Atherosclerotic Cardiovascular Disease and Health-Related Quality of Life Among Adults in the United States: National Health Interview Survey 2013-2017

25

A brief health-related quality of life (HRQoL) tool with construct validity for 26 atherosclerotic cardiovascular disease (ASCVD) may facilitate integration into healthcare 27 delivery. We examined ASCVD-related changes in the Health and Activity Limitation Index 28 (HALex), a generic HRQoL measure comprising perceived health and activity limitation. 29

Using data of 155,130 respondents of the National Health Interview Survey 2013-30 2017, we evaluated HALex scores by ASCVD (angina, heart attack, and stroke). Lower HALex 31 scores reflected worse HRQoL and a 0.03 change was regarded as the threshold for clinical 32 significance. Multivariable two-part models were used to assess HALex changes (β , 95%CI) 33 associated with ASCVD overall and in sex, age, and race/ethnicity groups. 34

Overall, participants with ASCVD – 6.8%, representing 15.7 million adults – had lower 35 HALex scores (0.67) than those without ASCVD (0.87). Females, age ≥ 65 years, and non-36 Hispanic Blacks had the lowest HALex scores. Overall, ASCVD was associated with a 37 substantial decrement in HALex (-0.10, [-0.10, -0.09]). Interactions between ASCVD and sex, 38 and race/ethnicity were both significant (p < 0.001). ASCVD-associated decrement in HALex 39 was clinically greater in: females (-0.11, [-0.12, -0.10]) than in males (-0.08, [-0.09, -0.07]); and 40 non-Hispanic Black (-0.13, [-0.15, -0.1]) than in non-Hispanic White (-0.09, [-0.10, -0.08]). 41 Though ASCVD impact on HALex was greater in age 18-64 years (-0.09, [-0.10, -0.08]), it was 42 not statistically different from the elderly (-0.06, [-0.07, -0.06]). 43

ASCVD was consistently associated with lower HRQoL, as measured by HALex, 44 across major demographics. HALex presents a feasible HRQoL tool to implement in healthcare. 45 . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprintthis version posted September 1, 2022. ; https://doi.org/10.1101/2022.08.31.22279420doi: medRxiv preprint 3

46 ASCVD; atherosclerotic cardiovascular disease; quality of life; Health and Activity Limitation 47 Index; HALex 48 . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprintthis version posted September 1, 2022. ; https://doi.org/10.1101/2022.08.31.22279420doi: medRxiv preprint 4 Non-standard Abbreviations and Acronyms 49 ASCVD atherosclerotic cardiovascular disease 50 COPD chronic obstructive pulmonary disease 51 GED General Educational Development 52 HALex Health and Activity Limitation Index 53 HRQoL health-related quality of life 54 NHIS National Health Interview Survey 55 OR odds ratio 56 PROM patient-reported outcome measures 57 SE standard error 58 . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprintthis version posted September 1, 2022. ; https://doi.org/10.1101/2022.08.31.22279420doi: medRxiv preprint 5 FULL TEXT 59

60 Atherosclerotic cardiovascular disease (ASCVD) is associated with a high burden of debilitating 61 symptoms and emotional stress related to decompensation and recurrent events, and invasive 62 management. With a plenitude of life-extending treatments and a shift in healthcare quality 63 assessment to value-based care, there is a growing call for the routine use of patient-reported 64 outcome measures (PROMs) like health-related quality of life (HRQoL) in the clinic beyond 65 trials (1). HRQoL not only summarizes wellbeing from a patient’s perspective, but is associated 66 with subsequent cardiovascular events, mortality, and the patterns of healthcare utilization and 67 expenditure (2,3). 68 However, the implementation of HRQoL assessment in clinical workflow is limited by factors 69 including the lengthy surveys in commonly used research tools such as the 36-item Short Form 70 Survey (or its shortened version, Short Form 12) and other utility-based tools (4). On the other 71 hand, the Health and Activity Limitation Index (HALex) is a generic HRQoL measure that 72 combines two items – perceived health and physical functioning – into a global score for 73 wellbeing (5). It is mathematically derived and is not subject to the variations other HRQoL tools 74 with community rated utility scores have owing to differing measuring techniques and 75 geographic variation (6–8). With a low administrative burden, HALex may be useful in the clinic 76 for assessing the impact of chronic conditions including ASCVD. 77 In this cross-sectional analysis, we described the differences in HRQoL associated with the 78 presence of ASCVD among U.S. adults using HALex. As a convenient and reliable alternative to 79 . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprintthis version posted September 1, 2022. ; https://doi.org/10.1101/2022.08.31.22279420doi: medRxiv preprint 6 more detailed tools, we expected ASCVD-related decrements in HALex in the total population 80 and within sex, age, and race/ethnicity groups. 81

