Abstract
This study examines whether habitual emotional suppression—a response-focused emotion regulation strategy documented to be more prevalent among men adhering to traditional U.S. masculine norms—predicts chronic physiological dysregulation across the life course. Using a cross-sectional sample of 412 men aged 25–70 years recruited in a single U.S. metropolitan region, we measured habitual suppression with the four-item Suppression subscale of the Emotion Regulation Questionnaire (ERQ), diurnal salivary cortisol collected across three weekdays as an index of hypothalamic–pituitary–adrenal (HPA) axis regulation, and inflammatory burden as a z-standardized composite of serum high-sensitivity C-reactive protein (hsCRP) and interleukin-6 (IL-6). Masculine norm conformity was assessed using the Emotional Control subscale of the Conformity to Masculine Norms Inventory-46 (CMNI-46), and stressor exposure was indexed by a life events checklist. Structural equation models (SEMs) indicate that habitual suppression significantly predicts flatter diurnal cortisol slopes (β = 0.34) and elevated inflammatory burden (β = 0.28) after controlling for demographics, health behaviors, and medical factors. Suppression also acts as a stress amplifier, magnifying the physiological effects of stressor exposure beyond additive contributions. Age-stratified analyses demonstrate that suppression effects strengthen across the life course, consistent with cumulative biological loading. Mediation analyses indicate that suppression explains 38%–43% of the association between masculine norm conformity and physiological dysregulation. These findings position emotional suppression as a biologically consequential risk pathway through which masculine gender socialization is physiologically embodied among men in the United States.
Keywords
Introduction
Male mortality rates exceed female mortality rates in nearly all developed countries, and it remains the case even after decades of governmental intervention in the health field. Although there is evidence of biological influences on sex differences about longevity, there is evidence that gender socialization, especially male norms of emotional stoicism, self-reliance, and invulnerability, may also turn into more physiological health risks via behavioral and biological mechanisms. However, the exact mechanisms by which masculine emotion standards are correlated with physiological performances are still poorly understood. Emotional suppression, which is the conscious inhibition of continuous emotion-expressive action, is another particularly consequential mechanism. The meta-analytic data confirm the fact that men tend to suppress emotions more than women, especially the emotions that indicate vulnerability, which include sadness, fear, and anxiety (Flynn et al., 2010). The present study is situated within the U.S. context, drawing on psychological and public health research on masculinity norms among men in the United States. Research in this tradition, rooted in both quantitative psychology and social science scholarship, has documented how culturally specific ideals of manhood shape men’s health-related behavior and wellbeing (Kimmel, 2006; Levant & Wimer, 2014). We acknowledge that masculinity norms are culturally variable and that findings may not generalize across national, racial/ethnic, or socioeconomic contexts. Psychological and public health research on masculinity in the United States reveals that adherence to the traditional masculine norms is associated with less emotional expression and help-seeking, and its effects on mental health outcomes are reported (Mokhwelepa & Sumbane, 2025; Wong et al., 2017). Nevertheless, there is very little research that looks at whether habitual suppression of emotion has any physiological health risks by way of chronic stress system dysregulation.
It is hypothesized that emotional suppression ought to produce physiological effects in several pathways as theorized. Emotion regulation process model postulates that suppression is cognitively expensive, and therefore, it takes long-term executive regulation to suppress the expression of emotions and keep arousal levels of emotions low in the body (Gross & John, 2003). This dissociates internal arousal and external expression that can be dysregulating of the biological systems that respond to stress. This mechanism is also empirically supported, with the laboratory research showing that instructed suppression in stressful behaviors results in an increase in the activity of the sympathetic nervous system relative to non-suppression states (Tyra et al., 2024). Yet critical gaps remain. To begin with, the majority of suppression studies use acute laboratory models that study instant physiological reactions to instructed suppression, leaving it unclear whether habitual suppression in everyday life initiates chronic physiological load. Second, suppression studies seldom focus on the effects of suppression on amplification of physiological responses to stressors other than the effect of the mere addition of suppression to other effects of stress. Third, the studies conducted on masculinity focused on physiological outcomes often rely on behavioral health practices (diet, exercise, health care usage) but not emotion regulation, which is a direct biological risk process. Fourth, developmental approaches that analyze the variation in effects of suppression through the life course are lacking, although there exist theoretical grounds that support the presence of cumulative effects.
This article fills these gaps by combining psychophysiological biomarkers and validated emotion regulation measuring in a heterogeneous sample of men (25–70 years). We will test four hypotheses based on allostatic load theory (McEwen & Stellar, 1993) and life course epidemiology (Kuh et al., 2003):
Our input is that emotional suppression is not only a psychological or behavioral process, but it is also a biological risk process where gender socialization is physiologically manifested. Through the use of intensive measurement of not only suppression (scale-validated self-report measures) but also physiological results (objective biomarker measures over multiple days) and structural equation modeling, which allows tests of both mediation and moderation, we demonstrate the existence of mechanisms between masculinity and physiological health that have not been previously studied due to the sole use of health behaviors.
Literature Review
Emotional Suppression as Strategy of Emotion Regulation
Emotional suppression is the process of suppressing continuing emotion-expressive behavior subsequent to the generation of an emotion, which is a response-oriented type of control in the process model of Gross and John (2003). Contrary to reappraisal, which entails altering the way of thinking about the situations involving emotion-inducing stimuli prior to the onset of emotion, suppression tries to cope with the emotions once they occur by regulating their external manifestation. This temporal difference is consequentially relevant since the suppression does not produce any alteration in the emotional arousal and avoids the expression of behavior, which produces physiological–behavioral dissociation.
The habitual suppression of individual differences is significantly varied and correlates meaningfully. The Emotion Regulation Questionnaire (ERQ) was used to assess habitual use of suppression and reappraisal and was developed by Gross and John (2003) to prove that suppression is associated with worse psychological outcomes, poorer social functioning, and worse wellbeing. John and Gross (2004) developed this to the lifespan, discovering that suppression is comparably steady throughout adulthood with demonstrating expenses in several adjustment domains.
