We analyzed publicly available, de-identified data from the sample adult core questionnaire of the 2018 National Health Interview Survey (NHIS). The NHIS is an annual, cross-sectional, in-person household interview survey of US noninstitutionalized civilians in all 50 states and the District of Columbia. The NHIS is among the primary data collection programs of the Centers for Disease Control and Prevention's (CDC's) National Center for Health Statistics and is a principal source of information on health outcomes, risk factors, and behaviors in the US. The NHIS uses a complex probability sampling strategy to select households and individuals, and estimates are weighted to represent the US adult civilian population. Respondents provided oral consent before participation, and the survey was approved by CDC's Research Ethics Review Board and the US Office of Management and Budget.
The sample adult component contained data for 25,417 respondents aged 18 years or older and had an unconditional final response rate of 53.1% in 2018. We excluded pregnant people and those missing data on self-reported CVD, CVD risk factors, and VI (n = 2,527), yielding a final analytic sample of 22,890 adults.
Our exposure was self-reported VI and was characterized as an affirmative response to the question: "Do you have difficulty seeing, even when wearing glasses?" The outcomes we investigated were self-reported CVD and 7 CVD risk factors. Self-reported CVD was ascertained by asking whether the respondent had ever been told by a doctor or other health professional that they had any of the following conditions: coronary heart disease, angina/angina pectoris, heart attack/myocardial infarction, stroke, or any kind of heart condition or heart disease. Using AHA's LS7 as a framework, we selected 7 self-reported CVD risk factors from the NHIS to examine cardiovascular health: current smoking, physical inactivity, excessive alcohol intake, obesity, hypertension, high cholesterol, and diabetes. Because NHIS does not regularly collect dietary data as part of its core survey content, dietary data were not collected in 2018 and could not be used; because the consumption of alcohol has complex effects on cardiovascular health, we included excessive alcohol intake in place of poor diet as a CVD risk factor. The self-reported CVD risk factors were separated into 2 categories: 1) risk behaviors: current smoking, physical inactivity, and excessive alcohol intake; and 2) health conditions: obesity, hypertension, high cholesterol, diabetes. The 3 risk behaviors were characterized as: current smoker (defined as those who had smoked more than 100 cigarettes in their lifetime and now smoke every day or some days), physical inactivity (defined as performing <10 minutes per week of light, moderate, or vigorous leisure-time physical activities), excessive alcohol intake (defined as consuming ≥12 drinks in their lifetime and >14 drinks/week in past year [for men] or >7 drinks/week in past year [for women]). Alcohol intake for the full adult sample was used for analyses; however, in the US the Minimum Legal Drinking Age (MLDA) has been 21 years since 1984. The 4 health conditions were obesity (body mass index >30 kg/m2, calculated using self-reported height and weight) and self-reported hypertension, high cholesterol, and diabetes, which were defined as an affirmative response to the question of whether the respondent had ever been told by a doctor or other health professional that they had hypertension or high blood pressure, high cholesterol, or diabetes or sugar diabetes, respectively. The NHIS does not directly measure blood pressure or collect biospecimens, so self-reported factors were used as proxy assessments.
Sociodemographic characteristics were age, sex, race and ethnicity (non-Hispanic Black, Hispanic, non-Hispanic White, and other racial/ethnic groups), education (less than high school, high school/general educational development, more than high school), marital status (married/domestic partnership, not married [including widowed, divorced, separated, or never married]), employment status (work for pay at job/business, not working for pay), health insurance (public, private, both, none), and family income-to-poverty threshold ratio (<1, 1 to <2, ≥2) based on the US Census Bureau federal poverty thresholds (https://www.census.gov/data/tables/time-series/demo/income-poverty/historical-poverty-thresholds.html).
Descriptive characteristics of the study population were tabulated, stratified by VI status. We used separate logistic regression models to generate adjusted prevalence ratios (aPRs) for those with VI (reference: no VI) for CVD and the 7 CVD risk factors. Models for each outcome controlled for age (as a continuous variable), sex, race and ethnicity, education level, marital status, employment status, income-to-poverty ratio, and health insurance. We examined the effect modification of the relationship between VI, CVD, and the 7 CVD risk factors by calculating the aPR for each age group (18–44 y, 45–64 y, ≥65 y) derived from a model that included an interaction term between VI and age group. We used χ 2 tests to examine whether the prevalence of VI varied by sociodemographic characteristics (differences considered significant at P < .05). We also determined the distribution of respondents by the number of CVD risk factors and VI status. All analyses accounted for complex survey design and sampling weights. Weighted analyses were performed using STATA version 16 (StataCorp LLC).
Prev Chronic Dis. 2022;19(7):e43 © 2022 Centers for Disease Control and Prevention (CDC)