Can conventional therapy (CT) be combined with herbal therapy (CT + H) in the management of Alzheimer’s disease (AD) to the benefit of patients? This was the question investigated by Chinese researchers in a recent retrospective cohort study funded by grants from China Ministry of Education, National Natural Science Foundation of China, Beijing Municipal Science and Technology Commission, and Beijing Municipal Commission of Health and Family Planning.

In total, 344 outpatients diagnosed as probable dementia due to AD were collected, who had received either CT + H or CT alone. The GRAPE formula was prescribed for AD patients after every visit according to TCM theory. It consisted mainly (what does ‘mainly’ mean as a description of a trial intervention?) of Ren shen (Panax ginseng, 10 g/d), Di huang (Rehmannia glutinosa, 30 g/d), Cang pu (Acorus tatarinowii, 10 g/d), Yuan zhi (Polygala tenuifolia, 10 g/d), Yin yanghuo (Epimedium brevicornu, 10 g/d), Shan zhuyu (Cornus officinalis, 10 g/d), Rou congrong (Cistanche deserticola, 10 g/d), Yu jin (Curcuma aromatica, 10 g/d), Dan shen (Salvia miltiorrhiza, 10 g/d), Dang gui (Angelica sinensis, 10 g/d), Tian ma (Gastrodia elata, 10 g/d), and Huang lian (Coptis chinensis, 10 g/d), supplied by Beijing Tcmages Pharmaceutical Co., LTD. Daily dose was taken twice and dissolved in 150 ml hot water each time. Cognitive function was quantified by the mini-mental state examination (MMSE) every 3 months for 24 months.

The results show that most of the patients were initially diagnosed with mild (MMSE = 21-26, n = 177) and moderate (MMSE = 10-20, n = 137) dementia. At 18 months, CT+ H patients scored on average 1.76 (P = 0.002) better than CT patients, and at 24 months, patients scored on average 2.52 (P < 0.001) better. At 24 months, the patients with improved cognitive function (△MMSE ≥ 0) in CT + H was more than CT alone (33.33% vs 7.69%, P = 0.020). Interestingly, patients with mild AD received the most robust benefit from CT + H therapy. The deterioration of the cognitive function was largely prevented at 24 months (ΔMMSE = -0.06), a significant improvement from CT alone (ΔMMSE = -2.66, P = 0.005).


The authors concluded that, compared to CT alone, CT + H significantly benefited AD patients. A symptomatic effect of CT + H was more pronounced with time. Cognitive decline was substantially decelerated in patients with moderate severity, while the cognitive function was largely stabilized in patients with mild severity over two years. These results imply that Chinese herbal medicines may provide an alternative and additive treatment for AD.

Conclusions like these render me speechless – well, almost speechless. This was nothing more than a retrospective chart analysis. It is not possible to draw causal conclusions from such data.


Because of a whole host of reasons. Most crucially, the CT+H patients were almost certainly a different and therefore non-comparable population to the CT patients. This flaw is so elementary that I need to ask, who are the reviewers letting such utter nonsense pass, and which journal would publish such rubbish? In fact, I can be used for teaching students why randomisation is essential, if we aim to find out about cause and effect.

Ahhh, it’s the ! I think the funders, editors, reviewers, and authors of this paper should all go and hide in shame.

7 Responses to Chinese herbs for Alzheimer’s disease? An excellent example of how fatally flawed research misleads us all

  • Sorry.
    You produce apparent evidence of improvement in CT+H patients over CT ones, and then conclude that, due to a “a whole host of reasons” – none of which you detail, except that “CT+H patients were almost certainly a different and therefore non-comparable population” (again for reasons unspecified) – the conclusions “render you speechless”.

    Either I’m very stupid, or unfamiliar with the vast part of your argument, since the graphs you posted – if true – suggest a different conclusion to yours.
    Of course, not knowing how the Chinese Ministry of Eduction conducts its studies, or whether they are internationally recognised obviously makes a big difference but, on the basis of this post, you have NOT proved your point, sir.

    • yes, quite possibly you are ‘very stupid’. comparing apples with pears leads to false conclusions. this flaw alone is fatal, no need to mention all the others [but feel free to detect them yourself, in case your self-diagnosis of stupidity was wrong].

    • Let me try to help clear it up a bit, Richard. Let me quote a couple of sentences from the study:

      It’s a retrospective cohort study and patients diagnosed as Alzheimer’s dementia were collected [8, 9]. Medical records between May 2011 to August 2016 were accessed by using administrative datasets of memory clinic.

      So, random people over 5 years… It’s a retrospective cohort study… That pretty much says it all. And there are significant differences between the groups in some important parameters. Also, the CT+H group contained >10% (29 of 243) patients with severe dementia, whereas the CT group contained only 1% (1 of 101) patients with severe dementia. I propose that this means that in the CT group there was more room for deterioration than in the CT+H group. There was a natural tendency for greater amount of deterioration and, so, it showed up in the end results. What makes this clearer is that the baseline MMSE score was lower in the CT+H group than in the CT group, so there is even less room for deterioration in the CT+H group. This effect is important to take into account and the final difference is small anyway.

      And, to avoid all unnecessary analysis in the first place, let me remind you that A+B vs. A studies only confirm that B is at least a placebo (or nocebo, or may even have negative effects, according to the corresponding results). For further efficacy considerations, B must be tested in isolation. But since this is not possible for ethical reasons, tons of (useless)ambiguous research keeps getting published, unethically exploiting this, often neglected, caveat.

      I’m open to disagreement of course…

      • As shown in the charts above, the authors did compare apples to apples at least in regards to stages of Alzheimer’s disease: they compared people with mild Alzheimer’s disease taking Chinese herbs plus conventional therapy with conventional therapy alone and they compared people with moderate Alzheimer’s disease taking Chinese herbs plus conventional therapy with conventional therapy alone.

        In some respects, the individuals in the conventional therapy plus herb group should have progressed faster than people in the conventional group because they included more people with diabetes, who had hypertension, and who were smokers.

        In essence, conventional therapy served as the placebo group in this study. After about twelve months, the conventional therapy group responded little better than the historical decline in people receiving no treatment.

        Heo’s study using Korean red ginseng produced almost the same results at two years. That study was open label and did not distinguish between stages of Alzheimer’s disease, did not include low and high responders, and did not look at the effects of conventional treatments on cognition. This study then partially confirmed Heo’s results but with important added details.

        Without getting into too much detail, Alzheimer’s disease likely is the result of oxidation and nitration. Donepezil/Aricept and Namenda slow down oxidation and nitration and many of the compounds in Chinese herbs scavenge oxidants and nitro-oxidants. Combining the two, may stop the progression of early Alzheimer’s disease and slow down its progression for those with moderate Alzheimer’s disease.

  • One positive aspect of some open access journals is that they publish the Reviewer Reports – for this article you can find it here;

    • very interesting!
      one reviewer seems as blind as a bat, the other is ok.
      i cannot see the authors’ response.
      i fail to see why the journal published the article.

      • Quite so. If one wished to design a study to provide misleading results one could not have done better than this one. What were the authors thinking? What was the journal thinking?

Leave a Reply

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.

Subscribe via email

Enter your email address to receive notifications of new blog posts by email.

Recent Comments

Note that comments can be edited for up to five minutes after they are first submitted but you must tick the box: “Save my name, email, and website in this browser for the next time I comment.”

The most recent comments from all posts can be seen here.