Seeing Each Other Matters: Visual Contact Enhances Heart Rate Variability Synchrony Between Interaction Partners

Seeing Each Other Matters: Visual Contact Enhances Heart Rate Variability Synchrony Between Interaction Partners | Authorea try { document.documentElement.classList.add('js'); } catch (e) { } var _gaq = _gaq || []; _gaq.push(['_setAccount', 'G-8VDV14Y67G']); _gaq.push(['_trackPageview']); (function() { var ga = document.createElement('script'); ga.type = 'text/javascript'; ga.async = true; ga.src = ('https:' == document.location.protocol ? 'https://ssl' : 'http://www') + '.google-analytics.com/ga.js'; var s = document.getElementsByTagName('script')[0]; s.parentNode.insertBefore(ga, s); })(); Skip to main content Preprints Collections Wiley Open Research IET Open Research Ecological Society of Japan All Collections About About Authorea FAQs Contact Us Quick Search anywhere Search for preprint articles, keywords, etc. Search Search ADVANCED SEARCH SCROLL This is a preprint and has not been peer reviewed. Data may be preliminary. 24 February 2026 V1 Latest version Share on Seeing Each Other Matters: Visual Contact Enhances Heart Rate Variability Synchrony Between Interaction Partners Authors : Nina Volkmer 0009-0009-9628-9921 [email protected] , Stella Wienhold 0009-0003-8275-4673 , Bernadette Denk 0000-0002-5836-8815 , and Jens Prüssner Authors Info & Affiliations https://doi.org/10.22541/au.177191837.70605191/v1 282 views 100 downloads Contents Abstract Information & Authors Metrics & Citations View Options References Figures Tables Media Share Abstract Introduction: Interpersonal synchrony is a phenomenon occurring across a range of physiological processes during social interactions. While synchrony has been linked to various interpersonal and performance aspects, the mechanisms by which synchrony is established remain predominantly unclear. Our main aim of this study was to explore the impact of the visual system on physiological synchrony. We also incorporated a lie detection task as a performance measure into the study design. Methods: N =35 participants conducted the experiment in two different dyadic constellations. They completed two lie detection sequences per dyad, once while being able to see each other (VISION condition) and once while being separated by an opaque barrier (NO VISION condition). Heart rate variability (HRV) was recorded continuously and subsequently transformed to cross-wavelet power values for synchrony analysis for four distinct frequency ranges (0.5-0.25, 0.25-0.13, 0.13-0.06, 0.06-0.04 Hz). Results : Physiological synchrony was higher in the VISION than in the NO VISION condition for three of the four examined frequency ranges (0.5-0.25, 0.25-0.13, 0.13-0.06 Hz). No association could be established for the lowest frequency range (0.06-0.04 Hz). Lie detection did not differ significantly between the two conditions and on its own did not have a significant main effect on synchrony. In one frequency range a positive interaction between lie detection and condition emerged (0.25-0.13 Hz). Discussion : Results clearly demonstrate that the visual system plays a role in establishing physiological synchronisation. While no relation of lie detection with condition and synchrony on their own could be found, an interplay between lie detection and visual information on physiological synchrony could be observed in one frequency range. Our findings add to the literature describing physiological synchrony as a multifaceted phenomenon. Introduction Everyday social interaction requires individuals to quickly infer others’ internal states and intentions from relatively few external cues. Social encounters often lead interaction partners to coordinate across multiple dimensions, including motor behavior, autonomic responses, and brain activity - a phenomenon often referred to as interpersonal synchrony (Hoehl et al., 2021). In the field of physiology, interpersonal physiological synchrony (IPS) describes the temporal alignment of partners’ physiological signals, such as heart activity, autonomic nervous system (ANS) activity, and neural oscillations (Palumbo et al., 2017). Cardiac synchrony, measured for example by coupled fluctuations in heart rate (HR) and heart rate variability (HRV), has been observed in parent-child pairs, romantic partners, and strangers performing in shared tasks or having emotionally engaging conversations (e.g. Marzoratti & Evans, 2022; Mayo et al., 2021). Higher IPS has been associated with greater perceived closeness, better cooperation, and more effective joint performance (Mayo et al., 2021). Although more and more studies are examining IPS in various contexts, the underlying mechanisms remain elusive. Specifically, the situational conditions and sensory channels that promote or constrain IPS are not yet well understood. A better understanding of when and how IPS arises is crucial for the design and interpretation of psychophysiological research on social interactions. Cardiac synchrony is often derived from individual HRV recordings. HRV is an established quantitative marker of cardiovascular regulation by the ANS. In particular, higher HRV is typically associated with greater parasympathetic influence on the heart and greater autonomic flexibility (Sammito et al., 2024; Thayer & Lane, 2000). From this perspective, HRV-based IPS can serve as an index of the extent to which partners’ autonomic states are coordinated and dynamically regulated together. Such coupling is thought to support successful social exchange and bonding by enabling rapid coordination of affective and behavioral responses (Hoehl et al., 2021; Mayo et al., 2021). One possible mechanism by which IPS is achieved is the visual system. Pupil diameter is directly controlled by the autonomic nervous system (sympathetic dilatation and parasympathetic constriction, see McCorry, 2007). Accordingly, fluctuations in pupil diameter reflect autonomic rhythms and have been shown to correlate with HRV measured via heart monitors, suggesting a close relationship between the ANS and ocular indices of arousal (Parnandi & Gutierrez-Osuna, 2013). This relationship implies that visual stimuli, especially eye contact, may be an important driver of IPS (Sato & Sato, 2025). Consistent with this idea, hyperscanning studies using functional near-infrared spectroscopy (fNIRS) and functional magnetic resonance imaging (fMRI) have shown that synchronization between brains is often higher during face-to-face interactions than during non-face-to-face or online interactions, with coherence being particularly enhanced during tasks involving mutual eye contact or shared intentionality (De Felice et al., 2025; Hirsch et al., 2017; Sato & Sato, 2025). For example, a fNIRS study found higher synchronization between brain activations in the right temporo-parietal junction during face-to-face interaction, but not during interaction with a concealed face (Kinreich et al., 2017). Synchronization, interpreted as a mechanism for the high-level perception of rapidly evolving social signals, tends to be higher in genuine face-to-face interactions than viewing pre-recorded videos or static images of a partner (Czeszumski et al., 2020; Hakim et al., 2023). On the physiological level, increased synchronization in face-to-face interaction has also been observed for both skin conductance and heart rate (Behrens et al., 2020). In addition, the importance of eye contact extends beyond simple sensory and autonomic processes to social cognition and interpersonal assessment. Eye contact is a powerful signal that indicates moments of shared attention and interaction. In natural conversations, mutual glances indicate a shift in shared attention and are accompanied by rapid changes in pupil synchrony (Wohltjen & Wheatley, 2021). Eye contact is further associated with higher self-reported engagement ratings by conversation partners (Wohltjen & Wheatley, 2021). Within the framework of the Polyvagal theory, eye contact is considered a core component of the “social engagement system”: it is compatible with the state supported by the vagal brake, in which sympathetic activity is inhibited to enable calm, socially engaged behavior (Porges, 2007, 2022; see also Niedźwiecka, 2023). From this perspective, well-regulated individuals may benefit from eye contact by exhibiting improved cognitive and social performance, while individuals with poorer regulation may find eye contact intrusive or distracting. This line of research suggests that visual cues, particularly eye contact, are not only powerful synchronizers but may also improve performance on tasks involving social inference. Finally, eye contact also plays an important role in people’s ideas about deception. Across cultures, non-experts commonly believe that liars avoid eye contact and that averting one’s gaze is a diagnostic indicator of deception (The Global Deception Research Team, 2006). Similar beliefs are also common among professionals who routinely make credibility judgments, such as police officers and judges (Bogaard et al., 2016; Bond & DePaulo, 2006; Strömwall & Granhag, 2003). Meta-analytic studies confirm that such beliefs hardly correspond to the actual diagnostic value of eye contact cues. In many studies and for many cues, eye behavior is at best a weak and unreliable indicator of lying (DePaulo et