Article Text


Anxiolytic-like behavioural effects of head electroacupuncture in rats susceptible to stress
  1. Jia Chen1,
  2. Douglas W Barrett2,
  3. Yuxin He3,
  4. F Gonzalez-Lima2
  1. 1Department of Traditional Chinese Medicine, School of Medicine, Jinan University, Guangzhou, China
  2. 2Department of Psychology and Institute for Neuroscience, University of Texas at Austin, Austin, Texas, USA
  3. 3AOMA Graduate School of Integrative Medicine, Austin, Texas, USA
  1. Correspondence to Dr F Gonzalez-Lima, University of Texas at Austin, 108 E. Dean Keeton Stop A8000, Austin, TX 78712, USA; gonzalezlima{at}


Aim To evaluate the behavioural effects of head electroacupuncture (EA) using the Holtzman rat model, a genetic strain showing susceptibility to stress-evoked helplessness.

Methods Putative anxiolytic and antidepressant behavioural effects of head EA were investigated using the light-dark and forced swim tests, respectively. The open field test was used to investigate motor activity. A total of 28 rats were used in two experiments, each with two groups (n=7 rats each). Rats were restrained and randomised to handling only (control) or 2Hz EA on the midline head anteriorly (at Yintang) and posteriorly (at GV20) for 3 days (experiment 1) or 4 days (experiment 2).

Results One day of EA did not modify behaviour in any of the tests (p>0.1); however, 2 days of 2 Hz EA treatment to the head had anxiolytic-like effects, as indicated by an improvement in ambulatory time and average velocity in the light-dark test (experiment 2). Relative to the control group, the EA group demonstrated greater ambulatory time (37.0±3.7 vs 25.2±3.6 s, p<0.05) and lower average velocity (2.73±0.06 vs 3.08±0.13 cm/s, p<0.05). However, EA treatment had no significant effects on the open field and forced swim tests in either experiment.

Conclusions Two days of EA treatment using 2 Hz pulsating electrical current at midline anterior and posterior acupuncture points on the head induces behavioural effects suggestive of anxiolysis.


Statistics from


Electroacupuncture (EA), which involves the application of a weak pulsating electrical current to acupuncture needles as a means of stimulation, was first reported in China in the mid 1930s.1 EA has been used clinically for brain and behavioural disorders for the past 80 years, including GV20 (Baihui)-based scalp EA according to the principles of Traditional Chinese Medicine, an approach that is supported by animal research.2 In the last few years, an increasing number of research studies in Western countries have explored weak non-invasive transcranial alternating current stimulation (tACS) as a technique for influencing brain function.3 Modern studies of tACS use electrical stimulation parameters similar to those used in Chinese EA but delivered via large contact electrodes instead of fine needles.3 EA applied to the head may lead to entrainment of ongoing brain wave oscillations.4 It has been hypothesised that Yintang and GV20 may be particularly important locations for transcranial electrical stimulation (and thus also EA) in view of their close proximity to limbic brain regions linked to emotional regulation (medial prefrontal cortex and cingulate cortex, respectively).5

It is known that body EA has effects on the brain: 2 Hz EA enhances evoked potentials in the human cerebral cortex,6 retards the time until first demand of patient-controlled analgesia after surgery,7 and induces the release of opioid peptides in the rat brain.8 Several clinical studies have reported the anxiolytic effects of EA,9 manual acupuncture (MA),10 and acupressure.11 Dias et al9 applied EA at 2 Hz to the limbs, face and ear as well as the scalp (GV20); consequently the effects cannot specifically be attributed to scalp EA. In depression, the clinical data are less clear; EA is more effective than acupuncture in some studies but not in others.12 For example, an observational study of EA in 36 cases of depression showed a 94.4% effective recovery rate.13 The clinical effect of acupuncture alone appears to be as good as fluoxetine, which is effective in 60–70% of patients,14 and may even be superior, with reported clinically effective rates of 83.3%.15 However, Andreescu et al16 did not find any benefit of EA over MA for severe depression in a randomised controlled trial; both treatments improved symptoms.

With respect to animal models, Li et al17 showed that EA has anxiolytic and antidepressant effects in rats with chronic neuropathic pain. Furthermore, several studies have shown that EA may influence regulation of the stress response. Yang et al18 showed that EA has beneficial regulatory effects on the circadian rhythms of temperature and melatonin in a rat model of depression induced by chronic stress. Yang et al19 demonstrated that combined treatment with 2 Hz EA and a low dose of citalopram achieved better antidepressant effects in a chronic unpredictable stress rat model than either treatment alone. Acupuncture has been shown to have anxiolytic effects on stress-induced physiological and behavioural responses.18 ,20 ,21

The aim of the present study was to investigate the effects of 2 Hz EA (at acupuncture points on the head with putative anxiolytic effects) using restrained Holtzman rats, a genetic strain with susceptibility to stress-induced helpless behaviour that we have previously characterised using behavioural testing (open field test, shuttle box test, forced swim test) and antidepressant (fluoxetine) treatment.22–25 Our main objective was to investigate whether EA at Yintang and GV20 exerts behavioural effects in this animal model with a view to performing mechanistic studies in the future.



