Association between Breakfast Frequency and Atherosclerotic Cardiovascular Disease Risk: A Cross-Sectional Study of KNHANES Data, 2014–2016

The burden of cardiovascular disease (CVD) is increasing worldwide and one related lifestyle choice is breakfast consumption. This study examined the association between breakfast frequency and the 10-year risk of atherosclerotic CVD (ASCVD). The 10-year risk of ASCVD was defined as the risk of the first event of nonfatal myocardial infarction, coronary heart disease death, and nonfatal or fatal stroke within 10 years. Data from the Korean National Health and Nutrition Examination Survey, 2014–2016 were analyzed, and 7212 participants aged 40–79 years with no history of CVD were included. ASCVD risk was calculated according to the pooled cohort ASCVD equation, and participants with a score >7.5% were considered at high risk of ASCVD. The association between breakfast frequency and high ASCVD risk was confirmed by logistic regression analysis. Participants who never ate breakfast were more likely to be in the high-risk group compared to participants who ate breakfast >5 times per week (OR (adjusted odds ratio) = 1.46; 95% CI (confidence interval) = 1.12–1.89), and the risk was especially higher in female participants and those with a family history of CVD. Our study confirms that breakfast consumption even once per week may prevent CVD.


Introduction
Cardiovascular disease (CVD), including heart attack and stroke, has a high mortality rate worldwide [1]. In South Korea, an estimated 53,597 deaths occurred from CVD in 2017, making it the second most common cause of mortality, accounting for 19% of all deaths in the country [2]. Additionally, in 2016, the medical cost of CVD was reported to be about 8 trillion Korean won (approximately 7 billion US dollars) [3]. Obesity, hypertension, diabetes, and dyslipidemia are the main causes of CVD [4] and the increasing prevalence of these diseases in adults in modern society raises valid concerns regarding a drastic increase in the incidence of CVD in the future. However, many of the diseases that lead to CVD are lifestyle-related and can be prevented.
Accordingly, various methods for identifying groups at high risk of CVD have been proposed, with the aim of implementing lifestyle interventions and to reduce the number of CVD-related deaths [5]. In 2013, the American College of Cardiology/American Heart Association (ACC/AHA) Guidelines for Cardiovascular Risk Assessment proposed the concept of the "pooled cohort atherosclerotic anonymized and unidentified before analysis. Therefore, this study did not require informed consent from the respondents.

Dependent Variable
The main outcome for this study was the assessment of ASCVD risk. The 10-year risk of ASCVD was defined as the risk of the first event of nonfatal myocardial infarction, coronary heart disease death, and nonfatal or fatal stroke within 10 years [6]. The risk of CVD was assessed by applying the pooled cohort risk assessment equation from the 2013 ACC/AHA guidelines [6]. The 10-year ASCVD risk score considers factors of age; sex; ethnicity; high-density lipoprotein cholesterol (HDL) and total cholesterol (TC) levels; systolic blood pressure (SBP; whether hypertension is treated or not); the prevalence of diabetes; and smoking status. Since it is appropriate for Koreans to be considered as Caucasians, regression coefficients for the white race were applied to the equation in this study. Participants with a risk score of >7.5% were regarded as at high risk because this score is used to indicate whether a prescription of statins should be considered [6]. In order to identify absolute risk, the dependent variable was designed in the form of a binary division with a cut-off of 7.5% (i.e., high-risk group: ≥7.5%, normal group: <7.5%).

Independent Variable
The independent variable in this study was the frequency at which breakfast was eaten per week. The questionnaire about the frequency of meals per day in the KNHANES is divided into breakfast, lunch, and dinner, which could be defined according to when one eats. Breakfast can be defined as a meal taken in the morning [24]. The following KNHANES questionnaire item was used to determine the independent variable, "How many times per week did you have breakfast in the last year?" There are four possible responses to the question: 5-7 times per week, 3-4 times per week, 1-2 times per week, almost none (0 times per week). Participants were categorized according to their responses to this questionnaire item.

