Heart rate variability and emotion regulation: Multi-ethnic differences in reappraisal and suppression

Abstract

Recent work found ethnic differences in the association between resting heart-rate variability (HRV), an index of emotion regulation (ER) capacity, and ER difficulties. The present study examined whether this relationship exists among American adults from other marginalized ethnic backgrounds living in the United States, African American|AfAm and Hispanic or Latino Americans|Hispanics, which remains unexplored in the literature. We addressed this gap by investigating whether self-reported ethnicity differentially moderated the relationship between log-transformed high-frequency HRV and ER, indexed by suppression and reappraisal. A total of 1047 emerging adults (M_age = 19.7 yrs [1.7]) had complete ER and resting 10-min HRV data (82 AfAm [65% women], 183 Asians [52% women], 228 Hispanics [59% women], and 554 non-Hispanic White|NHW [60% women]). HRV was highest among AfAm and Hispanics and lowest among Asian individuals. Linear regression, adjusting for waist circumference and sex, revealed that the HRV-reappraisal associations were of greater magnitude among self-identified ethnic minorities than NHW, despite similar mean levels of reappraisal. AfAm and Hispanics showed statistically similar negative associations, while similar marginal positive associations were observed among Asians and NHW. In contrast, AfAm and Asian individuals showed a statistically similar positive HRV-suppression association, which differed marginally from NHW. NHW reported lower suppression compared to all groups and showed a unique negative HRV-suppression association. Findings suggest that neurovisceral capacity and sociocultural contexts may jointly establish culturally-patterned regulatory behavior in American cultural contexts. Additional social contexts, including geographical locations and discrimination, should be investigated as mediators of these associations in future work.

Emotions emerge within a specific context, leading individuals to attempt to shape their emotional experience by intentionally or automatically modifying their thoughts or behaviors associated with those emotions (Gross & John, 1998; Mauss et al., 2007). Two commonly studied emotion regulation (ER) strategies in American social contexts are reappraisal (i.e., cognitive reframing) and suppression (i.e., inhibiting emotional expression) (Gross & Ford, 2023; Gross & John, 2003). Resting vagally-mediated heart rate variability (HRV), defined as the fluctuation between consecutive heartbeats, is a dependable indicator of cognitive regulation (Spangler et al., 2021; Thayer & Lane, 2009; Williams et al., 2017) and ER capacity (Lane et al., 2009; Ottaviani et al., 2016; Thayer & Lane, 2009; Williams et al., 2015). Longitudinal and cross-sectional evidence indicate that higher HRV is associated with greater habitual reappraisal and lower suppression use (Makovac et al., 2022; Visted et al., 2017), highlighting the relevance of HRV for understanding these strategies. Moreover, although many studies show that ethnic groups differ in HRV and ER processes, limited research has examined ethnic differences in the relationship between HRV and ER (Pourmand et al., 2024; Watanabe et al., 2023). Thus, since ER strategies are established in various sociocultural contexts, this study builds on recent work regarding Asian American and European American differences in the HRV-ER difficulties relationship (Gratz & Roemer, 2004; Watanabe et al., 2023) by examining the link between HRV and habitual reappraisal and suppression among emerging adults in the U.S. who self-identified as Asian, Black or African American (AfAm), Hispanic or Latino, and non-Hispanic White (NHW) (Campos & Kim, 2017; Coon & Kemmelmeier, 2001; Gross & John, 2003; John & Gross, 2004; Mesquita et al., 2017; Raval & Walker, 2019; Wilson & Gentzler, 2021). The current work proposes that U.S. sociocultural contexts (i.e., ethnicity) influence the association between emotion-related physiology (e.g., HRV) and habitual ER. To our knowledge, this is the first study to investigate how HRV supports habitual ER differently across four major ethnic groups in the U.S.

Sociocultural context of emotion regulation

From a sociocultural psychology perspective, cultural and psychological factors are mutually interdependent and dynamically shape individuals’ behavior, such as ER (Kitayama & Cohen, 2007; Markus & Hamedani, 2007; Markus & Kitayama, 2010). Instead of defining culture as a collection of people, we, along with others, define culture as patterns of selected meanings (i.e., internalized ideas or attitudes) and practices (i.e., meaningful acts or behaviors) (Adams & Markus, 2001, 2004; Atran et al., 2005; Bruner, 1990; Kroeber & Kluckhohn, 1953; Markus & Hamedani, 2007; Shweder, 1990), including psychological processes like ER. Consistent with Bandura’s social learning theory (1977), ER strategies are culturally patterned practices developed during childhood socialization (Abu Salih et al., 2023; Dunsmore & Halberstadt, 1997; Eisenberg et al., 1998; Raval & Walker, 2019; Rochanavibhata & Marian, 2023; Zahn-Waxler et al., 1996) and throughout life as individuals engage with their social environments (Gross & Ford, 2023; John & Gross, 2004; Kim & Sherman, 2007; Kitayama et al., 1997; Kitayama & Uskul, 2011; Lebra, 1976; Markus & Kitayama, 2010; Matsumoto, 1993; Wilson & Gentzler, 2021). These culturally patterned meanings and practices are often linked to socially constructed categories, such as race and ethnicity (Markus & Hamedani, 2007). For example, in the United States, socialization patterns of AfAm caregivers may reflect an American sociocultural context where experiences of racism and discrimination are prevalent (Berlin et al., 2002; Coll et al., 1996; Lee et al., 2019; Wilson & Gentzler, 2021), while NHW caregivers may socialize children to be autonomous (Kitayama et al., 2000; Raval & Walker, 2019; Trommsdorff et al., 2012). However, it is important to note that these assumptions may hold in some geographical and cultural contexts but not in others. Moreover, while culturally socialized and embedded meanings establish behavioral and psychological tendencies, we want to emphasize that sociocultural contexts are dynamic, and individual differences exist in how people interact with their social worlds, form meanings, and engage in practices, such as ER. Additionally, as cultural meanings evolve within individuals, cultural contexts can intersect, allowing people to embody diverse social identities and engage in various patterns of ER (Brewer & Yuki, 2007; Markus & Hamedani, 2007). Still, we aim to highlight cultural patterns and ER practices identified in the extant literature as most prevalent among individuals in Asian, AfAm, Hispanic or Latino, and NHW sociocultural contexts.

Individual differences in reappraisal and suppression are central to understanding how people regulate emotions across sociocultural contexts. According to the process model of ER (Gross, 2015; Gross & John, 1998, 2003), reappraisal is an antecedent-focused strategy that involves cognitively reframing a situation before an emotional response has fully developed, occurring early in the emotion-generative process. In contrast, suppression is response-focused and requires the inhibition of emotional expression after the response has already been activated (Gross, 2001; Gross & John, 2003). Although the use of both strategies may be automatic or deliberate, culturally patterned behavior, such as habitual ER, is established within sociocultural contexts (Abu Salih et al., 2023; Dunsmore & Halberstadt, 1997; Eisenberg et al., 1998; Kitayama & Uskul, 2011; Markus & Hamedani, 2007; Mauss et al., 2007; Raval & Walker, 2019; Rochanavibhata & Marian, 2023; Zahn-Waxler et al., 1996). For example, habitual suppression may be less common in contexts where autonomous values and open emotional expression are favored (e.g., in NHW contexts; Rothbaum & Rusk, 2011; Wierzbicka, 2004), whereas suppression may be more normative in contexts where its use helps to achieve harmony and cohesion (e.g., East Asian cultural contexts) (Coon & Kemmelmeier, 2001; Dunsmore & Halberstadt, 1997; Halberstadt & Eaton, 2002; Hofstede, 1980; Kitayama et al., 1997, 2000; Kitayama & Salvador, 2024; Kyeong et al., 2021; Markus & Kitayama, 1991; Wierzbicka, 2004). Although Latin Americans and Mexicans also prioritize harmonious social relationships, Mexican Americans rely on positive emotional expression to facilitate those bonds, suppressing more negative emotions compared to Chinese Americans (Campos & Kim, 2017; Kitayama & Salvador, 2024; Matsumoto et al., 2008; Ramírez-Esparza et al., 2008; Ruby et al., 2012; Salvador et al., 2024; Senft et al., 2021, 2023; Soto et al., 2005, 2011; Su et al., 2015). Conversely, in AfAm sociocultural contexts, children are often taught to suppress their emotions to prepare them for racial adversities, with emerging adults showing a tendency to suppress emotions to avoid conflict and manage chronic racism-related stress (Berlin et al., 2002; Dorr et al., 2007; Eldesouky & English, 2019; Finley et al., 2024; Seaton et al., 2012; Thayer et al., 2020; Thayer & Koenig, 2019; Wilson & Gentzler, 2021). Importantly, although reappraisal is often linked with positive (i.e., adaptive) and suppression with negative (i.e., maladaptive) outcomes (Bebko et al., 2019; Garnefski et al., 2004; Reichler et al., 2025; Webb et al., 2012), these patterns are inconsistent across individuals from diverse backgrounds. For example, suppression may be adaptive in certain sociocultural contexts (Chen et al., 2020; Hu et al., 2014; Perez & Soto, 2011; Schunk et al., 2022; Soto et al., 2011, 2012; Su et al., 2015; Watanabe et al., 2024). Therefore, additional studies that examine the physiological underpinnings of ER could advance our understanding of the use of various ER strategies in different sociocultural contexts.

