Objective and Subjective Multidimensional Sleep Health in Women After Myocardial Infarction Authors: Olivia C Liu, Milla Arabadjian, Anais Hausvater, Chorong Park, Amanda J Shallcross, Jolaade Kalinowski, Dayna A Johnson, Harmony R Reynolds, Tanya M Spruill Published: 2026 ABSTRACT AbstractBackgroundPoor sleep after acute cardiovascular events is associated with increased mortality. Women are more likely than men to report sleep disturbances after cardiovascular events, but other sleep metrics linked to cardiometabolic health remain underexplored.ObjectivesWe characterize multiple dimensions of sleep and examine factors associated with poor sleep among women with prior myocardial infarction (MI).MethodsParticipants completed 7 nights of wrist actigraphy, which measured sleep duration, sleep efficiency, and wake after sleep onset (WASO), and sleep diary, which measured self-reported sleep quality using a Likert scale. Multivariable regression modeled associations between sociodemographic, clinical, and psychosocial characteristics and sleep measures.ResultsAmong 94 post-MI women (median days since MI 93; IQR: 69-709), mean age was 60 years, and 67.0% identified as non-Hispanic White. The mean sleep duration was 405.1 minutes (approximately 6.8 hours), sleep efficiency 86.3%, WASO 61.6 minutes, and self-reported sleep quality 3.3, indicating fair-to-good sleep quality. Overall, 64.9% of women had short (<7 hours) or long (>9 hours) sleep duration, 33% low sleep efficiency (<85%), and 88.3% long WASO (>30 minutes). In exploratory analyses, associations with poor sleep included identifying as other than non-Hispanic White (shorter sleep duration, lower efficiency, and longer WASO), being nonpartnered (shorter sleep duration), hypertension (lower sleep efficiency), and higher depressive symptoms (lower sleep quality).ConclusionsMany women with prior MI had suboptimal sleep (short or long sleep duration, low sleep efficiency, and/or long WASO). Our results highlight the importance of discussing multiple dimensions of sleep with women post-MI to optimize cardiovascular health.Key words: myocardial infarction, sleep, women Sleep is essential to cardiometabolic health.1,2 Sleep health encompasses multiple dimensions, such as sleep duration, efficiency, regularity, timing, satisfaction, and daytime alertness.1, 2, 3 Short and long sleep duration have each been linked to cardiovascular disease risk factors including diabetes and hypertension,4,5 as well as with cardiovascular disease itself including acute coronary syndrome.6,7 Evidence suggests that sleep health before and after acute cardiovascular events influences recovery and long-term outcomes. For example, subjective disturbed sleep (“including general sleep quality, difficulty falling asleep, repeated awakenings, disturbed/uneasy sleep, and premature awakenings”) in the year before myocardial infarction (MI) is associated with repeat major adverse cardiovascular events within 10 years for women.8 Short and long sleep duration 3 months after acute coronary syndrome are each associated with increased all-cause mortality at 3 years.9 Emerging data characterizing sleep health after acute cardiovascular events have shown that women are more likely than men to report disturbed sleep in the year following acute coronary syndrome.10 However, prior studies on sleep health after acute coronary syndrome have not examined women specifically and have focused on single, subjectively-assessed sleep metrics.9,10 Although sleep duration is included in the American Heart Association Life’s Essential 8, a more recent American Heart Association Scientific Statement emphasizes the importance of assessing multiple dimensions of sleep, including both objective and subjective measures, to improve understanding of sleep and cardiometabolic health.2 In this study, we characterized multiple dimensions of sleep health using both objective and subjective measures in women after MI. We also investigated associations between sociodemographic, clinical, and psychosocial characteristics and sleep measures. Methods Overview We analyzed data from the NYU Women’s Heart Attack Research Program (HARP), a multicenter prospective cohort study of women with a history of MI. A subset of participants were enrolled in the HARP Stress Management Trial, a randomized controlled trial that tested the effects of a mindfulness intervention on perceived stress and sleep in women after MI (NCT02914483).11 The present study is a subanalysis of the HARP Stress Management Trial using baseline data collected