Psychological costs of the phenomenon of suppression are supported by meta-analytic evidence. Synthesizing 306 studies of emotion regulation strategies’ efficacy, Webb et al. (2012) discovered that suppression has a small negative impact on emotional experience and expression with an uncertain impact on physiological responding. Aldao et al. (2010) meta-analyzed regulation strategies in psychopathology and found that suppression had a positive relationship with anxiety, depression, and other mental health issues, but the effects sizes were small. Cognitive overheads of suppression have been well documented. Richards and Gross (2000) show that suppression reduces the memory of information that is conceived during suppression, implying that executive resources that are necessary to prevent expression absorb cognitive ability to encode information. These results suggest that chronic suppression can induce chronic demands on cognitive resources that could cause chronic physiological stress due to repeated demands on executive control.
Acute Physiological Implication of Emotional Suppression
Laboratory studies relating to short-term physiological changes of taught suppression indicate that there is a stable sympathetic nervous system arousal. The initial quantitative synthesis of suppression and acute physiological stress responses among healthy populations was done by Tyra et al. (2024), who synthesized 27 experimental and correlational studies. Their meta-analysis showed that inhibition under laboratory stressors results in greater cardiovascular reactivity, elevated blood pressure, and greater sympathetic activation than control conditions characterized with non-suppression.
Nevertheless, Tyra et al. (2024) observe a great level of heterogeneity among the studies and methods of measuring, and the effects can be both small and large, depending on the types of stressors, timing, and traits of the samples. Importantly, in their review, a significant gap is found, as acute studies reveal that there are indeed immediate physiological costs, but practically no studies show that the chronic physiological dysregulation measured by biomarkers of stress occurs in everyday life as a result of the habitual use of suppression.
There is partial evidence of cortisol reactivity and emotion regulation research on how emotion regulation works in a chronic fashion. In their study, Lam et al. (2009) have investigated emotion regulation strategies in the context of a social-evaluative speech, demonstrating that more emotional control (but not confined to suppression) forecasted a blunt cortisol reactivity in a few instances. This is an indication of complicated interactions in which suppression might intensify or dull stress hormone reactions based on individual variations as well as environmental influences.
Biology of Stress: Cortisol and Inflammatory Markers
It is necessary to base the understanding of suppression on stress biology to comprehend the possible physiological implications of suppression. The hypothalamic–pituitary–adrenal (HPA) axis is the Lot No. 1 neuroendocrine system that is activated by psychological stress, and cortisol is the major hormone that it secretes. A normal functioning of the HPA axis shows high levels of diurnal rhythm; cortisol levels are high just after waking up (cortisol awakening response) and decrease throughout the day, with peaks in the evening. This diurnal slope gives an index of the HPA axis regulation, and the slopes were flatter, which indicates dysregulation (Adam et al., 2017).
In a meta-analysis of diurnal cortisol slope and mental and physical health outcomes in 72 studies, Adam et al. (2017) found that flatter slopes in the morning were predictors of a higher mortality risk, adverse cardiovascular outcomes, and higher levels of mental health symptoms. Notably, they also observe that exposure to chronic stress is a major contributor that leads to slope flattening, and thus, processes that increase the effect of stress such as emotional suppression can lead to dysregulation of cortisol rhythms.
Another important pathway is the stress-inflammation linkage. In thorough reviews, Steptoe et al. (2007) and Marsland et al. (2017) showed that acute psychological stress leads to the elevation of the circulating levels of inflammatory markers such as C-reactive protein (CRP) and interleukin-6 (IL-6). Exposure to chronic stress leads to persistent inflammatory response, which is a cause of cardiovascular disease, metabolic syndrome, and other chronic illnesses (Slavich & Irwin, 2014).
There are various mechanisms, which connect psychological stress with inflammation. Powell et al. (2013) show that social stress causes inflammatory gene expression in 2-adrenergic pathways that involve the activation of the sympathetic nervous system. As explained by O’Connor et al. (2009), behavioral and biological pathways can also mediate the effects of psychological factors on the inflammatory markers, such as regulation strategies. Ridker (2016) follows the path of inflammatory cascades of CRP to IL-6 to IL-1, making inflammation the key to atherosclerotic disease processes.
The life-course perspectives focus on the cumulative effects. Hostinar et al. (2015) show that childhood adversity and recent stressors do not add up to midlife inflammation; rather, they add to the physiological burden of stress exposures through the lifespan. Carroll et al. (2013) discover that childhood maltreatment forecasts multisystem biological risk in adulthood and long-term physiological imprinting of childhood stress. These results indicate that the cumulative physiological load could be more with age because of habitual suppression starting at young adulthood.
HPA Axis Dysregulation and Mental Health
The interconnection between the mental health and the functioning of HPA axis is a very important context. The authors suggest combined awareness of dysregulation of the biological stress system in depressive and anxiety disorders (Vinkers et al., 2021), indicating that both hyper- and hypo-cortisolism may be a symptom of pathological adaptation to chronic stress. George et al. (2025) assess the cortisol axis in psychiatric disorders, highlighting variability in the patterns of HPA dysfunction. Oldehinkel and Bouma (2011) discuss gender differences in HPA reactivity during adolescence and conclude that females are more reactive to social-evaluative stressors than males, whereas males are more blunted. This trend can be attributed to the different ways males may have been socialized to control emotional expression. If males are socialized to inhibit emotional expression from adolescence onward, this may gradually desensitize HPA reactivity.
Fiksdal et al. (2019) show that there are relationships between symptoms of depression/anxiety and cortisol response to acute stress, as well as recovery patterns. Importantly, they conclude that emotion regulation problems predict slower cortisol recovery and that the process of suppression, or inhibition of adaptive emotional processing, may be a reason for prolonged stress hormone elevations.
The attenuation hypothesis formulated by Susman (2006) provides the theoretical basis of the study of cumulative suppression effects. According to this framework, chronic stress exposure leads to progressive impairment of the HPA axis, shifting the original hyper-reactivity to ultimate hypocortisolism and loss of regulatory capacity. When stress systems are chronically burdened by suppression, the same patterns of attenuation could appear in adulthood.