Twenty-eight adult male Holtzman rats weighing 205±9 g (mean±SEM) were purchased from Harlan (Madison, Wisconsin, USA) and maintained at the University of Texas Animal Resources Center, a facility accredited by the Association for the Assessment of Laboratory Animal Care International, under a 12 h/12 h light/dark cycle. Only males were used to avoid the behavioural variability observed in female rats during the 4–5 day oestrus cycle. Rats were housed in pairs during acclimatisation (including handling) for 1 week before the beginning of the experiment, from which point on they were singly housed. Experiments took place during daylight between 09:00 and 17:00. Experiments were performed in accordance with the National Institutes of Health guidelines for the use of experimental animals and were approved by the University of Texas Institutional Animal Care and Use Committee (reference no. AUP-2014-00400).

Justification for use of animals

The use of Holtzman rats allows both the genetic susceptibility and the degree of stress to be well controlled, which would otherwise be impossible with humans. This strain is a selectively bred animal model that is genetically susceptible to stress,22–25 thus only minimal restraint stress needs to be applied to be able to test for anxiolytic and antidepressant effects. Other rat strains that are not genetically susceptible to stress would require more chronic and prolonged stress to produce similar anxiety and depressive behaviour. Aside from the restraint stress, every possible effort was made to minimise discomfort to the animals and to reduce suffering, and the rats were treated with great care throughout the experiments. To minimise the number of animals used, a power analysis was conducted on the primary outcome measure of ambulatory time. Assuming between animal variability of 5 s, it was determined that seven animals per group would be required to detect a difference between the groups of 8 s (Cohen's d=1.6) in ambulatory time with 80% power at a two-tailed α level of 0.05.

Experimental design

The 28 rats were divided between two experiments, each with two groups (n=7 rats per group). Rats were randomised to restraint only (control) or 2 Hz EA at Yintang (anterior midline) and GV20 (posterior midline) on the head for 3 days (experiment 1) or 4 days (experiment 2). We aimed to examine for dose–response effects by varying the number of EA sessions in this way. Within each experiment, rats received the same behavioural testing and restraint procedures. The EA groups received EA, and the control groups were handled without needle insertion or electrical stimulation. All animals were subjected to the open field test,26 light-dark test27 and forced swim test,28 as previously described29 and according to the timeline outlined in table 1. The 10 min open field test was used to determine if the intervention had non-specific effects on general motor behaviour.26 The light-dark test was identical to the open field test, but included a dark-box insert that occupied half of the field. A small hole allowed the animal access to the lit part of the field. The animal was initially placed inside the dark side and allowed to explore freely. Increased ambulatory time in the light-dark test can be used to evaluate anxiolytic effects as rats show species-typical avoidance of bright open spaces.27 Finally, a reduction of immobility in the forced-swim test was used to evaluate the antidepressant effects of the intervention.29 Two forced swim sessions were conducted on consecutive days in each experiment; however, scoring was only performed during the second session (day 3 and 5 for experiments 1 and 2, respectively). Behavioural testing was conducted in Plexiglas chambers with infrared arrays of motion detectors mounted outside of them (Med Associates, St Albans, Vermont, USA). An automated computerised activity monitoring system (Med Associates) tracked and recorded the frequency, duration, speed, location and path of each subject's movements. To reduce risk of bias, the investigators recording the behavioural outcomes were blind to treatment and the same outcome measurements were used for all subjects.

Table 1

Timeline and procedures for experiments 1 and 2

Restraint and EA stimulation

During EA treatment (or handling of controls), the trunk of each rat was kept immobile using a sling-like apparatus (Jinan University, Guangzhou, China) consisting of a neck collar and jacket attached by a flexible cord to screws. The sling was suspended using the screws, thus providing trunk support while allowing limb and head mobility as per a sling, as there was no rigid restraint of the neck. In the EA groups, sterile disposable stainless steel needles (length 9 mm, diameter 0.2 mm; Huatuo, Suzhou Medical Supplies Factory Co, Ltd, China) were inserted to a depth of 3 mm at Yintang (above the frontonasal suture) and GV20 (at the midpoint of the interaural line). The two needles were connected to the output terminals of a standard EA apparatus (model no. LH202H, Beijing Huawei Medical Instrument, Beijing, China) and an alternating current with a pulse width of 0.3 ms was applied at 2 Hz frequency and 0.6 mA intensity for 30 min/day. The electrical stimulation parameters were verified using a two-channel digital real-time oscilloscope (model no. TDS 220, Tektronix, Beaverton, Oregon, USA).