before randomization and intervention delivery. The trial design and other results have been previously published.11,12 NYU Langone Health and all participating sites’ Institutional Review Boards approved the trial. Participants We included English-speaking women aged ≥21 years with a confirmed diagnosis of MI based on the Fourth Universal Definition,13 who were at least 2 months after MI and had elevated stress levels (defined as Perceived Stress Scale-4 ≥6).14 Exclusion criteria included moderately severe to severe depressive symptoms (Patient Health Questionnaire-9 [PHQ-9] ≥15),15 active suicidal ideation (PHQ-9 item 9 ≥1),15 significant cognitive impairment noted in the medical record or during screening, history or current diagnosis of psychosis, and active participation in another behavioral trial. All participants’ sex assigned at birth was female. Enrollment Participants were recruited for this study via 2 approaches: enrollment around the time of acute MI (median 78 days post-MI) and self-referral any time after MI (median 1,063 days post-MI). Research staff approached acute MI participants around the time of cardiac catheterization referral after MI and asked to contact them in 2 months for the HARP Stress Management Trial. Self-referral patients were recruited through community advertising, and research staff reviewed medical records to confirm eligibility, including a history of MI at least 2 months prior. All study participants provided informed consent. Study measures Participants completed questionnaires about sociodemographic, clinical, and psychosocial characteristics. Sociodemographic characteristics included age, race and ethnicity, marital and cohabitation status, educational attainment, and perceived financial strain (income covers needs poorly or not very well). Clinical characteristics included days since MI, history of hypertension, history of diabetes, tobacco use, and use of over-the-counter or prescription sleep-aid medications during the sleep monitoring period (“Did you take any over-the counter or prescription medications to help you sleep?” answered on a sleep diary). Disease-specific health status was assessed with the Seattle Angina Questionnaire-7 (SAQ-7) (reference range 0-100, with higher scores indicating better health status).16 Depressive symptoms were assessed with the PHQ-9 (reference range 0-15 [due to trial exclusion criteria], with higher scores indicating greater depression severity).15 Sleep was assessed with wrist actigraphy and a sleep diary. Participants wore a wrist actigraph (wGT3X-BT; ActiGraph Corp) on the nondominant wrist for 7 consecutive nights,17 and they completed a sleep diary during the same period. Wrist actigraphy data were processed using the ActiLife software (version 6.13.3; ActiGraph Corp). The primary sleep period was identified from the raw actigraphy data and consultation with the sleep diary. Editing of ActiLife data was performed by 1 of 3 raters following training and demonstration of inter-rater reliability. The Cole-Kripke algorithm, adapted to the ActiLife software, was used to identify minute-by-minute sleep/wake status and yielded the sleep measures described subsequently.18 Sleep measures included sleep duration, sleep efficiency, wake after sleep onset (WASO), and sleep quality. Sleep duration, measured by actigraphy, was operationalized as total number of minutes asleep during the primary sleep period (ie, excluding naps).19 Actigraphy-measured sleep efficiency was defined as total minutes asleep divided by total minutes in bed, multiplied by 100 (presented as a percentage).19,20 Actigraphy-measured WASO was defined as total minutes awake between sleep onset and final awakening.19 Sleep quality was self-reported on the sleep diary. In accordance with the Consensus Sleep Diary, participants rated their nightly sleep quality on a 5-point Likert scale (1 = very poor, 2 = poor, 3 = fair, 4 = good, and 5 = very good).21 We defined suboptimal sleep health in this study as duration <7 hours or >9 hours, efficiency <85%, and WASO more than 30 minutes based on the recommendations of the American Academy of Sleep Medicine and Sleep Research Society for sleep duration and prior research on optimal thresholds for sleep efficiency and WASO.2,22,23 We did not apply thresholds for sleep quality, as the Consensus Sleep Diary was developed to standardize sleep diary content rather than establish normative cut points, and there are limited data supporting an optimal sleep quality threshold using this tool.2,21 Statistical analyses The