The impact of PTSD on associations between sex hormones and cardiovascular disease symptoms

ABSTRACT

Background: Women have twice the lifetime prevalence of posttraumatic stress disorder (PTSD) relative to men, and PTSD is a known risk factor for cardiovascular disease (CVD). Two sex hormones – estradiol and progesterone – have been found to impact both PTSD and CVD symptomatology, but the way in which sex hormones influence cardiovascular physiology among individuals with PTSD is not well understood.

Objective: This study sought to clarify the association between sex hormones, PTSD, and CVD among trauma-exposed women.

Method: Sixty-six trauma-exposed women (M age = 31.45, SD = 8.92) completed a clinical interview for PTSD and self-reported CVD symptoms; estradiol and progesterone were assayed from blood samples. The association between each sex hormone and CVD symptoms was analyzed, controlling for age, systolic blood pressure (BP), and diastolic BP.

Results: Neither estradiol nor the PTSD-by-estradiol interaction was significantly associated with CVD symptoms. Higher progesterone and, relatedly, progesterone-to-estradiol ratio (PE ratio) were each significantly associated with greater CVD symptom severity, but only for individuals with lower relative PTSD severity.

Conclusions: The findings indicate that PTSD moderates the relationship between progesterone and CVD symptoms, and further research is warranted to reconcile findings in existing literature regarding the direction of and mechanisms behind this relationship.

HIGHLIGHTS

• Posttraumatic stress disorder (PTSD) is a risk factor for cardiovascular disease (CVD) and sex hormones have been implicated in their link.

• The current study examined associations between sex hormones, PTSD, and CVD symptoms among 66 trauma-exposed women.

• Estradiol was not significantly associated with CVD symptoms, but higher progesterone was significantly associated with greater CVD symptom severity, but only for individuals with lower relative PTSD severity.

Antecedentes: Las mujeres tienen una prevalencia del doble de trastorno de estrés postraumático (TEPT) a lo largo de la vida en comparación con los hombres, y el TEPT es un factor de riesgo conocido para la enfermedad cardiovascular (ECV). Se ha encontrado que dos hormonas sexuales, el estradiol y la progesterona, influyen tanto en el TEPT como en la sintomatología de la ECV, pero no se comprende bien la forma en que las hormonas sexuales influyen en la fisiología cardiovascular entre las personas con TEPT.

Objetivos: Este estudio buscó aclarar la asociación entre hormonas sexuales, TEPT y ECV en mujeres expuestas a traumas.

Método: Sesenta y seis mujeres expuestas a traumas (Edad media = 31.45, DE = 8.92) completaron una entrevista clínica para el TEPT y síntomas de ECV autoinformados; se analizaron estradiol y progesterona a partir de muestras de sangre. Se analizó la asociación entre cada hormona sexual y los síntomas de ECV, controlando por edad, presión arterial sistólica (PAS) y presión arterial diastólica (PAD).

Resultados: Ni el estradiol ni la interacción TEPT- estradiol se asociaron significativamente con los síntomas de ECV. Niveles más altos de progesterona y, relacionado con ello, la proporción progesterona-estradiol (PE ratio) se asociaron significativamente con una mayor gravedad de los síntomas de ECV, pero solo para individuos con una gravedad relativa más baja de TEPT.

Conclusiones: Los hallazgos indican que el TEPT modera la relación entre la progesterona y los síntomas de ECV, y se justifica una investigación adicional para reconciliar los hallazgos en la literatura existente sobre la dirección y los mecanismos detrás de esta relación.