Despite robust evidence that ER strategies are socialized differently across cultural contexts, few studies have examined how these habitual tendencies align with physiological regulatory capacity across multiple ethnic groups. Resting HRV is a well-established biomarker of the interplay between inhibitory cortical and subcortical circuits that support physiological, affective, and cognitive regulation, including ER (Jennings et al., 2016; Sakaki et al., 2016; Thayer et al., 2012; Thayer & Lane, 2000, 2009). Notably, higher resting HRV is associated with better situational awareness (Hansen et al., 2003; Saus et al., 2006; Thayer & Lane, 2009) and inhibitory control, enabling individuals to deploy appropriate ER strategies based on environmental demands (Appelhans & Luecken, 2006; Beauchaine & Thayer, 2015; Mather & Thayer, 2018; Melzig et al., 2009; Spangler et al., 2021; Thayer & Lane, 2009; Williams et al., 2015, 2017). Furthermore, higher HRV is also a reliable index of unwanted memories and thought suppression, indicating greater ER capacity and abilities (ER: Aldao & Mennin, 2012; Gillie et al., 2014, 2015; Ingjaldsson et al., 2003; Lane et al., 2009; Ottaviani et al., 2016; Thayer & Lane, 2000; Williams et al., 2015). Importantly, although HRV and ER are understudied across diverse ethnic groups, cross-sectional and meta-analytic data suggest meaningful differences among these groups (Hill et al., 2015; Kemp, Koenig, et al., 2016; Pourmand et al., 2024; Thayer & Koenig, 2019; Watanabe et al., 2023). AfAm and Brazilian Latinos tend to show higher HRV compared to European Americans (Kemp, Koenig, et al., 2016; Pourmand et al., 2024; Thayer & Koenig, 2019), who, in turn, exhibit higher HRV than Asian Americans (Eyre et al., 2013; Hall et al., 2013; O’Neill et al., 2019; Watanabe et al., 2023). Although Asian Americans report more difficulties with ER strategies compared to European Americans, both groups display fewer ER difficulties at higher resting HRV (Pourmand et al., 2024; Watanabe et al., 2023). Additionally, East Asian individuals show greater HRV reactivity during emotion-evoking stressors than Western Europeans, indicating adaptive ER (Liddell & Williams, 2019). Among middle-aged adults, higher HRV was associated with greater habitual reappraisal and suppression among AfAm and less use of these strategies among European Americans (Thayer & Koenig, 2019). The existing research suggests that higher HRV not only enhances ER capabilities and regulatory capacity but may also enable individuals to regulate emotions in culturally patterned ways. For example, in cultural contexts in which emotional suppression is socially adaptive or sanctioned (Cole et al., 1994; Kitayama et al., 1997; Markus & Kitayama, 1991; Matsumoto, 1993; Salvador et al., 2024; Senft et al., 2021, 2023; Wilson & Gentzler, 2021), higher HRV could facilitate automatic upregulation of suppression (Dorr et al., 2007; Finley et al., 2024; Thayer & Koenig, 2019). However, no study has investigated HRV among American Hispanics or Latinos or its relationship with specific ER strategies among Asian individuals, who are two of the largest racial/ethnic groups in the U.S. (U.S. Census Bureau, 2025). A key question we aim to address is whether higher HRV, which is consistently linked to greater capacity for voluntary and automatic regulatory functions, universally supports the same ER strategies across different sociocultural contexts or whether sociocultural context influences the link between HRV and habitual reappraisal or suppression.

Theoretical foundation

The idea that sociocultural context influences the link between emotion-related physiology and habitual ER is supported by two complementary frameworks, the Neurovisceral Integration Model (NIM) and the Neuro-Culture Interaction Model (NCI). Together, these models offer insight into individual differences in ER patterns and physiology across socially constructed groups, such as ethnicity (Kitayama & Uskul, 2011; Smith et al., 2017; Thayer et al., 2021; Thayer & Lane, 2000, 2009; Wilson et al., 2024). The NIM describes how, at rest, the central autonomic network, particularly the prefrontal-amygdala circuit via the vagus nerve (i.e., indexed by HRV) dynamically supports cardiac autonomic function, inhibitory control, and flexible, adaptive ER (Benarroch, 1993; Smith et al., 2017; Thayer et al., 2012; Thayer & Lane, 2000). The NCI suggests that repeated engagement with cultural values, beliefs, and practices shapes the structure and function of neural systems, enabling individuals to behave in culturally patterned ways (Chiao et al., 2013; Kitayama & Park, 2010; Kitayama & Uskul, 2011). For example, as noted, emotions are socialized and may be embedded during early childhood (Abu Salih et al., 2023; Kim & Sherman, 2007; Kitayama et al., 1997; Lebra, 1976; Markus & Kitayama, 2010; Matsumoto, 1993; Pomerantz & Thompson, n.d.; Rochanavibhata & Marian, 2023; Zahn-Waxler et al., 1996), potentially leading to distinct neural patterns across cultural groups (Deng et al., 2023; Gee et al., 2014; Han & Ma, 2014a; Kitayama et al., 2017a; Li et al., 2018; Luo et al., 2020; Reindl et al., 2018; Varnum et al., 2014; Wang et al., 2017). From a sociocultural psychology perspective, these models indicate that individuals are biological, social, and cultural beings, with each factor shaping their worlds and themselves (Kitayama & Park, 2007; Markus & Hamedani, 2007).

Building on and expanding these perspectives, we previously proposed a Sociocultural Neurovisceral Model (Watanabe et al., 2025) that combines the NIM and NCI frameworks. We suggest that sociocultural patterns shape psychological processes, such as ER tendencies, through a dynamic process by which individuals engage with their social environment, assign meanings to life experiences, and participate in patterned thoughts, behaviors, or actions (i.e., practices) (Bruner, 1990; Kitayama & Park, 2007; Markus & Hamedani, 2007; Shweder, 1990). These culturally-specific meanings and practices become embedded in neural processes, leading to different neural patterns, autonomic regulation, and behaviors (Chiao et al., 2013; Deng et al., 2023; Gee et al., 2014; Han & Ma, 2014a; Kitayama et al., 2017b; Kitayama & Park, 2010; Kitayama & Uskul, 2011; Li et al., 2018; Luo et al., 2020; Reindl et al., 2018; Varnum et al., 2014; Wang et al., 2017, 2022). For example, in cultural contexts that emphasize relational harmony and discourage emotional expressiveness, the physiological impacts of negative affect or habitual suppression are attenuated (Curhan et al., 2014; Miyamoto et al., 2013; Park et al., 2020). Indeed, prior work has found a stress-buffering effect of suppression on cardiovascular disease risk among Japanese, whereas suppression was harmful for Americans primarily from NHW backgrounds (Watanabe et al., 2025). Building on these insights, we propose that measures of neurovisceral control, such as resting HRV, may support different ER strategies depending on sociocultural context. Since cultural patterns are often linked with socially constructed categories like ethnicity (Markus & Hamedani, 2007), our Sociocultural Neurovisceral Model offers a theoretical foundation for investigating how HRV, as an indicator of neural and emotional regulation, may be differentially associated with regulatory strategies like reappraisal and suppression across individuals from diverse ethnic backgrounds.