82 Study design and sampling 83 This study used the 2013-2017 National Health Interview Survey (NHIS) data. The NHIS is an 84 annual household interview survey of the United States (U.S.) civilian, non-institutionalized 85 population sponsored by the National Center for Health Statistics and under the auspices of the 86 Centers for Disease Control and Prevention (9). Interviews are conducted using a complex, 87 multistage probability design to reflect changes in the distribution of the U.S. population and 88 produce nationally representative estimates. The NHIS has four core components – Household 89 Composition, Family, Sample Child, and the Sample Adult – which provide information on 90 respondents’ sociodemographic characteristics, health status and activity limitations, behavior 91 indicators, and health care access and utilization. In addition, NHIS data are supplemented with 92 survey weights to account for selection probabilities and non-response. Since NHIS data files are 93 publicly available and de-identified, this study was exempt from Institutional Review Board 94 review. 95 We used respondents aged ≥ 18 years from the Family Core, Sample Adult Core, and Person files 96 merged and pooled over five years (2013-2017). Since NHIS data files are publicly available and 97 de-identified, this study was exempt from the purview of Houston Methodist’s Institutional 98 Review Board. 99 . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprintthis version posted September 1, 2022. ; https://doi.org/10.1101/2022.08.31.22279420doi: medRxiv preprint 7 Study variables 100 Atherosclerotic Cardiovascular Disease 101 Respondents who affirmed ever receiving a clinician diagnosis of “angina pectoris”, “heart 102 attack (or myocardial infarction)”, “coronary heart disease”, or “stroke” were classified as having 103 ASCVD. Respondents with no information on their ASCVD status (1,052 [0.7%]) were excluded 104 from the analysis . All other NHIS participants included in the analysis were considered to not 105 have ASCVD. 106 The Health and Activity Limitation Index 107 HALex is based on two routinely assessed health indicators included in the NHIS: perceived 108 health status and levels of physical activity limitation (5). Responses to perceived health status 109 are subjective and include “excellent”, “very good”, “good”, “fair”, and “poor”. The assessment 110 of physical activity limitation is objective and the responses include: 1) not limited, 2) limited in 111 other activities, 3) limited in major activity, 4) unable to perform a major activity, 5) unable to 112 perform instrumental activities of daily living, and 6) unable to perform activities of daily living 113 (personal care needs). The mathematical derivation of HALex has been described extensively 114 (10). In brief, responses to the two items are combined in a matrix of 30 health states 115 (Appendix). For persons alive, HALex scores range from 0.10 to 1.00 (11). 116 Two outcomes were assessed – our primary outcome, mean HALex score, and a secondary 117 binary outcome, poor HALex performance. We categorized patients with HALex scores less than 118 0.84 (i.e., 20th percentile of the study sample distribution) as having poor HALex (12). There are 119 no widely established cut-off values for HALex, but the threshold for a clinically relevant impact 120 . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprintthis version posted September 1, 2022. ; https://doi.org/10.1101/2022.08.31.22279420doi: medRxiv preprint 8 is suggested to be 0.03 (13). We interpreted our results with this threshold. In sum, data of 121 155,130 participants with complete information on ASCVD and HALex, and non-zero HALex 122 scores (i.e., alive) were used in this study. 123 Covariates 124 Other variables included in this study were age (linear) and age group (18-64 years and ≥ 65 125 years); sex (male and female); race/ethnicity (non-Hispanic White, non-Hispanic Black, 126 Hispanic, non-Hispanic Asian, Hispanic, and Other); educational attainment (no high school 127 diploma, high school diploma/GED equivalent, some college, and ≥ college degree); health 128 insurance plan (uninsured, any private plan, Medicare, Medicaid, and other plans); household 129 income; obesity (body mass index ≥ 30 kg/m2); psychological distress within 30 days before the 130 survey; and comorbidities, which were all self-reported. Household income categories were 131 based on the percentage of family income relative to the federal poverty limit from the U.S. 132 Census Bureau – high income (≥ 400%), middle income (200% to < 400%), and low income (< 133 200%). Psychological distress was ascertained with the Kessler-6 Distress Scale score (14). For 134 comorbidities, respondents were asked if they had ever received a clinician diagnosis of arthritis, 135 cancer, chronic obstructive pulmonary disease (COPD), diabetes mellitus, hypertension, or 136 failing/weak kidneys. 