KEYWORDS:

• Cardiovascular
• autonomic
• sex
• estrogen
• estradiol
• progesterone
• PTSD

PALABRAS CLAVE:

• Cardiovascular
• autonómico
• sexo
• estrógeno
• estradiol
• progesterona
• TEPT

1. Introduction

Posttraumatic stress disorder (PTSD) is considered a significant risk factor for the development of cardiovascular disease (CVD) (Hargrave et al., 2022; Krantz et al., 2021; Šagud et al., 2017). Evidence suggests that PTSD causes dysregulation of the biological stress response via sustained alterations in the hypothalamic–pituitary–adrenal (HPA) axis and the autonomic nervous system (Dunlop & Wong, 2019; Sherin & Nemeroff, 2011; Yehuda et al., 2015). Dysregulation in these processes may underlie the increased rates of hypertension (Burg & Soufer, 2016; Kibler et al., 2009), type-II diabetes (Levine et al., 2014), atherosclerosis (O’Donnell et al., 2021), venous thromboembolism (Sumner et al., 2015), and acute cardiac events observed in PTSD (Edmondson et al., 2012; Turner et al., 2013). Recent meta-analyses have shown that PTSD is an independent risk factor for CVD development even when controlling for depression and coronary risk factors (e.g. high systolic blood pressure [BP], dyslipidemia), and that PTSD is associated with a 49% increase in risk for myocardial infarction onset and hospitalisations or cardiac-related death (Edmondson et al., 2013; Jacquet-Smailovic et al., 2022). In terms of PTSD symptom clusters, higher re-experiencing and hyperarousal symptoms have been associated with cardiac event recurrence and mortality (Edmondson et al., 2011; Presciutti et al., 2020).

Given that women experience twice the lifetime prevalence of PTSD relative to men (Kessler et al., 1995), there has been an increasing focus on sex effects in PTSD. While various social, cultural, and biological factors may influence the higher rate of PTSD among women, prior literature supports the effect of sex hormones, specifically estradiol and progesterone, on PTSD symptomatology (Garcia et al., 2018; Green & Graham, 2022; Seligowski et al., 2020). During the beginning of a typical 28-day menstrual cycle (the follicular phase), both estradiol and progesterone levels are low. Estradiol levels start to rise in the mid-to-late follicular phase and peak during the early-to-mid luteal phase, where progesterone also rises. In the late luteal phase, both estradiol and progesterone levels lower again. Lower levels of estradiol and progesterone have been associated with higher PTSD symptomology in women with PTSD, suggesting a protective effect of these sex hormones against PTSD development (Garcia et al., 2018; Green & Graham, 2022; Nillni et al., 2021). These studies also found, however, that flashbacks or re-experiencing symptoms were higher among individuals with PTSD during the luteal phase, when levels of estradiol and progesterone are higher. Although the high covariation of estradiol and progesterone levels throughout the menstrual cycle obscures the unique association of each hormone with PTSD symptoms, the discrepancy between these findings may be explained by the association of higher progesterone in the luteal phase with greater stress memory retention in PTSD (Garcia et al., 2018; Green & Graham, 2022; Nillni et al., 2021). Fear conditioning studies (i.e. paradigms that model the fear learning process that occurs in PTSD) have shown a strong relationship between lower estradiol levels and worse fear inhibition, fear extinction, and extinction recall in healthy samples and those with PTSD, indicative of a protective effect of estradiol (Glover et al., 2012, 2013; Milad et al., 2010; Wegerer et al., 2014). In contrast, higher progesterone levels have been associated with better extinction retention among women without PTSD, but worse extinction retention among women with PTSD (Pineles et al., 2016). A follow up study demonstrated that women with PTSD had deficits in the synthesis of progesterone metabolites, suggesting that deficient metabolite processing, and not necessarily low progesterone, may have led to worse extinction retention in this group (Pineles et al., 2018).

Cardioprotective effects of estradiol and progesterone have been studied in general populations; however, little is known about how sex hormones influence cardiovascular physiology among individuals with PTSD. Studies in general populations show that circulating estradiol and progesterone play a role in lowering BP via vasodilation, upregulation of low density lipoprotein (LDL) receptors to help lower cholesterol, promotion of endothelial repair after vascular injury, and the slowing of cardiac hypertrophy (Niță et al., 2021; Palmisano et al., 2017; Regitz-Zagrosek & Kararigas, 2017; Trivedi et al., 2022; Xue et al., 2014). Randomised controlled clinical trials have shown that starting hormone replacement therapy with synthetic estradiol and progestin in early post-menopausal women led to significantly decreased risk of myocardial infarction and heart failure at 10 years follow-up, as well as slowed progression of atherosclerosis (Hodis et al., 2016; Schierbeck et al., 2012). The cardioprotective effects of estradiol and progesterone have been implicated as contributing to the higher CVD risk in men compared to pre-menopausal women (Pérez-López et al., 2010). Despite beneficial effects of estradiol and progesterone in CVD, women with PTSD have a three-fold higher risk of developing coronary heart disease compared to women without PTSD symptoms (Kubzansky et al., 2009). Thus, further research into the role of sex hormones in the development of CVD among individuals with PTSD, and particularly women, is warranted.

