Research note 001 — 2026-08-30 (expanded 2026-09-08)
Tingling Vibrates at About 50 Hz
That feeling of your lips buzzing and going tingly when you eat huajiao (花椒, Sichuan pepper). A research team at the University of London applied huajiao to the lower lip of human subjects and had them compare it against mechanical vibration delivered to a finger, matching the frequencies. The frequency of the 「しびれ」 (in translation: "tingling") that subjects felt was consistently about 50 Hz — matching the band that the tactile RA1 afferent fibers are tuned to. In other words, that tingling is not a figure of speech: at the level of perception, it really is a 「振動」 (in translation: "vibration").
What this page establishes
The perceived frequency of huajiao's tingling has been measured, in a frequency-matching experiment against a finger, at about 50 Hz.
KCNK, a 2P-domain potassium channel, has been proposed as a candidate channel underlying tingling, but this is an inference from human psychophysical results, and the channel itself has not been measured in humans.
A reduction in tactile sensitivity caused by sanshool derivatives has been measured, but the same measure has not yet been taken using natural sanshool.
The finding that samples with stronger huajiao pungency are also rated more bitter has been reported at the abstract level of a single study.
Claims and evidence
tingling has a perceived frequency of About 50 Hz. Matches the frequency band of RA1 tactile afferent fiberssupported
Scope: Huajiao-induced tingling on human lips. Judged by frequency matching against vibration on the finger
Note: RA1 adaptation by prolonged vibration reduced the perceived tingling frequency. The 50.0 Hz figure comes from a single report with n=8; as of 2026-08, no independent group has replicated it or reported a different frequency (only follow-up studies from the same group).
Verbatim: "On average, the tingle induced by Szechuan pepper on the lips was matched by a mechanical vibration frequency of 50.0 Hz on the finger." Crushed huajiao was mixed into a 40% ethanol solution at a concentration of 0.05 g/ml and applied to the lips. Experiment 1 had 12 participants; the frequency-matching experiments had 8 participants each. Verbatim: "Eight new participants judged the tingling/vibration frequency on their lips as in experiments 2 and 3." In the adaptation condition, verbatim: "80 Hz mechanical vibration was applied on the lips (adapting stimulus)," after which, verbatim: "the perceived frequency of pepper-induced tingle was significantly reduced when preceded by an adapting mechanical vibration on the lips (t7 = 3.1, p = 0.017)." A control using ethanol alone, without huajiao, was also carried out.
Limitations: Each frequency-matching experiment had only 8 participants, a small sample. The standard error of the 50.0 Hz estimate (2.4 Hz) was obtained in the first extraction pass, but was not verbatim-confirmed in a narrower-scope second pass, so only the mean value can be cited as verified. The stimulus was crushed spice in 40% ethanol, not purified hydroxy-alpha-sanshool, so this frequency is a property of that spice preparation. The matching control (mechanical lip vibration matched at 46.4 Hz) shows a slight underestimation bias in the matching procedure itself.
Treats sanshool-induced tingling independently, as abnormal activation of flutter-range vibration channels. Verbatim: "The perceived sanshool-induced tingling frequency on the lips is reduced by adapting the RA channel using prolonged mechanical vibration, paralleling the reduction in perceived frequency of mechanical vibration." It further adds that sustained mechanical pressure reduced tingling intensity to a mean of 24.7% +/- s.d. 34.0 of baseline (t9 = 7.00, p < 0.001, dz = 2.21).
Limitations: Not a fully independent replication: Hagura and Haggard are co-authors of both this study and Hagura 2013. The description of RA adaptation here restates a prior finding rather than a new frequency measurement, and this paper does not report a new numerical estimate of the tingling frequency.
hydroxy-alpha-sanshool (sanshool alkylamides) — currently proposed mechanisms: Inhibition of KCNK 2P-domain potassium channels on mechanosensitive neurons. Separate from the TRPV1/TRPA1 pathwaycontested
Scope: Human psychophysics using the stable sanshool derivative IBA, and the neurophysiology that supports it
Note: The mechanism is contested: the KCNK (2P-type K+ channel) inhibition hypothesis (neural response unchanged even in TRPV1/TRPA1 double KO) versus the TRPV1-main-pathway hypothesis (aversion nearly eliminated in KO). Differences in method and temperature dependence are candidate causes of the disagreement. The phenomenon itself (the tingling, the prickling) is not disputed.