137 Statistical analysis 138 All statistical analyses incorporated the complex survey design and weighting for selection 139 probabilities and non-response. Variance estimation for the entire pooled cohort was obtained 140 from the Integrated Public Use Microdata Series (https://www.nhis.ipums.org) (15). We assessed 141 . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprintthis version posted September 1, 2022. ; https://doi.org/10.1101/2022.08.31.22279420doi: medRxiv preprint 9 the statistical significance of our estimates with a two-tailed alpha significance level of 5%. We 142 used Stata version 16 software (StataCorp, College Station, Texas) for all analyses. 143 We summarized the distribution of individual characteristics – mean (SD) for age; median (IQR) 144 for Kessler 6 score; sample frequency with weighted proportion for discrete variables) in the 145 total population and by ASCVD status. Chi-squared, t-test, and Mann-Whitney u test statistics 146 were used to compare the descriptive statistics between the ASCVD groups for discrete 147 variables, linear age, and Kessler 6 score, respectively. For HALex, we provided age-and-sex 148 adjusted least-square mean scores (with SE) for the total population and by ASCVD status. 149 For our primary multivariable analysis, we evaluated the associations between ASCVD and 150 HALex scores using two-part models (16). All models used accounted for linear age, sex, 151 race/ethnicity, insurance status, household income level, educational attainment, obesity, Kessler 152 6 score, and the individual comorbidities. For the two-part modeling of HALex scores, we first 153 performed a simple negative linear transformation (X = 1 – U, where U = HALex) to move the 154 utility index to a “health-utility decrement” scale with right skewness (16). Next, we fitted a 155 first-part logit model for the likelihood of positive (vs. zero) scores, and an ordinary least squares 156 regression in the second part for the predicted score conditioned on having a positive score. The 157 same set of covariates were used in both models. Finally, a simple reconversion of the estimates 158 and 95% confidence limits (U = 1 – X) was performed to obtain results on the original HALex 159 scale (16). About 86% of participants had complete data on all variables of interest and the 160 highest missingness was 7.5% for household income levels with missingness for other variables 161 ranging from 0.8 to 3.3%. Consequently, we did not impute values for any missing data. 162 . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprintthis version posted September 1, 2022. ; https://doi.org/10.1101/2022.08.31.22279420doi: medRxiv preprint 10 In subgroup analyses by age group (18-64 years and ≥ 65 years), sex, and race/ethnicity, we 163 repeated the fully adjusted analysis of HALex scores and tested for potential interactions 164 between these characteristics and ASCVD status separately. 165 In a secondary analysis of poor HALex, we used odds ratios (OR, 95% CI) to assess the 166 association between ASCVD and poor HALex in the overall population and in subgroups 167 defined by age, sex, and race/ethnicity. 168 169

170 In our sample of 155,130 NHIS participants with complete HALex data, the estimated 171 prevalence of ASCVD was 6.8%, representing about 15.7 million adults annually (Table 1). 172 Individuals with ASCVD were more likely to be men, older, have Medicare insurance plan, and 173 reside in low-income households. They also reported a greater burden of comorbidities. 174 Summary of HALex scores 175 Overall, adults with ASCVD had lower mean HALex scores (0.67 [SE 0.01]) than those without 176 ASCVD (0.87 [SE 0.00]). This trend was observed across all study characteristics (Table 2). 177 With ASCVD, females averaged a significantly lower HALex score than males (0.62 vs. 0.69), a 178 contrast to the similar scores observed in the absence of ASCVD. With ASCVD, persons aged 179 18-64 years averaged a lower HALex score than those aged ≥ 65 years. The reverse – higher 180 HALex scores among persons 18-64 years of age – was observed in the absence of ASCVD. 181 Non-Hispanic Blacks averaged the lowest HALex scores, irrespective of ASCVD status. Lower 182 educational status, lower household income and higher psychological distress were all associated 183 . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprintthis version posted September 1, 2022. ; https://doi.org/10.1101/2022.08.31.22279420doi: medRxiv preprint 11 with significantly lower HALex scores than the respective counterparts in both ASCVD groups. 184 Of the comorbidities, COPD and kidney failure were associated with the lowest HALex scores 185 reflecting their debilitating nature. 