Masculinity Standards and Men’s Health
Masculine gender expectations that focus on emotional control, independence, and invincibility exert cultural pressures that force men to withhold emotions. Mahalik et al. (2007) show that norms of masculinity are predictive of health risk behavior such as delay in seeking help, substance abuse, and lack of preventive health care. A meta-analysis of the links between masculine norm conformity and mental health outcomes by Wong et al. (2017) also revealed that the traditional masculinity ideologies predict depression, anxiety, and psychological distress. Levant and Wimer (2014) explain the constructs of masculinity as having a protective buffer and a risk factor to health among men. This is because some masculine values such as goal-orientation and self-efficacy can facilitate wellbeing and other ones such as emotional restriction and help-seeking avoidance pose dangers. Such a subtle approach does not ignore the masculinity as simplistically harmful in its models.
Nonetheless, there is little research on masculinity that considers physiological results. Behavioral health practices as mediators between masculine norms and health outcomes are the subject of most of the studies, including diet quality, exercise frequency, and health care utilization (Mahalik et al., 2007). Direct study of emotion regulation as a biological risk process whereby masculine socialization is physiologically instantiated is a vital gap that is filled in this study.
It is important to note that masculinity norms in the United States have not been static. Historical scholarship has documented the evolution of American ideals of manhood across the 20th century, including shifts from producer-based to consumer-based masculine identities and the gradual emergence of more emotionally expressive masculine scripts among certain generational and demographic groups (Horlacher & Floyd, 2017; Kimmel, 2006). More recently, particularly in the aftermath of the COVID-19 pandemic and the broader mainstreaming of mental health discourse, there has been a notable shift in public conversation around men’s emotional wellbeing in the United States. Anti-stigma campaigns, celebrity disclosures, and social media movements have helped reduce some of the social penalties associated with men seeking help or expressing vulnerability, at least among younger cohorts (Gough & Robertson, 2010; Mokhwelepa & Sumbane, 2025). Against this backdrop, the finding that emotional suppression remains a robust predictor of physiological dysregulation across all age groups—including younger men who have been socialized in an ostensibly more emotionally permissive cultural climate—is particularly striking. It suggests that even as cultural discourses around masculinity shift, the habitual practice of suppression may persist as a deeply ingrained behavioral pattern with measurable biological consequences.
The intersectional viewpoints are focused on the fact that the norms of masculinity are different in racial and socioeconomic contexts. Griffith et al. (2012) address the concept of measuring masculinity in men of color by pointing out that unique cultural definitions and manifestation of manhood need culturally based measurement. The principles of suppression as a biological risk should consider the different masculine ideologies and varying levels of stress exposures under different social standpoints.
Life Course Epidemiology
Life course epidemiology offers the much-needed models that can be used to comprehend how effects of emotional suppression can be cumulative throughout adulthood. According to Kuh et al. (2003), life course epidemiology is the research of the long-term impacts of exposures to physical and social factors in gestation, childhood, adolescence, and adult life on risk of diseases. This view focuses on how health is experienced at any stage of life and is not only an indicator of the current situation but also the sum total of exposures throughout the life span.
There are a number of life course models that can be used to explain suppression effects. Cumulative disadvantage models are based on the assumption that exposures add up with the added exposure increasing the exposure to the health risk (Ferraro & Shippee, 2009). Critical period models define periods of development when exposures have exceptionally powerful effects. Pathway models follow the lineage of risk with initial exposures determining future courses of action and results. In the case of emotional suppression, the best models seem to be the cumulative disadvantage models, which are habitual suppression being implemented daily over a number of decades, having the potential to dysregulate stress biology as it becomes repetitive and apparent in physiological activation.
Halfon and Hochstein (2002) express the frameworks of life course development of health with the focus on the dynamic association between the time, human development, and context. They mention that there is plasticity in biological systems that allows them to adapt to changing conditions yet are susceptible to progressive damage. This idea is formalized in allostatic load theory, which holds that physiological wear and tear result from chronic activation of stress systems (McEwen & Gianaros, 2010; McEwen & Stellar, 1993). Cooper et al. (2010) demonstrate that objectively measured physical capability predicts mortality, with capabilities declining progressively across later adulthood. If emotional suppression accelerates physiological aging through chronic stress system dysregulation, similar age-graded patterns should emerge—suppression effects intensifying in older cohorts with longer cumulative exposure durations.
Methods
Study Design and Participants
We recruited 412 men aged 25–70 years (M = 46.8, SD = 12.4) through community health centers, workplace wellness programs, and online advertisements in a large metropolitan area (population >2 million) in the southeastern United States. Recruitment occurred from January 2021 through March 2023. Inclusion criteria required: (1) self-identified male gender, (2) age 25–70 years, (3) English fluency, (4) no current use of corticosteroid medications affecting cortisol assessment, (5) no diagnosed inflammatory conditions (rheumatoid arthritis, inflammatory bowel disease), and (6) no self-reported history of cardiovascular events (myocardial infarction, stroke).
Sample size was determined via power analysis for structural equation models (SEMs) testing mediation effects. Assuming small-to-medium effect sizes (β = 0.15–0.25) for suppression→biomarker pathways, α = 0.05, power = 0.80, and accounting for 15% missing data on biomarker variables, simulations indicated N = 380 as the minimum sample size. We overrecruited to N = 412 to ensure adequate power after data cleaning.
To conduct life course analyses, we stratified recruitment in three age groups: young adulthood (25–39 years, n = 142), middle age (40–54 years, n = 148), and older adulthood (55–70 years, n = 122). In each cohort, we selected broad samples in terms of race/ethnicity (48: White, 31: Black, 15: Latino, 6: other), education (28: high school or less, 41: some college/associate degree, 31: bachelor’s degree or higher), and income (between $18,000 and 180,000 annually). Full demographic characteristics could be found in Table 1.
Demographic Characteristics and Descriptive Statistics (N = 412).
Note. Descriptive statistics presented by age cohort and total sample. Continuous variables tested using ANOVA (F-statistic); categorical variables tested using chi-square. Cortisol slope values are absolute values where higher (less negative) values indicate flatter, dysregulated slopes. hsCRP = high-sensitivity C-reactive protein; IL-6 = interleukin-6; inflammatory burden = z-standardized composite of log-transformed hsCRP and IL-6. PSQI = Pittsburgh Sleep Quality Index (higher scores = worse sleep). Table 1 shows that age cohorts did not differ in suppression, masculine norms, or stressor exposure, enabling clean age-graded analyses, although biomarkers showed expected age differences with older adults exhibiting flatter cortisol slopes and higher inflammation.