In 68 subjects aged 20-59 across 5 experiments (Exp 1 self-desensitization n = 25; Exp 2 cross-desensitization n = 30 capsaicin and n = 30 mustard oil; Exp 3 time-intensity n = 28; Exp 4 temperature-tingling n = 29; Exp 5 tingling-touch n = 31), IBA-induced tingling was not reduced by pretreatment with either the TRPV1 or TRPA1 agonist: the effect of capsaicin pretreatment was non-significant (p = 0.585; tingling rating p = 0.104), and the same was true for mustard oil (p = 0.200; p = 0.845). Time-intensity used a modified procedure in which subjects rated on a 0-10 scale every 60 seconds for 32 minutes while selecting descriptor words. The authors conclude, verbatim: "alkylamides activate mechanosensitve neurons via blockade of KCNK two-pore potassium channels to induce the novel tingling sensation."
Limitations: The conclusion of KCNK involvement is an inference from a pattern of psychophysical dissociation; the channel itself was not measured in this study. The cross-desensitization argument rests on a null result, and its interpretation depends on whether the pretreatment concentration achieved effective desensitization. IBA is a stable derivative, not natural sanshool. An independent rodent study (Riera 2009) reports alpha-sanshool as a TRPV1/TRPA1 agonist and found that aversion was nearly eliminated in TRPV1 knockouts, so the divergence from these channels is contested at the mechanism level.
Context: The paper does not stop at the level of the spice, but names the compound explicitly (hydroxyl-alpha-sanshool, hereafter sanshool); the frequency and adaptation results (matching at 50.0 Hz; a reduction in perceived tingling frequency after 80 Hz lip adaptation, t7 = 3.1, p = 0.017) place this perception within the RA1 flutter pathway, complementing a mechanistic account via mechanoreceptive channels.
Limitations: This study used crushed Sichuan pepper, not isolated sanshool; the attribution to hydroxy-alpha-sanshool is not something verified here but the authors' identification of the active compound. The psychophysical channel assignment does not identify a molecular target; it is consistent with KCNK blockade but does not demonstrate it.
Verbatim: "We found that sanshool (100 uM) serves as an excitatory agent for a specific subgroup (52.3%; n = 2,063) of sensory neurons." Verbatim: "The greatest potency was observed with KCNK3 and KCNK18 (IC50 = 30.3 +/- 4.9 and 50.2 +/- 1.9 uM, respectively); KCNK9 was much less sensitive (IC50 = 450 +/- 30.1 uM)." Verbatim: "sanshool-evoked responses were not altered (in prevalence or magnitude) in sensory neurons cultured from mutant mice lacking both TRPV1 and TRPA1 channels."
Limitations: Rodent, not human. It provides a channel-level mechanism that human psychophysics can only infer. The 1 mM drinking-test observation is behavioral, not perceptual.
Contradicts: Reports hydroxy-alpha-sanshool (alpha-SOH) as an agonist of both TRP channels, giving the EC50 values verbatim as alpha-SOH: EC50 = 1.1 uM (TRPV1), alpha-SOH: EC50 = 69 uM (TRPA1), and states, verbatim: "1 mM alpha-SOH was markedly aversive to WT animals but, in KO animals, the aversion was almost entirely eliminated (P < 0.001)." The authors distance themselves clearly from the two-pore potassium channel hypothesis, stating, verbatim: "However, our results showing that alpha-SOH did not activate DRGs, which did not respond to capsaicin, diverge from those of Bautista et al. (2008), who identified sanshool responses in both capsaicin-sensitive and insensitive neurons," and, verbatim: "This finding deviates from the behavioural results presented by Bautista et al. (2008) where their TRPV1/TRPA1 double KO mice remained sensitive to the aversive effect of 1 mM alpha-SOH."
Limitations: Rodent and heterologous expression systems, not human perception. The authors themselves state, verbatim: "However, to assess taste preference we used a different testing paradigm from that used by Bautista et al. (2008)," and suggest, verbatim: "In this regard, KCNK channels may be less sensitive to sanshool at higher temperatures." Because aversion in a drinking test is not the same measure as human perception of tingling, the disagreement concerns the mechanism, not the existence of the sensation.
