Thirty-three eligible participants were recruited at Chengdu University of Traditional Chinese Medicine from 2 October 2021 to September 2022. All eligible participants met the inclusion criteria and provided written informed consent before participation.

To avoid the influence of individual differences, a randomized crossover protocol was used in this study. According to previous acupuncture neuroimage studies (Bai et al., 2009; Rheu et al., 2011; Claunch et al., 2012), participants received SSH and TTL on separate days, with a washout of 1 day. The order of acupuncture manipulation was determined by the computer-generated random number table. All acupuncture manipulations were performed by one licensed acupuncturist with clinical experience of over 5 years. SSH and TTL were performed at Quchi (LI11, left side) using the disposable needle (size 0.25 × 40 mm, Hwatuo, Suzhou Medical Supplies Factory Co., Ltd., Suzhou, China). The acupuncture experiment paradigm was shown in Figure 1A. The details of the following acupuncture manipulations were shown in Figure 1B.

The depth of needling is divided into three layers, shallow, middle, and deep, corresponding to the subcutaneous layer (fat), loose connective tissue layer (superficial fascia), and dense connective tissue (deep fascia and muscle; Chen, 2017).

The fNIRS scan lasted for 720 s (Figure 1C). During the scan, participants were asked to close their eyes and keep relaxed without falling asleep. A non-invasive multi-channel continuous wave fNIRS instrument (NIRx Medical Technologies, New York) was utilized to record the oxyhemoglobin (oxy-Hb) and deoxyhemoglobin (deoxy-Hb) signals. The absorption of near-infrared light was measured at the wavelengths of 785 and 830 nm and the sampling frequency was 3.91 Hz. Same as our recent study (Chen et al., 2022), the sensorimotor cortex (SMC) and prefrontal cortex (PFC) was selected as regions of interest (ROI). As shown in Figure 1C, two 2 × 3 optode probe sets consisting of seven channels (CHs) were placed at the bilateral SMC, and one 3 × 5 optode probe set consisting of 22 CHs was placed at the bilateral PFC. The fNIRS optode was placed according to the international 10–10 EEG system (Tsuzuki et al., 2017), with Fpz and Cz as reference points. The separation distance of sources and detectors was set to 2.5–3 cm. The corresponding anatomical positions of fNIRS CHs were measured by an electromagnetic 3D digitizer system (PATRIOT; Polhemus) and registered on the Montreal Neurological Institute MNI brain space using a virtual registration method (Supplementary Table S1).

The statistical analysis consists of two parts: the within-group comparison of SSH/TTL and resting period and the between-group comparison of SSH and TTL. The within-group comparison was conducted with paired t-test. The between-group comparison was conducted with two sampled t-test based on the delta value (Δ) of cortical activity and FC (Δ = acupuncture manipulation − pre-acupuncture rest). The significance threshold was set to p < 0.05. To minimize the risk of type I errors, the p-values were corrected with false discovery rate (FDR) correction for multiple comparisons.

Thirty-three subjects completed SSH and TTL manipulations. Four subjects in SSH and three subjects in TTL were excluded due to poor data quality. There were no significant differences between SSH and TTL in all metrics (Supplementary Table S2).

In SSH, the bilateral primary motor cortex (MI), the left primary somatosensory cortex (SI), and the right somatosensory association cortex were activated (Figure 2A, paired t-test, p < 0.05, FDR correction); the FC between the right and left PFC,

The PFC and SI (Figure 2C, paired t-test, p < 0.05, FDR correction). In TTL, the bilateral PFC and the bilateral MI were activated (Figure 2B, paired t-test, p < 0.05, FDR correction); the FC between the PFC and SI, the left PFC and left MI were increased (Figure 2D, paired t-test, p < 0.05, FDR correction). However, there was no difference in cortical activation and FC between SSH and TTL under the corrected threshold.