analytic sample comprised enrolled participants who completed the wrist actigraphy protocol with at least 3 nights of valid actigraphy data.24,25 Descriptive statistics were used to summarize sociodemographic, clinical, and psychosocial characteristics and sleep measures. Categorical variables were described with frequencies and percentages, and continuous variables were described with means and SDs or medians and IQRs. To calculate cohort-level mean sleep duration, sleep efficiency, WASO, and sleep quality, we used a two-step approach. First, we calculated individual means of sleep duration, sleep efficiency, WASO, and sleep quality across all valid nights for each participant. These participant-level means were then averaged to produce cohort-level means, thereby avoiding overweighting participants with more valid nights of sleep data. For sleep measures that demonstrated non-normal distributions based on visual inspection and Shapiro-Wilk testing (sleep efficiency and WASO), we additionally reported cohort-level medians and IQRs. In supplemental analyses, we compared characteristics between participants who did and did not complete the wrist actigraphy protocol with at least 3 nights of valid data, using chi-square test or Fisher exact test for categorical variables and Student’s t-test or Mann-Whitney U test for continuous variables, as appropriate based on variable distributions. In addition, in supplemental analyses, we compared characteristics and sleep measures between acute MI participants and self-referral participants, as well as sleep measures between participants who completed the protocol in the winter vs all other seasons. In exploratory analyses, we used multivariable linear regression to model associations between sociodemographic, clinical, and psychosocial characteristics and sleep measures (sleep duration, sleep efficiency, WASO, and sleep quality). Factors associated with sleep measures were chosen a priori, based on theory and prior literature linking each factor to sleep health.4,5,10,26, 27, 28, 29 The number of factors were limited to reduce the risk of overfitting given the modest sample size. Sociodemographic characteristics selected were age, race and ethnicity, educational attainment (bachelor’s degree or graduate degree vs less than bachelor’s degree), partner status (married or cohabiting vs unmarried and not cohabitating), and financial strain (perceived financial strain vs no perceived financial strain). Although the sample included diverse racial and ethnic backgrounds (non-Hispanic White, non-Hispanic Black, Hispanic, Asian, more than 1 race and ethnicity, and other race and ethnicity), the race and ethnicity variable was dichotomized as non-Hispanic White vs race and ethnicity other than non-Hispanic White in regression models due to small sample sizes of individual groups. Clinical characteristics were days since MI, history of hypertension, history of diabetes, tobacco use (current vs former or never), any use of sleep-aid medications during the sleep monitoring period, and disease-specific health status assessed via SAQ-7. Depressive symptoms were assessed via PHQ-9. Given the exploratory nature of these analyses and modest sample size, a prespecified alpha level of 0.10 was used as a threshold to guide identification of potential factors associated with sleep measures.30,31 Regarding ordinary least squares assumptions, we assessed multicollinearity, heteroscedasticity, and normality of residuals for all models, as described further in the Supplemental Methods. For models with non-normally distributed residuals (sleep efficiency and WASO), sensitivity analyses were performed using natural log transformation of the outcomes. All analyses were conducted with Stata Version 18 (StataCorp LLC). Results Among 130 women enrolled in the HARP Stress Management Trial, 94 (72.3%) completed the wrist actigraphy protocol with at least 3 nights of valid data and were included in the present study (Figure 1). The mean age was 60 years and 67.0% of participants identified as non-Hispanic White (Table 1). Almost half of the participants had a bachelor’s or graduate degree (48.9%), were married or cohabitating with a partner (46.8%), and reported having financial strain (41.5%). The median number of days from MI to trial enrollment was 93 (IQR: 69-709). Overall, 53.2% of participants had hypertension, 21.3% had diabetes, and 11.7% currently used tobacco. During the sleep-monitoring period, 10.6% of participants used over-the-counter or prescription sleep-aid medications on some