Given the varying findings on the impact of estradiol and progesterone on PTSD and CVD, the goal of the present study was to examine associations among PTSD symptoms, sex hormones, and CVD symptoms (e.g. chest pain, shortness of breath) in a sample of trauma-exposed women. Considering that prior research demonstrated different associations between sex hormone levels and sympathetic arousal by PTSD status (Pineles et al., 2016), we also sought to determine whether the associations between sex hormones and CVD symptoms differed by PTSD symptom severity. We hypothesised that (1) PTSD symptom clusters would be significantly positively correlated with CVD symptomology based on findings from existing literature (Edmondson & Cohen, 2013), and (2) estradiol and progesterone would be more strongly associated with CVD symptoms among women with higher total PTSD symptom severity compared to those with lower total PTSD symptom severity. We did not have a-priori hypotheses for the PE ratio given that very limited research has been conducted using this hormone measure in trauma-exposed samples.

2. Methods

Sixty-six women (M age = 31.45, SD = 8.92) were recruited from the community in Boston and the surrounding area between July 2020 and June 2023. All participants endorsed lifetime Criterion A trauma exposure, which was determined using the Life Events Checklist. (LEC-5; Weathers et al. 2018). Participants were not selected on the basis of CVD, and none of the participants endorsed current or lifetime CVD events (stroke, heart attack, or heart failure). In terms of race, 53 participants identified as White (80.3%), four (6.1%) as Asian or South-Asian, and five (7.6%) as Black or African American; four participants (6.1%) indicated that their race was not listed or chose not to respond. Eight participants (12.3%) identified as being of Latino, Hispanic, or Spanish origin. Most participants endorsed having completed four years of college (n = 30, 46.2%), single marital status (n = 48, 73.8%), being employed full-time (n = 37, 56.9%), and having a household income ≥ $50,000 (n = 32, 48.5%).

Participants completed several self-report measures evaluating PTSD symptoms and CVD symptoms, all of which are described below. Participants also completed a five-minute resting state BP assessment, with an average of three measurements per participant. Participants provided a blood sample of 44 ml at the start of the visit, from which estradiol and progesterone were assayed. All procedures were conducted in accordance with the ethical standards of the Institutional Review Board, and the United States Federal Policy for the Protection of Human Subjects. After an explanation of study procedures, written informed consent was obtained from all participants. Participants received $100 as compensation for completion of the study.

2.1. Measures

2.1.1. Life Events Checklist (LEC-5; Weathers et al., 2013)

The LEC-5 is a self-report measure that screens for various types of potentially traumatic events and the level of exposure to those events in a respondent’s lifetime. The measure includes seventeen types of potentially traumatic events, such as natural disaster, physical assault, sexual assault, combat, and life-threatening illness. For each type of potentially traumatic event, the assessment asks what kind of exposure the participant had to such an event, specifically: (1) ‘Happened to me’, (2) ‘Witnessed it’, (3) ‘Learned about it’, (4) ‘Part of my job’, (5) ‘Not sure’, and/or (6) ‘Doesn’t apply’.

