Acupuncture has a long history in treating various diseases and physiological malfunctions in Eastern medicine and is gaining increasing popularity worldwide (1,2). According to a review by the World Health Organization Consultation on Acupuncture, stroke ranked highly among the ailments for which acupuncture had proved to be effective (3). In addition, acupuncture is effective in a broad spectrum of diseases and disorders, and is therefore regarded as a nonspecific therapy in clinical practice (4). It is important to understand the mechanism of acupuncture within the context of Traditional Chinese Medicine (TCM). According to the principles of TCM, ‘qi’ is an important constituent of the human body that is transported by the meridian system (5). The meridian system is a complex network of neurovascular bundles that are characterized by distributed conductance, resistance, inductance and capacitance as functions of time and frequency (6). In other words, qi interacts with time and flows through the body. Based on this model, acupuncture works by regulating the circulation of qi (7). Given the relationship among qi, time and meridian, the method of Zi-Wu-Liu-Zhu is useful in calculating the optimum time to administer acupuncture at a particular meridian (8). For example, qi is strongest on the stomach meridian of Foot-Yangming between 7:00 a.m. and 9:00 a.m., which indicates that the optimum time for acupuncture at the Foot-Yangming is between 7:00 a.m. and 9:00 a.m. (9). Converging evidence from animal studies (10,11) and clinical investigations (12–14) indicates that the time to administer acupuncture determines the effectiveness and success of the procedure. In the last five years, time-related acupuncture has received increased focus and has been developed as an interdiscipline of chronoacupuncture (8).

It has been suggested that the majority of acupuncture effects are mediated via the brain (15). With the development of neuroimaging techniques, functional magnetic resonance imaging (fMRI) is becoming a useful method to investigate the neural mechanism of acupuncture. An increasing number of fMRI studies have demonstrated different brain networks for acupuncture performed at different acupuncture points (16). Even for a particular acupuncture point, differences were revealed between healthy subjects and patients (17,18). To the best of our knowledge, the neural mechanism underlying the effect of timing on acupuncture has not been investigated. The present study aimed to investigate brain activation patterns during acupuncture at different time periods and to consider the timing effects in stroke patients and healthy subjects. The acupuncture point of Zusanli (ST36), which has been frequently used in the treatment of stroke patients in clinical practice in China (17,19,20), was selected. In the literature, acupuncture stimulation at ST36 in healthy individuals demonstrated effects in the somatosensory (5,21) and motor (21–23) areas, cerebellum (5,17,24), limbic system (17,21,23–28) and higher cognitive areas (5,17,21,23,28). Compared with healthy controls, a weaker activation pattern of a similar network was revealed for stroke patients (17). In the present study the timing effect was tested by comparing acupuncture at the optimum time of between 7:00 a.m. and 9:00 a.m. (the AM condition) and a non-optimum time of between 3:00 p.m. and 5:00 p.m. (the PM condition). It was hypothesized that acupuncture stimulation at ST36 would differ during the time periods of 7:00–9:00 a.m. and 3:00–5:00 p.m., and the timing effect would differ between stroke patients and healthy subjects.

The present study investigated brain activation patterns in healthy and stroke subjects receiving acupuncture at the acupuncture point of ST36 during two time periods, one in the morning (the AM condition) and one in the afternoon (the PM condition). The main finding was a stronger activation in the AM than in the PM condition in both subject groups. To the best of our knowledge, this is the first neuroimaging report providing evidence for the mechanism of chronoacupuncture.

Acupuncture of ST36 has consistently been shown to elicit reduced neural activity in the limbic system (24,37). Similar limbic reductions in neural activity have also been found with other acupuncture points, such as LI4 (38), LV3 (39) and ST44 (23). These findings suggest a general modulation effect of acupuncture on the limbic system. Given the widespread anatomical and functional connections of the limbic system (40), it is expected that acupuncture affects the limbic system. The present results for the AM condition revealed activations of the hippocampus in the healthy subjects and putamen in the stroke subjects. The reversed pattern of BOLD response could be attributed to different control conditions. The studies mentioned previously used tactile stimulation as the baseline condition (24,37), whereas the present study used the resting state as the control condition. In addition to limbic activations, the present results showed activations for the AM condition in the superior temporal gyrus, prefrontal cortex and supplementary motor area in both healthy and stroke subjects. Prefrontal activations have been found previously in healthy (37) and stroke (41) subjects, suggesting possible higher cognitive modulation effects. This is in agreement with the theory that the cerebro-cerebellar system (42) could account for acupuncture effects.

Notably, the present results revealed significantly different activations following acupuncture of ST36 between the AM and PM conditions. Stronger activations were found in the middle and superior frontal gyrus, cingulum, thalamus and cerebellum in the AM condition than those in the PM condition in healthy subjects. Similar results of stronger activations in the AM condition than those in the PM condition were revealed for the stroke subjects: the regions included the left cuneus, right supplementary motor area and right inferior parietal gyrus. The results were in agreement with the theory of chronoacupuncture stating that there is an optimum time for acupuncture to achieve the most curative effect (8), and this appears to apply to both healthy subjects and stroke patients. The factor of timing in conducting acupuncture has been largely ignored in previous neuroimaging studies, which may have contributed to their heterogeneous results (16,43) and low test-retest reliability (44). Although the factors of acupuncture points, acupuncture manipulation and stimulation methods have been investigated, the consideration of timing in conducting acupuncture is also important.

The present results demonstrated different fMRI activation patterns of acupuncture at ST36 at different time periods in stroke patients and healthy subjects. Future studies with larger sample sizes should further explore the timing effects of acupuncture at different acupoints and in different populations.