nights, and 25.6% used sleep-aid medications on most or all nights. Self-reported sleep-aid medications are listed in Supplemental Table 1. The mean SAQ-7 score was 80.7, corresponding with “excellent” disease-specific health status.32 Almost thirty-percent of participants had PHQ-9 scores ≥10 (29.8%), indicating moderate depressive symptoms. Women who completed the wrist actigraphy protocol with at least 3 nights of valid data were less likely to have hypertension than those that did not (53.2% vs 72.2%, P = 0.049); other characteristics were comparable between groups (Supplemental Table 2). Compared with participants recruited through self-referral, acute MI participants were less likely to have a bachelor’s or graduate degree (36.7% vs 70.6%, P = 0.002) and more likely to report financial strain (51.7% vs 23.5%, P = 0.006), whereas other characteristics were similar (Supplemental Table 3). Flow Chart of Sample Sample sizes for participants enrolled in Heart Attack Research Program Stress Management Trial, included in the present study, and included in multivariable regression analysis. HARP = Heart Attack Research Program; PHQ-9 = Patient Health Questionnaire-9. Table 1Characteristics of Sample (N = 94)Age, y, mean ± SD59.9 ± 12.8Race and ethnicity, n (%) Non-Hispanic White63 (67.0%) Non-Hispanic Black18 (19.2%) Hispanic7 (7.5%) Asian3 (3.2%) More than 1 race and ethnicity1 (1.1%) Other2 (2.1%) Missing0 (0.0%)Education, n (%) Bachelor’s or graduate degree46 (48.9%) Less than bachelor’s degree47 (50.0%) Missing1 (1.1%)Partner status, n (%) Married or cohabitating44 (46.8%) Unmarried and not cohabitating49 (52.1%) Missing1 (1.1%)Financial strain, n (%) Yes39 (41.5%) No54 (57.5%) Missing1 (1.1%)Days since MI, median (IQR)93 (69-709) Hypertension, n (%) Yes50 (53.2%) No44 (46.8%) Missing0 (0.0%)Diabetes, n (%) Yes20 (21.3%) No74 (78.7%) Missing0 (0.0%)Tobacco use, n (%) Current11 (11.7%) Former38 (40.4%) Never43 (45.7%) Missing2 (2.1%)Use of sleep-aid medications, n (%) None55 (58.5%) Some nights10 (10.6%) Most nights7 (7.5%) Every night17 (18.1%) Missing5 (5.3%)SAQ-7 total scorea, mean ± SD80.7 ± 17.8PHQ-9 score ≥10b, n (%) Yes28 (29.8%) No65 (69.2%) Missing1 (1.1%)MI = myocardial infarction; PHQ-9 = Patient Health Questionnaire-9; SAQ-7 = Seattle Angina Questionnaire-7.aMissing 1 observation (n = 93).bIndicates moderate depression severity. Characteristics of Sample (N = 94) MI = myocardial infarction; PHQ-9 = Patient Health Questionnaire-9; SAQ-7 = Seattle Angina Questionnaire-7. Missing 1 observation (n = 93). Indicates moderate depression severity. Cohort-level mean sleep duration was 405.1 minutes (SD 63.3 minutes), or approximately 6.8 hours per night (Table 2, Central Illustration). The mean sleep efficiency was 86.3% (SD 6.3%), with a median of 88.2% (IQR: 82.6% to 90.2%). The mean WASO was 61.6 minutes (SD 32.0 minutes), with a median of 52.6 minutes (IQR: 41.7-75.6 minutes). Among the 94 participants who completed the wrist actigraphy protocol with at least 3 nights of valid sleep data, 89 (94.7%) also completed the sleep diary. The mean sleep quality was 3.3 (SD 0.7), corresponding with fair-to-good sleep quality. Sleep duration, sleep efficiency, WASO, and sleep quality were similar between acute MI and self-referral participants (Supplemental Table 3). In addition, sleep parameters did not differ significantly in the winter compared with other seasons (Supplemental Table 4). Based on participant-level means, almost two-thirds of participants in the cohort (64.9%) did not meet the recommended sleep duration of 7 to 9 hours, mostly due to short (61.7%) rather than long sleep (3.2%). One-third of participants (33.0%) had low sleep efficiency (<85%), and 88.3% of participants had long WASO (>30 minutes). Among the participants with normal sleep duration (n = 33), 21.2% had low sleep efficiency and 87.9% had long WASO (Supplemental Table 5). Among the participants with normal sleep efficiency (n = 63), 58.7% had short or long sleep duration, and 82.5% had long WASO. Table 2Sleep Parameters, Measured Via Actigraphy and Self-Reported DiaryActigraphyn = 94Sleep durationa (minutes), mean ± SD405.1 ± 63.3Sleep efficiencyb (percentage), mean ± SD86.3 ± 6.3Sleep efficiencyb (percentage), median (IQR)88.2 (82.6-90.2)Wake after sleep onsetc (minutes), mean ± SD61.6 ± 32.0Wake after sleep onsetc (minutes), median (IQR)52.6 (41.2-75.6)Average sleep duration 7-9 hours, n (%)33 (35.1%)Average sleep duration <7 hours, n (%)58 (61.7%)Average sleep duration >9 hours, n (%)3 (3.2%)Average sleep efficiency <85%, n (%)31 (33.0%)Average wake after sleep onset >30 minutes, n (%)83 (88.3%)Sleep diaryn = 89Sleep qualityd, mean ± SD (1 [very poor] to 5 [very good])3.3 ± 0.7aTotal number of minutes asleep, defined by actigraphy.bTotal minutes asleep divided by total minutes in bed, multiplied by 100, defined by actigraphy.cTotal minutes awake between sleep onset and final awakening, defined by actigraphy.dSelf-reported on 5-point scale (1 = very poor, 2 = poor, 3 = fair, 4 = good, and 5 = very good). Sleep Parameters, Measured Via Actigraphy and Self-Reported Diary Total number of minutes asleep, defined by actigraphy. Total minutes asleep divided by total minutes in bed, multiplied by 100, defined by actigraphy. Total minutes awake between sleep onset and final awakening, defined by actigraphy. Self-reported on 5-point scale (1 = very poor, 2 = poor, 3 = fair, 4 = good, and 5 = very good). Sleep Measures and Factors Associated With Poor Sleep in Women After Myocardial Infarction Sleep measures included sleep duration, sleep efficiency, wake after sleep onset, and sleep quality. Exploratory analyses explored associations between participant sociodemographic, clinical, and psychosocial characteristics and sleep measures. PSS-4 = Perceived Stress Scale-4. Race and ethnicity other than non-Hispanic White was associated with shorter sleep duration (B = −26.32; 90% CI: −52.04 to −0.60), lower sleep efficiency (B = −3.40; 90% CI: −6.02 to −0.78), and longer WASO (B = 15.69; 90% CI: 1.67-29.72) (Table 3). Unmarried and not cohabitating partner status was associated with shorter sleep duration (B = −37.74; 90% CI: −60.97 to −14.51). Hypertension was associated with lower sleep efficiency (B = −2.88; 90% CI: −5.24 to −0.52). The use of over-the-counter or prescription sleep-aid medications was associated with longer sleep duration (B = 23.20; 90% CI: 0.58-45.81) and higher sleep efficiency (B = 2.76; 90% CI: 0.46-5.07). Higher PHQ-9 scores were associated with poorer sleep quality (B = −0.05; 90% CI: −0.08 to −0.02). Sensitivity analyses of regression models using natural log-transformed sleep efficiency and WASO supported main analyses (Supplemental Table 6). Table 3Associations Between Participant Characteristics and Sleep Duration, Sleep Efficiency, Wake After Sleep Onset, and Sleep QualitySleep Duration, Minutes (n = 86)Sleep Efficiency, Percentage (n = 86)Wake After Sleep Onset, Minutes (n = 86)Sleep Quality Score (n = 86)Age0.04 [−0.83 to 0.92]−0.04 [−0.13 to 0.05]0.27 [−0.21 to 0.75]0.01 [0.00-0.01]Race and ethnicity other than non-Hispanic Whitea−26.32 [−52.04 to −0.60]b−3.40 [−6.02 to −0.78]c15.69 [1.67-29.72]b0.02 [−0.27 to 0.30]Bachelor’s or graduate degreea−18.71 [−41.62 to 4.19]0.21 [−2.13 to 2.54]−2.75 [−15.24 to 9.75]0.25 [0.00-0.50]Unmarried and not cohabitatinga−37.74 [−60.97 to −14.51]d−1.24 [−3.61 to 1.13]1.06 [−11.61 to 13.72]−0.15 [−0.41 to 0.11]Financial straina7.23 [−15.99 to 30.45]−2.33 [−4.70 to 0.04]11.77 [−0.89 to 24.43]−0.24 [−0.50 to 0.02]Days since MI0.00 [−0.01 to 0.00]0.00 [0.00-0.00]0.00 [0.00-0.01]0.00 [0.00-0.00]Hypertension−11.60 [−34.75 to 11.56]−2.88 [−5.24 to −0.52]c11.08 [−1.55 to 23.71]−0.14 [−0.40 to 0.11]Diabetes9.23 [−19.21 to 37.66]−0.44 [−3.34 to 2.46]5.52 [−9.98 to 21.03]0.14 [−0.18 to 0.46]Current tobacco use−24.31 [−57.77 to 9.15]−2.21 [−5.62 to 1.20]15.31 [−2.94 to 33.55]0.05 [−0.32 to 0.42]Any use of sleep-aid medicationsa23.20 [0.58-45.81]b2.76 [0.46-5.07]b−6.16 [−18.49 to 6.17]−0.21 [−0.47 to 0.04]SAQ-7 score0.60 [−0.11 to 1.32]0.06 [−0.02 to 0.13]−0.29 [−0.68 to 0.09]0.00 [0.00-0.01]PHQ-9 score−1.18 [−3.02 to 1.56]−0.06 [−0.34 to 0.22]−0.01 [−1.51 to 1.48]−0.05 [−0.08 to −0.02]dValues are unstandardized regression coefficient (95% CI). The presented coefficients represent multivariable regression estimates from 4 independent models.Abbreviations as in Table 1.aReference groups are non-Hispanic White, less than bachelor’s degree, married or cohabitating, no financial strain, and no use of sleep-aid medications during the sleep monitoring period, respectively.bP < 0.10.cP < 0.05.dP < 0.01. Associations Between Participant Characteristics and Sleep Duration, Sleep Efficiency, Wake After Sleep Onset, and Sleep Quality Values are unstandardized regression coefficient (95% CI). The presented coefficients represent multivariable regression estimates from 4 independent models. Abbreviations as in Table 1. Reference groups are non-Hispanic White, less than bachelor’s degree, married or cohabitating, no financial strain, and no use of sleep-aid medications during the sleep monitoring