2.1.2. Clinician-Administered PTSD Scale for DSM-5 (CAPS-5; Weathers et al., 2018)

The CAPS-5 is a 30-item clinician-administered structured interview used for the diagnosis of PTSD. Twenty items relate to the 20 DSM-5 PTSD symptoms, and the ten additional items relate to factors such as onset and duration of symptoms and subjective distress. Each of the twenty items related to DSM-5 PTSD symptoms is scored on a severity scale of 0–4 (i.e. 0 = absent, 1 = mild/subthreshold, 2 = moderate/threshold, 3 = severe/markedly elevated, and 4 = extreme/incapacitating). A total CAPS-5 symptom severity score is calculated as the sum of the twenty individual item scores. Symptom cluster scores are calculated as the sum of the severity scores within each cluster. The CAPS-5 was used to diagnose PTSD based on DSM-5 criteria, which includes Criterion A trauma exposure, one Re-experiencing and one Avoidance symptom each rated two or higher, two symptoms rated two or higher for Negative Alterations in Cognition and Mood, as well as Alterations in Arousal and Reactivity, duration of symptoms ≥ 30 days, and significant distress and/or functional impairment.

2.1.3. Mini International Neuropsychiatric interview (MINI; Sheehan et al., 1998)

The MINI is a clinician-administered screening measure of psychiatric symptoms and diagnoses. Specific MINI modules included: major depressive episode, panic disorder, social anxiety disorder, and obsessive-compulsive disorder. Four additional modules were administered in order to screen for exclusion criteria: alcohol use disorder, substance use disorder, psychosis, manic episode.

2.1.4. Health Questionnaire

A Health Questionnaire was used as a comprehensive assessment of participant physical health (e.g. perceived health, medical history, menstrual cycle information). The questionnaire included assessment of six CVD symptoms that were adapted from the Women's Ischemia Syndrome Evaluation (WISE) symptom history form: ankle swelling, calf pain, chest pain, shortness of breath, irregular heart rate, and leg pain (Gomez et al., 2020; Krantz et al., 2006; Merz et al., 1999). Participants were asked whether they were currently experiencing any of these symptoms; scores ranged from 0 to 6, with higher scores indicating a greater number of symptoms.

2.2. Physiological data collection and processing

2.2.1. Blood pressure

Omicron Corporation’s IntelliSense monitor was used to collect BP data. The monitor measured systolic and diastolic BP three times on average for each participant, and the average of the three readings was displayed. BP measurements were taken after five minutes of seated rest for each participant.

2.2.2. Blood sample

One blood draw of 44 ml was taken at the start of the visit. Blood samples were collected in EDTA tubes, wet iced immediately, centrifuged, and stored in a −80C freezer. Serum was quantified for estradiol (pg/mL) and progesterone (ng/mL) levels using mass spectrometry. In pre-menopausal naturally cycling women, the expected range of estradiol is 21–649 pg/mL and the expected range of progesterone is <1–19 ng/mL (Stricker et al., 2006).

2.3. Data analysis

Ten participants had estradiol levels outside of the expected range (below 21 pg/mL). Given that removal of these participants did not affect study outcomes, they were included in all analyses. No participants had progesterone levels outside of the expected range. Bivariate correlations were used to test our first hypothesis that higher PTSD symptoms would be associated with worse CVD symptoms. Three regression models were conducted to determine the association between total PTSD symptoms and sex hormones on CVD symptoms, as well as the interactions between PTSD symptoms and (1) estradiol (hypothesis 2), (2) progesterone (hypothesis 2), and (3) the PE ratio (exploratory). Sex hormone data were not normally distributed and thus we did not fulfil regression assumptions of normality or homoscedasticity. However, our regressions did fulfil assumptions of linearity (based on scatter plots) and multicollinearity (based on bivariate correlations). While bivariate correlation analyses included PTSD symptom clusters, regression analyses were limited to total PTSD symptoms in order to limit the number of tests conducted, particularly considering our small sample size. To create interaction terms, PTSD symptoms and sex hormone variables were first mean centred. In Model 1, CVD symptoms were the dependent variable and PTSD symptoms, estradiol, age, trauma exposure (from LEC-5), systolic BP, diastolic BP, and the PTSD-by-estradiol interaction were included as independent variables. In Model 2, CVD symptoms were the dependent variable and PTSD symptoms, progesterone, age, trauma exposure, systolic BP, diastolic BP, and the PTSD-by-progesterone interaction were included as independent variables. In Model 3, CVD symptoms were the dependent variable and PTSD symptoms, the progesterone-to-estradiol ratio (PE ratio), age, trauma exposure, systolic BP, diastolic BP, and the PTSD-by-PE ratio interaction were included as independent variables. The PE ratio was calculated as (progesterone*1000)/estradiol. Simple slopes analyses were conducted only for significant interactions. In these analyses, two regression models are conducted for each potential moderation. For example, a model of the association between estradiol and CVD symptoms at high and low PTSD severity would be calculated by subtracting and adding one standard deviation from the mean CAPS-5 score. Moderation would be supported if the association between estradiol and CVD symptoms was significant at only the high or low level of PTSD severity (the proposed moderator; Aiken & West, 1991). All analyses were conducted in SPPS v.28 with alpha = .05.