Statistical analysis

Data were analysed using SPSS v.20 (IBM, Chicago, Illinois, USA) and are presented as mean±SEM. Differences between the two groups within each experiment were assessed using one-way analysis of variance (ANOVA) and a two-tailed p value <0.05 was considered to be statistically significant.


Experiment 1

Table 2 shows the results of the open field test, which reflects general motor behaviour. Compared with control rats, a single EA treatment did not modify behaviour in the open field test when performed on day 1. Moreover, there were no differences between the EA and control groups in the light-dark test (a behavioural test reflecting anxiety) on day 1 with respect to ambulatory time (33.8±4.1 vs 36.0±6.8 s, p=0.77) or average velocity (2.75±0.07 vs 2.95±0.18 cm/s, p=0.29). Similarly, 3 days of EA did not modify rat behaviour in the forced swim test on day 3, a behavioural test reflecting antidepressant effects of interventions (table 3).

Table 2

Results of the open field test on day 1 (experiment 1) or day 3 (experiment 2)

Table 3

Results of the forced swim test on day 3 (experiment 1) or day 5 (experiment 2)

Experiment 2

The open field test results on day 3 (after 2 days of EA stimulation) were similar to experiment 1 with no significant group differences between the EA and control groups (table 2), suggesting that EA had no demonstrable effect on ambulatory behaviour. However, in the light-dark test on day 3, ambulatory time was greater in the EA group compared to the control group (37.0±3.7 vs 25.2±3.6 s, p=0.04; figure 1) and average velocity was lower (2.73±0.06 vs 3.08±0.13 cm/s, p=0.04; figure 2), suggesting an anxiolytic effect of EA. After 2 days of treatment, EA-treated rats spent significantly more time exploring in the light and exhibited a slower pace than controls. However, as shown in table 3, there were no significant differences between the two groups in the forced swim test on day 5 (after 4 days of EA stimulation).

Figure 1

Ambulatory time (s) during a light-dark test in 14 Holtzman rats subjected to restraint stress followed by two 30 min sessions of 2 Hz electroacupuncture (EA) at Yintang/GV20 (EA group, n=7) or no treatment (control group, n=7). *p<0.05 EA versus control.

Figure 2

Average velocity (cm/s) of ambulation during a light-dark test in 14 Holtzman rats subjected to restraint stress followed by two 30 min sessions of 2 Hz electroacupuncture (EA) at Yintang/GV20 (EA group, n=7) or no treatment (control group, n=7). *p<0.05 EA versus control.


The principal finding of this study is that midline head EA at 2 Hz for 2 days (but not 1 day) exerted behavioural changes that can be interpreted as anxiolytic-like effects in an animal model with genetic susceptibility to stress—the Holtzman rat.22–25 The extent to which an intervention can increase exploratory activity in the light-dark test has been interpreted as a measure of anxiolysis, and is dependent on the activity level of a control group subjected to the same stress conditions.30 The light-dark test is based on the natural aversion of rodents to brightly illuminated spaces and on the spontaneous exploratory behaviour of rodents in novel environments. Both the EA and control groups were subjected to the same level of restraint stress and behavioural testing, and showed the same levels of activity in the open field and forced swim tests. Together these findings suggest that the significant differences between groups in the light-dark test in experiment 2 cannot simply be discounted as non-specific changes.

Anxiolytic, antidepressant, analgesic, and positive physiological effects of EA have been reported previously in other rat models. Inconsistences in the results from various laboratories may be accounted for by differences in the types and severity of stressors as well as EA location and stimulation frequency. For example, Li et al17 showed that EA at GV20 (as used in the present study) and GB34 (Yanglingquan), located near the knee joint, produced anxiolytic and antidepressant effects in a rat model of chronic neuropathic pain induced by chronic constriction injury of the sciatic nerve, which is significantly more aversive and chronic than the model used in the present study. EA at acupuncture points in the chest had parasympathetic physiological effects in a different restraint stress rat model, which may be mediated by vagal stimulation.18 Furthermore, MA has been demonstrated to increase markers of innate immunity after restraint stress in mice.31

In terms of stimulation frequency, Han8 reviewed multiple body EA studies and demonstrated that stimulation at 2 Hz has different effects on opioid neuropeptide release than stimulation at 15 or 100 Hz. For example, 2 Hz (but not 100 Hz) EA induces release of β-endorphin and enkephalin in the brain, which may have anxiolytic as well as analgesic effects under stress conditions. Yao et al21 showed that head EA has beneficial regulatory effects on circadian rhythms in rats subjected to chronic unpredictable stress. Although they used the same acupuncture points (Yintang and GV20) and frequency of stimulation (2 Hz) as the present study, it should be noted that EA was given for 20 min, once a day for 21 days, in a different rat model.