Research questions and hypotheses

Prior evidence in a racially diverse, unstratified sample has shown that higher HRV was associated with both reappraisal and suppression (Butler et al., 2006). However, in European samples, HRV predicts greater use of reappraisal and reduced suppression over time (Makovac et al., 2022) and cross-sectionally (Visted et al., 2017). Thus, it remains unclear whether these associations hold in more diverse sociocultural contexts. Building on this work, the present study examined whether these patterns generalize across sociocultural contexts in the United States. Moreover, since no previous study has examined the relationship between HRV and habitual reappraisal and suppression while distinguishing among and between individuals who self-identified as having Asian, AfAm, Hispanic or Latino, and NHW backgrounds, our study aimed to fill this gap. We tested whether HRV and ER would differ among Asians, AfAm, and NHW individuals and whether the association between HRV and ER would differ among these groups. Our a priori hypotheses were that Asians and AfAm would report greater suppression use (Kemp, Fráguas, et al., 2016; Thayer & Koenig, 2019), with AfAm exhibiting higher HRV and Asian individuals lower HRV (Eyre et al., 2013; Hall et al., 2013; O’Neill et al., 2019; Pourmand et al., 2024; Watanabe et al., 2023). Additionally, we hypothesized that HRV would be inversely associated with suppression and positively with reappraisal in Asian and NHW individuals, and that positive relationships would be observed among AfAm (Pourmand et al., 2024; Sörman et al., 2022; Thayer & Koenig, 2019; Watanabe et al., 2023; Zelkowitz & Cole, 2016). We focused on interpreting effect sizes and confidence intervals (Amrhein, Greenland, et al., 2019; Amrhein, Trafimow, et al., 2019; Lipsey et al., 2012; Miller, 2023; Pek & Flora, 2018).

Method

Transparency and Openness

Following the Transparency and Openness Promotion guidelines (Nosek et al., 2015) and JARS (Appelbaum et al., 2018), we report how we determined our sample size, all data exclusions (if any), all manipulations, and all measures in the study. Data and analysis codes are available upon reasonable request from the first author. While this study’s design and analysis were not preregistered, hypotheses were generated a priori.

Participants

The present study was conducted at a university in central Texas, USA. The Institutional Review Board at the university approved all procedures and the study adhered to the ethical standards of the 1964 Declaration of Helsinki (protocol numbers: 1149692–3; 1355632–7; 1437574–10; 1948842–7). Written informed consent was obtained from all participants. Recruitment, enrollment, and procedures are briefly outlined below as they are described in detail elsewhere (Tyra et al., 2023, 2025). There was no overlap among participants who were pooled across four studies conducted in the same laboratory. Although laboratory tasks differed across the studies, the experimental protocols were consistent.

For studies 1, 2, and 4, the university’s online subject pool was used to recruit participants and screen for eligibility (i.e., over 18 years old, no current illness or infection, and no history of cardiovascular disease). Participants were asked to refrain from consuming alcohol and engaging in vigorous exercise (12 h prior), drinking caffeine (2 h prior), and consuming food or liquids except water (2 h prior) before their laboratory visit. Participants in study 3, were recruited from the surrounding community via flyers and advertisements. Additional criteria for Study 3 participants included “no history of a chronic medical or neurological disorder, no aneurysm clips, pacemaker, cochlear implant, colorblindness, metal exposure, or claustrophobia. Additionally, participants were between the ages of 18 and 30, with no risk of pregnancy and no conditions that would prohibit participation in physical exercise testing. Participants provided written informed consent before data collection. Upon study completion, participants in study 3 received $50 for their time. Participants in studies 1, 2, and 4 received two to three hours of research credits toward their psychology and/or neuroscience course requirements.

We based our sample size estimation on prior work on HRV and ER conducted with emerging adults who self-identified as Asian Americans and European Americans (Watanabe et al., 2023). The effect size in this study was considered moderate (r = −.43). With an alpha of 0.05 and power of 0.80, the projected sample size needed to achieve this effect size using GPower (Erdfelder et al., 2009) is approximately n = 29 per group. With a total sample size of n = 1047 participants (183 Asian Americans; 82 AfAm; 228 Hispanic Americans, and 554 NHW), power was deemed sufficient for the present study.

Measures

Participants’ perceived social status was assessed using the MacArthur Scale of Subjective Social Status Ladder (Adler & Epel, 2000). Additionally, participants self-identified their race as either Asian, Black or African American, Native Hawaiian or Other Pacific Islander, Mixed/Other, or White and their ethnicity as either Hispanic or Latino or not Hispanic or Latino. Participants who identified their ethnicity as Hispanic or Latino, regardless of their race, were recoded as “Hispanic or Latino” with no overlap. For reporting simplicity, we use the term Hispanic to refer to individuals in the Hispanic or Latino race/ethnicity category (Flanagin et al., 2021). We use the term ethnicity to refer to this constructed category, which is based on participants’ self-reported race and ethnicity (Wilson et al., 2024). Due to the historical harms of race-based biomedical research and health care (Cerdeña et al., 2020; Hernandez-Boussard et al., 2023; Wright et al., 2022), we attempt to use race- and racism-conscious language throughout, as recommended by the National Academies of Sciences, Engineering, and Medicine (Wilson et al., 2024).

Heart rate variability was measured continuously throughout the protocol at a sampling rate of 500 Hz (MindWare Technologies LTD, Westerville, OH). The resulting raw interbeat intervals were exported into a single text file for each participant and further processed using Kubios HRV analysis software version 3.5.0, using automated R-peak detection (Tarvainen et al., 2013). A threshold-based piecewise cubic interpolation method was used to remove artifacts. Data were detrended using smoothness priors (Tarvainen et al., 2002, 2013). Mean resting time- and frequency-domain indices of HRV were calculated. Based on prior work on ethnic differences in HRV, we considered the frequency-domain index, autoregressive high-frequency HRV (ms2; Pourmand et al., 2024; Watanabe et al., 2023).

Emotion regulation was measured using the Emotion Regulation Questionnaire (ERQ) on a 7-point Likert scale ranging from 1 (strongly disagree) to 7 (strongly agree) (Gross & John, 2003). Reappraisal showed good internal consistency by ethnic group (AfAm: α = 0.86, Asians: α = 0.81, Hispanics: α = 0.85, NHW: α = 0.85). Suppression showed acceptable internal consistency by ethnic group (AfAm: α = 0.74, Asians: α = 0.75, Hispanics: α = 0.75, NHW: α = 0.78).

Procedure

After providing written informed consent, participants’ height and weight were measured. Next, a blood pressure cuff was attached to the participant’s non-dominant arm. Then, three electrocardiography (ECG) electrodes were placed on their upper and lower torso (Bio Protech Inc., Chino, CA). A 10-minute period was provided for the participants to adjust to the laboratory environment. Then, a 10-minute formal resting baseline period occurred, followed by an acute psychological stress task. Participants were given 10 minutes of relaxed sitting to recover from the stress task. Trait questionnaires were administered at the conclusion of the visit. Other than the type of stress task, the laboratory protocols were consistent across each study.