Procedures
The participants followed a two-phase protocol. The first phase consisted of a face-to-face baseline visit, during which the participants gave an informed consent and filled out questionnaires measuring emotion regulation and masculinity norms, demographic data and health history, and were trained to collect saliva samples. Blood samples in which the inflammatory markers were measured were taken using venipuncture by trained phlebotomists. Phase 2 included home saliva collection over three successive typical weekdays, whereby the participants were expected to self-sample saliva on four occasions per day (as soon as they have woken up, at 30 minutes after waking up, early afternoon, at bedtime) using the Salivette collection instruments.
In people, compensation was given to them in the form of monetary compensation; they were given 125 dollars to compensate them (baseline visit, 50 dollars to complete the saliva collection protocol). Saliva sampling adherence was also assessed by the use of sample logs filled by participants that captured collection times, which were confirmed by electronic time-stamped containers to ensure a subsample of 30% was random (with 94% adherence to prescribed timing within the span of 20 minutes).
Measures
Emotional Suppression
This was measured with the four-item Suppression subscale of the ERQ (Gross & John, 2003). The respondents rated the statements such as “I control my emotions by not expressing them” and “When I am feeling negative emotions, I make sure not to express them” on seven-point scales (1 = strongly disagree, 7 = strongly agree). Internal consistency was excellent (Cronbach’s α = 0.81; Gross & John, 2003). Scores were computed as the mean of all four items, with higher scores indicating greater habitual emotional suppression (possible range = 1–7).
Salivary Cortisol
Salivary cortisol provides a non-invasive measure of HPA axis activity, and its diurnal rhythm serves as a validated index of stress system regulation. Saliva was collected using Salivette® synthetic swabs and stored at −80°C pending assay. Samples were analyzed in duplicate using high-sensitivity enzyme-linked immunosorbent assay (ELISA; Salimetrics, State College, PA, USA), following standard protocols recommended by the manufacturer, with intra-assay coefficient of variation (CV) below 5% and inter-assay CV below 8%. Cortisol values at each time point were averaged across the three collection days. The diurnal cortisol slope was calculated by regressing log-transformed cortisol values on hours since waking. Higher (less negative) slope values indicate flatter, more dysregulated diurnal rhythms; for interpretive clarity, we report absolute slope values so that higher values reflect greater dysregulation.
Inflammatory Markers
Inflammatory markers were assayed to index systemic inflammatory burden, a key biological pathway linking chronic stress to cardiovascular disease, metabolic syndrome, and other adverse health outcomes. High-sensitivity C-reactive protein (hsCRP) and IL-6 were measured in serum samples using commercial ELISA (R&D Systems, Minneapolis, MN, USA). Values were log-transformed prior to analysis to correct positive skew. A composite index of inflammatory burden was constructed by averaging the z-standardized log-transformed values of hsCRP and IL-6, which were moderately correlated (r = .52, p < .001). Higher composite scores indicate a greater inflammatory burden; the standardized index has a mean of 0 and standard deviation of 1 in the present sample, with an observed range of −1.8 to +3.4.
Stressor Exposure
Measured using a Life Events Scale adapted from Dohrenwend et al. (1978), which assessed 18 significant life events occurring in the past 12 months (e.g., job loss, divorce/separation, death of a close family member, major illness, financial hardship). Respondents indicated the occurrence of each event; the number of endorsed events was summed to yield a total stressor count (range = 0–18). In addition, respondents rated the subjective severity of each endorsed event (1 = mildly stressful, 5 = extremely stressful), and these ratings were averaged across endorsed events to index perceived stress intensity.
Masculine Norm Conformity
Masculine norm conformity was assessed using the Conformity to Masculine Norms Inventory-46 (CMNI-46; Mahalik et al., 2007), a validated abbreviated measure of adherence to traditional masculine ideology across multiple domains. We selected the Emotional Control subscale (five items; Cronbach’s α = 0.76) as the subscale most directly relevant to emotional suppression. Sample items include “I never share my feelings” and “I tend to keep my feelings to myself.” Items are rated on a four-point Likert-type scale (0 = strongly disagree, 3 = strongly agree). The subscale score was computed as the mean across all five items, with higher scores reflecting greater conformity to restrictive emotional control norms.
Health Covariates
Health covariates were assessed via a combination of objective measurement and validated self-report. Body mass index (BMI, kg/m2) was calculated from measured height and weight. Health behaviors were assessed using validated self-report items (O’Connor et al., 2009) including cigarette smoking status (never/former/current), alcohol use (drinks per week), and physical activity (minutes per week of moderate-to-vigorous intensity activity). Medication use (antihypertensives, antidepressants, statins) was coded as a dichotomous variable. Chronic health conditions (hypertension, diabetes, depression) were assessed by self-report and cross-checked against medication lists. Sleep quality was assessed using the Pittsburgh Sleep Quality Index (PSQI; Buysse et al., 1989), a validated 19-item self-report instrument; higher scores indicate poorer sleep quality.
Analytical Strategy
Missing Data
Any rate of missing data was low (4.2% questionnaires, 8.7% biomarkers majorly as a result of insufficient volume of saliva or blood draw challenges). We used full information maximum likelihood (FIML) in SEMs with all available data and no implication. The results of missingness analysis showed no systematic trends based on the study variables (Missing Completely At Random [MCAR] test by Little 2 34.2 = 2.41, p = .41).
Descriptive and Preliminary Analyses
We calculated descriptive statistics and zero-order correlations between all the variables of the study. We tested distributions of variables of interest of biomarkers, to which we used the logarithmic transformations to correct the positive skew. To compare demographic variation between cohorts based on age, we used the analysis of variance (ANOVA) and chi-square tests to test differences between groups.
SEM
Major analyses were done with SEM in Mplus 8.7 (Muthén & Muthén, 2017) with robust maximum likelihood estimation. SEM allows testing of many regression pathways at a time and correcting measurement error and giving a global model fit measure. Model fit is measured with several metrics: 2/df ratio less than 3, comparative fit index (CFI) more than .90, Tucker–Lewis index (TLI) more than .90, root mean square error of approximation (RMSEA) less than .08, and standardized root mean square residual (SRMR) less than .08 (Kline, 2023).