Show datacompound-hydroxy-alpha-sanshool —has_currently_proposed_mechanisms→ "Inhibition of KCNK 2P-domain potassium channels on mechanosensitive neurons. Separate from the TRPV1/TRPA1 pathway"
Human experiments with a sanshool derivative (IBA) also show that 「触覚が鈍る」 (in translation: "tactile sensitivity drops") during the tingling. This does not happen with capsaicin, the pungent compound in chili peppers, which suggests that tingling involves channels in the mechanosensory system rather than the pain system.
Scope: Human psychophysics using IBA applied to skin and lip sites
Note: Suggests that tingling is mediated through mechanosensory channels rather than pure nociception.
On the front third of the tongue, IBA significantly raised the tactile threshold: the R-index fell from 93.8 +/- 1.2 to 84.3 +/- 2.1 with the 0.02 N von Frey filament (p < 0.001), and from 86.9 +/- 1.7 to 75.2 +/- 2.1 with the 0.008 N filament (p < 0.001). Capsaicin, tested as a control stimulus, did not affect tactile sensitivity (p = 0.456; p = 0.687). Experiment 5 had 31 subjects (16 male, 15 female; ages 25-56).
Limitations: The stable derivative IBA was used, not natural hydroxy-alpha-sanshool, and testing was on the front of the tongue. The measure is the R-index of discriminability, not an absolute force threshold. The divergence from capsaicin is a null result for the capsaicin condition, showing that no detectable capsaicin effect occurred at the concentration used — it does not prove that no such effect exists.
Context: Demonstrates the reverse direction of the same tingling-touch interaction. Sustained mechanical pressure suppressed sanshool-induced tingling (to a mean of 24.7% +/- s.d. 34.0 of baseline; t9 = 7.00, p < 0.001, dz = 2.21), and in a graded-load experiment, the probability of reporting a stronger sensation on the loaded side of the lip decreased progressively as the force increased (F1.6,11.2 = 12.09; p = 0.002). The authors attribute this to an inhibitory interaction between mechanoreceptive channels, rather than to nociceptive or C-tactile pathways.
Limitations: Uses natural sanshool (ZANTHALENE), not the IBA derivative from Albin 2010, and measures the suppression of tingling by touch pressure rather than the rise in tactile threshold caused by alkylamides. It therefore supports a bidirectional interaction, but does not itself reproduce the von Frey threshold result.
The huajiao stimulus that causes tingling also moves together with bitterness. In a study that divided 9 samples of Chinese huajiao into a high-SPU group (pericarp Scoville Pungency Unit above 150,000) and a low-SPU group (below 150,000), and rated oral sensory attributes by descriptive analysis with a trained human panel, the intensity of attributes including bitterness was reported to be higher in the high-SPU group than in the low-SPU group (reported at the abstract level). The conditions that strongly induce tingling also overlap with the conditions under which bitterness is felt more strongly.
Sichuan pepper (huajiao) — perceived bitterness is positively associated with: A high Scoville Pungency Unit (SPU) value of the pericarp. When 9 samples (SPU 38,000-460,000) were divided into a high-SPU group (>150,000) and a low-SPU group (<150,000), the intensity of oral sensory attributes including bitterness, rated by descriptive analysis, was higher in the high-SPU group than in the low-SPU group.preliminary
Scope: Layer: food. 9 samples of Chinese huajiao (Zanthoxylum bungeanum), evaluated by a trained human panel using descriptive analysis, time-intensity (TI), and temporal dominance of sensations (TDS). SPU is a value measured by the Scoville test.
Note: Bitterness is one of the sensory attributes measured by descriptive analysis, alongside tingling, vibration, salivation, burning, astringency, and numbing. This is an abstract-level group comparison; the abstract does not report correlation coefficients or the significance level for each individual attribute.
The source states: "The intensities of tingling, vibrating, salivating, burning, astringency, numbing, and bitterness in high-performing SPU samples (measured at more than 150,000) were found to be much higher than in low SPU samples (less than 150,000)."
Limitations: This is an abstract-level check; the significance level and correlation coefficient for bitterness alone are not reported. It is a 2-group comparison of high-SPU versus low-SPU groups, and a continuous dose-response relationship is not shown.