Data Analysis

Participants with complete data on all variables of interest were included in analyses (n = 1047). The Statistical Package of Social Sciences (SPSS ver. 28, IBM, Chicago, IL, USA) was used to conduct all analyses unless otherwise specified. As no prior studies have examined HRV-ER associations among Hispanic adults, we used two-tailed tests for this group to allow for effects in either direction. One-tailed tests were conducted for all comparisons between Asians and AfAm with NHW for whom prior empirical work provided a priori directional hypotheses Pourmand et al., 2024; Sörman et al., 2022; Thayer & Koenig, 2019; Watanabe et al., 2023). Although one-tailed tests are less common, this approach was guided by clearly specified a priori hypotheses and aligns with recommendations for maintaining logical consistency between theoretical and statistical hypotheses in directional testing (Cho & Abe, 2013). Moreover, while one-tailed testing “increases the likelihood of rejecting the null hypothesis” (Friedman, 1968, p. 250), it does not alter the interpretation of the magnitude of the effect, which is our focus in the present study (Friedman, 1968; Rosnow & Rosenthal, 1996; Amrhein, Greenland, et al., 2019; Amrhein, Trafimow, et al., 2019; Lipsey et al., 2012; Miller, 2023; Pek & Flora, 2018). We report 95% confidence intervals (CI) for two-tailed and 90% CIs for one-tailed tests. Plots were constructed using RStudio (available at https://www.rstudio.com/; R Core team, 2022) the ggplot2 (v. 3.5.1; Wickham, 2023), dplyr (v. 1.1.1; Wickham & François, 2023), and tidyr packages (v. 1.3.0; Wickham & Girlich, 2023). Chi-square tests were conducted to examine the proportion of women and men by ethnicity. Natural log-transformed HRV was used to fit assumptions of linear analyses. To examine ethnic differences between our variables of interest, we conducted independent means t-tests by ethnicity and report r-effect sizes and CIs (Rosenthal et al., 2000; Rosnow & Rosenthal, 2009), obtained using the Psychometrica effect size calculator (Ellis, 2010; Rosenthal et al., 2000). Pearson’s zero-order correlations were computed separately for each ethnic group, and the magnitude of difference between the associations were assessed with Fisher’s r to z transformation (Lenhard & Lenhard, 2016; Lowry, n.d.; Meng et al., 1992; Rothman, 1990, 2014; Steiger, 1980).

SPSS PROCESS Model 1 was used for the moderation analyses (IBM SPSS Statistics for Macintosh, 2021), controlling for sex and waist circumference (Koenig & Thayer, 2016; Windham et al., 2012). A total of three Asians, three Hispanics, and three NHW participants had extreme HRV values. However, these participants were retained as their HRV values were physiologically plausible relative to the participants’ other physiological measures. Pre-planned simple slope tests were performed (Lowry, n.d.; Meng et al., 1992). Extreme studentized residual scores and DFFITS, a global index of influence, were evaluated in regression diagnostics. No participant exerted an undue influence on the regression line or overall model fit (Belsley et al., 1980). As subjective social status is differentially associated with sensitivity to environmental and emotional cues (Kraus et al., 2011, 2012) and objective socioeconomic status with HRV (Hill et al., 2017; Sloan et al., 2005; Trisha et al., 2016), sensitivity analyses with subjective social status as a covariate were conducted. Sensitivity analyses with tobacco, alcohol, and physical activity as covariates were also conducted (Laborde et al., 2017; Sammito et al., 2024).

Results

Demographic characteristics

Our total sample (n = 1047) included individuals of Asian (n = 183, M_age = 19.7 [1.6] years, 96 women), AfAm (n = 82, M_age = 19.8 [1.8] years, 53 women), 228 Hispanic individuals (M_age = 19.7 [1.9] years, 134 women), and NHW backgrounds (n = 554, M_age = 19.7 [1.7] years, 334 women). Subjective social status was significantly different among the ethnic groups (F (1, 1043 = 47.44, r = .23 [0.17, 0.29], p < .001), with higher scores among NHW (7.4 [1.4]) compared to Asian (7.0 [1.6]), AfAm (6.6 [1.5]), and Hispanic individuals (6.6 [1.8]). Across the studies, there were no significant differences in reappraisal or suppression (ps > .27). Participants in Study 2 had lower HRV than those in Study 1 (t (640) = −2.01, r = −.08 [−0.16, 0.00], p = .04), Study 3 (t (580) = −5.39, r = .21 [0.13, 0.30], p < .001), and Study 4 (t (673) = −2.36, r = .09 [0.01, 0.17], p = .02). Participants in study 4 had lower HRV than those in study 3 (t (403) = −2.95, r = −.15 [−0.24, −0.05], p = .003). The distribution of women and men did not differ significantly across ethnic groups (X² [1047] = 4.69, p = .20). Regardless of ethnicity, women had significantly lower suppression scores than men (Asians: (t [181] = −3.80, r = −.27 [−0.42, −0.13], p < .001; AfAm: (t [80] = −2.28, r = −.25 [−0.47, −0.03], p = .03; Hispanics: (t [226] = −1.95, r = −.13 [−0.26, 0.00], p = .05; NHW: t [552] = −4.13, r = −.17 [−0.26, −0.09], p < .001). HRV was marginally lower in NHW women (t [552] = −1.44, r = −.06 [−0.14, 0.02], p = .08). No other sex differences were notable (ps > .34, rs < .03).

AfAm had significantly higher HRV than Asian (t (263) = 4.47, r = .27 [0.14, 0.39], p <.001), Hispanic (t [308] = 3.30, r = .18 [0.07, 0.30], p = .001), and NHW individuals (t [634] = 3.96, r = .16 [0.09, 0.22], p < .001). NHW had significantly lower suppression scores than Asian (t [735] = −2.74, r = −.10 [−0.16, −0.04], p = .003) and Hispanic individuals (t [780] = −2.08, r = −.07 [−0.14, 0.00], p = 0.04) and marginally lower scores than AfAm individuals (t [634] = −1.52, r = −.06 [−0.13, 0.01], p = .07). No other notable ethnic differences emerged (see Table 1).

Table 1.

Sample characteristics

	Sex (% Women/Men)	Waist (in.)	HRV (ms2)	Reappraisal	Suppression
	X2 (1047) = 4.69, p = .20	M (SD) rt [CI], p	M (SD) rt [CI], p	M (SD) rt [CI], p	M (SD) rt [CI], p
Asian Am, (n = 183)	52/41%	31.14 (5.30)a	6.35 (1.05)	30.73 (5.57)	17.36 (4.83)a
v. Blacks		−.09 [−0.21, 0.03] .13	−.27 [−0.39, 0.14] <.001	−.01 [−0.13, 0.11] .87	.03 [−0.10, 0.15] .67
v. Hispanics		−.17 [−0.27, 0.08] <.001	−.09 [−0.19, 0.01] .04	.02 [−0.08, 0.12] .70	.03 [−0.06, 0.13] .48
v. NHW		−.04 [−0.11, 0.03] .16	−.06 [−0.12, 0.00] .03	.03 [−0.03, 0.09] .22	.10 [0.04, 0.16] .003
AfAm (n = 82)	65/35%	32.14 (4.23)	6.97 (1.01)	30.85 (6.46)	17.10 (4.90)a
v. Hispanics		−.08 [−0.19, 0.03] .16	.18 [0.07, 0.30] .001	.03 [−0.09, 0.14] .66	.01 [−0.10, 0.12] .90
v. NHW		.05 [−0.03, 0.12] .12	.16 [0.09, 0.22] <.001	.03 [−0.04, 0.09] .23	.06 [−0.01, 0.13] .07
Hispanics (n = 228)	59/41%	33.12 (5.82)a	6.54 (1.02)	30.50 (6.25)a	17.02 (5.03)
v. NHW		.15 [0.08, 0.22] <.001	−.02 [−0.09, 0.05] .29	.01 [−0.06, 0.08] .74	.07 [0.00, 0.14] .04
NHW (n = 554)	60/40%	31.54 (4.34)a	6.49 (1.02)	30.34 (5.93)	16.18 (5.13)a

Note: Table 1 provides Pearson Chi-Square (X²) for differences between women and men by ethnicity. Effect size (r), associated confidence intervals for r’s (CI) for sex differences, and p-values (p) for waist in inches, HRV indexed by natural log-transformed high-frequency HRV (ms2), and cognitive reappraisal and expressive suppression measured using the Emotion Regulation Questionnaire (Gross & John, 2003). Ethnic differences between Asian Americans|Asian, African Americans|AfAm, Hispanic or Latino Americans|Hispanics, and Non-Hispanic White Americans (NHW) backgrounds are also presented. One-tailed tests were conducted for Asian, AfAm, and NHW individuals, and for all comparisons between these groups corresponding to our directional hypotheses (90% CIs reported). Comparisons between Asian and AfAm individuals were two-tailed. All tests with Hispanics were two-tailed.