186 Multivariable ASCVD-HALex score analysis 187 We present our multivariable analysis of HALex scores. Overall, ASCVD was associated with a 188 clinically relevant decrease in HALex performance (β = -0.10; 95% CI [-0.10, -0.09]) (Figure 1). 189 We observed, in separate models, significant interactions between ASCVD and sex (p < 0.001), 190 and race/ethnicity (p < 0.001). The decrement in HALex scores associated with ASCVD was 191 greater in females (β = -0.11, 95% CI [-0.12, -0.10]) than in males (β = -0.08, 95% CI [-0.09, -192 0.07]). For age groups, the negative impact of ASCVD on HALex scores was observed to be 193 greater in the younger age group (β = -0.09, 95% CI [-0.10, -0.08]) than those aged ≥ 65 years (β 194 = -0.06, 95% CI [-0.07, -0.06]). For racial/ethnic subgroups, the greatest ASCVD impact was 195 observed with non-Hispanic Blacks (β = -0.13, 95% CI [-0.15, -0.11]). The decrement in HALex 196 associated with ASCVD was not clinically different between non-Hispanic White (β = -0.09, 197 95% CI [-0.10, -0.08]) and Hispanics (β = -0.10, 95% CI [-0.12, -0.08]) groups. 198 Secondary analysis of poor HALex 199 Overall, 17.7% (95% CI [17.4, 18.1]) of the study population performed poorly on HALex. A 200 significantly greater proportion in the ASCVD group had poor HALex than those without 201 ASCVD (40.3% v. 15.3%; p < 0.001; Appendix). The distribution of poor HALex performance 202 across study variables for the ASCVD groups are presented as well in the Appendix. 203 . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprintthis version posted September 1, 2022. ; https://doi.org/10.1101/2022.08.31.22279420doi: medRxiv preprint 12 From the multivariable analysis, ASCVD was associated with increased odds of poor HALex 2.5 204 times that observed with the group without ASCVD (OR = 2.46, 95% CI [2.28, 2.66]). 205 Interactions between ASCVD and age group and ASCVD and race/ethnicity were statistically 206 significant (p < 0.001), but there was no significant interaction observed between ASCVD and 207 sex (p = 0.05). The increment in odds of poor HALex associated with ASCVD was higher in age 208 18-64 years (OR = 2.84; 95% CI [2.52, 3.20]) than in age ≥ 65 years. Similarly, the increment in 209 odds of poor HALex associated with ASCVD was highest in the non-Hispanic Black group (OR 210 = 2.33; 95% CI [2.74, 4.04]). Hispanic and non-Hispanic White groups had similar increments in 211 odds of poor HALex associated with ASCVD presence. 212 213

214 In a large survey of adults in the U.S., reporting any ASCVD condition was associated with a 215 significantly lower HRQoL (as measured with HALex). This clinically significant difference was 216 independent of demography, education, insurance, household income, and comorbidity burden. 217 Further, the impact of ASCVD on lower HALex scores was significantly greater in females, 218 adults 18-64 years of age, and non-Hispanic Black adults. The impacts of ASCVD on HALex 219 scores and poor HALex performance were similar in the Hispanic and non-Hispanic White 220 groups. We did not intend to establish clinical values of HALex for persons with ASCVD. 221 Rather, we set out to corroborate the construct validity of a HRQoL tool that could be 222 implemented in the clinical workflow without much challenge. 223 The sex differences in the impact of ASCVD on HALex is congruent with the established trend 224 that women have poorer health outcomes in conditions like coronary heart disease (17). Women 225 . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprintthis version posted September 1, 2022. ; https://doi.org/10.1101/2022.08.31.22279420doi: medRxiv preprint 13 may present as a diagnostic dilemma in the clinic due to confounding atypical presentations or 226 less common pathophysiological processes. Consequently, they are less likely to receive prompt 227 and/or appropriate management to realize optimal outcomes (18,19). 228 We also observed worse HRQoL and the largest ASCVD decrement in non-Hispanic Black 229 compared to the other groups. A similar trend was previously observed in a national survey of 230 adults 35-89 years of age which assessed HRQoL with various indices (20). Interestingly, the 231 impact of ASCVD on HRQoL in Hispanic was comparable to non-Hispanic White, despite the 232 former averaging lower HALex in the presence of ASCVD. This paradox, commonly observed 233 in population studies on other health outcomes (21), could be related to a more resilient 234 subjective construct of health in the face of health challenges owing to cultural values and 235 supportive social relationships among Hispanics (22). 