period, respectively. P < 0.10. P < 0.05. P < 0.01. Discussion To our knowledge, this is the first study that describes multiple dimensions of sleep using objective and subjective measures in women after MI. About 65% of women had short or long sleep duration, 33% had low sleep efficiency, and about 88% had long WASO. In exploratory analyses, race and ethnicity other than non-Hispanic White was associated with shorter sleep duration, lower sleep efficiency, and longer WASO. Unmarried and not cohabitating partner status was associated with shorter sleep duration, hypertension was associated with lower sleep efficiency, and higher depressive symptoms were associated with lower self-reported sleep quality. Our results emphasize the importance of discussing sleep health with patients after MI in the clinical setting. Women in our cohort had suboptimal sleep based on multiple measures, even months after MI. Furthermore, about a quarter of women took sleep-aid medications most or all nights during the monitoring period, a rate higher than that of women in the general population (10.2%) based on the 2020 National Health Interview Survey from the Center for Disease Control and Prevention.33 Unfortunately, in cohort studies of U.S. patients and physicians, between one-third to two-thirds of patients reported having never discussed sleep with their medical providers,34,35 and less than half of primary care physicians counseled their patients on sleep hygiene.36 Given the importance of sleep to cardiometabolic health and the prevalence of poor sleep health in cardiovascular populations, cardiologists have a critical role in addressing sleep with their patients, beyond assessing for common sleep disorders. Our findings reinforce that multiple dimensions should be addressed in addition to sleep duration, which is the only current metric for sleep health in Life’s Essential 8.1 Patients may have adequate sleep by 1 dimension but suboptimal sleep by other dimensions. Illustratively, approximately one-fifth of patients with normal sleep duration had low sleep efficiency. Moreover, although sleep efficiency and WASO are both markers of sleep continuity (the ability to initiate and maintain sleep),3 the majority of participants with normal sleep efficiency had long WASO, reflecting falling asleep quickly but having fragmented sleep after sleep onset.37 Low sleep efficiency and long WASO each independently confer increased risk of incident major adverse cardiovascular events,38 and further data are needed to understand how sleep efficiency, WASO, and sleep latency may interact in affecting cardiovascular health. With the growing evidence linking multiple dimensions of sleep independently to cardiometabolic health,2 providers must discuss sleep health through a multidimensional lens when talking to patients. Providers may ask about patients’ sleep duration, ability to fall asleep, wakefulness after falling asleep, satisfaction with sleep, and daytime alertness after sleeping. One tool for doing so is the RU-SATED scale, a short, validated questionnaire for multidimensional sleep health.3,39 Cardiac rehabilitation programs offer another possible setting for using the RU-SATED scale and other sleep assessments, which could be incorporated into program enrollment and follow-up visits.40,41 This approach could help identify patients with suboptimal sleep during the post-MI recovery period, allowing for early sleep health education and counseling, referrals, and interventions if needed. When assessing sleep health in patients, clinicians can also consider subgroups at higher risk of poor sleep after MI. Exploratory results suggested that identifying with a race and ethnicity other than non-Hispanic White, unmarried and not cohabitating partner status, and higher depressive symptoms were each associated with worse sleep outcomes. These associations show consistency with prior literature on risk factors for poor sleep. Among people with subclinical cardiovascular disease, Black individuals have higher odds of sleep apnea syndrome, short sleep, poor sleep quality, and daytime sleepiness, and Hispanic and Chinese individuals have higher odds of sleep disordered breathing and short sleep, compared with non-Hispanic White individuals.42 Similarly, among patients after MI, Black patients have poorer sleep quality scores than White patients.43 Future studies with large, diverse cohorts of post-MI patients are needed to investigate potential variations in sleep metrics across racial and ethnic groups. Literature also supports the