Superscipt^a identifies sex differences p < .05. All p < .05 are bolded.

Ethnic differences in HRV associations with reappraisal and suppression

A marginally significant positive unadjusted HRV-reappraisal association in Asians (r = .13 [−0.02, 0.27], p = .09) and a negative HRV-reappraisal association in AfAm (r = −.19 [−0.41, 0.03], p = .09) were observed. Statistical difference tests of the magnitude of the associations showed that the negative HRV-reappraisal association in AfAm was significantly different than the positive associations in NHW (z = 1.96, p = .03) and Asian individuals (z = 2.35, p = .02). The positive HRV-reappraisal association in Asians was significantly different than the negative association in Hispanic individuals (z = 2.12, p = .03). The negative Hispanic association was marginally different relative to the positive association in NHW individuals (z = 1.65, p = .10). No other notable ethnic differences were observed, including suppression (zs < 1.25, ps > .21).

Moderation analyses by ethnicity

Statistically controlling for waist circumference and sex, pre-planned simple slope analyses showed that higher HRV was associated with marginally higher reappraisal scores in Asians (r_t = .12 [−0.01, 0.24], p = .06) and significantly lower reappraisal scores in AfAm (r_t = −.20 [−0.38, −0.02], p = .03) (Table 3, Figure 1). In contrast, among Hispanics, higher HRV was associated with lower reappraisal scores with a non-trivial effect size (r_t = −.09 [−0.22, 0.04]), comparable to a marginal trend in p-values and similar to or exceeding effect sizes reported in previous studies (Steiger, 2004; Thayer & Koenig, 2019; Watanabe et al., 2023). Similarly, among AfAm, higher HRV was associated with greater suppression scores and this relationship was non-trivial (r_t = .09 [−0.09, 0.27], p = .21; Table 4, Figure 2). Specifically, these effect sizes align with prior work on ER difficulties (Thayer & Koenig, 2019; Watanabe et al., 2023). Among NHW, higher HRV was associated with marginally less suppression use (r_t = −.06 [−0.13, 0.01], p = .07). Sensitivity analyses adjusting for social status yielded nearly identical results for both reappraisal and suppression models (r_z’s < |−.04|). Sensitivity analyses adjusting for tobacco, alcohol, and physical activity yielded nearly identical results for reappraisal and suppression (r_z’s < |−.03|).

Table 3.

Linear regression of reappraisal by HRV and ethnic group

	Ref = Asian						Group Differences
	b	SE	95%CI	t	p			z	r	CI	p
HF-HRV	0.67	0.42	−0.16, 1.50	1.59	.11		Asians v AfAm	2.37	.14	0.02, 0.26	.02
AfAm	12.86	5.36	2.34, 23.38	2.40	.02		Asians v. Hispanics	2.07	.10	0.01, 0.20	.04
Hispanics	7.51	3.75	0.16, 14.87	2.00	.045		Asians v. NHW	0.88	.03	−0.03, 0.09	.19
NHW	2.34	3.18	−3.90, 8.58	0.73	.46		AfAm v. Hispanics	−0.86	−.05	−0.16, 0.06	.39
Waistin	0.01	0.04	−0.06, 0.09	0.37	.71		AfAm v. NHW	−2.03	−.08	−0.14, 0.01	.02
Sex (W)	−0.39	0.39	−1.16, 0.38	−1.00	.32		Hispanics v. NHW	−1.66	−.06	−0.13, 0.01	.10
HRV × AfAm	−1.88	0.78	−3.41, −0.35	−2.41	.02
HRV × Hispanics	−1.2	0.58	−2.33, −0.08	−2.09	.04
HRV × NHW	−0.43	0.49	−1.39, 0.53	−0.88	.38
Conditional Effects
	b	SE	t	r	CI	p
Asians	0.67	0.42	1.59	.12	−0.01, 0.24	.06
AfAm	−1.21	0.66	−1.85	−.20	−0.38, 0.02	.03
Hispanics	−0.53	0.39	−1.37	−.09	−0.22, 0.04	.17
NHW	0.24	0.25	0.97	.04	−0.03, 0.11	.17

Note: Table X shows the regression of the interaction between heart-rate variability (HRV) indexed by natural log-transformed high-frequency HRV and ethnic group (e.g., HRV × Hispanic) on cognitive reappraisal and expressive suppression indexed by the Emotion Regulation Questionnaire (Gross & John, 2003), controlling for waist circumference in inches (Waistin). Unstandardized regression coefficients (b) and their associated standard errors (SE), t-values (t) p-values (p). Effect size (r) and their associated confidence intervals (CI) for the conditional effects of the ethnic groups on reappraisal are presented separately for each ethnic group. Group differences test the difference between slopes for ethnic groups relative to Asian Americans|Asians (e.g., Asian v. Hispanic); z-statistics (z), p, r, and associated CIs are reported. One-tailed tests were conducted for comparisons between Asians, African Americans|AfAm), and Non-Hispanic White Americans|NHW, corresponding to our directional hypotheses (90% CIs reported). Comparisons between Asian and AfAm individuals were two-tailed. All tests with Hispanic or Latino Americans|Hispanics were two-tailed. All p < .05 are bolded.

Figure 1.

Reappraisal Moderation by Ethnicity

Note: These figures show the conditional role of ethnicity on the relationship between HRV and reappraisal Asian Americans|Asians, African Americans|AfAm), Hispanic or Latino Americans|Hispanics, and Non-Hispanic White Americans|NHW.

Table 4.

Linear regression of suppression by HRV and ethnic group

	Ref = Asian						Group Differences
	b	SE	95%CI	t	p			z	r	CI	p
HF-HRV	0.18	0.35	−0.51, 0.87	0.51	.61		Asians v AfAm	−0.32	−.02	−0.14, 0.10	.75
AfAm	−2.05	4.44	−10.77, 6.67	−0.46	.65		Asians v. Hispanics	0.43	.02	−0.08, 0.12	.67
Hispanic	0.69	3.11	−5.41, 6.78	0.22	.83		Asians v. NHW	1.27	.05	−0.01, 0.11	.10
NHW	2.08	2.63	−3.09, 7.25	0.79	.43		AfAm v. Hispanics	0.66	.04	−0.07, 0.15	.51
Waistin	0.07	0.03	0.00, 0.13	2.08	.04		AfAm v. NHW	1.27	.05	−0.01, 0.12	.10
Sex (W)	−1.74	0.32	−2.37, −1.10	−5.36	< .001		Hispanics v. NHW	0.83	.03	−0.04, 0.10	.40
HRV × AfAm	0.26	0.65	−1.01, 1.53	0.40	.69
HRV × Hispanic	−0.17	0.48	−1.10, 0.77	−0.35	.73
HRV × NHW	−0.49	0.41	−1.29, 0.31	−1.20	.23
Conditional Effects
	b	SE	t	r	CI	p
Asians	0.18	0.35	0.61	.05	−0.10, 0.19	.31
AfAm	0.44	0.54	0.81	.09	−0.13, 0.31	.21
Hispanics	0.01	0.32	0.03	.002	−0.13, 0.13	.97
NHW	−0.31	0.21	−1.51	−.06	−0.15, 0.02	.07

Note: Table X shows the regression of the interaction between heart-rate variability (HRV) indexed by natural log-transformed high-frequency HRV and ethnic group (e.g., HRV × Hispanic) on expressive suppression indexed by the Emotion Regulation Questionnaire (Gross & John, 2003) controlling for waist circumference in inches (Waistin). Unstandardized regression coefficients (b) and their associated standard errors (SE), t-values (t) p-values (p). Effect size (r) and their associated confidence intervals (CI) for the conditional effects of the ethnic groups on suppression are presented separately for each ethnic group. Group differences test the difference between slopes for ethnic groups relative to Asian Americans|Asians (e.g., Asian v. Hispanic); z-statistics (z), p, r, and associated CIs are reported. One-tailed tests were conducted for comparisons between Asians, African Americans|AfAm), and Non-Hispanic White Americans|NHW, corresponding to our directional hypotheses (90% CIs reported). Comparisons between Asian and AfAm individuals were two-tailed. All tests with Hispanic or Latino Americans|Hispanics were two-tailed. All p < .05 are bolded.