236 On average, we observed lower HALex scores in those 18-64 years of age with ASCVD than 237 their older counterparts. Though not statistically significant, we also observed a greater 238 decrement in HALex with ASCVD in the group 18-64 years of age (-0.09 vs -0.06). While 239 younger age is associated with better health and a greater chance of recovering from ASCVD 240 and its management, younger individuals are more likely to have maladaptive and adverse 241 psychological experiences after their diagnosis (23). Another potential explanation to this 242 observation is the phenomenon that aging populations may report better subjective health due to 243 lowered expectations rather than ‘actual’ better health (24). 244 Implications 245 The adoption into clinical workflow of a HRQoL tool like HALex – which is easy to administer 246 and has construct validity – would be in line with the call to routinely assess PROMs in the 247 . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprintthis version posted September 1, 2022. ; https://doi.org/10.1101/2022.08.31.22279420doi: medRxiv preprint 14 clinic. HALex is impacted by the burden of chronic disease, more so than other preference-based 248 measures (25,26). This is in part due to the ability of the constituent perceived-health-status item 249 to discriminate well among levels of functioning (27). As physical functioning is strongly 250 associated with cardiovascular disease prognosis, HALex assessment could be relevant in patient 251 management, especially in the holistic evaluation of rehabilitation (28). 252 Additionally, HALex may have utility in longitudinally assessing the impact of treatment on 253 HRQoL (29). Under a chronic care model with self-management support strategy, patients are 254 reinforced to assume an active and shared responsibility of their health with providers. They are 255 educated to better recognize their health issues, acknowledge the need for health behavior 256 modifications, and initiate and maintain such modifications. In utilizing this context of care, 257 perceived health status (and HALex) would be based on a more comprehensive 258 conceptualization of health standards set and evaluated by patients themselves. Consequently, for 259 patients who may not have a change in physical functioning, a change in their valuation of their 260 own health could signal deficiencies in care standards. 261

262 HALex as a HRQoL measure is not without limitations. Firstly, domains of health such as 263 emotional, mental, and social functioning, which are relevant in HRQoL (30), are omitted in 264 HALex derivation. This may its discrimination ability, especially for populations with clustering 265 at the highest level of health (5). Nevertheless, we accounted for psychological distress using a 266 valid and reliable tool in the Kessler 6 distress scale, which minimizes the bias from this health 267 domain on our estimates. 268 . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprintthis version posted September 1, 2022. ; https://doi.org/10.1101/2022.08.31.22279420doi: medRxiv preprint 15 Secondly, the reliance of HALex on a subjectively perceived health status raises the question of 269 how much of the difference in HALex scores may be related to differential perception. Although 270 perceived health status tends to be congruous with objective measures of health, differences exist 271 between the two health assessments owing to varying health expectations (or preoccupations), 272 and the relevance of physical function to major lifetime occupation (31). Nevertheless, it is well 273 known that the manner in which people account for the many dimensions of health when rating 274 their overall health is relatively stable across disease populations. 275 Finally, while the NHIS is designed to be representative of the US population, the reliance on 276 self-reports for health information including medical conditions and anthropometric 277 measurements (i.e., weight and height) implies some potential for misclassification of study 278 variables. However, previous studies have found a high correlation between the self-reported 279 information in NHIS and the verified information found in other national datasets (32). 280 281

282 In conclusion, ASCVD is associated with lower HALex and this association was observed to be 283 greater in female adults, younger age group, and non-Hispanic Black persons. HALex retains a 284 construct validity and has a low administration burden, making it a potentially useful HRQoL 285 tool to implement in healthcare delivery. 286 . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprintthis version posted September 1, 2022. ; https://doi.org/10.1101/2022.08.31.22279420doi: medRxiv preprint 16