al., 2003; Sporer & Schwandt, 2007; Vrij et al., 2019). Overall, the accuracy of lie detection is usually only slightly above chance, regardless of whether observers have access to visual cues or must rely solely on auditory information (Bond & DePaulo, 2006; Loy et al., 2018). Accordingly, there are very few studies that have directly examined whether the mere possibility of eye contact with an interaction partner improves the observer’s ability to decide whether that partner is lying or telling the truth, compared to otherwise equivalent interactions without eye contact. Bringing this literature together reveals the possibility that eye contact may be one mechanism through which IPS influences social judgments, such as lie detection. If eye contact facilitates the alignment of attention and emotional contagion, it may also increase the extent to which partners’ autonomic responses are coupled. Conversely, it has been suggested that greater IPS reflects a deeper bond with the partner and more accurate tracking of their internal states (Mayo et al., 2021; Palumbo et al., 2017; Tomashin et al., 2022; see also Porges, 2007 for HRV as an index of social engagement ability). Such engagement could, in principle, provide richer information for assessing the truthfulness of a partner’s statements, for example, by making subtle changes in arousal or affect more detectable. To our knowledge, no study has directly examined whether HRV-based IPS during face-to-face interaction predicts accuracy in assessing truth and lies on an experimental basis, or whether such a relationship depends on the presence of eye contact. The present study fills these gaps by combining a dyadic deception paradigm with continuous recording of heart activity under controlled variations of eye contact (VISION vs. NO VISION condition). Based on the theoretical background and as pre-registered in the Open Science Framework (OSF; https://doi.org/10.17605/OSF.IO/DHGKP, Wienhold et al., 2025), we formulated three hypotheses regarding the relationships between HRV-based IPS, eye contact, and lie detection. First, we hypothesized that HRV synchrony would be higher under the VISION condition than under the NO VISION condition (H1). Second, we predicted higher accuracy when partners could see each other (H2). Third, we tested whether greater physiological synchrony is associated with more accurate truth assessment, regardless of visual condition (H3). By experimentally manipulating visual access while simultaneously assessing HRV synchrony, and deception judgments, this study aims to clarify whether eye contact is a necessary or facilitating condition for physiological synchrony and whether synchrony itself contributes to accurate social evaluation. Participants and Recruitment This study adhered to the ethical principles of the Declaration of Helsinki and was approved by the Institutional Review Board of the University of Konstanz, Germany (IRB statement 28/2022). It was conducted as part of an experimental lab course. All students partaking in the course were invited to participate voluntarily and without compensation. Participants could also object to their data being used outside the scope of the course. The final sample consisted of N = 35 psychology students with a mean age of 21.80 ( SD = 2.44, Range = 19-31). Participants predominantly indicated female sex (88.57%) and gender affiliation (see Table S1 in the Supplementary Information for participant characteristics). Three groups (two groups with n=12 and one group with n=11) were tested on three separate time points occasions. Because an odd group number did not allow for every participant to be allocated to a dyad. This setup resulted in 12 and 10 dyadic observations per group, respectively. Design We implemented a 2x2 experimental design. We manipulated whether participants were able to see each other (VISION vs. No VISION) by varying the presence of wooden separators between them during the interaction phases. The order of conditions was counterbalanced across the three groups. Participants were randomly matched with two dyadic partners within their group. With their first dyadic partner, they completed the interaction phase, starting with their predetermined condition order (VISION or NO VISION). With the second partner, they would then engage in the opposite VISION condition order. In each interaction phase, participants were asked to tell a truth or a lie, whereby they were free to choose the order of presentation. The order of speaking was also randomized. As participants were involved in the study’s conceptualisation, no blinding was involved. Procedure and Material In line with recommendations (Laborde et al., 2017), participants were asked to follow their regular sleep schedule, as well as not to consume alcohol or partake in strenuous physical activity for 24 hours, nor consume caffeinated beverages for two hours before experimental start. Additionally, they were advised to wake up at least one hour before data collection so that they would be able to arrive on time without needing to hurry. After providing informed consent, participants were equipped with a 2-channel ECG measurement device Polar Team Pro (Polar Electro Oy). ECG measurements were recorded with a temporal resolution of 1ms (≈ 10000 Hz sampling rate) and monitored over an iPad (Apple Inc.) and the Polar Team Pro App (Polar Electro Oy). Participants completed a three-minute baseline measurement sitting spread out across the testing room while looking at the room’s edges. Thereafter, they sat down at a desk in front of their first dyadic partner sitting on the other side, either seeing (VISION condition) or not seeing (NO VISION condition) the interaction partner through the use of a wooden barrier (roughly 60 x 60 cm) between them . A study-accompanying questionnaire (Questionstar, 2011) was distributed on an iPad (Apple Inc., n.d.). Participants were asked to provide information about day-specific control variables for HRV measurement and their current affect (Positive and Negative Affect Scale; Breyer & Bluemke, 2016). After completion, they took part in two interaction phases of two minutes each. Each interaction phase consisted of one participant providing either a true or false statement and subsequent follow-up questions by the other participant. Thereafter, participants answered questions regarding the truth content of their statement as well as whether they believed their counterparts’ claim. This was followed by another set of interaction phases, with the second statement. Thereafter, dyads performed the same interaction procedure again under the reversed VISION condition. Finally, they provided information on their relationship with their dyadic partner, followed by a three-minute discussion about the truth content of their statements. Participants then returned to their original seats to conduct a mid-session baseline measurement for three minutes. The interactional procedure was subsequently repeated with the second dyadic partner, starting with the reversed VISION condition order. The experiment ended with another three-minute resting baseline (see Figure 1). Demographic and anamnestic data as well as more stable state questionnaires (Perceived Stress Scale; Schneider et al., 2020) were assessed separately. Figure 1 Experimental Procedure Note . Experimental procedure displaying the two possible condition orders (Variant A and Variant B). Dyads are either able to see each other (VISION) or separated by an opaque barrier (NO VISION), indicated by a black line. Participants are allocated to to two different dyadic constellations separated by baseline phases (indicated by a shift in colour). In Variant A dyads start with the VISION condition and then move to the NO VISION condition. In Variant B dyads start with the NO VISION condition and then move to the VISION condition. Defined length of experimental phases is indicated in minutes. Preprocessing and Computation of Synchrony Preprocessing, statistical analysis and graphical representation were performed in R (Posit team, 2025; R Core Team, 2025b), additionally relying on the packages RHRV (Rodriguez-Linares et al., 2024), parsedate (Csárdi & Torvalds, 2024), haven (Wickham et al., 2025), foreign (R Core Team, 2025a), dplyr (Wickham et al., 2023), lubridate (Grolemund & Wickham, 2011), stringr, ggplot2 (Wickham et al., 2016), imputeTS (Moritz & Bartz-Beielstein, 2017), tclck (R Core Team, 2025b), corrplot (Wei & Simko, 2024), svDialogs (Grosjean, 2025), reshape2 (Wickham, 2007), tidyr (Wickham et al., 2024), lavaan (Rosseel et al., 2025) and WaveletComp (Roesch & Schmidbauer, 2025), tidyverse Wickham et al., 2024, nlme (Pinheiro et al., 2025; MuMIn Bartoń, 2025), rstatix (Kassambara, 2023), sjPlot (Lüdecke, 2025), flextable (Gohel & Skintzos, 2026), ggthemes (Arnold, 2025), ggbeeswarm (Clarke et al., 2025), ggdist (Kay, 2025) and jtools (Long, 2022). Our physiological data was preprocessed using in-house scripts. We first corrected outliers and interpolated missing values from raw RR-data. Outliers were defined as RR-intervals differing between 25 to 50% from the previous recorded value using a variable threshold, depending on absolute RR-intervals and HRV magnitude, with the aim of correcting less than 10% of the original data. If a threshold greater than 50% was necessary