3. Results

See Table 1 for descriptive statistics and bivariate correlations. A total of 65 women completed all parts of the CAPS-5 (one participant became ill during the study session). Of those 65 women, 37 (56.9%) met diagnostic criteria for PTSD. In terms of trauma exposure, participants reported sexual assault (n = 48, 72.3%), accidents (n = 47, 71.2%), physical assault (n = 36, 54.5%), natural disasters/fire (n = 15, 22.7%), illness/injury (n = 13, 19.7%), assault with a weapon (n = 9, 13.6%), toxic substance exposure (n = 3, 4.5%), and sudden violent death (n = 2, 3.0%). The average number of traumatic events endorsed on the LEC-5 was 4.11 and the average age of exposure for the CAPS-5 index trauma was 17.89 years (SD = 9.79). As with most trauma-exposed individuals, our sample included participants with co-occurring disorders, including major depressive episode (n = 19, 28.8%), social anxiety disorder (n = 16, 24.2%), panic disorder (n = 7, 10.6%), and obsessive-compulsive disorder (n = 11, 16.7%). There were no significant differences in sex hormones or CVD symptoms when comparing those with and without PTSD (based on CAPS-5 diagnosis), and there were no significant differences in sex hormones, PTSD symptoms, or CVD symptoms among those who were naturally cycling compared to those who were not naturally cycling.

Table 1. Correlations and descriptive statistics for study variables.

	1	2	3	4	5	6	7	8	9	10	11	12
1. Estradiol	–
2. Progesterone	−.006	–
3. Progesterone-to-estradiol ratio	.122	.956**	–
4. PTSD total severity	.070	.073	.180	–
5. PTSD Criterion B severity	.020	.073	.287	.843**	–
6. PTSD Criterion C severity	.046	.155	.143	.651**	.465**	–
7. PTSD Criterion D severity	.165	−.006	.133	.918**	.657**	.540**	–
8. PTSD Criterion E severity	−.064	.065	.057	.845**	.654**	.451**	.667**	–
9. Trauma exposure	−.145	−.183	−.033	.021	.023	−.028	.020	.030	–
10. Systolic blood pressure	−.240	−.004	−.282	.070	.017	.086	.036	.127	.154	–
11. Diastolic blood pressure	−.230	−.168	−.278	.065	.072	.121	.032	.044	.152	.820**	–
12. Cardiovascular symptoms	−.239	.023	.040	.277*	.304*	−.079	.245	.303*	−.319*	−.009	−0.047	–
Mean	94.39	1.35	13.39	25.48	8.82	3.77	13.03	9.80	4.11	119.34	78.89	0.73
SD	99.17	2.67	25.48	13.30	4.92	2.28	7.39	5.34	2.31	15.01	12.73	1.05
Minimum	2.96	0.04	0.35	0.00	0.00	0.00	0.00	0.00	0.00	90.33	53.00	0.00
Maximum	529.00	13.00	130.00	53.00	20.00	8.00	28.00	24.00	12.00	154.00	106.00	4.00

Note: *p < .05; **p < .01.