287 1. Kornowski R. Patient-reported outcome measures in cardiovascular disease. Eur Heart J - 288 Qual Care Clin Outcomes [Internet]. 2021 Aug 9 [cited 2021 Nov 16];(qcab051). Available 289 from: https://doi.org/10.1093/ehjqcco/qcab051 290 2. Reuben DB, Tinetti ME. Goal-Oriented Patient Care — An Alternative Health Outcomes 291 Paradigm. N Engl J Med. 2012 Mar 1;366(9):777–9. 292 3. Hostetter M, Klein S. Using Patient-Reported Outcomes to Improve Health Care Quality 293 [Internet]. The Commonwealth Fund. [cited 2021 Nov 16]. Available from: 294 https://www.commonwealthfund.org/publications/newsletter-article/using-patient-reported-295 outcomes-improve-health-care-quality 296 4. Blumenthal DM, Strom JB, Valsdottir LR, Howard SE, Wagle NW, Ho KKL, et al. Patient-297 Reported Outcomes in Cardiology. Circ Cardiovasc Qual Outcomes. 2018 298 Nov;11(11):e004794. 299 5. Erickson P. Evaluation of a population-based measure of quality of life: the Health and 300 Activity Limitation Index (HALex). Qual Life Res Int J Qual Life Asp Treat Care Rehabil. 301 1998 Feb;7(2):101–14. 302 6. Hammerschmidt T, Zeitler HP, Gulich M, Leidl R. A comparison of different strategies to 303 collect standard gamble utilities. Med Decis Mak Int J Soc Med Decis Mak. 2004 304 Oct;24(5):493–503. 305 7. Lam CLK, Brazier J, McGhee SM. Valuation of the SF-6D Health States Is Feasible, 306 Acceptable, Reliable, and Valid in a Chinese Population. Value Health J Int Soc 307 Pharmacoeconomics Outcomes Res. 2008 Apr;11(2):295–303. 308 8. Abellán Perpiñán JM, Sánchez Martínez FI, Martínez Pérez JE, Méndez I. Lowering the 309 ‘Floor’ of the Sf-6d Scoring Algorithm Using a Lottery Equivalent Method. Health Econ. 310 2012;21(11):1271–85. 311 9. National Center for Health Statistics. NHIS - Data, Questionnaires and Related 312 Documentation [Internet]. Centers for Disease Control and Prevention. 2021 [cited 2021 Nov 313 17]. Available from: https://www.cdc.gov/nchs/nhis/data-questionnaires-documentation.htm 314 10. Livingston EH, Ko CY. Use of the Health and Activities Limitation Index as a Measure of 315 Quality of Life in Obesity. Obes Res. 2002;10(8):824–32. 316 11. Asada Y. Assessment of the health of Americans: the average health-related quality of life 317 and its inequality across individuals and groups. Popul Health Metr. 2005 Jul 13;3(1):7. 318 12. Kachan D, Fleming LE, Christ S, Muennig P, Prado G, Tannenbaum SL, et al. Health Status 319 of Older US Workers and Nonworkers, National Health Interview Survey, 1997-2011. Prev 320 Chronic Dis. 2015 Sep 24;12:E162. 321 . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprintthis version posted September 1, 2022. ; https://doi.org/10.1101/2022.08.31.22279420doi: medRxiv preprint 17 13. Mouelhi Y, Jouve E, Castelli C, Gentile S. How is the minimal clinically important 322 difference established in health-related quality of life instruments? Review of anchors and 323 methods. Health Qual Life Outcomes. 2020 May 12;18(1):136. 324 14. Kessler RC, Andrews G, Colpe LJ, Hiripi E, Mroczek DK, Normand SLT, et al. Short 325 screening scales to monitor population prevalences and trends in non-specific psychological 326 distress. Psychol Med. 2002 Aug;32(6):959–76. 327 15. Blewett LA, Drew JAR, King ML, Williams KCW. IPUMS Health Surveys: National Health 328 Interview Survey, Version 6.4 [dataset] [Internet]. Minneapolis, MN; 2019. Available from: 329 https://doi.org/10.18128/D070.V6.4 330 16. Wolowacz SE, Briggs A, Belozeroff V, Clarke P, Doward L, Goeree R, et al. Estimating 331 Health-State Utility for Economic Models in Clinical Studies: An ISPOR Good Research 332 Practices Task Force Report. Value Health. 2016 Sep 1;19(6):704–19. 333 17. Gao Z, Chen Z, Sun A, Deng X. Gender differences in cardiovascular disease. Med Nov 334 Technol Devices. 2019 Dec 1;4:100025. 335 18. Gelber A, Drescher M, Shiber S. Sex Differences in Identifying Chest Pain as Being of 336 Cardiac Origin Using the HEART Pathway in the Emergency Department. J Womens Health 337 2002. 2022 Jul;31(7):926–31. 338 19. Zhao M, Woodward M, Vaartjes I, Millett ERC, Klipstein/i3 Grobusch K, Hyun K, et al. Sex 339 Differences in Cardiovascular Medication Prescription in Primary Care: A Systematic 340 Review and Meta/i3 Analysis. J Am Heart Assoc. 2020 Jun 2;9(11):e014742. 341 20. Pereira CCA, Palta M, Mullahy J, Fryback DG. Race and preference-based health-related 342 quality of life measures in the United States. Qual Life Res Int J Qual Life Asp Treat Care 343 Rehabil. 2011 Aug;20(6):969–78. 344 21. Ruiz JM, Steffen P, Smith TB. Hispanic Mortality Paradox: A Systematic Review and Meta-345 Analysis of the Longitudinal Literature. Am J Public Health. 2013 Mar;103(3):e52–60. 346 22. Gallo LC, Penedo FJ, Espinosa de los Monteros K, Arguelles W. Resiliency in the Face of 347 Disadvantage: Do Hispanic Cultural Characteristics Protect Health Outcomes? J Pers. 348 2009;77(6):1707–46. 349 23. Okunrintemi V, Benson EMA, Derbal O, Miedema MD, Blumenthal RS, Tibuakuu M, et al. 350 Age-specific differences in patient reported outcomes among adults with atherosclerotic 351 cardiovascular disease: Medical expenditure panel survey 2006–2015. Am J Prev Cardiol. 352 2020 Sep 1;3:100083. 