to correct the recording, that dataset was excluded. In addition, missing values resulting from outlier deletion were interpolated based on surrounding values. If more than 30 consecutive seconds were missing, the dataset was also excluded from data analysis. From the cleaned RR-data, we then interpolated one HR value per second to obtain time series of equal length across participants within the same dyad. We then computed wavelet power for each participant and cross-wavelet power (CWP) for each dyad (WaveletComp package:(Roesch & Schmidbauer, 2025). Maximal values were extracted in 30-second intervals and by four frequency bands – upper high frequency (UHF; 0.50-0.25 Hz/ 2-4 seconds per cycle), lower high frequency (LHF; 0.25-0.125 Hz/ 4-8 seconds per cycle), upper low frequency (ULF, 0.125-0.063 Hz/ 8-16 seconds per cycle), and lower low frequency (LLF; 0.063-0.042 Hz/ 16-24 seconds per cycle). We obtained these different frequency ranges by splitting the traditional high and low frequency bands in half. We did this to take advantage of the wavelet method’s ability to provide a more fine grained temporal and frequency resolution (Morlet et al., 1982). We can therefore capture the dynamic relationship in interpersonal settings for smaller frequency bands and time intervals (Issartel et al., 2015). The 30-second intervals are additionally grouped by experimental phase to be able to perform subsequent analyses. To allow comparison across dyads, CWP values were z-standardized per frequency band and per session prior to statistical analysis. Further preparation of physiological, anamnestic, protocol and experimental data involved calculation of mean and sum scores as well as the creation of combined variables for the testing of control variables. Due to our dyadic set-up control variables for dyadic HRV synchrony were either separated into a higher and lower column (e.g., higher age and lower age present in the dyad) or by establishing the relation between the two dyad partners (e.g., both female). From N =34 available dyadic recordings n =9 dyads were excluded due to excessive amounts of missing data during the recording, or failure to correctly preprocess, leading to a final dyadic sample size of N =25 for hypotheses 1 and 3. One participant was excluded from hypothesis 2, because of not providing one lie detection assessment, leading to N =34 for this hypothesis. Statistical Analysis All hypotheses were tested at a significance value of p <.05. Statistical analysis was conducted in line with the preregistered plan, except when indicated otherwise. The exploratory analysis has not been performed yet. Hypothesis 1: Association between HRV Synchronization and Visual Contact To account for our nested data structure, we set up separate linear mixed-effects models for our frequency bands of interest (UHF, LHF, ULF, LLF). We compared more complex models with the previous best model fit by reverting to ANOVAs including likelihood-ratio tests (p <.05). Where ANOVA comparisons are not possible, we rely on AIC and BIC comparisons. The model incorporated HRV data of the interactional phases in which participants were providing their statements and had to decide whether to believe their counterpart. Our data consisted of one z-transformed CWP coefficient per 30-second interval, nested within dyads (1042 intervals across 25 dyads). We built a basic model representing the data structure first before adding our predictor of interest, starting with random intercepts for dyads. We subsequently fitted an AR(1) residual correlation structure to account for temporal autocorrelation across repeated 30-second intervals within the dyads. For our basic model structure, we then checked whether time interval exerted a significant fixed effect and, if significant, also a random effect. For reasons of simplicity, we deviated from the preregistered analysis plan insofar that we did not insert any random effects apart from the dyadic random intercept, and restricted analysis to the interactional phases. We additionally inserted a temporal autocorrelation structure in order to properly account for our repeated measurements prompting us to revert to the nlme (Pinheiro et al., 2025) for the computation of mixed-effect models. We subsequently added potential confounding variables (see Laborde et al., 2017; Sammito et al., 2024 for overviews) individually to our model. A range of demographic, health, and lifestyle-related covariates (e.g., age, BMI, diseases, medication, physical activity, sleep), as well as interactional and experimental factors (e.g., relation between participants, group) were tested separately. Control variables were retained only if they significantly improved model fit, and retested in combination. Thereafter, the fixed effects of absolute HRV of participants (mean wavelet power over the course of the experiment, z-standardized within frequency band) were tested, to account for individual activity contributing to synchrony estimates, before accounting for condition order. After establishing our control model, the experimental condition (VISION vs. NO VISION, Reference = NO VISION) was entered as a fixed effect. Hypothesis 2: Lie Detection and Vision To test if participants were significantly better in truth assessment when seeing each other, we created scores for individual detection per condition by summing lie detection accuracy across both dyads. We then compared the sum of correctly assessed statements per participant in the VISION condition with those in the NO VISION condition by means of a one-sided paired t-Test (rstatix package: Kassambara, 2023). To control for counterbalancing order, we subsequently conducted an ANCOVA. Contrary to the preregistration and due to simplicity, we did not additionally perform a mixed-effects model. Hypothesis 3: Lie Detection and HRV Synchronisation For Hypothesis 3, we set up basic and control linear mixed-effects models. Due to one missing detection statement, models were built on a slightly reduced dataset (1032 intervals across 25 dyads). To account for evoked differences during the two experimental conditions, we also added condition before inserting our dyadic detection variable. Dyadic detection was computed for each interactional sequence with a subsequent rating. Detection accuracy for that sequence was summed across both participants in a dyad. Dyadic detection was added as a fixed main effect and exploratively in interaction with condition. Changes from the preregistered analysis plan mirrored those of hypothesis 1. From the preregistered analysis plan we chose to compute model 1 because of its increased resolution in comparison to model 2. Due to working with dyadic intercepts, we additionally computed dyadic detection scores instead of working with individual values per participant. Sample size and power Sensitivity analyses in G*Power (version 3.1) indicated that medium-to-large effects could be detected for all hypotheses while the majority of small effects would not be found. Results As proof of principle, we tested whether z-standardized CWP captured event-related changes in HRV synchrony throughout our experimental procedure by modeling events (individual baseline and interaction) in mixed-effects models at the dyad level with an AR(1) residual correlation structure to account for temporal autocorrelation across repeated 30-s intervals. In all bands, the addition of experimental phase significantly improved model fit compared to the pure autocorrelation model (all likelihood ratio tests p < .001), indicating that z-standardized CWP synchrony values captured systematic fluctuations in dyadic HRV synchrony over the course of the procedure. Across all dyads, z-standardized CWP showed robust increases from the baseline phase, where participants sat isolated in the room facing the window or a wall, to interaction phases in the slower bands (ULF: b = 0.75, p < .001 ; and LLF: b = 0.53, p < .001) and in the lower portion of the high frequency band (LHF: b = 0.54, p < .001). In contrast, higher high frequnency band showed the opposite pattern, with the interacting phase exhibiting lower synchrony compared to baseline (UHF: b = -0.17, p = .002). Across all bands, the addition of the event variable significantly improved model fit compared to the pure autocorrelation model (all likelihood ratio tests p < .001), confirming the suitability of the z-standardized CWP approach for detecting experimentally induced, event-related changes in dyadic HRV synchrony. The corresponding contrasts within the dyads are shown in Figure 2. Figure 2 Event-dependent HRV Synchrony Estimated by Cross-Wavelet Power. Note. Dots represent dyad-level means of HRV synchrony (z-standardized CWP) for the contrast Interaction - Individual baseline within each frequency band, with black point indicating grand mean, and error bars showing 95% bootstrapped confidence intervals. Positive values indicate higher synchrony during interaction relative to baseline. Hypothesis 1: Association between HRV Synchronization and Visual Contact Hypothesis 1 stated that HRV synchrony between interaction partners would be higher in the VISION than in the NO VISION condition. To test this, we quantified cardiac IPS as z-standardized CWP between partners’ heart rate time series (z-standardized within each frequency band). Higher z-standardized