Show dataorganism-zanthoxylum-bungeanum —has_perceived_bitterness_positively_associated_with→ "A high Scoville Pungency Unit (SPU) value of the pericarp. When 9 samples (SPU 38,000-460,000) were divided into a high-SPU group (>150,000) and a low-SPU group (<150,000), the intensity of oral sensory attributes including bitterness, rated by descriptive analysis, was higher in the high-SPU group than in the low-SPU group."
This kind of 「時間とともに立ち上がって消える感覚」 (in translation: "sensation that rises and fades over time") is measured using the time-intensity (TI) method, a standardized sensory evaluation technique. For huajiao, 7 sensations, including tingling, numbing, and salivation, have been measured with a human panel.
time-intensity sensory evaluation is used as A method for characterizing time-dependent oral chemesthetic sensations, such as the tingling sensation of Sichuan peppersupported
Scope: Human trained-panel studies on pungent/irritant spices
Note: 2026-08-31: Calibration downgrade from established to supported, because the supporting evidence does not derive from a full-text check. Can be re-upgraded after a full-text check.
9 Sichuan pepper samples were characterized using Scoville units, descriptive analysis, time-intensity (TI), and temporal dominance of sensations (TDS) across 7 oral sensations.
Limitations: Panel size is not stated in the abstract (unchanged). However, the abstract itself does state the comparison figures between the high and low groups for SPU and TI (AUC/Imax/Ttot), which had previously been missed in extraction. The full text (Methods/Results) has not been reached.
A different research group applied the same combination of time-intensity and temporal-dominance-of-sensations methods to a different pungent/irritant material, characterizing the pungency of baijiu and its subqualities. In TI, a significant difference was found in the rate of release of pungency: aged baijiu had a higher release rate, with a significantly decreased area under the curve and duration. In TDS, the dominant subqualities differed — burning and numbness for young baijiu, versus burning, tingling, and dryness for aged baijiu.
Limitations: Panel size and scale details could not be obtained from the abstract. The full text has not been obtained (Elsevier). This shows that TI/TDS are used for chemesthetic attributes in general, and is not specific to Sichuan pepper.
2-AFC, time-intensity, and TDS were applied to hydroxy-sanshool compounds dissolved in water-ethanol, sucrose, NaCl, and MSG. The detection threshold for pungency was suppressed in sucrose, while NaCl and MSG had no effect; the area under the pungency curve increased with NaCl and MSG, and decreased significantly in sucrose. TDS resolved the spatial and temporal movement, within the mouth, of tingling, astringency, vibration, and numbness, with significant differences found in maximum rate, peak time, and duration.
Limitations: The first author shares a surname with the first author of another 2018 source in this collection on the same compound class, so independence from existing evidence is limited. Panel size is not stated in the abstract; the full text has not been obtained.
Verbatim: "Temporal oral burn profiles for cinnamaldehyde, piperine and capsaicin were quantified using time-intensity (TI) parameters." Compared with cinnamaldehyde, capsaicin was 1.5 times, 2.0 times, and 3.0 times greater in time to maximum, plateau time, and total time, respectively. This shows that, 20 years earlier than the Sichuan pepper and baijiu studies, an already-unrelated group had applied TI to oral chemesthesis as early as 1993.
Limitations: The number of assessors is not stated in the abstract. It covers only burning/irritation and does not address tingling-type chemesthesis.
The TI results showed that in the high-SPU group (>150,000 SPU), AUC, maximum intensity (Imax), and duration (Ttot) were more than 13,482.91, 42.05, and 900 s respectively, whereas in the low-SPU group (<150,000 SPU) these were lower, at about 4,201.06, 16.61, and 700 s respectively. Verbatim: "TI results showed that the high area under the curve (AUC), high intensity of maximum (Imax) and long duration time (Ttot) in high SPU samples were more than 13,482.91, 42.05, and 900 s, respectively, while these values were as low as 4,201.06, 16.61, and 700 s, respectively, in the low SPU samples."
Limitations: Panel size and statistical-test details are not stated in the abstract. The full text (Methods/Results) has not been confirmed. The abstract was quoted verbatim, but the units and scale definitions of AUC/Imax/Ttot themselves can only be confirmed in the full text.