Figure 2.

Suppression Moderation by Ethnicity

Note: These figures show the conditional role of ethnicity on the relationship between HRV and suppression among Asian Americans|Asians, African Americans|AfAm), Hispanic or Latino Americans|Hispanics, and Non-Hispanic White Americans|NHW.

Simple slope difference tests showed that the negative relationship between HRV and reappraisal scores in AfAm was significantly different than the positive relationships in Asians (r_z = −.14 [−0.26, −0.02], p = .02) and NHW (r_z = −.08 [−0.14, 0.01], p = .02). The positive relationship in Asians was significantly different than the negative relationship in Hispanics (r_z = .10 [0.01, 0.20], p = .04). The Hispanic association was marginally different from the positive relationship in NHW (r_z = −.06 [−0.13, 0.01], p = .10). Moreover, the negative relationship between HRV and suppression in NHW was marginally different from the positive relationships in Asians (r_z = .05 [0.01, 0.11], p = .10) and AfAm (r_z = .05 [−0.01, 0.12], p = .10). No other notable ethnic differences emerged.

Discussion

Previous studies have examined the role of ethnicity (e.g., Asian Americans and European Americans) in the relationship between HRV and ER difficulties (Watanabe et al., 2023). However, the present study adds new evidence demonstrating that, in American sociocultural contexts, HRV is differentially associated with two habitual ER tendencies: reappraisal and suppression. Notably, we provide the first evidence of the relationship between HRV and ER among individuals from Latino and Hispanic backgrounds living in the U.S. Our findings reveal that AfAm and Hispanic individuals exhibited a unique and similar pattern of higher HRV and lower reappraisal, while Asians and NHW showed a tendency for more habitual reappraisal use with higher HRV. This partially confirms existing literature (Soto et al., 2012; Thayer & Koenig, 2019; Watanabe et al., 2023). The reappraisal and HRV relationships were of greater magnitude among individuals who self-identified as U.S. ethnic minorities (i.e., Asians, AfAm, and Hispanics) compared to NHW. For AfAm and NHW, the reappraisal relationships with HRV were significantly different compared to the suppression relationships. NHW individuals displayed a trend toward lower suppression, while Asian and AfAm individuals showed a trend toward higher suppression with higher HRV. These findings lend additional empirical support to the Sociocultural Neurovisceral Model, suggesting that neurovisceral capacity (i.e., HRV) and culturally patterned ideas and practices may jointly influence ER strategies, such as reappraisal and suppression in American sociocultural contexts (Watanabe et al., 2025). Although much more work is needed in this domain, our findings enhance our understanding of individual differences in the physiological underpinnings of ER among individuals who self-identified as Asian, AfAm, and Latino or Hispanic, who may frequently regulate their emotions in the face of unfair treatment and racial discrimination in the U.S. (Han et al., 2023; Hwang & Goto, 2008; Lee et al., 2019; Umaña-Taylor & Updegraff, 2007).

Consistent with meta-analytic and cross-sectional findings, AfAm had the highest resting HRV and Asian individuals showed the lowest among all ethnic groups (Hill et al., 2015; Kemp, Koenig, et al., 2016; Pourmand et al., 2024; Thayer & Koenig, 2019). Interestingly, although reappraisal use was similar across all ethnic groups, averaging approximately 71% of the maximum score with less than one point difference, the relationship between HRV and reappraisal varied meaningfully. Specifically, the magnitude of this effect was greater among individuals who self-identified as U.S. ethnic minorities (i.e., Asians, AfAm, and Hispanics) compared to NHW. Importantly, in line with our expectations, higher HRV was associated with higher reappraisal in both Asians and NHW individuals. This finding aligns with previous research conducted with primarily Asian and NHW samples in the U.S. (Butler et al., 2006) and with European samples (Makovac et al., 2022; Visted et al., 2017). Moreover, our finding supports the notion that higher HRV is associated with fewer difficulties with ER strategies in these groups (Watanabe et al., 2023; Williams et al., 2015), which in turn is associated with greater reappraisal (Sörman et al., 2022; Zelkowitz & Cole, 2016). Notably, the effect size of HRV on reappraisal was three times stronger among Asian individuals than among NHW individuals. This disparity is even more pronounced than in prior work, where the effect of higher HRV on fewer difficulties with ER strategies was only twice as strong among Asian Americans compared to European Americans (Watanabe et al., 2023).

In contrast, higher HRV was linked with lower reappraisal among AfAm and Hispanics, even though they exhibited the highest HRV among the ethnic groups. This suggests that even with a greater capacity to regulate their emotions, it may be particularly challenging to alter the meaning of negative situations before an emotional response occurs, especially in specific situational contexts (Ford & Troy, 2019; Gross & John, 2003; John & Gross, 2004; Troy et al., 2018). For instance, reappraisal may be used less frequently during high-intensity negative situations, such as in contexts of racial discrimination, an experience common among AfAm and Hispanic individuals (Lee et al., 2019; Sheppes et al., 2011). Notably, in American sociocultural contexts, AfAm individuals report the highest levels of racial discrimination (Lee et al., 2019), and the point estimate for the effect size of higher HRV on lower reappraisal was twice as large in this group compared to Hispanic individuals. However, the compatibility intervals overlapped considerably, implying that shared experiences of racial discrimination may partly explain the lower reappraisal tendencies observed among individuals in these groups. On the other hand, our study’s cross-sectional design raises the possibility of reverse causality; frequent unsuccessful reappraisal attempts could negatively impact psychological health (Ford et al., 2017; Ford & Troy, 2019; Perez & Soto, 2011; Soto et al., 2012; Wheaton et al., 2018), potentially leading to lower HRV. This idea is supported by research indicating that racism, a significant determinant of health in the U.S. (Boyd et al., 2020), is associated with depression (Wheaton et al., 2018) and lower resting HRV (Hill et al., 2017), characteristics associated with major depressive disorder (for meta-analyses, seeBrown et al., 2018; Koch et al., 2019; Koenig & Thayer, 2016). Indeed, in contexts of high oppression and discrimination, reappraisal may be less beneficial for psychological outcomes among Latinos (Perez & Soto, 2011; Soto et al., 2012). Furthermore, the perceived controllability of situational factors can also influence ER tendencies (Nolen-Hoeksema & Aldao, 2011; Nolen-Hoeksema & Jackson, 2001; Tamres et al., 2002; Troy et al., 2013). Thus, higher HRV may be especially relevant in situational contexts like discrimination, where early appraisals of situational control influence the deployment of reappraisal (Saus et al., 2006; Thayer et al., 2009). Future research should investigate these possibilities, including whether discrimination, a common experience among ethnically marginalized groups in the U.S. (Lee et al., 2019), mediates the relationship between HRV and reappraisal, as discrimination was not measured in the current study.