We have approximated three major models including (1) direct effects model that tests suppression-to-biomarker relationships adjusted by covariates, (2) interaction model that tests suppression-to-stressor interaction predicting biomarkers, and (3) mediation model that tests whether suppression is a mediator of masculine norm-to-biomarker relationships.
To do mediation analyses, we used the bootstrapping methods with 10,000 resamples to obtain the bias-corrected 95% confidence intervals (CIs) of indirect effects (Preacher & Hayes, 2008). Those effects that were deemed to be significant in this case where 95% CI did not include zero. We calculated the proportion mediated (indirect effect)/(total effect) according to MacKinnon et al. (2004).
Age-Graded Analyses
We hypothesized the age-moderating effect of suppression by performing a multigroup SEM to compare model parameters across different age groups. Our initial estimates were unconstrained models in which all parameters were free to vary between groups, followed by constrained models in which suppression = biomarker groupwise equal. It was found that significant χ2 difference tests showed moderation, and then the parameter estimates in each group were examined.
Sensitivity Analyses
We performed several robustness tests: (1) recreating the models with hsCRP and IL-6 as independent variables, not a composite index of inflammation, (2) testing the effects of cortisol awakening response and area under the curve as other HPA axis indices, (3) testing the quadratic effects of age, (4) testing the interaction of suppression and health behaviors, and (5) eliminating the participants diagnosed with depression because they had altered HPA axis in mood disorders.
Results
Descriptive Statistics and Preliminary Analyses
Table 1 provides demographic and descriptive statistics of all the variables of the study. Mean scores of the participants were moderate in suppression (M = 4.2, SD = 1.3 on a seven-point scale), and there was a significant range (1.0–7.0) between best and worst scores. The average slope of cortisol was −0.52 µg/dl/hour (SD = 0.18), which is in the healthy range of a declining diurnal cortisol but with a significant range of deviations. The composite of inflammatory burden had a design mean of M = 0.0, SD = 1.0 due to z-standardization, and the range of −1.8 to +3.4 observed indicated that there would be a great deal of variation in the inflammatory status of individuals.
The average exposure to stressors in the last year was 2.8 life events (SD = 2.1), and 89% of the participants said they had at least one major stressor. Masculine norm conformity on the Emotional Control subscale had a mean of M = 1.8, SD = 0.7 on 0–3 scale, showing a moderate level of support of restrictive emotional norms.
The expected trends were found in the correlations of the first order. Habitual suppression was found to have a positive relationship with flatter cortisol slopes (r = .31, p < .001), higher inflammatory burden (r = .24, p < .001), increased stressor exposure (r = .19, p < .001), and higher masculine emotional control norms (r = .41, p < .001). The flatter slopes (r = .22, p < .001) and increased inflammation (r = .28, p < .001) were related to age, which is physiological aging. The exposure to stressors predicted slopes of flatter (r = .26, p < .001) and high inflammation (r = .21, p < .001).
There were no significant differences in suppression levels (F = 1.8, p = .17) or stressor exposure (F = 2.1, p = .13) or masculine norm endorsement (F = 0.9, p = .41) between the age cohorts, making it possible to perform clean age-graded analyses without confounding by differences in exposure. Nevertheless, cohorts varied in biomarkers, as anticipated old people exhibited steeper slopes and inflammation than the young people.
Direct Effects: Suppression Predicting Physiological Dysregulation
Our first SEM model tested direct effects of habitual suppression on cortisol slopes and inflammatory burden, controlling for age, race/ethnicity, education, income, BMI, smoking, alcohol use, physical activity, sleep quality, medications, and chronic conditions (see Table 2). Model fit was acceptable (χ2/df = 2.31, CFI = 0.93, RMSEA = 0.06, SRMR = 0.05).
Structural Equation Model Results—Suppression Predicting Physiological Dysregulation.
Note. N = 412. Standardized coefficients (β) from a structural equation model with robust maximum likelihood estimation. Cortisol slope is coded so higher values = flatter (more dysregulated) slopes. Inflammatory burden = z-standardized composite of log-transformed hsCRP and IL-6. Model simultaneously predicts both outcomes with correlated errors (r = .34, p < .001). All continuous predictors are grand-mean centered. Table 2 demonstrates that emotional suppression significantly predicted both flatter cortisol slopes and elevated inflammatory burden after controlling for demographics, health behaviors, and medical factors, with effect sizes comparable to established risk factors.
p < .001, **p < .01, *p < .05.
Suppression significantly predicted flatter cortisol slopes (β = 0.34, SE = 0.05, p < .001), indicating that each standard deviation increase in habitual suppression is associated with 0.34 SD flatter diurnal decline. This effect remained robust after controlling for all covariates, including health behaviors that might otherwise explain suppression-health associations. To contextualize effect magnitude, participants in the highest suppression quartile showed cortisol slopes averaging 0.38 μg/dl/hour flatter than those in the lowest quartile—a difference comparable to adding 15 years of age.
Suppression also significantly predicted elevated inflammatory burden (β = 0.28, SE = 0.05, p < .001). Each SD increase in suppression is associated with 0.28 SD higher inflammation. When examining hsCRP and IL-6 separately, effects remained significant for both (hsCRP: β = 0.24, p = .001; IL-6: β = 0.26, p < .001), indicating broad inflammatory activation rather than marker-specific effects.
These findings support H1, demonstrating that habitual emotional suppression predicts physiological dysregulation across stress-responsive biological systems even after accounting for numerous potential confounds. The magnitude of suppression effects proved comparable to or exceeding the effects of established risk factors like smoking and physical inactivity.
Stress Amplification: Suppression × Stressor Interactions
To test H2 regarding suppression as a stress amplifier, we estimated interaction models including suppression × stressor exposure product terms predicting biomarkers (covariates included as in main effects models). For cortisol slopes, the suppression × stressor count interaction proved significant (β = 0.19, SE = 0.06, p = .003), indicating that suppression intensified stress effects beyond additive contributions.