The 2 Hz frequency of weak electrical stimulation overlying the midline cranium may be an important parameter, considering the hypothetical neurophysiological mechanisms underlying the present results. We speculate that head EA is comparable to tACS, which has been the subject of recent Western studies exploring weak current at low frequencies (near to 2 Hz) as a tool for affecting brain function in animals and humans.3 However, mechanistically, this hypothesis has not been tested with head EA. On the other hand, tACS and EA stimulation are fundamentally different from transdermal electrical stimulation of the head, which uses very high frequencies (pulse-modulated at 7–11 kHz) designed to stimulate peripheral afferent fibres of the trigeminal and cervical spinal nerves.32 According to Reato et al,3 the effects of tACS (using a weak current and low frequency) result from weak but simultaneous polarisation of a large number of neurons, leading to an entrainment of brain wave oscillations. For example, entrainment of ongoing neuronal activity to weak alternating current at low frequencies, resembling those of cortical slow oscillations (1–2 Hz), has been shown across multiple cortical areas by Ozen et al33 in anaesthetised rats. Weak alternating current applied to the head can entrain slow-wave oscillations in a frequency-dependent manner.34 This is particularly the case for 2 Hz, a frequency matched to endogenous slow-wave oscillations.3 Fröhlich and McCormick35 used a computational model of slow-wave oscillations to show that these effects could be explained by the weak polarisation of the membrane acting simultaneously on multiple synaptic neurons. This prediction has been demonstrated empirically for low frequency-specific entrainment of multi-unit activity in vivo.36 Furthermore, neurobiological studies of EA at Yintang and GV20 in rats have shown that 2 Hz electrical stimulation leads to differential neuronal gene expression,37 antidepressant effects,38 altered N-methyl-D-aspartate receptor (NMDA) signalling,17 and increased expression of brain-derived neurotrophic factor (BDNF) and its receptor tropomyosin receptor kinase B (TrkB),19 especially in the hippocampus, which is one of the brain areas involved in the stress response. Hence it may be speculated that using EA at 2 Hz frequency to transcranially modulate limbic areas may have anxiolytic effects in stress models.

Strengths and limitations

The use of Holtzman rats was a particular strength of this study because this genetic strain allows for minimal restraint stress to be applied in order to test for anxiolytic and antidepressant effects. Furthermore, we previously validated all the behavioural tests used that have been tested in this strain.22 ,23 Other rat strains that are not genetically susceptible to stress require more chronic, prolonged stress to produce similar anxiety and depressive behaviours.17 One limitation is that the number of EA sessions was relatively low, and may therefore have failed to achieve a sufficient ‘dose’ of EA. It remains possible that a greater number of EA treatments than the four daily treatments we used before the forced swim test may be needed to observe antidepressant effects. While we have previously shown efficacy of the antidepressant fluoxetine in the Holtzman rat strain using the same tests,24 it is also possible that EA does not have antidepressant efficacy in this particular model. Another potential limitation is the small sample size used. The minimal number of animals and the very large effect size (d=1.6) factored into our sample size calculation ultimately increases the likelihood that negative results represent type 2 error (‘false negative’) and that positive results represent type 1 error (‘false positive’).

In summary, the behavioural findings of the present study, combined with the results of previous research into tACS and EA, suggest that the putative anxiolytic effects of EA may be linked to slow-wave entrainment produced by 2 Hz alternating current stimulation of the head. Since there is no clear evidence that scalp EA directly stimulates the brain, further mechanistic research is required to prove this hypothesis. The animal model that we chose may have some advantages compared to other rat models of stress that are relatively more chronic and/or severe, given that behavioural effects were found after only two EA daily treatments. Further studies in this model may help to contribute to the understanding of EA at the cellular, metabolic and network levels, and thereby help to provide insights into its apparent anxiolytic effects in rats and humans.


We would like to thank Professor Xiaoyin Chen, Department of Traditional Chinese Medicine, School of Medicine, Jinan University, China.


View Abstract


  • Contributors JC performed the experiment at the University of Texas at Austin. DWB performed the statistical analysis and figures and helped with the experiment. YH participated in the discussion. FG-L designed the study and prepared the manuscript. All the authors have read and approved the final manuscript.

  • Funding This work was supported by the National Natural Science Foundation of China (grant no. 81273616 and 81473557) and the Guangdong Natural Science Foundation (grant no. S2013010013434).

  • Competing interests None declared.

  • Ethics approval Experiments were approved by the University of Texas Institutional Animal Care and Use Committee, Austin, Texas, USA.

  • Provenance and peer review Not commissioned; externally peer reviewed.

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.