Context: The Scoville Pungency Units (SPU) of the 9 samples ranged from 38,000 to 460,000. Verbatim: "The SPU of samples was determined as ranging from 38,000 to 460,000."
Limitations: Details of the SPU measurement method (dilution series, panel) are not stated in the abstract.
Show dataconcept-time-intensity-method —is_used_as→ "A method for characterizing time-dependent oral chemesthetic sensations, such as the tingling sensation of Sichuan pepper"
The same time-intensity method is also used for subjects other than huajiao. In a study of the bitterness of theasinensin A, a tea catechin dimer, aqueous solutions at concentrations of 200, 400, 600, 800, and 1000 µmol/L were held in the mouth for 10 seconds and then expectorated, with bitterness recorded at 10-second intervals up to 240 seconds; it is reported that the duration of bitterness lengthened as concentration rose. The application to huajiao itself remains confirmed only at the abstract level, but the procedure of this method has been confirmed in the full text.
time-intensity sensory evaluation is used in: Bitterness measurement of the tea catechin dimer theasinensin A (concentrations 200, 400, 600, 800, and 1000 µmol/L; held in the mouth for 10 seconds then expectorated, with bitterness intensity recorded at 10-second intervals up to 240 seconds, at 25°C)supported
Scope: Human panel trial by Zhang et al. (2023)
Note: The source reports that the duration of bitterness lengthened as concentration increased. This study covers bitterness alone; there is no description connecting it to sweetness or huigan.
3 mL of sample solution was held in the mouth for 10 seconds and then expectorated, with bitterness intensity recorded every 10 seconds for up to 240 seconds. Concentrations of 200, 400, 600, 800, and 1000 µmol/L, at 25°C. The source reports that the duration of bitterness lengthened as concentration increased.
Limitations: Covers bitterness alone; there is no description connecting it to sweetness or huigan. The number of panelists could not be clearly confirmed in the text.
Show dataconcept-time-intensity-method —is_used_in→ "Bitterness measurement of the tea catechin dimer theasinensin A (concentrations 200, 400, 600, 800, and 1000 µmol/L; held in the mouth for 10 seconds then expectorated, with bitterness intensity recorded at 10-second intervals up to 240 seconds, at 25°C)"
NIGAMI LAB has confirmed the about-50-Hz perceived frequency, the proposed sanshool mechanism, and the reduction in tactile sensitivity in the full text of the original papers. The method of characterizing huajiao's stimulus by the time-intensity method has been confirmed only at the abstract level, across multiple papers.
「約50Hz」 (in translation: "about 50 Hz") is the perceived frequency under the condition of huajiao applied to human lips, and this has not been generalized to every tingling stimulus.
Frequently asked
Why does huajiao's tingling happen?
The mechanism of huajiao's tingling itself has still not been fully identified. Experiments using rat sensory neurons report that sanshool strongly inhibits KCNK, a 2P-domain potassium channel. Human experiments, meanwhile, have measured that the perceived frequency of tingling matches between lip and finger, and that tactile sensitivity drops during tingling — but neither of these measures the channel itself. So the molecular mechanism of tingling in humans remains at the level of an inference from animal experiments.
Is huajiao's tingling related to bitterness?
In a study comparing 9 samples of Chinese huajiao by stimulus strength, samples with a higher pericarp Scoville Pungency Unit were also reported to be rated with higher intensity for oral sensations including bitterness. However, this result is only a comparison between 2 groups — high-SPU and low-SPU — and the strength of the correlation or the significance level for bitterness alone has not been confirmed. Within the scope of this material, no measurement is found that directly links the tingling frequency itself to the strength of bitterness.
What is not known
The involvement of the KCNK 2P-domain channel has not been measured in humans and remains confined to experiments in rat sensory neurons.
Whether the perceived tingling frequency comes out at the same about-50-Hz for pungent substances other than huajiao could not be confirmed from this material.
The number of panelists in the panel test that measured huajiao's pungency by the time-intensity method is not stated in the abstract and is unknown.
The correlation coefficient and significance level between bitterness intensity and stimulus strength remain at the level of an abstract-level description and have not been measured.
An experiment that used natural sanshool to confirm, at the same time, both the tingling frequency and the reduction in tactile threshold has not yet been found.