3.1. Bivariate correlations

There was a significant positive correlation between CVD symptoms and trauma exposure (r = .25, p = .045), PTSD total score (r = .28, p = .027), PTSD Re-experiencing (r = .30, p = .015) and PTSD Hyperarousal (r = .30, p = .015). There were no significant correlations between trauma exposure or PTSD symptoms with sex hormones, nor sex hormones with CVD symptoms. As a sensitivity analysis, we also examined correlations among the subset of participants who were naturally cycling (n = 37) given that much of the prior literature on PTSD and sex hormones has been conducted with naturally cycling individuals. Among the n = 37 participants who were naturally cycling, there were significant positive correlations between PTSD Re-experiencing symptoms and progesterone (r = .37, p = .035), as well as the PE ratio, (r = .47, p = .021). We were not powered to run regression models with this sub-sample given its small size. As another sensitivity analysis, we examined correlations in participants with PTSD (n = 37, 19 of whom were naturally cycling). Among those with PTSD, there were significant negative correlations between estradiol and systolic BP (r = −.35, p = .034) and diastolic BP (r = −.36, p = .028).

3.2. Regression models

There was no significant effect of estradiol on CVD symptoms (β = −.11, t = −.72, p = .476), and the PTSD-by-estradiol interaction was also not significant (β = −.28, t = −1.77, p = .083). A significant association was observed between progesterone level and CVD symptoms (β = .45, t = 3.03, p = .004), such that higher progesterone levels were associated with higher CVD symptom severity. The PTSD-by-progesterone interaction was significant, β = −.52, t = −3.53, p < .001, suggesting that the association between progesterone and CVD symptoms varied by PTSD symptom severity. A significant association was also observed between PE ratio and CVD symptoms (β = .74, t = 3.04, p = .005), such that higher PE ratio levels were associated with higher CVD symptom severity. The PTSD-by- PE ratio interaction was significant as well, β = −.94, t = −4.04, p < .001, suggesting that the association between PE ratios and CVD symptoms varied by PTSD symptom severity.

3.3. Simple slopes analyses

Simple slopes analyses were conducted to determine the nature of the significant interactions. For progesterone, two simple slopes regression models were conducted: (1) at high PTSD severity and (2) at low PTSD severity. These high and low PTSD variables were calculated by subtracting and adding one standard deviation (13.30) from the mean CAPS-5 score (25.48), respectively. Analyses revealed that the association between progesterone levels and CVD symptoms was significant at low PTSD severity, β = 1.08, t = 3.85, p < .001, but not at high PTSD severity, β = −.17, t = −1.03, p = .310 (Figure 1, left panel). For the PE ratio, two simple slopes regression models were conducted in the same manner. As was observed for progesterone, the association between PE ratios and CVD symptoms was significant at low PTSD severity, β = 1.74, t = 3.79, p < .001, but not at high PTSD severity, β = −.27, t = −1.53, p = .138 (Figure 1, right panel) (Table 2).

Figure 1. Simple slopes analyses.

Note: CVD = cardiovascular disease; PE = progesterone-to-estradiol.

Table 2. Linear regression models.

Dependent Variable: CVD Symptoms	β	t	p
Estradiol	−.11	−.72	.476
PTSD severity	.24	1.85	.070
PTSD-by-estradiol interaction	−.28	−1.77	.083
Systolic blood pressure	.16	.69	.495
Diastolic blood pressure	−.25	−1.13	.262
Trauma exposure	.19	1.19	.241
Age	−.11	−.70	.489
R^2= .199
Progesterone	.45	3.03	.004*
PTSD severity	.02	.12	.909
PTSD-by-progesterone interaction	−.52	−3.53	<.001*
Systolic blood pressure	.03	.11	.913
Diastolic blood pressure	.14	.60	.554
Trauma exposure	.10	.62	.537
Age	.10	.60	.551
R^2= .297
Progesterone-estradiol ratio	.74	3.04	.005*
PTSD severity	.03	.16	.877
PTSD-by-progesterone-estradiol ratio interaction	−.94	−4.04	<.001*
Systolic blood pressure	−.18	−.71	.486
Diastolic blood pressure	.26	1.08	.291
Trauma exposure	.20	1.09	.283
Age	.12	.57	.572
R^2= .414

Note: *p < .05; CVD = cardiovascular disease.

