessions and 27 846 who were not. PSM was applied and the result was 2 matched cohorts of 6394, one of which had received at least 6 sessions of acupuncture within a year of diagnosis and the other which had received no acupuncture.
The primary endpoint was mortality, and individuals were followed for 6 years from their index date. This endpoint reminds me of a potential form of bias in this sort of trial, which I first came across and wrote about on the blog in 2019 – Immortal time bias.
In this instance, the immortal time is from the index date to the 6th acupuncture session in the cohort who received acupuncture. During this period, it is impossible for anyone in this cohort to die, and hence they are immortal, or appear to be so due the selection parameters (inclusion criteria). I did not find any mention of immortal time in this paper, but that is likely to be because one of the exclusion criteria was death within one year of the index date. This means that both groups were essentially immortal for their first year, so any potential advantage in terms of immortal time related to having to receive 6 sessions of acupuncture was no longer relevant. There is something about the concept of immortal time that still makes me chuckle…
So, moving to the opposite, that is mortality, what was the result? There was a modest, but very highly significant, reduction in all-cause mortality in the cohort exposed to 6 or more sessions of acupuncture (aHR 0.887). A risk reduction of just over 11% does not seem like a huge effect, but I guess any reduction in mortality if preferable to the alternative. Interestingly, the major effect appeared to be in the cancer category (aHR 0.764), although there was also significantly reduced mortality in digestive and unclassified categories. In some categories the mean risk reduction was more dramatic, but due to low numbers the CIs were wide and so these were not statistically significant. Both mental health and ‘other’ categories had mean risk reductions of c50%.
A much smaller retrospective cohort study from Korea in IPD demonstrated a risk reduction of 50% in admission to hospital with pneumonia or sepsis associated with the addition of integrated medicine.[3] The latter included acupuncture in over 98% of patients.
This will now be the 7th blog post included in my PD category. Reading back over the others I see that I a not particularly convinced by the value of acupuncture in this condition as yet. Perhaps the most interesting blog describes a reduction in the risk of developing IPD (or PD) associated with acupuncture in patients with depression. Depression is associated with a 3 fold increased risk, and acupuncture appears to almost normalise this with a 61% risk reduction – see Acupuncture PD and depression.
References
1 Hwang Y-C, Lee J, Kang D, et al. A nationwide retrospective cohort study of the association between acupuncture exposure and clinical outcomes of idiopathic Parkinson’s disease using health insurance claim data in South Korea. Integr Med Res. 2025;14:101146. doi: 10.1016/j.imr.2025.101146
2 Jeon MJ, Lee J, Lee WH, et al. Prevalence of Parkinson disease in Korea: a comprehensive analysis and regional comparison. J Korean Soc Geriatr Neurosurg. 2023;19:56–61. doi: 10.51638/jksgn.2023.00052
3 Woo Y, Hyun MK. Effectiveness of Integrative Therapy for Parkinson’s Disease Management. Front Aging Neurosci. 2019;11:40. doi: 10.3389/fnagi.2019.00040
The BMAS Blog 
You must be logged in to post a comment.