CWP values reflect stronger HRV synchronization between partners. The effect of the experimental condition on synchronization was tested using multilevel mixed-effects modeling. Across bands, only a small number of covariates improved fit and were therefore retained in the final models (see Supplementaries Table S2-S5). . Supporting Hypothesis 1, visual contact led to a significant increase in HRV synchrony in three of the four frequency bands, with higher z-standardized CWP in Vision than in No Vision (UHF: b = 0.32, p < .001; LHF: b = 0.32, p < .001; ULF: b = 0.22, p = .009). In contrast, the addition of the experimental condition in the LLF did not improve model fit, and was not statistically significant, b = 0.01, p = .950, suggesting that there is no evidence of increased HRV synchrony due to vision contact in this frequency range (see Figure 3). Table 1 contains the predictor models for UHF, LHF, ULF and LLF. The full model comparison statistics are reported in Tables S2-S5 in the Supplementary Information. Beyond the condition effect, a consistent pattern emerged across all frequency bands. Dyads in which partners exhibited greater individual HRV activity also showed greater synchrony. Specifically, both wavelet powers were positively associated with stronger HRV synchronization, as dyads in which the partners exhibited greater individual variability in a given band (higher WP) also exhibited higher z-standardized CWP in that band (e.g., UHF: b WP1 = 0.28, p < .001; b WP2 = 0.21, p = .003 LHF; similar positive correlations were observed for the other three frequency bands). No systematic linear time trends in synchrony were observed once temporal autocorrelation via the AR(1) structure was taken into account. Table 1 Mixed-effects models predicting dyadic HRV synchrony (Hypothesis 1) Frequency band Predictors b 95% CI p UHF Intercept -0.20 -0.37 – 0.71 <.001 lower BMI (z) 0.01 -0.09 – 0.12 .809 higher mean WP within dyad (z) 0.27 0.17 – 0.37 <.001 lower mean WP within dyad (z) 0.13 0.04 – 0.22 .008 condition [Vision] 0.32 0.18 – 0.47 <.001 LHF Intercept 0.66 0.36 – 0.96 <.001 higher alcohol consumption: two to four times a month -0.62 -0.92 – -0.32 <.001 higher alcohol consumption: two to three times a week -0.60 -0.96 – -0.24 .002 higher exercise hours -0.00 -0.02 – 0.01 .661 mean WP of participant 1 (z) 0.16 0.06 – 0.26 .004 mean WP of participant 2 (z) 0.25 0.15 – 0.35 <.001 condition [Vision] 0.32 0.17 – 0.48 <.001 ULF Intercept 0.19 -0.04 – 0.42 .112 lower alcohol consumption: less than once a month -0.18 -0.48 – 0.12 .232 lower alcohol consumption: two to four times a month -0.26 -0.65 – 0.12 .170 higher dyad exercise hours 0.01 -0.01 – 0.04 .304 mean WP of participant 1 (z) 0.14 -0.03 – 0.31 .093 mean WP of participant 2 (z) 0.36 0.16 – 0.55 .001 condition [Vision] 0.22 0.06 – 0.38 .009 LLF Intercept 0.15 0.00 – 0.30 .049 lower age (z) 0.05 -0.08 – 0.19 .428 hurry to study [one participant] 0.08 -0.22 – 0.38 .586 hurry to study [both participants] -0.19 -0.77 – 0.38 .488 mean WP of participant 1 (z) 0.19 0.04 – 0.35 .014 mean WP of participant 2 (z) 0.11 0.00 – 0.23 .050 condition [Vision] 0.01 -0.17 – 0.18 .950 Frequency band ΔLR extended vs. control variable model p σ² τ₀₀ (Dyad) ICC Marginal R² / Conditional R² UHF 18.75 <.001 0.88 0.00 0.01 0.137 / 0.141 LHF 16.62 <.001 0.94 0.00 0.00 0.169 / 0.171 ULF 6.86 .009 0.86 0.00 0.00 0.170 / 0.170 LLF 0.00 .950 0.86 0.00 0.00 0.098 / 0.098 Note . Overview of final predictor models for the frequency ranges upper high (UHF, lower high (LHF), upper low (ULF) and lower low (LLF). Outcome: dyadic HRV synchrony (z-standardized CWP) measured during interaction phases. Estimates: fixed-effect coefficients (standardized predictors where indicated by a z). Provided are fixed slopes (b), 95% confidence intervals (CI) and p-values (p). ΔLR values reflect likelihood-ratio tests comparing extended versus control variable models, σ² denotes residual variance, τ₀₀ the dyad-level random-intercept variance, ICC the intraclass correlation coefficient; and R² values are marginal and conditional R². N = 25 dyads, 1042 observations. Figure 3 HRV Synchrony (Cross-Wavelet Power) by Condition Across Frequency Bands. Note. Points represent 30-s intervals nested within dyads. Violin shapes depict the kernel density of observations; embedded boxplots indicate the median and interquartile range. For visualization, raw (non-z-standardized) cross-wavelet power values are displayed to illustrate differences in absolute magnitude across frequency bands. Statistical analyses were conducted on within-band z-standardized CWP values. In all cases, higher CWP values indicate stronger HRV synchrony. Hypothesis 2: Lie Detection and Vision We expected participants to be better at detecting inaccurate statements when seeing each other. While detection scores were generally higher in the VISION ( M = 2.41, SD = 1.16) compared to the NO VISION ( M = 2.06, SD = 1.22) condition, this difference was not significant, t (31) = 1.15, p = .130. When adding the control variable counterbalancing, no significant effect of the variable itself (p > .05) or significant changes in the effect of the condition variable were found. Due to slightly skewed normality and outliers, we additionally performed a Wilcox signed-rank test, which yielded comparable results, V =164, p =.112. Hypothesis 2 is therefore not supported. Figure 4 Individual Lie Detection Score by Condition Note. Boxplot (mean and interquartile ranges) of individual lie detection score summed across both dyadic interactions and separated by condition (Range 0-4). Higher values signify better lie detection. Mean by condition is denoted as a black point. Individual data points are denoted in grey. Shading displays the distribution of datapoints. Hypothesis 3: Lie Detection and HRV Synchronisation For hypothesis 3, we expected better lie detection to be linked to greater physiological synchronisation within a dyad. Due to missing one lie detection statement, sample size for this analysis was slightly decreased. The basic and control variable models were set up analogously to hypothesis 1, but on the smaller dataset. Condition was additionally tested as a covariate. All frequency bands contained an autocorrelational factor for the repeated measurement of cross-wavelet power and additionally incorporated the mean wavelet power of one (LLF) or both (UHF, LHF, ULF) individuals. Individual control variables improved model fit significantly. The final models included minimal BMI in the dyad and condition for UHF, maximal exercise hours, maximal alcohol frequency, and condition for LHF, maximal exercise hours, minimal alcohol frequency, and condition for ULF, as well as minimal age for LLF. Other control variables significantly improved model fit when added to the model individually but not once the control variable with the best model fit was already added. Cardiovascular and psychological diseases also exerted significant effects for the ULF and LLF model but were not pursued due to missing data. Individual wavelet power values were added for all frequencies. Across all frequency bands, there was no main effect of lie detection on HRV synchrony (UHF: b = -0.03, p = .438; LHF: b = 0.01, p = .737; ULF: b = 0.04, p = .317; LLF: b = 0.02, p = .667). This changed somewhat when the interaction between condition and lie detection was added to all models. In the LHF band this interaction effect was significant (UHF: b VISION x detection dyadic = 0.07, p = .346; LHF: b VISION x detection = 0.20, p = .013; ULF: b VISION x detection = 0.12, p = .144 ), but did not improve model fit. Dyads detecting at least one lie in the VISION condition appear to be synchronising more, with some dyadic CWP values surpassing a score of 0.75 in comparison to all dyadic CWP values remaining under 0.50 when not correctly detecting any lie or truth (see Figure 5a). Conversely, in the NO VISION condition, the CWP pattern when neither partner detected the statement correctly was more variable with some CWP values surpassing 0.50 (see Figure 5b). Overall, we did not find support for a positive link between lie detection and physiological synchrony across conditions and therefore no support for our hypothesis. In the LHF band we found a positive effect only in the VISION and a negative effect in the NO VISION condition. For reasons of space, Table 2 reports the final interaction models for UHF, LHF, and ULF only; the corresponding main-effects-only models for all four frequency bands, and the full model comparison statistics are reported in Tables S6-S10 in the Supplementary Information. Table 2 Mixed-effects models predicting dyadic HRV synchrony (Hypothesis 3) Frequency band Predictors b 95% CI p UHF Intercept -0.20 -0.31 – 0.09 <.001 lower dyad BMI (z) 0.01 -0.10 – 0.12 .821 higher partner mean WP within dyad (z) 0.27 0.16 – 0.38 <.001 lower partner mean WP within dyad (z) 0.13 0.04 – 0.23 .009 condition [Vision] 0.33 0.19 – 0.48 <.001 dyadic detection (z) -0.07 -0.17 – 0.04 .224 condition [Vision] x dyadic detection (z) 0.07 -0.08 – 0.22 .346 LHF Intercept 0.60 0.28 – 0.93 <.001 higher partner alcohol consumption: two till four times a month -0.59 -0.91 – -0.26 .001 higher partner alcohol consumption: two to three times a week -0.58 -0.97 – -0.20 .005 higher dyad exercise hours (z) -0.00 -0.02 – 0.02 .880 mean WP of participant 1 (z) 0.16 0.05 – 0.27 .008 mean WP of participant 2 (z) 0.25 0.15 – 0.36 <.001 condition [Vision] 0.33 0.17 – 0.48 <.001 dyadic detection -0.09 -0.20 – 0.02 .111 condition [Vision] x dyadic detection 0.20 0.04 – 0.35 .013 ULF Intercept 0.18 -0.05 – 0.401 .129 lower partner alcohol consumption: less than once a month -0.06 -0.37 – 0.26 .711 lower partner alcohol consumption: two till four times a month 0.01 -0.33 – 0.34 .971 higher dyad exercise hours (z) 0.00 -0.02 – 0.02 .893 higher partner mean WP within dyad (z) 0.20 0.04 – 0.36 .018 lower partner mean WP within dyad (z) 0.25 0.13 – 0.37 <.001 condition [Vision] 0.19 0.03 – 0.35 .018 dyadic detection -0.02 -0.13 – 0.09 .756 condition [Vision] x dyadic detection 0.12 -0.04 – 0.28 .144 Frequency band ΔLR extended vs. control variable model p σ² τ₀₀ (Dyad) ICC Marginal R² / Conditional R² UHF 1.48 .478 0.88 0.01 0.01 0.140 / 0.148 LHF 5.98 .050 0.93 0.01 0.01 0.178 / 0.187 ULF 3.16 .206 0.82 0.00 0.00 0.171 / 0.171 Note . Overview of final predictor models for the frequency ranges upper high (UHF), lower high (LHF), and upper low (ULF). Outcome: dyadic HRV synchrony ( z-standardized CWP) measured during interaction phases. Estimates: fixed-effect coefficients (standardized predictors where indicated by a z). Provided are fixed slopes ( b ), 95% confidence intervals ( CI ) and p-values ( p ). ΔLR values reflect likelihood-ratio tests comparing extended versus control variable models, σ² denotes residual variance, τ₀₀ the dyad-level random-intercept variance, ICC the intraclass correlation coefficient; and R² values are marginal and conditional R². N = 25 dyads, 1032 observations. Figure 5 HRV Synchrony Regressed from Lie Detection in the LHF band a) VISION b) NO VISION Note . HRV synchrony as z-standardized Cross-wavelet power (CWP) regressed from the dyadic detection score for each of the four interactional sequences per dyad. Interactional sequences are separated in VISION (a) and NO VISION (b). The dyadic detection score is created by summing the lie detection score of both individuals (either 0 for not detecting or 1 for detecting) in a dyad (range 0-2). The detection score used for graphical illustration is not standardized, while the detection score used in statistical analysis was z-standardized. Dots represent dyadic CWP extracted per 30-second interval. Grey shading marks a 95% confidence interval. Discussion While an increasing body of research underscores the role of IPS in social interactions, the mechanisms of how IPS emerges remain under-investigated (Gordon et al., 2025). Eye contact and information obtained from interaction partners’ eye movements are crucial components of everyday interactions (e.g. Wohltjen & Wheatley, 2021). For example, people are often convinced that they can reliably identify lying based on eye-related cues (The Global Deception Research Team, 2006). Here, we investigated whether visual contact increases IPS at the level of the heart using a cross-wavelet power-based analysis of HRV synchronization, and whether this was related to accurately detecting lying behavior. We found that seeing the interaction partner increased synchrony on most of the frequency bands selected here. However, participants could not determine lying behavior more accurately when seeing each other. Better lie detection was also not linked to greater physiological synchronisation on its own, however, in interaction with the Vision condition, it became significant in one (LHF) frequency band. Proof of Principle: Cross-Wavelet Power as a Well-Suited Index for HRV Synchrony Before testing our main hypotheses, we first examined whether z-standardized CWP reliably captured the expected changes in dyadic HRV synchrony during the experimental phases. To this end, we compared CWP across the predefined phases (individual baseline versus shared interaction phases). Across all four frequency bands, synchrony varied systematically with event structure. Phases in which partners interacted exhibited higher CWP than the individual baseline, with particularly pronounced differences observed in the lower frequency bands (LHF, ULF, LLF), and in the opposite direction in the highest frequency band (UHF). The decrease in UHF synchrony from baseline to interaction could potentially reflect a general reduction in high-frequency HRV during active engagement, consistent with vagal withdrawal and decline in vagally mediated HRV from rest to cognitive or emotional challenge (Shaffer & Ginsberg, 2017; Task Force of the ESC & NASPE, 1996). These findings support the use of CWP as a sensitive proxy for dynamic event-related changes in cardiac IPS, and are in line with the idea that time-frequency approaches are well suited to capturing non-stationarity. Visual Contact and Physiological Synchrony We hypothesized that physiological synchrony between interaction partners would be higher when partners could see each other than when eye contact was blocked. Using CWP between partners’ HR time series as an index of HRV synchrony, we found significant evidence for this prediction in three of the four here implemented frequency bands. In the UHF, LHF, and ULF bands, dyads in the vision condition showed significantly higher CWP than dyads in the no-vision condition, even after controlling for partners’ individual WP of the partners in the corresponding band and accounting for temporal autocorrelation across repeated measures. Only the LLF band showed no reliable condition effect, adding condition did not improve model fit, and the estimate was essentially zero. Thus, the pattern of results largely supports Hypothesis 1 and suggests that seeing each other is an important mechanism for establishing cardiac IPS. Conceptually, this finding is consistent with the idea that mutual eye contact and visual access to the partner provide important cues for IPS, such as facial expressions and micro-movements that support the prediction of the other’s behavior and state, emotional convergence, and behavioral coordination (Hoehl et al., 2021; Prochazkova & Kret, 2017). In addition to providing behavioural information, eye contact may influence IPS more directly via visuocardiac pathways. Autonomic fluctuations in pupil size are associated with HRV (Parnandi & Gutierrez-Osuna, 2013), and face-to-face interactions involving mutual gaze reliably enhance neural synchrony (De Felice et al., 2025; Hirsch et al., 2017; Sato & Sato, 2025). These processes are consistent with broader theoretical accounts that explain physiological coupling in dyads in terms of dynamic interpersonal coordination, affective alignment, and mutual regulation (Feldman, 2012; Helm et al., 2014; Mayo et al., 2021; Palumbo et al., 2017). In our study, the mere presence or absence of visual information, with all other aspects of the interaction held constant, was sufficient to modulate HRV synchrony, strengthening the argument that social perceptual inputs themselves and not just shared task structures or shared environments contribute to autonomic coupling. The frequency-specific pattern adds nuanced insights. The fact that the effect of vision was strong and consistent in UHF, LHF, and ULF but absent in LLF suggests that not all components of HRV are equally sensitive to visual social cues. One possibility is that the LLF band in our analysis captures slower oscillatory processes that are less directly modulated by momentary social exchanges. Another, more methodological explanation is that our here implemented 30 sec intervals are not optimal for the LLF band, which would reduce statistical power for this band. Since we used the same analysis pipeline for all bands, the more parsimonious interpretation is that visual contact preferentially influences those HRV components that are more labile and more closely related to ongoing interaction and attention, consistent with previous work linking higher-frequency HRV changes to social and emotional processes. A second consistent pattern was that higher individual WP in a given band was associated with stronger CWP in that band. Since absolute WP is a metric of HRV, dyads whose individuals had higher absolute HRV also tended to produce stronger synchronization. This can be interpreted in at least two non-mutually exclusive ways. On the one hand, it could reflect a genuine physiological phenomenon where individuals with greater autonomic flexibility or stronger rhythmicity may simply have a greater ability to attune to others. On the other hand, it could partly reflect a signal-to-noise problem: when each individual’s signal is stronger, common fluctuations are more easily recognized as synchrony. Importantly, controlling for this influence did not eliminate the effect of the condition, suggesting that the vision effect cannot be reduced to simple differences in individual HRV magnitude. We also tested a wide range of demographic and health-related covariates (age, BMI, self-reported illnesses, medications, physical activity, sleep, substance use, and others). None of these variables consistently improved model fit. This robustness argues against the assumption that the observed differences in synchrony between “VISION” and “NO VISION” were due to systematic differences in physical health, lifestyle, or basic demographic characteristics of the couples. Instead, it supports