4. Discussion

This study sought to clarify the role of sex hormones in self-report CVD symptoms among trauma-exposed women, specifically by (1) examining associations between PTSD symptoms, sex hormones, and CVD symptoms and (2) by assessing whether the associations between sex hormones and CVD symptoms differed by PTSD symptom severity. We did not observe significant findings for estradiol either at the correlation level or based on PTSD symptom severity. For progesterone, higher levels were associated with worse CVD symptoms, particularly among women with low but not high PTSD severity.

When examining associations between PTSD symptoms and CVD symptoms, we found that two PTSD symptom clusters – Re-experiencing and Hyperarousal – were significantly positively correlated with CVD symptoms, but the Avoidance and Negative Cognitions and Mood symptom clusters were not. While literature examining the association between PTSD symptom clusters and cardiovascular health is limited, these results support early findings that the association between PTSD and cardiovascular health differs by symptom cluster. For example, a recent study in trauma-exposed women found that increased risk of hypertension was associated with PTSD symptoms from the Re-experiencing and Avoidance clusters, but not those from the Negative Cognitions and Mood cluster (Sumner et al., 2020). In contrast, higher Negative Cognitions and Mood symptoms have been associated with decreased cardiovascular reactivity in response to a psychosocial stressor in a study of trauma-exposed subjects, which the authors attributed to the dampening effect of depressive symptoms on cardiovascular response (Lee et al., 2022). Furthermore, a recent study found that more Avoidance symptoms in a trauma-exposed community sample were associated with parasympathetic dominance (i.e. lower low- to high- frequency ratio; Lee et al., 2022). While dysregulation of the HPA axis, indicating sympathetic rather than parasympathetic dominance, is among the biological mechanisms hypothesised to increase CVD risk in PTSD, parasympathetic dominance may indicate a sub-segment of individuals with PTSD with lower CVD risk. Alternatively, the findings may indicate a slower-acting mechanism for increased CVD risk, potentially through psychosocial risk factors related to avoidance symptoms, such as increased social phobia or decreased social support (Edmondson & Cohen, 2013). Overall, our results support the limited body of literature indicating that PTSD symptoms clusters are differentially correlated with CVD symptoms.

In assessing the role of PTSD symptom severity in the association between sex hormones and CVD symptoms, we found that neither estradiol nor the PTSD-by-estradiol interaction were significantly associated with CVD symptoms. This contradicted our hypothesis, as well as some prior literature indicating that lower levels of estradiol are associated with higher levels of PTSD (Garcia et al., 2018; Green & Graham, 2022; Nillni et al., 2021) and CVD risk (Hargrave et al., 2022; Krantz et al., 2022; Šagud et al., 2017). However, a systematic review found that the strong evidence for an association between high estradiol levels and better fear extinction recall in healthy samples did not translate to women with PTSD, such that higher estradiol was associated with worse symptom severity (Garcia et al., 2018). As the authors suggested, differences in the type and timing of hormone measurement (e.g. plasma versus saliva, time of day), as well as sample differences may impact study findings and reproducibility. For example, the majority of studies in the fear learning literature used healthy samples rather than samples of women who were trauma-exposed or had PTSD, and among the five studies included in the review that utilised clinical samples, the PTSD samples were small, and the findings related to menstrual or hormonal status and fear acquisition, extinction, and inhibition were mixed (Garcia et al., 2018). In our sensitivity analyses among those with PTSD (n = 37), lower estradiol levels were associated with higher systolic and diastolic BP, which is consistent with prior research suggesting that estradiol is cardioprotective in premenopausal women (Pérez-López et al., 2010). The inconsistency in these findings and the paucity of existing research on the impact of estradiol on PTSD and CVD symptoms using clinical samples suggest the need for further research on the impact of estradiol in clinical samples of trauma-exposed women.