353 24. Sarkisian CA, Prohaska TR, Wong MD, Hirsch S, Mangione CM. The relationship between 354 expectations for aging and physical activity among older adults. J Gen Intern Med. 2005 355 Oct;20(10):911–5. 356 . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprintthis version posted September 1, 2022. ; https://doi.org/10.1101/2022.08.31.22279420doi: medRxiv preprint 18 25. Hanmer J, Cherepanov D, Palta M, Kaplan RM, Feeny D, Fryback DG. Health Condition 357 Impacts in a Nationally Representative Cross-Sectional Survey Vary Substantially by 358 Preference-Based Health Index. Med Decis Mak Int J Soc Med Decis Mak. 2016 359 Feb;36(2):264–74. 360 26. Stewart ST, Cutler DM, Rosen AB. Comparison of trends in US health-related quality of life 361 over the 2000s using the SF-6D, HALex, EQ-5D, and EQ-5D visual analog scale versus a 362 broader set of symptoms and impairments. Med Care. 2014 Dec;52(12):1010–6. 363 27. Idler EL, Kasl SV. Self-ratings of health: do they also predict change in functional ability? J 364 Gerontol B Psychol Sci Soc Sci. 1995 Nov;50(6):S344-353. 365 28. Shepherd CW, While AE. Cardiac rehabilitation and quality of life: a systematic review. Int 366 J Nurs Stud. 2012 Jun;49(6):755–71. 367 29. Health Services Delivery Programme, Division of Health Systems and Public Health. 368 Integrated care models: an overview [Internet]. World Health Organization: Regional Office 369 for Europe; 2016 Oct p. 1–31. Available from: 370 https://www.euro.who.int/__data/assets/pdf_file/0005/322475/Integrated-care-models-371 overview.pdf 372 30. McDowell I. Measuring Health: A guide to rating scales and questionnaires [Internet]. 3rd 373 ed. New York: Oxford University Press; 2006 [cited 2021 Nov 18]. 764 p. Available from: 374 https://oxford.universitypressscholarship.com/10.1093/acprof:oso/9780195165678.001.0001/375 acprof-9780195165678 376 31. Maddox GL. Some Correlates of Differences in Self-assessment of Health Status Among the 377 Elderly1. J Gerontol. 1962 Apr 1;17(2):180–5. 378 32. Nelson DE, Powell-Griner E, Town M, Kovar MG. A Comparison of National Estimates 379 From the National Health Interview Survey and the Behavioral Risk Factor Surveillance 380 System. Am J Public Health. 2003 Aug;93(8):1335–41. 381 . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprintthis version posted September 1, 2022. ; https://doi.org/10.1101/2022.08.31.22279420doi: medRxiv preprint 19 FIGURE LEGENDS 382 Figure 1. Association between HALex and ASCVD from the National Health Interview Survey 383 2013-2017. 384 [Caption] 1 Adjusted for sex, age (linear), race/ethnicity, insurance status, household income 385 level, educational attainment, obesity, Kessler-6 score, and comorbidities. 386 2 Adjusted for age (linear), race/ethnicity, insurance status, household income level, educational 387 attainment, obesity, Kessler-6 score, and comorbidities. 388 3 Adjusted for sex, race/ethnicity, insurance status, household income level, educational 389 attainment, obesity, Kessler-6 score, and comorbidities. 390 4 Adjusted for sex, age (linear), insurance status, household income level, educational attainment, 391 obesity, psychological distress, and comorbidities. 392 * Other race/ethnicity group not shown due to limited sample size. 393 † p-value for the test of interaction with Bonferroni correction. 394 ‡ Statistically significant. 395 . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprintthis version posted September 1, 2022. ; https://doi.org/10.1101/2022.08.31.22279420doi: medRxiv preprint 20 DECLARATIONS 396 Ethics approval and consent to participate 397 Not applicable. 398 Consent for publication 399 Not applicable. 400 Availability of data and materials 401 The datasets analyzed in the current study are available from the corresponding author on 402 reasonable request. 403 Competing interests 404 K.N. is on the advisory board of Amgen and Novartis, and his research is partly supported by the 405 Jerold B. Katz Academy of Translational Research. K.N. and M.CA. are on the Steering 406 Committee of the PAK-SEHAT Study, partially funded by an unrestricted research grant from 407 Getz Pharma. A.A.H. declares current funding from the United States National Institutes of 408 Health, the World Bank, and the World Health Organization. The other authors report no 409 conflicts of interest relevant to this work. 410 Funding 411 Not applicable 412 Authors’ contributions 413 K.K.H., K.N., and Z.J. contributed to the study's conception and design. Material preparation and 414 data analysis were performed by K.K.H. and J.VE. The first draft of the manuscript was prepared 415 by K.K.H. and Z.J. S.L. prepared all figures. All authors reviewed the manuscript. All authors 416 read and approved the final manuscript. 417