The interaction pattern was shown by the analysis of simple slopes (see Figure 1). In low habituation (−1 SD) participants, cortisol slopes had significant correlations with stressor exposure (0.12, p = .08). The effects of stressors were substantially increased among the high suppression participants (+1 SD), who had a 0.43 with p = .001. This triple difference shows that suppression is the amplifier of stress—exactly the same exposure to stressors produces vastly different physiological effects when it comes to emotion regulation habits.
Suppression as stress amplifier—interaction effects on cortisol dysregulation.
In the case of inflammatory burden, the interaction between the suppression and stressor exhibited the same patterns (=0.16, SE = 0.06, p = .009). The inflammatory levels of high suppressors who were exposed to various stressors were found to be higher by 0.71 SD above the low suppressors who were also exposed to the same stressors, and this was in contrast to the 0.19 SD difference between high stressor exposure and low stressor exposure.
These results are a solid argument in favor of H2, which proves that suppression does not simply contribute to the additional risk but truly changes stressor-to-physiological effect mechanisms. This amplification process can be used to understand why some people respond with resilient physiological reactions to stress and others experience characterized dysregulation—emotion regulation strategies moderate stress embodiment.
Effects by Age: Life Course Accumulation
Multigroup SEM testing age cohort moderation of suppression effects showed some significant differences (two-difference test between unconstrained and constrained models: 2,872.4, p = .001). Analysis of the cohort-specific parameters supported H3 in relation to the strength of effects that increase with age.
Suppression effects were greater in older adults compared to young adulthood: old age (55–70 years) 3 = 0.42, SE = 0.08, p = .001; young adulthood (25–39 years) 3 = 0.18, SE = 0.08, p = .04. This progressive escalation shows progressive physiological stress load—one more decade of lifelong habitual stress leads to a successively greater degree of HPA axis dysregulation.
In the case of inflammatory burden, the patterns closely resembled those of age-grade: young adulthood β = 0.19, p = .03; middle age β = 0.27, p < .001; older adulthood β = 0.34, p < .001. The fact that the effects of suppression were found even among young adults implies that physiological effects are early manifested, and that the cumulative effects are manifested over decades.
Importantly, the age-graded patterns were still there when the cumulative exposure to stressors and age-related changes in health status were held constant, which suggested that the effects of suppression were moderated by chronological age itself (a proxy of exposure duration) and not just the influence of age-correlated factors.
Mediation: Masculine Norms to Physiology Suppression
To test H4, we approximated mediation frameworks exploring the existence of relationships in suppression between the norms of masculine emotional control and physiological results. The overall impact of masculine norms on inflammatory burden was substantial (= .31, SE = 0.05, p < .001). With the introduction of suppression as an inhibitory variable, the indirect effect mediated by suppression was substantial (0.13, 0.03, 0.08, 0.19), and it explained 43% of the overall effect. The other direct effect of masculine norms on inflammation lessened; nevertheless, it was significant (0.18, SE = 0.05, p < .001) suggesting mediation.
In the case of cortisol slopes, there was a similarity in mediation patterns. Sum total of masculine norms: β = 0.29, p < .001. Indirect effect by suppression: 8. = 0.11, 95% CI = [0.06, 0.17], which explains 38% of the total effect. The direct effect was still important (= 0.18, p < .001).
These results confirm H4 the fact that emotional suppression is a considerable channel according to which masculine socialization is transformed into a physiological one. Nevertheless, there are still significant direct effects, which indicate that other unknown mechanisms (e.g., masculine norms of health behaviors, health care avoidance, workplace exposures) are also involved in physiology–masculinity relationships.
Sensitivity Analyses
Findings were strong in a number of sensitivity analyses. In analyzing the cortisol awakening response instead of diurnal slope, the suppression was the predictor of blunted awakening response (−0.24, p = .002) as expected in the dysregulation of the HPA axis but through different time patterns. Suppression predicted both high overall cortisol exposure (0.19, p = .006) and disrupted circadian rhythm (0.28, p < .001) in the analysis of the area under the curve.
Suppression effects were slightly reduced but remained significant (cortisol slopes: 0.29, p = .001; inflammation: 0.24, p = .001) when the respondents with diagnosed depression were excluded (n = 47), which proved that the results can be applied to clinical groups. The interaction of suppression and health behavior showed no effect, meaning that suppression influences are independent of lifestyle.
Comparing the results of the inflammatory markers individually, hsCRP demonstrated an even greater suppression relations (0.28, p = .001) compared to IL-6 (0.23, p = .001), but both of them were significant. This trend indicates widespread inflammatory stimulation and not pathway-specific effects.
Discussion
Suppression as Biological Risk Process
This article offers research finding that habitual emotional suppression is a process of biological risk where masculine gender socialization is physically instantiated. We have shown that dysregulated stress biology with flattened diurnal cortisol slopes and high levels of systemic inflammation, which are known to predict morbidity and mortality, is manifested in men with a habit of suppressing emotions (Adam et al., 2017; Ridker, 2016). These effects remained even after adjusting for many possible confounds such as health behaviors, chronic health conditions, and socioeconomic factors, meaning that there are direct physiological effects of suppression, not just a correlation with other health risks.
The effect sizes of suppression were found to be large and clinically significant. There was cortisol dysregulation in participants in the upper-most suppression covariance as large as 15 years of chronological age and an inflammatory burden as large as that of obesity or smoking. Such effect sizes dispute the view that emotion regulation is a soft psychological factor, which has little physiological impact. Rather, our findings place suppression and predetermined behavior risk factors (smoking, physical inactivity, poor diet) as a by-product of health determinant.
We build upon prior laboratory studies that have shown the existence of acute physiological costs of instructed suppression (Tyra et al., 2024) by proving that habitual suppression in everyday life produces chronic dysregulation of biomarkers. This difference is critical; although anybody may occasionally suppress emotions with short-term physiological activation, people who suppress regularly in a variety of circumstances and over long periods of time have a physiological load that is translated into disease risk. This cumulative interpretation is supported by the life-course perspective that we have incorporated into our age-graded analyses.
Mechanisms of Stress Amplification
Perhaps the most theoretically significant finding of ours is the suppression that works as a stress amplifier and not as an additive risk factor. The large suppression-by-stressor interactions indicate that the same exposure to stressor is accompanied by very different physiological effects under different habits of emotion regulation. Cortisol dysregulation was demonstrated by high suppressors who were expressed to various life stressors threefold in comparison to low suppressors who were as well exposed to the same stressors.
