Stimulated by Liu et al 2026.[1]
HT – hypertension
RCT – randomised controlled trial
IF – impact factor
BP – blood pressure
SBP – systolic BP
DBP – diastolic BP
XNKQ – Xing Nao Kai Qiao (needling method and formula for stroke)
HXSF – Huo Xue San Feng (needling method and formula for HT)– key to acronyms
This is a large (n=480) open prospective RCT with 2 parallel arms. It aims to see if acupuncture treatment following a first ischaemic stroke in patients with hypertension (HT) reduces the risk of stroke recurrence. It was published in Frontiers in Neurology (IF 2.8) at the beginning of April 2026.
Stroke comes up a lot in the large Retrospective Cohorts highlighted on this blog. To select a few of the most relevant, we have seen that acupuncture is associated with a reduced risk of stroke in patients with:
- fibromyalgia – see FM and stroke risk [2]
- migraine – see Stroke risk in migraine 2023 [3]
- rheumatoid arthritis – see Stroke risk in RA 2024 [4]
- insomnia – see Acupuncture and stroke risk in insomnia.[5]
Hypertension (HT) does not come up on this blog often. I made a few spikey comments on the Cochrane review from 2018 after reading it temporarily worsened my own HT – see Hypertension 2018. We have also seen a reduced risk of HT associated with acupuncture in patients with chronic spontaneous urticaria – see Hypertension risk in CSU 2023.
Back to the current paper from Tianjin, the geography of which I have mentioned before – see Acupuncture for PSA 2024. Here we have a prospective test of some of these associations that we have read about in the large retrospective cohorts. The latter show associations but do not provide evidence of a causative link – for that we need prospective research such as this.
The group from Tianjin recruited 480 patients with their first ever stroke and hypertension. The stroke had to be ischaemic and between 2- and 6-weeks duration. The patients had to be 35 to 70 years old. The BP cut offs seem to cater for use of at least one antihypertensive. The SBP needed to be between 160 and 140 – less than 160 was one of the inclusion criteria and less than 140 was one of the exclusion criteria. Similarly, for DBP we have the values 100 and 90 respectively.
The mean SBP and DBP were below 140 and 90 at baseline, but these baseline measures were 24-hour ambulatory BP means, so it is understandable that they might be different to a single reading during screening for potential entry to an RCT.
So, what about the acupuncture intervention? I was scanning the paper and adding bookmarks for navigation during the webinar, and I often bookmark the authors’ conclusion, but the latter left me confused. It seemed to suggest that two different acupuncture protocols, each with 4-letter acronyms were being compared. To my horror, when scrutinizing the methods and then checking in the published protocol,[6] I found that acupuncture with manual stimulation was indeed used in both groups.
Acupuncture is relatively routine in stroke rehabilitation in the Far East, at least in traditional medicine hospitals, so to perform acupuncture against no acupuncture (as I had assumed this paper was about) might prove difficult for recruitment in that context.
Patients in both groups received the XNKQ needling method, which is a formula for stroke. This involves needling PC6 bilaterally (manual stimulation for 1 to 3 minutes), GV26 (with ‘sparrow pecking’ until tears flow from the eyes), and 4 points on the side affected by the stroke – SP6, HT1, LU5, BL40. The later points are needled quite vigorously until the relevant limb twitches 3 times, or (at LU5) the sensation is felt from the elbow to the entire hand).
The HXSF formula includes ST9, LI4, LR3, LI11, ST36. So, big points in all 4 limbs plus ST9, which is a point at the level of the thyroid cartilage and described as being over the carotid pulse. I have discussed needling at ST9 previously – see Acupuncture for CAD 2024.
Acupuncture sessions were for 5 (in paper) or 6 (in protocol) days a week for the first 6 weeks, then 3 times a week up to 12 weeks.
From a WMA perspective there is not a huge difference between these formulae. The difference is probably greater in the mind and intention of the practitioner who has been taught to use one approach for stroke and the other for HT. Having said that, there is some difference in dose of needling by combining both formulae in one group, which would result in more than double the number of needle insertions. Given the rather vigorous needling stimulation performed in the XNKQ method, there could be a ceiling effect in terms of the general (non-segmental) effects of needle stimulation.
Out of the 480 patients randomised, there was a dropout of just over 20% leaving 182 in the experimental group and 189 in the control. The absolute numbers for stroke recurrence at one year was 5 and 7 respectively. That corresponds to 2.7% and 3.7%, which are not significantly different, but considerably lower than the estimate I got from ChatGPT 5.4 for the one-year recurrence rate following the first ischaemic stroke in patients with hypertension, which was 10% to 15%. I guess the latter estimate is based on the Western populations, so I asked Dseek (formerly DeepSeek) as well, and it came up with 5% to 15%, but it was very quick to come up with that. CoPilot, which uses ChatGPT (but not 5.4 currently) came up with a figure of 18%.
If you want to try yourself, the exact text I used was as follows:
Ischaemic stroke in patients with hypertension – what is the recurrence rate at 1 year following the first stroke?
Secondary outcomes included 24-hour ambulatory BP as well as home BP measurements. There were modest (~7 mmHg) but statistically highly significant reductions in BP in both groups with no differences between groups.
What can I say? I would not have asked that particular research question, but I understand the approach from a purely TCM point of view. The main positive message is that this group is not afraid to publish negative or neutral findings. It is likely that acupuncture had a protective effect on recurrent stroke, although that is an uncontrolled observation, so we have not really progressed far from our hopeful speculation resulting from the large retrospective cohorts I mentioned above.
References
1 Liu W, Li B, Rong M, et al. Effect of acupuncture on ischemic stroke patients with hypertension: a randomized clinical trail. Front Neurol. 2026;17:1717706. doi: 10.3389/fneur.2026.1717706
2 Huang M, Yen H-R, Lin C, et al. Acupuncture decreased the risk of stroke among patients with fibromyalgia in Taiwan: A nationwide matched cohort study. PLOS ONE. 2020;15:e0239703. doi: 10.1371/journal.pone.0239703
3 Liao C-C, Chien C-H, Shih Y-H, et al. Acupuncture Is Effective at Reducing the Risk of Stroke in Patients with Migraines: A Real-World, Large-Scale Cohort Study with 19-Years of Follow-Up. Int J Environ Res Public Health. 2023;20:1690. doi: 10.3390/ijerph20031690
4 Huang C-Y, Huang M-C, Liao H-H, et al. Effect of acupuncture on ischaemic stroke in patients with rheumatoid arthritis: a nationwide propensity score-matched study. BMJ Open. 2024;14:e075218. doi: 10.1136/bmjopen-2023-075218
5 Huang C-H, Lin S-K, Lin H-J, et al. Clinical effects of acupuncture treatment for prevention of insomnia-induced stroke: A large-scale cohort study. J Tradit Complement Med. 2025;15:51–61. doi: 10.1016/j.jtcme.2024.07.003
6 Du Y-Z, Gao X-X, Wang C-T, et al. Acupuncture lowering blood pressure for secondary prevention of stroke: a study protocol for a multicenter randomized controlled trial. Trials. 2017;18:428. doi: 10.1186/s13063-017-2171-5
The BMAS Blog 
You must be logged in to post a comment.