relationship between unmarried and not cohabiting partner status and poor sleep,44,45 as well as the likely bidirectional relationship between depression and poor sleep,10,46,47 These results highlight the opportunity to address co-occurring and potentially modifiable factors associated with sleep. One proposed mechanism linking unmarried and not cohabitating partner status to poor sleep is lower perceived social support,48, 49, 50 and interventions aimed at enhancing social support may promote better sleep.51,52 Identifying and addressing depressive symptoms may also improve sleep outcomes.53 Study limitations Our study has several limitations. The HARP Stress Management Trial enrolled women with a history of MI, so the present study does not include a control group without MI, limiting comparisons of sleep parameters between patients with and without MI. Given that the trial included women after MI with Perceived Stress Scale-4 score ≥6 and excluded women with PHQ-9 score ≥15, active suicidal ideation, cognitive impairment, and psychosis, our study’s sample may not be representative of women with a history of MI broadly. The reported sleep measures should be interpreted in the context of the study’s enrichment for elevated stress and exclusion of participants with moderately severe to severe depressive symptoms. The prevalence of elevated depressive symptoms in our cohort may be underestimated, and the restricted range of PHQ-9 scores may have attenuated the observed association between depressive symptoms and sleep measures. Although fairly diverse, the sample did not include large enough numbers of racial and ethnic minority groups for robust between-group comparisons. Wrist actigraphy does not provide accurate estimates of sleep latency,54 so this aspect of sleep could not be characterized. Some participant characteristics that could influence sleep health, such as body mass index, obstructive sleep apnea, restless leg syndrome, heart failure, atrial fibrillation, chronic kidney disease, and alcohol use, were not part of the questionnaires and were unavailable for participants who received care outside of NYU. Thus, they were not included in regression analyses. We identified several participant characteristics associated with our sleep measures that were consistent with prior literature; however, given the modest sample size, our study may have been underpowered to detect additional associations. Some participants enrolled in the HARP Stress Management Trial did not complete the wrist actigraphy protocol with sufficient valid data, although the completion of the protocol was high at 72%. Conclusion In this cohort of women after MI, about 65% had short or long sleep duration, 33% had low sleep efficiency, and 88% had long WASO. In addition, many women reported using medications for sleep. In exploratory analyses, race and ethnicity other than non-Hispanic White race and ethnicity, unmarried and not cohabitating partner status, hypertension, and higher depressive symptoms were related to poorer sleep health. Given the importance of sleep to cardiometabolic health and the prevalence of poor sleep across multiple domains in this cohort of women post-MI, clinicians have an important role in discussing sleep health with this patient population. Clinical trials are needed to determine how best to improve multidimensional sleep health among women with a history of MI. COMPETENCY IN MEDICAL KNOWLEDGE: Suboptimal sleep health across multiple domains is common among women with a history of MI. It is important to discuss sleep health in women with MI using a multidimensional lens. TRANSLATIONAL OUTLOOK: Future researches focused on interventions to improve sleep health are needed among women after MI. Funding support and author disclosures This work was funded by the American Heart Association Go Red For Women Strategically Focused Research Network grant #16SFRN28850003, 16SFRN27810006, and 812162. The authors have reported that they have no relationships relevant to the contents of this paper to disclose. Footnotes The authors attest they are in compliance with human studies committees and animal welfare regulations of the authors’ institutions and Food and Drug Administration guidelines, including patient consent where appropriate. For more information, visit the Author Center. Appendix For an expanded Methods section and supplemental tables, please see the online version of this paper. Supplementary material Supplemental Material References Associated Data Supplementary Materials Supplemental Material