This process of amplification assists in the elimination of the long-standing mysteries in the study of the stress–health relationship. There are also significant personal differences in physiological responses to apparently comparable exposures to stressors—some people demonstrate resilient adaptation, whereas others develop severe dysregulation (McEwen & Gianaros, 2010). Emotion regulation strategies are a highly important, although widely overlooked, source of this heterogeneity according to our results. Suppression can inhibit the adaptive processing of emotions that would otherwise reduce physiological recovery and extend the activation of the stress system and not allow a restoration to a homeostatic baseline.
Theoretical allostatic load models stress that allostatic chronic stress experiments cause physiological wear and tear due to recurrent stress without sufficient rest (McEwen & Stellar, 1993). Suppression can disrupt recovery since it causes emotional arousal within an individual despite the termination of behavioral expression. This results in dissociation between current arousal and external alleviation, and stress systems remain active longer than they need. This pattern can eventually lead to dysregulation of the sensitivity of stress systems, with the resultant flattening of cortisol slope and long-term inflammatory stimulation that we had found.
The findings of the inflammation have been shown to be especially consequential due to the proven role of inflammation in cardiovascular disease, diabetes, dementia, and other causes of mortality that are most common (Ridker, 2016; Slavich & Irwin, 2014). In the event that suppression facilitates prolonged inflammatory reacting, this furnishes a realistic biologic route between the masculine emotional norms and the high mortality rates in men. Future studies need to be done on whether suppression predicts prospective clinical cardiovascular events and mortality.
Life Course Accumulation and Age-Graded Effects
Our age-based studies of increasing effects of suppressions through adulthood would give important data on cumulative physiological load, which is in line with life course epidemiological models (Ferraro & Shippee, 2009; Kuh et al., 2003). At young adulthood, there was a moderate yet significant effect of suppression on cortisol. The doubling of effects occurred over middle age. Effects in old adulthood almost tripled the effects of young adults. This gradual escalation shows that the effects of suppression multiply over the decades instead of staying the same.
This trend fits allostatic load theory that postulates that cumulative wear-and-tear of regulatory systems occurs as a result of repeated exposures to chronic stress (McEwen & Gianaros, 2010). Suppression in each case may lead to small amounts of immediate dysregulation, but thousands of suppression events over decades gradually burn out stress system regulatory capacity. This adds up to fatal compensatory processes, leading to the pronounced dysregulation that we have found in older adults.
The evidence that the effects of suppression were observed even at the young adult age implies that some physiological effects must start early before the individual develops clinical disease. This suggests intercession potentials—aiding the young men to come up with healthier response methods in managing their emotion may be a way of saving years of accrued physiological weight. But the effects that increase with age also indicate that even late-life psychotherapeutic interventions that encourage emotional expression and processing may benefit physiological health but may need more intense intervention to change entrenched patterns.
Such age cohort results contrast our study with cross-sectional masculinity studies incapable of distinguishing the effects of cohort groups and the effects of development. Explicit in recruiting through the broad age range and moderating age test gives us evidence of age-related intensification as cumulative exposure and not merely because of differences in cohort socialization to masculinity. The next-generation studies ought to utilize the designs with the true longitudinal characteristics, which follow the individual over decades, to clearly determine the within-person patterns.
Physiological Embodiment and Masculinity
Our mediation results show that emotional suppression is an important route in which masculine norms are physiologically encoded. Habitual suppression practices predicted physiological dysregulation with significant parts of the total effects, 38%–43% by masculine emotional control ideology. This offers empirical evidence to theoretical arguments according to which gender socialization gets under the skin to influence health and does so via behavioral and biological mechanisms (Mahalik et al., 2007).
Nevertheless, there were significant direct influences of masculine norms on physiology even when the effect of suppression was considered, showing that there are other unmeasured mechanisms. The possible causes are most likely masculine impacts on health practices (diet, exercise, substance use), health care refusal and late help-seeking (Mokhwelepa & Sumbane, 2025), work exposures in the male-dominated hazardous sectors, and social isolation due to masculine norms of self-reliance. Comprehensive mediation models that involve all of these multiple pathways should be tested in the future.
Our results have implications for masculinity scholarship. They propose an exit out of taking masculine norms as mere beliefs about culture to conceptualizing them as practices with physiological implications that can be measured. The embodiment view acknowledges that recurring behavioral reenactments of gender, such as the silencing of emotions to execute masculinity, physically transform biological systems that govern health and illness. This makes masculinity not an abstract ideology but a practice that is lived, which constitutes physiology.
Notably, our findings do not mean that masculinity is universally bad or that males should renounce masculinity factors. Instead, they opine that certain masculinity ideals with a focus on emotional restraint pose a health threat, and that there are other masculine ideals that can support wellbeing (Levant & Wimer, 2014). Interventions could also operate in terms of masculinity and rebrand emotional sensitivity and expression as masculine assets of being strategic enough to pursue goals or being responsible enough to have dependents and have to stay healthy. These would help to suppress less without making men denounce masculine identities.
Clinical and Public Health Implications
We believe that there are several opportunities to be found through intervention. To begin with, the process of mental health care intervention on men must be explicit regarding emotion regulation, as opposed to symptom reduction. A variety of approaches to emotion regulation such as cognitive-behavioral therapy, acceptance and commitment therapy, and mindfulness-based interventions are usually aimed at men but not explicitly. Masculine-friendly language of these interventions (strategic emotional management, performance optimization) could make more males participative (Addis & Mahalik, 2003; Vogel et al., 2011).
Second, emotion regulation training should be included in the stress management programs instead of stressor reduction or relaxation training. We have found reasons to believe that altering the response of men to stressors (lessening suppression, enhancing adaptive processing) could be as significant as the exposure to stressors themselves under amplification processes.
Third, traditionally, behavioral risk factors could be screened at primary care by physicians as well as habitual suppression. The ERQ suppression subscale is a simple brief measure that could determine men at high physiological risk. The providers would be able to normalize emotional expression and send high suppression men to the relevant interventions before the physiological consequences accrue.