the interpretation that the experimental manipulation of visual information itself was responsible for the observed modulation of HRV synchrony. Lie detection and Visual Contact We expected lie detection to be better when participants could see each other in comparison to when being separated by an opaque barrier. Although lie detection values were descriptively higher in the VISION condition, the difference between conditions was not significant. Although this was not in line with our hypothesis, it fits into the literature not finding a particular advantage of lie detection in settings where participants can see each other (Hartwig & Bond, 2014). While some potential effective cues of deception are observable distinctively when seeing someone, others as those relating to how a statement is portrayed should also be noticeable in the NO VISION condition (compare DePaulo et al., 2003). Lie Detection and Physiological Synchrony Additionally, we expected lie detection to be linked to physiological synchrony across both conditions. We tested this by looking at the effect of detection on heart rate variability synchrony, operationalised as CWP, for four frequency ranges. We did not find a main effect of lie detection on physiological synchrony in any frequency range. In the LHF frequency band, we did however, observe differential effects of lie detection on physiological synchrony, depending on the respective condition. A greater lie detection score was linked to increased physiological synchrony in the VISION and decreased physiological synchrony in the NO VISION condition. Overall, our results at this point add to a somewhat more inconsistent relation of physiological synchrony to outcome measures found across studies (Gordon & Bartsch, 2026). Better lie detection being linked to IPS when being able to see each other fits with our general assumption that synchrony may help evaluate deception judgements by reducing complexity and providing additional information regarding for example their affect or authenticity (Hoehl et al., 2021; Kret, 2015; Prochazkova & Kret, 2017). Pupil dilation could be a potential mechanism, as it relates to HRV (compare Parnandi & Gutierrez-Osuna, 2013) and has also been noted as a deception cue (DePaulo et al., 2003). The results for the NO VISION condition seem more puzzling, in contrast. If physiological synchrony were merely indicative in the VISION condition because more cues to synchronise are available, then we would generally expect there to not be any effect in the NO VISION condition. However, physiological synchrony appears to be a dynamic construct which is not per se connected to only positive outcomes, so that in some situations it may be preferable to disengage (Mayo & Gordon, 2020). In this regard, the negative association which we observed in the LHF band could indicate that more physiological synchrony in absence of the appropriate cues might even be misleading. Since we did not manipulate physiological synchrony itself, this assertion remains merely speculative. Limitations and Outlook While our study offers many new insights, a couple of limitations should be mentioned at this point. Here, the majority of considerations revolve around the sample and contextual circumstances. For one, our sample size was relatively small, not allowing us to draw any final conclusion for hypothesis 2. A sensitivity analysis showed that this design had 80% power to detect medium-to-large effects. However, smaller effects may have gone undetected. Due to repeated measurement and a two-dyad constellation per participant sample size was appropriate for hypotheses 1 and 3. Due to contextual circumstances of the experiment forming part of a lab course, participants were familiar with each other as well as not blind to the experimental conditions and aims. The question of generalizability also comes into play, as synchronisation can differ depending on whether one is familiar with another person or not (Bizzego et al., 2019). Relationship status could also be relevant for the lie detection task itself, although the majority of participants indicated knowing each other solely through their studies. Furthermore, the lie detection task may be too variable in itself to allow for conclusions regarding performance of an individual or the dyad. Participants may have relied on different tactics (for example, telling a story that happened to someone they know as if it happened to them) that we did not assess or control for. Finally, we did not take gender into account at all, which is likely to play a role in this line of research and should be systematically looked at in future studies, but could not be investigated due to a very imbalanced gender distribution in our groups. Despite these limitations, the present findings contribute to a growing body of evidence that social context and sensory access to the partner shape the physiological connection between individuals. Here, the visual system emerged as a particularly strong modulator of HRV synchrony across multiple frequency bands, beyond individual HRV characteristics and a range of demographic and health factors. Future work could build on this by combining manipulations of vision with manipulations of other communication channels (e.g., touch; see Goldstein et al., 2017, 2018; Nguyen et al., 2021; Reddan et al., 2020), by investigating whether similar patterns emerge for other modalities. Furthermore, effects of the visual system on physiological synchrony should also be investigated in more detail. For one, they could be explored in more naturalistic settings or clinical groups, and by linking frequency-specific synchrony more directly to subjective experiences such as rapport, empathy, or perceived connectedness. Additionally, the exact driver behind this effect (e.g., eye contact, additional information from posture, respiration, etc.) should be further elaborated on by incorporating eye tracking and varying the visual input available (e.g., by using different types of barriers). Conclusion Physiological synchrony is a multifaceted phenomenon occurring during social interactions. To date, the exact mechanisms of establishing physiological synchrony are not sufficiently explored. We propose that different sensory modalities could provide the needed input for people to synchronise in social settings. In this study, we explored the visual system as a potential mechanism of interpersonal physiological synchrony. Results indicate that the visual system plays a prominent role in establishing synchrony across high frequency and the adjacent low-frequency ranges. While we did not find any connection between lie detection accuracy and physiological synchrony across both the VISION and the NO VISION conditions, we did observe frequency-specific differential effects within the two conditions. Consent for publication All participants provided consent for publication of anonymized data. Competing interests The authors declare no competing interests. Funding The project was supported by the Deutsche Forschungsgemeinschaft (DFG, German Research Foundation) under Germany’s Excellence Strategy (EXC 2117–422037984) via the Centre for the Advanced Study of Collective Behaviour at the University of Konstanz. Code availability All analysis code used in this study is publicly available on OSF at https://doi.org/10.17605/OSF.IO/FB2P5 (Wienhold et al., 2025) including scripts to reproduce the reported results and figures. Data availability The dataset used for the primary analyses reported in this article (i.e., main variables required to test the main hypotheses) is available on OSF at https://doi.org/10.17605/OSF.IO/FB2P5 (Wienhold et al., 2025). Sensitive variables (e.g., health-related covariates) are not shared publicly but can be made available upon reasonable request to the authors. Author contributions NV, SW, and JCP designed the study. NV and SW collected, preprocessed and analyzed the data. NV and SW drafted the manuscript. NV, JCP, BFD, and SW revised the manuscript. NV and SW contributed equally to this work and share first authorship. All authors approved the final manuscript. Acknowledgement We would like to thank the students of the experimental lab courses for their contribution to this research. We would also like to thank Ben Waldschmidt for helping with data collection. References Apple Inc. (n.d.). Apple iPad [Computer software].Arnold, J. B. (2025). ggthemes: Extra Themes, Scales and Geoms for “ggplot2” (p. 5.2.0) [Computer software]. https://doi.org/10.32614/CRAN.package.ggthemesBartoń, K. (2025). MuMIn: Multi-Model Inference (p. 1.48.11) [Computer software]. https://doi.org/10.32614/CRAN.package.MuMInBehrens, F., Snijdewint, J. A., Moulder, R. G., Prochazkova, E., Sjak-Shie, E. E., Boker, S. M., & Kret, M. E. (2020). Physiological synchrony is associated with cooperative success in real-life interactions. Scientific Reports , 10 (1), 19609. https://doi.org/10.1038/s41598-020-76539-8Bizzego, A., Azhari, A., Campostrini, N., Truzzi, A., Ng, L. Y., Gabrieli, G., Bornstein, M. H., Setoh, P., & Esposito, G. (2019). Strangers, Friends, and Lovers Show Different Physiological Synchrony in Different Emotional States. Behavioral Sciences , 10 (1), 11. https://doi.org/10.3390/bs10010011Bogaard, G., Meijer, E. H., Vrij, A., & Merckelbach, H. (2016). Strong, but Wrong: Lay People’s and Police Officers’ Beliefs about Verbal and Nonverbal Cues to Deception. PLOS ONE , 11 (6), e0156615. https://doi.org/10.1371/journal.pone.0156615Bond, C. F., & DePaulo, B. M. (2006). Accuracy of Deception Judgments. Personality and Social Psychology Review , 10 (3), 214–234. https://doi.org/10.1207/s15327957pspr1003_2Breyer, B., & Bluemke, M. (2016). Deutsche Version der Positive and Negative Affect Schedule PANAS (GESIS Panel). Zusammenstellung sozialwissenschaftlicher Items und Skalen (ZIS) . https://doi.org/10.6102/ZIS242Clarke, E., Sherrill-Mix, S., & Dawson, C. (2025). ggbeeswarm: Categorical Scatter (Violin Point) Plots (p. 0.7.3) [Computer software]. https://doi.org/10.32614/CRAN.package.ggbeeswarmCsárdi, G., & Torvalds, L. (2024). parsedate: Recognize and Parse Dates in Various Formats, Including All ISO 8601 Formats (p. 1.3.2) [Computer software]. https://doi.org/10.32614/CRAN.package.parsedateCzeszumski, A., Eustergerling, S., Lang, A., Menrath, D., Gerstenberger, M., Schuberth, S., Schreiber, F., Rendon, Z. Z., & König, P. (2020). Hyperscanning: A Valid Method to Study Neural Inter-brain Underpinnings of Social Interaction. Frontiers in Human Neuroscience , 14 , 39. https://doi.org/10.3389/fnhum.2020.00039De Felice, S., Di Ciò, F., Tompkins, D., Hakim, U., Pinti, P., Vigliocco, G., & Hamilton, A. F. D. C. (2025). Learning with others: Teacher–learner brain synchrony depends on mutual gaze and joint attention. Cerebral Cortex , 35 (12), bhaf323. https://doi.org/10.1093/cercor/bhaf323DePaulo, B. M., Lindsay, J. J., Malone, B. E., Muhlenbruck, L., Charlton, K., & Cooper, H. (2003). Cues to deception. Psychological Bulletin , 129 (1), 74–118. https://doi.org/10.1037/0033-2909.129.1.74Feldman, R. (2012). PARENT–INFANT SYNCHRONY: A BIOBEHAVIORAL MODEL OF MUTUAL INFLUENCES IN THE FORMATION OF AFFILIATIVE BONDS. Monographs of the Society for Research in Child Development , 77 (2), 42–51. https://doi.org/10.1111/j.1540-5834.2011.00660.xGohel, D., & Skintzos, P. (2026). flextable: Functions for Tabular Reporting (p. 0.9.11) [Computer software]. https://doi.org/10.32614/CRAN.package.flextableGoldstein, P., Weissman-Fogel, I., Dumas, G., & Shamay-Tsoory, S. G. (2018). Brain-to-brain coupling during handholding is associated with pain reduction. Proceedings of the National Academy of Sciences , 115 (11). https://doi.org/10.1073/pnas.1703643115Goldstein, P., Weissman-Fogel, I., & Shamay-Tsoory, S. G. (2017). The role of touch in regulating inter-partner physiological coupling during empathy for pain. Scientific Reports , 7 (1), 3252. https://doi.org/10.1038/s41598-017-03627-7Gordon, I., & Bartsch, R. P. (2026). Correlates of interpersonal physiological synchrony and sources of empirical heterogeneity. Nature Reviews Psychology . https://doi.org/10.1038/s44159-026-00535-4Gordon, I., Tomashin, A., & Mayo, O. (2025). A theory of flexible multimodal synchrony. Psychological Review , 132 (3), 680–718. https://doi.org/10.1037/rev0000495Grolemund, G., & Wickham, H. (2011). Dates and Times Made Easy with lubridate . Journal of Statistical Software , 40 (3). https://doi.org/10.18637/jss.v040.i03Grosjean, P. (2025). svDialogs: “SciViews::R” - Standard Dialog Boxes for Windows, MacOS and Linuxes (p. 1.1.2) [Computer software]. https://doi.org/10.32614/CRAN.package.svDialogsHakim, U., De Felice, S., Pinti, P., Zhang, X., Noah, J., Ono, Y., Burgess, P. W., Hamilton, A., Hirsch, J., & Tachtsidis, I. (2023). Quantification of inter-brain coupling: A review of current methods used in haemodynamic and electrophysiological hyperscanning studies. NeuroImage , 280 , 120354. https://doi.org/10.1016/j.neuroimage.2023.120354Hartwig, M., & Bond, C. F. (2014). Lie Detection from Multiple Cues: A Meta‐analysis. Applied Cognitive Psychology , 28 (5), 661–676. https://doi.org/10.1002/acp.3052Helm, J. L., Sbarra, D. A., & Ferrer, E. (2014). Coregulation of respiratory sinus arrhythmia in adult romantic partners. Emotion , 14 (3), Article 3. https://doi.org/10.1037/a0035960Hirsch, J., Zhang, X., Noah, J. A., & Ono, Y. (2017). Frontal temporal and parietal systems synchronize within and across brains during live eye-to-eye contact. NeuroImage , 157 , 314–330. https://doi.org/10.1016/j.neuroimage.2017.06.018Hoehl, S., Fairhurst, M., & Schirmer, A. (2021). Interactional synchrony: Signals, mechanisms and benefits. Social Cognitive and Affective Neuroscience , 16 (1–2), 5–18. https://doi.org/10.1093/scan/nsaa024Issartel, J., Bardainne, T., Gaillot, P., & Marin, L. (2015). The relevance of the cross-wavelet transform in the analysis of human interaction – a tutorial. Frontiers in Psychology , 5 . https://doi.org/10.3389/fpsyg.2014.01566Kassambara, A. (2023). rstatix: Pipe-Friendly Framework for Basic Statistical Tests (p. 0.7.3) [Computer software]. https://doi.org/10.32614/CRAN.package.rstatixKay, M. (2025). ggdist: Visualizations of Distributions and Uncertainty (p. 3.3.3) [Computer software]. https://doi.org/10.32614/CRAN.package.ggdistKinreich, S., Djalovski, A., Kraus, L., Louzoun, Y., & Feldman, R. (2017). Brain-to-Brain Synchrony during Naturalistic Social Interactions. Scientific Reports , 7 (1), 17060. https://doi.org/10.1038/s41598-017-17339-5Kret, M. E. (2015). Emotional expressions beyond facial muscle actions. A call for studying autonomic signals and their impact on social perception. Frontiers in Psychology , 6 . https://doi.org/10.3389/fpsyg.2015.00711Laborde, S., Mosley, E., & Thayer, J. F. (2017). Heart Rate Variability and Cardiac Vagal Tone in Psychophysiological Research – Recommendations for Experiment Planning, Data Analysis, and Data Reporting. Frontiers in Psychology , 08 . https://doi.org/10.3389/fpsyg.2017.00213Long, J. A. (2022). jtools: Analysis and Presentation of Social ScientificData [Computer software]. https://doi.org/10.21105/joss.06610Loy, J. E., Rohde, H., & Corley, M. (2018). Cues to Lying May be Deceptive: Speaker and Listener Behaviour in an Interactive Game of Deception. Journal of Cognition , 1 (1), 42. https://doi.org/10.5334/joc.46Lüdecke, D. (2025). sjPlot: Data Visualization for Statistics in Social Science (p. 2.9.0) [Computer software]. https://doi.org/10.32614/CRAN.package.sjPlotMarzoratti, A., & Evans, T. M. (2022). Measurement of interpersonal physiological synchrony in dyads: A review of timing parameters used in the literature. Cognitive, Affective, & Behavioral Neuroscience , 22 (6), 1215–1230. https://doi.org/10.3758/s13415-022-01011-1Mayo, O., & Gordon, I. (2020). In and out of synchrony—Behavioral and physiological dynamics of dyadic interpersonal coordination. Psychophysiology , 57 (6), e13574. https://doi.org/10.1111/psyp.13574Mayo, O., Lavidor, M., & Gordon, I. (2021). Interpersonal autonomic nervous system synchrony and its association to relationship and performance – a systematic review and meta-analysis. Physiology & Behavior , 235 , 113391. https://doi.org/10.1016/j.physbeh.2021.113391McCorry, L. K. (2007). Physiology of the Autonomic Nervous System. American Journal of Pharmaceutical Education , 71 (4), 78. https://doi.org/10.5688/aj710478Moritz, S., & Bartz-Beielstein, T. (2017). imputeTS: Time Series Missing Value Imputation in R. The R Journal , 9 (1), 207. https://doi.org/10.32614/RJ-2017-009Morlet, J., Arens, G., Fourgeau, E., & Giard, D. (1982). Wave propagation and sampling theory; Part II, Sampling theory and complex waves. Geophysics , 47 (2), 222–236. https://doi.org/10.1190/1.1441329Nguyen, T., Abney, D. H., Salamander, D., Bertenthal, B. I., & Hoehl, S. (2021). Proximity and touch are associated with neural but not physiological synchrony in naturalistic mother-infant interactions. NeuroImage , 244 , 118599. https://doi.org/10.1016/j.neuroimage.2021.118599Niedźwiecka, A. (2023). Eye contact effect: The role of vagal regulation and reactivity, and self-regulation of attention. Current Psychology , 42 (4), 3284–3290. https://doi.org/10.1007/s12144-021-01682-yPalumbo, R. V., Marraccini, M. E., Weyandt, L. L., Wilder-Smith, O., McGee, H. A., Liu, S., & Goodwin, M. S. (2017). Interpersonal Autonomic Physiology: A Systematic Review of the Literature. Personality and Social Psychology Review , 21 (2), 99–141. https://doi.org/10.1177/1088868316628405Parnandi, A., & Gutierrez-Osuna, R. (2013). Contactless Measurement of Heart Rate Variability from Pupillary Fluctuations. 2013 Humaine Association Conference on Affective Computing and Intelligent Interaction , 191–196. https://doi.org/10.1109/ACII.2013.38Pinheiro, J., Bates, D., & R Core Team. (2025). nlme: Linear and Nonlinear Mixed Effects Models (p. 3.1-168) [Computer software]. https://doi.org/10.32614/CRAN.package.nlmePolar Electro Oy. (n.d.-a). Polar Team Pro App [Computer software]. Retrieved https://apps.apple.com/de/app/polar-team-pro/id97762078Polar Electro Oy. (n.d.