Control Variables
Covariates were age, sex, socioeconomic status (SES), health-related factors, and survey year. SES included marital status, education level, household income level, occupation (white collar, pink collar, blue collar, and others), and rural/urban region. Health-related factors were alcohol status, perceived stress level, perceived health status, physical activity, body mass index (BMI), nutrition status, total energy, fat and carbohydrate intake per day, and family history of CVD. Physical activity was categorized into 3 groups, considering the amount and intensity of physical activity according to the metabolic equivalent task (MET). MET was calculated by considering strength, time, and the number of days of physical activity (MET range <600, low; 600-2999, moderate; ≥3000, high) [25]. Family history was categorized as 'Yes' if the respondent indicated they had a family history of CVD (including stroke, myocardial infarction, or angina) and 'No' if they did not. Nutrition status was categorized by mean nutrient adequacy ratio [MAR (9)]. MAR(9) calculates the nutrient adequacy ratio (NAR) of nine nutrients (protein, calcium, phosphorus, iron, vitamin A, vitamin B1, vitamin B2, niacin, and vitamin C), adds them all together, and divides by nine [26]. A MAR(9) score of >0.75 was categorized as good and a score <0.75 was categorized as poor [27].

Statistical Analysis
The chi-square test, t-test, and logistic regression were used for the analysis of the data. The chi-square test was used to examine the significance of differences in the ASCVD risk by the frequency of breakfast after adjusting for covariates. p-values <0.05 were considered significant. The association and significance of differences between ASCVD risk (dependent variable) and the number of breakfasts per week (independent variable) were analyzed with logistic regression analysis for adjusted odds ratios (ORs) and 95% confidence intervals (CIs). Subgroup analysis was performed by sex, MET, and family history. In addition, we tested for multicollinearity in this statistical model. The variance inflation factors (VIFs) were all less than 10, indicating no excessive correlation between independent variables. All statistical analyses were performed using SAS software, version 9.4 (SAS Institute Inc., Cary, NC, USA).

Study Population
The general characteristics of the study population are provided in Table 1. The total number of participants was 7205, of which 2786 were considered ASCVD high-risk, and 4419 were considered ASCVD normal. In both the high-risk and normal groups, the frequency of breakfast consumption was 5-7 per week for most participants. With the exception of survey year (p = 0.7889), the differences in all other variables between ASCVD risk groups were statistically significant. In the ASCVD high-risk group, participants were most likely to eat breakfast 5-7 times per week (43.2%; n = 2414), followed by 0 times per week (28.5%; n = 169).  Table 2 shows the association between variables and ASCVD high-risk. Logistic regression analyses adjusted for covariates revealed that breakfast frequency was related to ASCVD risk. Individuals who did not eat breakfast (0 times per week) were significantly more likely to belong to the high-risk group than those who ate breakfast 5-7 times per week (reference group) (OR = 1.46; 95% CI = 1.12-1.89). The risk for women was lower than that for men (OR = 0.06; 95% CI = 0.05-0.07). A family history of CVD had no statistical impact on ASCVD risk (OR = 0.88; 95% CI = 0.75-1.03). 1.00 -* OR: adjusted odds ratio. All covariates were adjusted; ASCVD risk: 10-year ASCVD risk was measured using the pooled cohort equation; 10-year risk <7.5%: normal group, ≥7.5%: high-risk group; ASCVD: atherosclerotic cardiovascular disease, adjusted OR: odds ratio, 95% CI: confidence interval; BMI: body mass index. Figure 1 shows the results of subgroup analyses regarding the impact of sex and family history on ASCVD high-risk according to the number of breakfasts consumed per week. Both men and women who did not eat breakfast were more likely to be ASCVD high-risk compared to the 5-7 times reference group (men: OR = 1.47, 95% CI = 1.04-2.09; women: OR = 1.55, 95% CI = 1.03-2.34), with women having a higher likelihood of being in the high-risk group compared to men. Compared to individuals without a family history of CVD, those who reported a family history of CVD tended to have higher ORs for high ASCVD risk when the frequency of breakfast consumption was ≤4 times per week; in particular, individuals who did not eat breakfast at all were significantly more likely to be at high risk (OR = 2.10; 95% CI = 1.26-3.50). reference group (men: OR = 1.47, 95% CI = 1.04-2.09; women: OR = 1.55, 95% CI = 1.03-2.34), with women having a higher likelihood of being in the high-risk group compared to men. Compared to