The relationships between HRV and suppression were similar in magnitude to those found in cross-sectional studies and meta-analyses (Hu et al., 2014; Thayer & Koenig, 2019; Webb et al., 2012). Notably, a substantial portion of our confidence intervals suggested that NHW individuals showed a trend toward lower suppression (Makovac et al., 2022; Visted et al., 2017), whereas Asians and AfAm individuals exhibited trends toward higher suppression with higher HRV. This inverse association is not entirely unexpected. In Butler et al. (2006), a racially diverse sample, including primarily European, Asian, African, and Latin American individuals, women instructed to either suppress or reappraise during a social interaction both exhibited increases in RSA (HF-HRV) compared with uninstructed controls. These findings suggest that increased parasympathetic activity may accompany active regulatory effort regardless of whether the strategy is inhibitory (suppression) or cognitive (reappraisal). Thus, the contrasting patterns observed in our study imply that higher HRV may enhance one’s capacity to engage in inhibitory control, supporting the automatic suppression of unwanted thoughts and emotional expressions (Beauchaine & Thayer, 2015; Butler et al., 2006; Finley et al., 2024; Gillie et al., 2014, 2015; Gross, 2001; Gross & John, 2003; Melzig et al., 2009; Spangler et al., 2021; Thayer & Lane, 2009), in ways that align with internalized cultural values (Eisenberg, 2020; Eisenberg et al., 1998; Kitayama et al., 2000; Mauss et al., 2007; Saus et al., 2006; Thayer et al., 2009; Thayer & Lane, 2009). For example, the higher average suppression levels seen among Asian and Hispanic individuals compared to NHW may reflect cultural values that prioritize interconnection and harmony (Deng et al., 2019; Kitayama et al., 2000; Kitayama & Salvador, 2024; Ramírez-Esparza et al., 2008; Soto et al., 2005; Su et al., 2015; Wierzbicka, 2004). Conversely, lower suppression use may reflect autonomous values that promote open emotional expression in NHW American cultural contexts (Campos et al., 2014; Coon & Kemmelmeier, 2001; Hofstede, 1980; Kitayama et al., 1997; Kitayama & Salvador, 2024; Markus & Kitayama, 1991; Raval & Walker, 2019; Rothbaum & Rusk, 2011; Trommsdorff et al., 2012; Wierzbicka, 1994). Furthermore, the relatively higher HRV and suppression use among AfAm individuals compared to NHW may reflect regulatory strategies shaped by ongoing experiences of sociocultural stress, including chronic racism, which may be particularly salient in American social contexts (Berlin et al., 2002; Brondolo et al., 2009; Coll et al., 1996; Dorr et al., 2007; Seaton et al., 2012; Sheppes et al., 2011; Thayer et al., 2020). Our results suggest that in these American sociocultural contexts, higher HRV may facilitate the upregulation of suppression when culturally appropriate or accepted, or the downregulation of suppression where emotional expression is normative (e.g., among NHW) (Cole et al., 1994; Coon & Kemmelmeier, 2001; Dunsmore & Halberstadt, 1997; Halberstadt & Eaton, 2002; Kitayama et al., 1997; Kyeong et al., 2021; Markus & Kitayama, 1991; Matsumoto, 1993; Salvador et al., 2024; Senft et al., 2021, 2023; Watanabe et al., 2025; Wilson & Gentzler, 2021). Interestingly, despite differences in HRV and similar levels of suppression use, higher HRV was linked to suppression at a similar magnitude across all minoritized ethnic groups (Gross & John, 2003). Notably, this pattern was most similar among the groups with the highest and lowest HRV, AfAm and Asian individuals. These findings support research indicating that individuals in these groups tend to strongly identify with their ethnicity (Charmaraman & Grossman, 2010; Pellebon, 2000; Seaton et al., 2012), share similar cultural orientations (Coon & Kemmelmeier, 2001), experiences of discrimination (Han et al., 2023; Hwang & Goto, 2008; Lee et al., 2019), and, despite differences in HRV, a biomarker of goal-related ER capacity, may exhibit similar ER tendencies, particularly in American sociocultural contexts. Trending differences observed between NHW participants and those of Asian and AfAm backgrounds suggest that larger samples might uncover significant ethnic differences. Further research with a larger AfAm sample is necessary to determine whether HRV is differently associated with suppression among this group and other ethnic groups in the U.S. Additionally, exploring individuals’ desired emotional states (i.e., ideal affect; Tsai, 2007) and ensuring cross-cultural validity of ER measures could clarify these findings.

Given previous findings, we expected that higher HRV would be associated with less frequent emotional suppression among Asian individuals (Sörman et al., 2022; Watanabe et al., 2023; Zelkowitz & Cole, 2016). The differing results observed among Asian participants in the present study may be attributed to variations in sample composition. Specifically, the inverse relationships between the ERQ and DERS were demonstrated in primarily NHW samples (Gratz & Roemer, 2004; Sörman et al., 2022; Zelkowitz & Cole, 2016). Therefore, future studies should explicitly compare these associations across various ethnic groups in the U.S. and other geographical and cultural contexts. Additionally, differences in social status are linked to varying sensitivities to environmental and emotional cues (Kraus et al., 2011, 2012; Troy et al., 2017) and HRV (Hill et al., 2017; Sloan et al., 2005; Trisha et al., 2016). In our study, Texas college students rated their subjective social status as above average, unlike studies conducted at a university in Ohio where social status was not reported (Watanabe et al., 2023; Williams et al., 2019). However, when we controlled for social status in our sensitivity analyses, the results with and without social status adjustments were largely consistent for both reappraisal and suppression. This suggests that, although social status was a significant covariate in the reappraisal models, social status-related differences had minimal impact on the relationships between suppression, reappraisal, and HRV. Still, variations in sample composition could limit the comparability of our findings with studies that did not stratify by ethnicity (Sörman et al., 2022; Zelkowitz & Cole, 2016) or those with lower, unmeasured, or objective measures of social status (Watanabe et al., 2023). Consequently, future research should investigate various measures of ER and objective measures of socioeconomic status within multiethnic samples to further clarify these findings.

From a theoretical perspective, our findings emphasize the value of the Sociocultural Neurovisceral Model in American and East Asian sociocultural contexts (Watanabe et al., 2025). According to this model, sociocultural contexts and neurovisceral capacity, indexed by vagally mediated HRV, can jointly influence regulatory behavior. This model, grounded in the perspectives of the NIM and the NCIM, suggests that sociocultural contexts shape culturally patterned emotion regulation tendencies, strengthening specific neurovisceral pathways over time (Kitayama & Uskul, 2011; Smith et al., 2017; Thayer et al., 2021; Thayer & Lane, 2000; Watanabe et al., 2025). These neural adaptations reinforce habitual emotion regulation patterns, creating a dynamic, reciprocal loop between culture, physiology, and regulation (Watanabe et al., 2025). Through repeated interactions with their social environment, individuals internalize attitudes and values, engaging in culturally patterned and meaningful practices like ER (Berlin et al., 2002; Coll et al., 1996; Kitayama et al., 2000; Lee et al., 2019; Raval & Walker, 2019; Trommsdorff et al., 2012; Wilson & Gentzler, 2021). For example, AfAm caregivers may socialize their children to suppress emotional expressions within an American context of prevalent experiences of racism and discrimination (Berlin et al., 2002; Coll et al., 1996; Lee et al., 2019; Wilson & Gentzler, 2021). As a result, levels of suppression of negative emotions tend to be higher among AfAm adolescents (e.g., Steele et al., 2005) and emerging adults, as seen in our study. Similarly, suppression among Asian individuals may serve adaptive, culturally accepted functions, such as maintaining harmonious relationships and reducing conflict, particularly in American sociocultural contexts (Cole et al., 1994; Coon & Kemmelmeier, 2001; Dunsmore & Halberstadt, 1997; Halberstadt & Eaton, 2002; Kitayama et al., 1997; Kyeong et al., 2021; Markus & Kitayama, 1991; Matsumoto, 1993; Salvador et al., 2024; Wilson & Gentzler, 2021). Since individuals are social, cultural, and biological beings, these repeated, culturally patterned behaviors become embedded at the neural level over time, consistent with the Hebbian principle of long-term potentiation (Anderson, 2010; Kitayama & Cohen, 2007; Kitayama & Uskul, 2011; Markus & Hamedani, 2007). Evidence from cross-sectional and meta-analytic fMRI studies indicates differences in task-based and resting neural patterns between individuals from cultures that emphasize harmonious social relationships and those with more autonomous values (Deng et al., 2023; Han & Ma, 2014b; Kitayama et al., 2017b; Kraus & Kitayama, 2019; Li et al., 2018; Luo et al., 2020; Wang et al., 2017). Therefore, HRV, a biomarker of inhibitory cortical-subcortical circuits that support adaptive ER (i.e., consistent with regulatory goals), may differently facilitate culturally patterned behaviors, such as habitual ER, depending on sociocultural contexts (Appelhans & Luecken, 2006; Beauchaine & Thayer, 2015; Jennings et al., 2016; Melzig et al., 2009; Sakaki et al., 2016; Thayer et al., 2012; Thayer & Lane, 2000, 2009). In our study, higher HRV was linked to lower reappraisal among AfAm and Hispanic individuals, who perceived their social status as lower. Conversely, Asian and NHW individuals perceived their social status as higher and reported greater use of reappraisal. The overlapping compatibility intervals suggest that sociocultural contexts, such as social status in American contexts, may influence habitual ER patterns and the physiological mechanisms underlying them (Kraus et al., 2011, 2012). Our results underscore the importance of examining strategy-specific pathways (e.g., antecedent-focused versus response-focused) to gain a deeper understanding of how the brain and body dynamically co-regulate emotions and behavior across diverse sociocultural contexts.