-b). Polar Team Pro Docking Station [Computer software]. Retrieved https://support.polar.com/de/polar-team-proPorges, S. W. (2007). The polyvagal perspective. Biological Psychology , 74 (2), 116–143. https://doi.org/10.1016/j.biopsycho.2006.06.009Porges, S. W. (2022). Polyvagal Theory: A Science of Safety. Frontiers in Integrative Neuroscience , 16 , 871227. https://doi.org/10.3389/fnint.2022.871227Posit team. (2025). RStudio: Integrated Development Environment for R [Computer software]. Posit Software, PBC. http://www.posit.co/Prochazkova, E., & Kret, M. E. (2017). Connecting minds and sharing emotions through mimicry: A neurocognitive model of emotional contagion. Neuroscience & Biobehavioral Reviews , 80 , 99–114. https://doi.org/10.1016/j.neubiorev.2017.05.013Questionstar. (2011). QUESTIONSTAR – Online Survey Tool [Computer software]. https://questionstar.comR Core Team. (2025a). foreign: Read Data Stored by “Minitab”, “S”, “SAS”, “SPSS”, “Stata”, “Systat”, “Weka”, “dBase”, … (p. 0.8-91) [Computer software]. https://doi.org/10.32614/CRAN.package.foreignR Core Team. (2025b). R: A language and environment for statistical computing (Version 2025.05.1+513) [Computer software]. R Foundation for Statistical Computing. https://www.R-project.org/Reddan, M. C., Young, H., Falkner, J., López-Solà, M., & Wager, T. D. (2020). Touch and social support influence interpersonal synchrony and pain. Social Cognitive and Affective Neuroscience , 15 (10), 1064–1075. https://doi.org/10.1093/scan/nsaa048Rodriguez-Linares, L., Vila, X., Lado, M. J., Méndez, A. J., Otero, A., & Garcia, A. (2024). RHRV: Heart rate variability analysis of ECG data (R package (Version 5.0.0) [Computer software]. https://CRAN.R-project.org/package=RHRVRoesch, A., & Schmidbauer, H. (2025). WaveletComp (Version 1.2) [R]. https://CRAN.R-project.org/package=WaveletCompRosseel, Y., Jorgensen, T. D., & De Wilde, L. (2025). lavaan: Latent variable analysis (Version 0.6-20) [Computer software]. https://CRAN.R-project.org/package=lavaanSammito, S., Thielmann, B., & Böckelmann, I. (2024). Update: Factors influencing heart rate variability–a narrative review. Frontiers in Physiology , 15 , 1430458. https://doi.org/10.3389/fphys.2024.1430458Sato, R., & Sato, H. (2025). Synchronization of brain activity associated with eye contact: Comparison of face-to-face and online communication. Scientific Reports , 15 (1), 608. https://doi.org/10.1038/s41598-024-84602-xSchneider, E. E., Schönfelder, S., Domke-Wolf, M., & Wessa, M. (2020). Measuring stress in clinical and nonclinical subjects using a German adaptation of the Perceived Stress Scale. International Journal of Clinical and Health Psychology , 20 (2), 173–181. https://doi.org/10.1016/j.ijchp.2020.03.004Shaffer, F., & Ginsberg, J. P. (2017). An Overview of Heart Rate Variability Metrics and Norms. Frontiers in Public Health , 5 , 258. https://doi.org/10.3389/fpubh.2017.00258Sporer, S. L., & Schwandt, B. (2007). Moderators of nonverbal indicators of deception: A meta-analytic synthesis. Psychology, Public Policy, and Law , 13 (1), 1–34. https://doi.org/10.1037/1076-8971.13.1.1Strömwall, L., & Granhag, P. A. (2003). How to Detect Deception? Arresting the Beliefs of Police Officers, Prosecutors and Judges. Psychology, Crime & Law , 9 (1), 19–36. https://doi.org/10.1080/10683160308138Task Force of the ESC & NASPE. (1996). Heart Rate Variability: Standards of Measurement, Physiological Interpretation, and Clinical Use. Circulation , 93 (5), 1043–1065. https://doi.org/10.1161/01.CIR.93.5.1043Thayer, J. F., & Lane, R. D. (2000). A model of neurovisceral integration in emotion regulation and dysregulation. Journal of Affective Disorders , 61 (3), 201–216. https://doi.org/10.1016/S0165-0327(00)00338-4The Global Deception Research Team. (2006). A World of Lies. Journal of Cross-Cultural Psychology , 37 (1), 60–74. https://doi.org/10.1177/0022022105282295Tomashin, A., Gordon, I., & Wallot, S. (2022). Interpersonal Physiological Synchrony Predicts Group Cohesion. Frontiers in Human Neuroscience , 16 , 903407. https://doi.org/10.3389/fnhum.2022.903407Vrij, A., Hartwig, M., & Granhag, P. A. (2019). Reading Lies: Nonverbal Communication and Deception. Annual Review of Psychology , 70 (1), 295–317. https://doi.org/10.1146/annurev-psych-010418-103135Wei, T., & Simko, V. (2024). R package “corrplot”: Visualization of a correlation matrix (Version 0.95) [Computer software]. https://github.com/taiyun/corrploWickham, H. (2007). Reshaping Data with the reshape Package. Journal of Statistical Software , 21 (12). https://doi.org/10.18637/jss.v021.i12Wickham, H. (2010). reshape2: Flexibly Reshape Data: A Reboot of the Reshape Package (p. 1.4.5) [Computer software]. https://CRAN.R-project.org/package=reshape2Wickham, H. (2023). stringr: Simple, Consistent Wrappers for Common String Operations (p. 1.6.0) [Computer software]. https://doi.org/10.32614/CRAN.package.stringrWickham, H., Chang, W., Henry, L., Pedersen, T. L., Takahashi, K., Wilke, C., Woo, K., Yutani, H., Dunnington, D., & Van Den Brand, T. (2016). ggplot2: Create Elegant Data Visualisations Using the Grammar of Graphics (p. 4.0.2) [Computer software]. https://doi.org/10.32614/CRAN.package.ggplot2Wickham, H., François, R., Henry, L., Müller, K., & Vaughan, D. (2023). dplyr: A Grammar of Data Manipulation (p. 1.2.0) [Computer software]. https://doi.org/10.32614/CRAN.package.dplyrWickham, H., Miller, E., & Smith, D. (2025). haven: Import and export “SPSS”, “Stata” and “SAS” files (Version 2.5.5) [Computer software]. https://haven.tidyverse.orgWickham, H., Vaughan, D., & Girlich, M. (2024). tidyr: Tidy messy data (Version 1.3.1) [Computer software]. https://CRAN.R-project.org/package=tidyrWienhold, S., Volkmer, N., & Pruessner, J. C. (2025). The role of eye contact in physiological synchrony . https://doi.org/10.17605/OSF.IO/DHGKPWohltjen, S., & Wheatley, T. (2021). Eye contact marks the rise and fall of shared attention in conversation. Proceedings of the National Academy of Sciences , 118 (37), e2106645118. https://doi.org/10.1073/pnas.2106645118 Information & Authors Information Version history V1 Version 1 24 February 2026 Copyright This work is licensed under a Non Exclusive No Reuse License. Authors Affiliations Nina Volkmer 0009-0009-9628-9921 [email protected] University of Konstanz View all articles by this author Stella Wienhold 0009-0003-8275-4673 University of Konstanz View all articles by this author Bernadette Denk 0000-0002-5836-8815 University of Konstanz View all articles by this author Jens Prüssner University of Konstanz View all articles by this author Funding Information Deutsche Forschungsgemeinschaft EXC 2117–422037984 Centre for the Advanced Study of Collective Behaviour Metrics & Citations Metrics Article Usage 282 views 100 downloads .FvxKWukQNSOunydq8rnd { width: 100px; } Citations Download citation Nina Volkmer, Stella Wienhold, Bernadette Denk, et al. Seeing Each Other Matters: Visual Contact Enhances Heart Rate Variability Synchrony Between Interaction Partners. Authorea . 24 February 2026. DOI: https://doi.org/10.22541/au.177191837.70605191/v1 If you have the appropriate software installed, you can download article citation data to the citation manager of your choice. Simply select your manager software from the list below and click Download. For more information or tips please see 'Downloading to a citation manager' in the Help menu . Format Please select one from the list RIS (ProCite, Reference Manager) EndNote BibTex Medlars RefWorks Direct import Tips for downloading citations document.getElementById('citMgrHelpLink').addEventListener('click', function() { popupHelp(this.href); return false; }); $(".js__slcInclude").on("change", function(e){ if ($(this).val() == 'refworks') $('#direct').prop("checked", false); $('#direct').prop("disabled", ($(this).val() == 'refworks')); }); View Options View options PDF View PDF Figures Tables Media Share Share Share article link Copy Link Copied! Copying failed. Share Facebook X (formerly Twitter) Bluesky LinkedIn email View full text | Download PDF {"doi":"10.22541/au.177191837.70605191/v1","type":"Article"} Now Reading: Share Figures Tables Close figure viewer Back to article Figure title goes here Change zoom level Go to figure location within the article Download figure Toggle share panel Toggle share panel Share Toggle information panel Toggle information panel Go to previous graphic Go to next graphic Go to previous table Go to next table All figures All tables View all material View all material xrefBack.goTo xrefBack.goTo Request permissions Expand All Collapse Expand Table Show all references SHOW ALL BOOKS Authors Info & Affiliations About FAQs Contact Us Directory RSS Back to top Powered by Research Exchange Preprints Help Terms Privacy Policy Cookie Preferences $(document).ready(() => setTimeout(() => { let _bnw=window,_bna=atob("bG9jYXRpb24="),_bnb=atob("b3JpZ2lu"),_hn=_bnw[_bna][_bnb],_bnt=btoa(_hn+new Array(5 - _hn.length % 4).join(" ")); $.get("/resource/lodash?t="+_bnt); },4000)); (function(){function c(){var b=a.contentDocument||a.contentWindow.document;if(b){var d=b.createElement('script');d.innerHTML="window.__CF$cv$params={r:'9fe2b00b1cbddfa9',t:'MTc3OTE4OTM2Nw=='};var a=document.createElement('script');a.src='/cdn-cgi/challenge-platform/scripts/jsd/main.js';document.getElementsByTagName('head')[0].appendChild(a);";b.getElementsByTagName('head')[0].appendChild(d)}}if(document.body){var a=document.createElement('iframe');a.height=1;a.width=1;a.style.position='absolute';a.style.top=0;a.style.left=0;a.style.border='none';a.style.visibility='hidden';document.body.appendChild(a);if('loading'!==document.readyState)c();else if(window.addEventListener)document.addEventListener('DOMContentLoaded',c);else{var e=document.onreadystatechange||function(){};document.onreadystatechange=function(b){e(b);'loading'!==document.readyState&&(document.onreadystatechange=e,c())}}}})();