While estradiol was not significantly associated with CVD symptoms, we found that progesterone, and relatedly the PE ratio, was significantly associated with CVD symptoms. Among the subset of participants who were naturally cycling, higher progesterone levels and PE ratios were associated with worse Re-experiencing symptoms. These findings align with existing literature indicating that higher progesterone is associated with greater stress memory retention, negative stimuli recall, and negative intrusive memories (Garcia et al., 2018; Green & Graham, 2022; Hsu et al., 2021; Nillni et al., 2021). Increased stress memory retention and recall may increase the frequency at which the stress response is triggered. Among individuals with PTSD, the stress response is characterised by dysfunction in the autonomic nervous system and HPA axis (e.g. increased sympathetic and decreased parasympathetic activity), and these responses to stress are thought to increase inflammation and metabolic dysregulation, both of which are associated with increased CVD (Michopoulos et al., 2017; Seligowski et al., 2022). The limited existing literature aligns with but does not fully explain our finding that the association between progesterone and CVD symptoms was only significant at low PTSD symptom severity. The association of progesterone and extinction learning retention has been found to differ between women with PTSD and trauma-exposed women without PTSD (Nillni et al., 2021), such that higher progesterone levels are associated with worse extinction retention among women with PTSD, but better retention among those without PTSD. Additionally, women with PTSD are more likely than trauma-exposed women without PTSD to have deficits in the conversion of progesterone into its metabolites allopregnanolone and pregnanolone (Pineles et al., 2020); however, deficits in the conversion of progesterone were found to decrease extinction recall for women with PTSD but not those without, which would suggest that lower progesterone could be associated with worse Re-experiencing symptoms and thus worse CVD symptoms for women with PTSD and not those without. This contradicts our finding that higher progesterone levels are associated with CVD symptoms only among those with low PTSD symptoms. These contradictory findings suggest the need for further research on the moderating role of PTSD in the impact of progesterone on CVD.

Certain limitations of the current study must be noted. In this study, CVD symptoms were evaluated via self-report rather than through clinical diagnosis. Given that none of our participants endorsed CVD events (stroke, heart attack, heart failure), the CVD symptoms that were endorsed could have been indicative of early risk or anxiety (e.g. chest pain). We attempted to mitigate this by measuring and controlling for BP, but future studies would benefit from additional objective measures of pre-clinical CVD risk (e.g. endothelial function). While a strength of this study was our measurement of plasma-based sex hormone levels, not all women were naturally cycling. Given that some women were on hormonal contraceptives, which can affect hormone levels in different ways depending on the contraceptive type, caution is warranted when interpreting our results in the context of prior literature (although it is important to note that there were no significant differences in sex hormones based on naturally cycling status in our study). Additionally, the current study did not collect data on whether participants had ever engaged in PTSD treatment. Given that currently undergoing PTSD treatment could affect symptom presentation, this is a limitation and should be considered in future studies. Finally, an a-priori power analysis was not conducted and post-hoc power analyses are not recommended given their high variability (Heinsberg & Weeks, 2022; Zhang et al., 2019). Given our small sample size, we may have been underpowered to detect the observed effects.

The current study improves our understanding of sex effects in the relationships between PTSD and CVD by finding a moderating effect of PTSD severity on the relationship between sex hormones and CVD symptoms. By using plasma estradiol and progesterone levels rather than estimates based on menstrual cycle phases, the current study advances the literature on the impact of sex hormones on PTSD and CVD symptoms, which has to date had inconsistent definitions and measurements of menstrual cycle phases and sex hormone levels. Further research is required to expand the current findings, and such studies should include larger samples of trauma-exposed women and conduct more comprehensive assessments of CVD symptoms. Additional research is also warranted to clarify how progesterone relates to increased CVD symptoms and why PTSD symptom severity moderates this relationship. Better characterisation of sex effects in the PTSD-CVD link will ultimately improve CVD risk identification in PTSD.

Data availability statement

The data that support the findings of this study are available from the corresponding author, [AVS], upon reasonable request.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Additional information

Funding

This work was supported by American Heart Association [grant number 20CDA35310031]; National Institutes of Health [grant number K23MH125920].