Fourth, men’s health campaigns on public health should deal directly with emotional wellbeing and not physical health behaviors only. Our evidence of significant physiological effects is ignored by the current campaigns of diet, exercise, and cancer screening without consideration of the emotion regulation. Messaging may reposition emotional awareness by keeping the engine (body) by keeping track of the dashboard (emotions).
Finally, working wellness programs are promising intervention environments, considering that men are exposed to occupational environments. Emotion regulation skills training could be part of stress management programming with the impact probably compounded by the delivery through the group modalities, allowing peer support and masculine identity reconstruction processes around the emotional openness.
Limitations and Future Directions
There are a number of constraints that should be mentioned. First, we have a cross-sectional design that does not allow us to have conclusive causality, regardless of the statistical controls and theoretically underpinned hypotheses. It is also possible that it is a reverse causation process, as physiological dysregulation may provoke greater suppressive action as men strive to cope with unpleasant body experiences. Yet, this possibility appears to have lower plausibility than our suggested course because of timing (suppression habits are probably developed earlier than biomarker dysregulation) and theoretical basis (emotion regulation before physiological consequences) and theory. A longitudinal design that measures suppression and biomarkers at several time points would be able to establish the temporal precedence unquestionably and measure reciprocal effects.
Second, we selected men who were willing to engage in a research with emotion evaluation and data collection of biological data, which might have chosen more convenient men to talk about emotions. The men who were the greatest proponents of restrictive masculinity may well have refused to take part, generating restriction of range that would underestimate actual suppression effects. It is not necessary that future research adopt different recruitment methods or embedded designs in existing cohort studies to reach the harder-to-reach populations.
Third, in our biomarker evaluation, although rigorous, two systems of stress were captured (HPA axis, inflammation). Additional pertinent pathways involve sympathetic nervous system functioning (measured through heart rate variations, blood pressure), immune functions (cellular immunity, antibody response), and metabolic imbalance (insulin resistance, lipid profiles). Further examination of various biological systems would be a more thorough evaluation of suppression-embodied outcomes.
Fourth, the evaluation of emotion regulation was based entirely on self-report surveys. Even though the ERQ presents good psychometric properties and validity, self-reporting suppression does not necessarily reflect the actual regulation behavior. Other additional methods such as daily diary analyses of momentary regulation, coding of instances of suppressions when exposed to laboratory stressors, or physiological evidence of decoupling in emotions and behavior would enhance the validity of measurement.
Fifth, our sample was focused in a single U.S. metropolitan region in the south-east area, so it was not geographically generalized. Masculine values and stress exposures are different in geographical areas, cities, and rural areas and cultural communities. Multisite designs that include a wide range of geographic and cultural settings would help to clear up generalizability and establish whether cultural differences in masculinity modulate suppression–physiology correlations.
Sixth, we targeted men only and avoided the gender comparisons. Although the theoretical reason provided to support the use of men-only sampling (masculine norms value suppression, especially among men) was that comparative studies looking at the relationship between suppression and physiology between genders would show the relationship between gender-specific and general mechanisms. Since women also repress emotions in certain situations, research into gender variations in situations where, why, and with what effects suppression happens would serve to further the cause.
Further studies must eliminate these limitations by: (1) prospective longitudinal studies to measure suppression and biomarkers across adulthood, better initiated in adolescence when masculine socialization is more intense, (2) daily diary studies to evaluate momentary suppression and physiological responses in natural settings, (3) intervention studies to measure whether lessening suppression is better than physiological health, (4) studying neural mechanisms of suppression–physiology relationships, and (5) intersectional studies that determine how suppression–physiology interactions differ among race, class, and.
Conclusion
This article shows how habitual emotional suppression is a biological risk process whereby masculine gender socialization gets embodied physiologically. Men with a history of habitually suppressed emotions have dysregulated cortisol oscillations and increased systemic inflammation, which are established biomarkers of morbidity and mortality. Suppressive functioning does not just act as an additive risk but as a stress amplifier, which further enhances physiological effects of stressor exposure beyond mere primary effects. Moreover, suppression effects strengthen progressively across the life course, consistent with a cumulative biological loading model in which decades of habitual suppression gradually erode the regulatory capacity of stress-responsive systems.
The results of our findings are empirical data to support theoretical arguments that gender penetrates the skin to influence health biologically. Emotional suppression is one of the substantial processes that connect the masculine emotional control norms to physiological dysregulation and mediate 38%–43% of masculinity–physiology correlations. This puts emotion regulation no longer at the fringe of psychological determinants but in the center through which social experience forms biological health.
The intensity of suppression effects, which is equal and comparable to such known behavioral risk factors as smoking and lack of physical activity, sets the public health and clinical practice in such a way that it is important to pay attention to emotion regulation specifically, not only to diet, exercises, and medical examination. Even basic interventions that assist men in utilizing more healthy emotion control strategies could stop decades of accrued physiological dues, especially when applied during the early adulthood while dysregulation has not yet been adopted as a habit.
As masculinity research advances, our study demonstrates the value of moving beyond correlational survey designs toward biologically informed investigations examining the physiological embodiment of gender socialization. By integrating validated psychological assessment with rigorous biomarker measurement and sophisticated SEM, we reveal mechanisms linking masculinity to health that have eluded prior behavioral research. Future work should build on this foundation through prospective longitudinal designs, experimental intervention studies, and comprehensive multisystem biological assessment to fully map pathways through which gender becomes biology and biology becomes disease.
Footnotes
Author Contributions
Yucui Pu and Zeming Kong contributed to methodology, investigation, conceptualization, formal analysis, data curation, writing—original draft, and writing—review and editing.
Funding
The authors disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This work was supported by the Science and Technology Research Program of Chongqing Municipal Education Commission (Grant No. KJZD-K202502903) and the 2024 Scientific Research Platform of Chongqing Preschool Education College (Project Name: Digital Elderly Care Service Big Data Application Research).
Declaration of Conflicting Interests
The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Data Availability Statement
The datasets used and/or analyzed during the current study are available from the corresponding author on reasonable request.
Declarations: Use of AI
The author used artificial intelligence (AI) tools to assist with language editing and improving the readability of this manuscript. The scientific content, data, analyses, interpretations, and conclusions are entirely the work of the author. The author takes full responsibility for the integrity and accuracy of the submitted work.