418 . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprintthis version posted September 1, 2022. ; https://doi.org/10.1101/2022.08.31.22279420doi: medRxiv preprint 21 Not applicable. 419 . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprintthis version posted September 1, 2022. ; https://doi.org/10.1101/2022.08.31.22279420doi: medRxiv preprint Table 1. Characteristics of adults ≥ 18 years from the National Health Interview Survey (2013- 2017). Characteristic Total No ASCVD ASCVD Sample N=155,130 142,866 12,264 Weighted population, n (%) 231,560,617 215,812,388 (93.2) 15,748,229 (6.8) Sociodemographic Sex* Male 73,825 (48.3) 65,607 (47.7) 8,218 (56.9) Female 90,525 (51.7) 82,985 (52.3) 7,540 (43.1) Mean age* 49.75 (SD 18.39) 48.04 (SD 17.93) 66.84 (SD 13.63) Age group* 18-64 years 121,956 (78.62) 116,295 (81.40) 5,661 (46.16) ≥ 65 years 33,174 (21.38) 26,571 (18.60) 6,603 (53.84) Race/Ethnicity* Non-Hispanic White 100,223 (65.2) 91,521 (64.6) 8,702 (72.8) Non-Hispanic Black 20,284 (12.1) 18,492 (12.1) 1,792 (12.5) Non-Hispanic Asian 8,880 (5.9) 8,523 (6.1) 357 (3.2) Hispanic 23,759 (15.8) 22,529 (16.3) 1,230 (10.3) Other 1,984 (1.0) 1,801 (1.0) 183 (1.2) Insurance status* Uninsured 18,953 (12.3) 18,277 (12.8) 676 (6.1) Private 78,944 (56.9) 76,628 (59.3) 2,316 (23.7) Medicare 29,864 (115.5) 24,124 (13.4) 5,740 (45.2) Medicaid 17,930 (11.0) 15,513 (10.5) 2,417 (17.7) Other 7,611 (4.3) 6,613 (4.1) 998 (7.4) Family income* High income 51,472 (40.5) 48,691 (41.3) 2,781 (29.8) Middle income 40,789 (28.7) 37,619 (28.6) 3,170 (29.9) Low income 51,546 (30.8) 46,261 (30.1) 5,285 (40.3) Education* ≥ College degree 65,124 (42.9) 61,444 (43.8) 3,680 (32.5) Some college 30,612 (19.7) 28,374 (19.8) 2,238 (18.0) High school/GED 38,533 (25.0) 34,867 (24.6) 3,666 (30.1) < High school 20,251 (12.4) 17,648 (11.9) 2,603 (19.4) Clinical characteristics Median Kessler 6 score* (psychological stress) 1 (0-4) 1 (0-4) 2 (0-6) Arthritis* 37,510 (21.5) 31,133 (19.4) 6,377 (50.0) Cancer* 14,195 (8.1) 11,618 (7.2) 2,577 (20.4) . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprintthis version posted September 1, 2022. ; https://doi.org/10.1101/2022.08.31.22279420doi: medRxiv preprint COPD* 5,156 (2.7) 3,327 (1.9) 1,829 (13.6) Diabetes* 14,910 (8.8) 11,353 (7.3) 3,557 (29.6) Failing/weak kidneys* 3,172 (1.7) 2,052 (1.2) 1,120 (8.6) Obesity* 49,863 (31.9) 45,059 (31.3) 4,804 (40.1) Hypertension* 49,960 (29.1) 40,998 (26.0) 8,962 (71.7) Column % weighted to the US population presented. * All tests of comparison by ASCVD status were statistically significant (p < 0.001). Abbreviations: ASCVD – atherosclerotic cardiovascular disease. COPD – chronic obstructive pulmonary disease. GED – general educational development. . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprintthis version posted September 1, 2022. ; https://doi.org/10.1101/2022.08.31.22279420doi: medRxiv preprint Table 2. Age- and sex-adjusted mean HALex scores of adults ≥ 18 years from the National Health Interview Survey (2013-2017). Characteristics Total ASCVD No ASCVD Overall 0.85 (0.00) 0.67 (0.01) 0.87 (0.00) Sociodemographic Sex Male 0.86 (0.00) 0.69 (0.00) 0.87 (0.00) Female 0.84 (0.00) 0.62 (0.01) 0.86 (0.00) Age group 18-64 years 0.85 (0.00) 0.61 (0.01) 0.87 (0.00) ≥ 65 years 0.81 (0.00) 0.64 (0.01) 0.82 (0.00) Race/ethnicity Non-Hispanic White 0.86 (0.00) 0.68 (0.01) 0.87 (0.00) Non-Hispanic Black 0.81 (0.00) 0.58 (0.01) 0.83 (0.00) Non-Hispanic Asian 0.87 (0.00) 0.71 (0.02) 0.89 (0.00) Hispanic 0.83 (0.00) 0.63 (0.01) 0.85 (0.02) Other 0.78 (0.01) 0.60 (0.03) 0.80 (0.01) Insurance status Uninsured 0.82 (0.00) 0.66 (0.01) 0.84 (0.00) Private 0.89 (0.00) 0.77 (0.01) 0.90 (0.00) Medicare 0.67 (0.00) 0.51 (0.00) 0.68 (0.00) Medicaid 0.68 (0.00) 0.49 (0.01) 0.70 (0.00) Other 0.81 (0.00) 0.65 (0.01) 0.83 (0.00) Education ≥ College degree 0.89 (0.00) 0.73 (0.01) 0.90 (0.00) Some college 0.84 (0.00) 0.67 (0.01) 0.86 (0.00) High school/GED 0.82 (0.00) 0.64 (0.01) 0.84 (0.00) < High school 0.77 (0.00) 0.58 (0.01) 0.78 (0.00) Household income High income 0.90 (0.00) 0.78 (0.01) 0.91 (0.00) Middle income 0.85 (0.00) 0.70 (0.01) 0.87 (0.00) Low income 0.76 (0.00) 0.57 (0.01) 0.79 (0.00) Comorbidities Arthritis 0.75 (0.00) 0.57 (0.01) 0.77 (0.00) Cancer 0.78 (0.00) 0.60 (0.01) 0.80 (0.00) COPD 0.56 (0.01) 0.44 (0.01) 0.60 (0.01) Diabetes 0.70 (0.00) 0.53 (0.01) 0.72 (0.00) Hypertension 0.78 (0.00) 0.61 (0.01) 0.80 (0.00) Kidney failure 0.58 (0.01) 0.45 (0.01) 0.63 (0.01) Obesity 0.80 (0.00) 0.62 (0.01) 0.82 (0.00) Abbreviations: ASCVD – atherosclerotic cardiovascular disease. COPD – chronic obstructive pulmonary disease. GED – general educational development. . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprintthis version posted September 1, 2022. ; https://doi.org/10.1101/2022.08.31.22279420doi: medRxiv preprint . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprintthis version posted September 1, 2022. ; https://doi.org/10.1101/2022.08.31.22279420doi: medRxiv preprint