Due to its complexity, diversity, and variation, investigating the central mechanism of SSH and TTL requires neuroimaging techniques with high spatiotemporal resolution. On the one hand, the changes elicited are extremely subtle. Neuroimaging techniques with excellent temporal resolution are sufficient to catch these subtle changes over time. On the other hand, studies have revealed that activations induced by acupuncture are mainly in deep brain tissues (Huang et al., 2012). Neuroimaging techniques with high spatial resolution are necessary to record changes induced by SSH and TTL.

Among acupuncture neuroimaging studies, electroencephalogram (EEG), fNIRS, functional magnetic resonance imaging (fMRI), and positron emission tomography (PET) are most frequently used. EEG provides the best temporal resolution (Cao et al., 2021), and fMRI has the optimal spatial resolution (Chou et al., 2020). Considering the temporal resolution, spatial resolution, and feasibility in a real clinical condition, fNIRS was used in this study (Figure 3). Its temporal resolution is next only to EEG (Li et al., 2022), and the spatial resolution is second to fMRI (Chen et al., 2020). Moreover, fNIRS has excellent ecology validity for use in real clinical settings. Unfortunately, this study encountered some limitations in utilizing fNIRS for SSH and TTL research. Consequently, combing higher temporal and spatial resolution is recommended for future neuroimaging studies on SSH and TTL.

Currently, fMRI-EEG can achieve the optimal spatiotemporal resolution (He and Liu, 2008). EEG-fNIRS can improve temporal resolution and feasibility in real clinical settings (Uchitel et al., 2021). At the same time, applications of multimodal neuroimaging may bring potential difficulties in the study design of SSH and TTL, such as the judgment of the needle sensation, acupuncture manipulations on the upper body, and the compatibility between neuroimaging techniques. Nevertheless, neuroimaging techniques with higher spatiotemporal resolution are still more applicable for investigating the central mechanism of SSH and TTL.

SSH and TTL play an important role in affecting acupuncture’s therapeutic effects. To understand SSH and TTL in depth, the relationship between these two manipulations and their therapeutic effects should be considered sufficiently. Though using healthy subjects in the research can receive good compliance, it is difficult to characterize the effects of ARRMs on treating diseases. Thus, there is a growing trend in acupuncture neuroimaging studies to change subjects from healthy subjects to patients (Cai et al., 2018; Zhang et al., 2021). Research showed that positive results were reported in acupuncture studies on patients, while negative results were found on healthy subjects only (Qiu et al., 2020). Health subjects might be the possible factor for no FC difference between SSH and TTL. Hence, it is better to use diseases, combing the therapeutic effects to investigate the mechanism of SSH and TTL. SSH can be used for treating conditions with deficiency and cold symptoms, and TTL can be selected for controlling disorders with excessive and heat symptoms. For example, rheumatoid arthritis with cold and heat types can be applied as a disease carrier for exploring SSH and TTL. Therefore, it is suggested that future studies combine with therapeutic effects to investigate the central mechanism of SSH and TTL.

Strict and detailed quality control is the guarantee of acquiring reliable results. In this study, rigorous quality control was performed from three aspects: participants, practitioners, and the research environment. For participants, all healthy subjects were recruited in accordance with the inclusion criteria. They were required to maintain a regular lifestyle, away from alcohol and caffeine for 24 h before fNIRS scans. Then they were asked to sit quietly with eyes closed, not thinking about specific things, and concentrate on the experience of needle sensations during acupuncture task scans. For practitioners, this study was conducted following the pre-established standard operating procedure (SOP). All acupuncture manipulations were completed by one trained acupuncturist. For the research environment, it was kept quiet and dark to avoid the interference of light and sound on fNIRS signals. All scans were carried out by the same fNIRS equipment.

For future studies, it is recommended to perform strict and detailed quality control. It requires subjects’ good compliance, including focusing on feeling the needle sensation during acupuncture task scans and complying with trial requirements. SOP should be made and practitioners should be trained in advance. Finally, the research environment is needed to avoid contraindications of neuroimaging techniques’ scans and manipulations of SSH and TTL.