Strengths, Limitations, and Future Directions

Several limitations of our work should be acknowledged. First, we lacked sufficient power to separate our ethnic groups by specific cultural heritage. As a result, our “Hispanic” group included individuals from Spanish-speaking countries, Latin-American countries, and Mexico, while our Asian group comprised both South and East Asians. Therefore, our findings should be interpreted with caution, as aggregating cultural groups may have introduced unintentional bias (M. R. Wilson et al., 2024). Nonetheless, this study is the first to investigate the HRV-ER relationship among Hispanic or Latino individuals and to compare how HRV relates to specific ER strategies across individuals with Asian, AfAm, Hispanic, and NHW backgrounds. Second, we only examined two ER strategies, which may be influenced by contextual factors such as the nature of the situation being appraised (e.g., academic or interpersonal) or cultural values, which are neither fixed nor uniform (Campos et al., 2014; Campos & Kim, 2017; Carver et al., 1989; Dixon-Gordon et al., 2015; Goubet & Chrysikou, 2019; Kitayama et al., 1997; Kitayama & Cohen, 2007; Kwon et al., 2013; Markus & Hamedani, 2007; Markus & Kitayama, 1991; Nolen-Hoeksema & Aldao, 2011; Nolen-Hoeksema & Jackson, 2001; Rogier et al., 2019; Tamres et al., 2002). These unmeasured contextual factors, including ER strategies such as acceptance (Wojnarowska et al., 2020), direct measures of cultural identity and practices, and levels of acculturation, could further clarify our findings, since ethnicity alone may not fully capture shared cultural or social experiences (Clauss-Ehlers et al., 2019; Comas-Díaz, 2011). Nonetheless, by examining both antecedent-focused (reappraisal) and response-focused (suppression) ER strategies, our design allowed us to identify potentially meaningful individual differences in how HRV may support distinct ER processes across ethnicities. Furthermore, future studies should explore whether the relationship between HRV and ER varies based on gender or sex (Williams et al., 2019). It is important to note that, although many of our effect sizes did not reach statistical significance, a substantial portion of the compatibility intervals support our conclusions, which enhances the interpretability and robustness of our results. Additionally, our effect sizes are comparable in magnitude to those found in meta-analyses examining suppression and reappraisal effects on physiological outcomes (Webb et al., 2012) and suppression effects on negative mental health outcomes like depression and anxiety among individuals with Eastern and Western cultural values (Hu et al., 2014). Nevertheless, our findings should be interpreted with caution due to the relatively healthy and physically active profile of our sample, our focus on an American sociocultural context, and the cross-sectional nature of our study, which may limit the generalizability of our results to other populations where cardiovascular health is more heterogeneous and clinically significant, as well as other geographical and cultural settings.

Conclusion

Together our findings align with existing literature that indicates individuals from various ethnic backgrounds exhibit both similar and divergent ER patterns (Miyamoto et al., 2017; Salvador et al., 2024; Senft et al., 2021, 2023; Watanabe et al., 2025; Watanabe et al., 2023). These patterns may be influenced by higher resting HRV, an index of emotional and cognitive regulation (Appelhans & Luecken, 2006; Thayer et al., 2012; Thayer & Lane, 2009; Williams et al., 2015; Wilson & Gentzler, 2021). The relationship patterns observed were contingent upon on whether the ER strategy employed was antecedent- or response-focused, and, importantly on the individual’s ethnicity. This suggests that both HRV and U.S. sociocultural contexts play a meaningful role in shaping the deliberate and automatic use of reappraisal and suppression (Cole et al., 1994; Gross & John, 2003; John & Gross, 2004; Seaton et al., 2012; Senft et al., 2023; Tamir & Hu, 2024; Watanabe et al., 2023; Wilson & Gentzler, 2021). While these findings enhance our understanding of how physiology, ethnicity, and ER intersect, substantially more research is required to achieve the ultimate goal of improving health outcomes among ethnically diverse individuals.

Table 2.

Pearson’s zero-order correlations with HRV for all variables stratified by ethnicity

Ethnicity	Reappraisal	Suppression		Asian Difference		AfAm Difference		Hispanic Difference		NHW Difference
	r [CI]	r [CI]		r of z [CI]	p	r of z [CI]	p	r of z [CI]	p	r of z [CI]	p
Asians	.13 [−0.02, 0.27]	.03 [−0.11, 0.18]	Reappraisal	-	-	−.03 [−0.15, 0.09]	.64	.01 [−0.08, 0.11]	.77	.03 [−0.04, 0.11]	.17	Suppression
AfAm	−.19 [−0.41, 0.03]	.10 [−0.12, 0.31]		.14 [0.02, 0.26]	.02	-	-	.04 [−0.07, 0.15]	.49	.05 [−0.03, 0.12]	.11
Hispanics	−.09 [−0.22, 0.04]	.01 [−0.12, 0.14]		.10 [0.01, 0.20]	.03	−.05 [−0.16, 0.07]	.42	-	-	.02 [−0.05, 0.09]	.50
NHW	.04 [−0.04, 0.13]	−.05[−0.13, 0.03]		.04 [−0.04, 0.11]	.17	−.08 [−0.15, 0.00]	.03	−.06 [−0.13, 0.01]	.10	-	-

Note: Table 2 shows Pearson’s zero−order correlations (r) with heart-rate variability (HRV) indexed by natural log-transformed high-frequency HRV stratified by self-identified ethnic group (Asian Americans|Asians, African Americans|AfAm), Hispanic or Latino Americans|Hispanics, and Non-Hispanic White Americans|NHW) between cognitive reappraisal and expressive suppression indexed by the Emotion Regulation Questionnaire (Gross & John, 2003). Correlation differences (r of z-score|z) and associated confidence intervals (CI) and p-values (p) for reappraisal and suppression appear below and above the diagonal, respectively. One-tailed tests were conducted for comparisons between Asians, AfAm, and NHW, corresponding to our directional hypotheses (90% CIs reported). Comparisons between Asian and AfAm individuals were two-tailed. All tests with Hispanics were two-tailed. All p < .05 are bolded.

Data availability

Data will be made available on request.