Discussion
This study examined the association between the number of times breakfast was eaten per week and the risk of ASCVD incidence in 10 years determined through the pooled cohort ASCVD equation, using nationally representative KNHANES data. Our results indicate that individuals who regularly fail to eat breakfast are statistically significantly more likely to belong to the high-risk group for ASCVD than those who eat breakfast more than 5 times per week. This result was similar for both sexes. Additionally, women who involve in intense physical activity or those with a family history of CVD were more likely to be at high-risk if they consistently skip breakfast.
A previous cross-sectional study on the relationship between the breakfast frequency and CVD incidence for middle-aged Korean men divided breakfast frequency into 3 categories: hardly ever (0-2 times per week), sometimes (3-5 times per week), and always (6-7 times per week). The study found that the "hardly ever" group had significantly higher triglyceride levels [22]. Another similar study using Japanese cohort data divided breakfast frequency into 4 categories (0-2, 3-4, 5-6, and 7 times per week) and showed that the incidence of CVD was significantly higher in groups that ate breakfast 0-2 times per week compared to groups that ate breakfast every day [23]. Together, these studies indicate that people who rarely eat breakfast (mainly less than 3 times per week) have a higher risk of CVD than those who regularly eat breakfast.
Our study found that women had a higher CVD risk than men when the frequency of breakfast consumption was ≤2 times per week. Many previous studies have reported that women who do not eat breakfast have health problems linked to CVD-related factors such as high TC or low-density lipoprotein-cholesterol, elevated blood pressure, low cortisol, and obesity [28][29][30][31][32][33]. In addition, some studies have reported more relevance to obesity when skipping breakfast in women than men [28][29][30].
A previous study has underscored the importance of preventing and managing CVD in women for the following reasons: lower attention to CVD due to past perceptions that CVD is a disease predominantly prevalent among males, and higher obesity rates in women than in men [34]. Our study supports the importance of CVD management in women.
Our study results support and extend findings from previous studies by indicating that eating breakfast once or twice per week can help prevent CVD compared to not eating breakfast at all. Breakfast is eaten at the beginning of the day and a healthy day that begins with breakfast may improve an individuals' overall healthy lifestyle. A previous study suggested that skipping breakfast can have an impact on other meals during the day [35] and can reflect an individual's general healthy behavior [10,11]. The present study is meaningful in terms of public health because it considers the preventative aspects of CVD, which has a high disease burden in modern societies. Moreover, it clarifies the association between CVD and breakfast frequency, a lifestyle habit that can be modified for disease prevention. The strengths of the study included the fact that it used a large, national data set compiled by national institutions, increasing the reliability of the original data. In addition, the study excluded individuals previously diagnosed with cardiovascular-related diseases to minimize the weaknesses of a cross-sectional study design.
However, several limitations should be taken into account when interpreting the results of this study. First, because this was a cross-sectional study, the results cannot identify a definite causal relationship; because breakfast is not an intervention, the relatively low risk of CVD in individuals who eat breakfast more often may be related to their generally healthier lifestyles. Second, despite the efforts of the surveying agency to reduce bias, the original data we analyzed may have been affected by response bias. In addition, the KNHANES data regarding the frequency of meals were obtained based on the timing of the meal (i.e., breakfast, lunch, and dinner) so the frequency of meals could be checked according to the meal constitution. However, it was not possible to determine the items consumed for breakfast, since nutritional data were only given for total daily intake and the corresponding nutritional content. Therefore, we considered the total intake of nutritional factors per day as control variables. Future studies need to identify the exact amount of calories and nutrients intakes for breakfast. Third, a 10-yr ASCVD risk assessment model (pooled cohort ASCVD equation) from the ACC/AHA guideline was used to estimate the risk of ASCVD in the study sample. The guideline is suitable for non-Hispanic Caucasians or African-Americans, the risk may be over-estimated for Asians. However, since this is a problem that is based on patients diagnosed with ASCVD, its use can also be tolerated in Asians from a preventive perspective [36]. In addition, an appropriate model for Asians has not been developed. Thus in many Asian countries, including South Korea, a 10-yr ASCVD risk assessment model has been used [37][38][39][40][41][42][43]. Finally, because the pooling cohort ASCVD equation is only applicable to individuals aged 40-79 years, this study could not assess the CVD risk for other age groups.

Conclusions
The results of this study suggest that daily breakfast skipping is associated with an increased risk of CVD, and indicates that eating breakfast even once per week may prevent CVD. In particular, the results highlight the role of a family history of CVD in CVD risk in women. Furthermore, the results emphasize the importance of eating breakfast and suggests a simple lifestyle change of eating breakfast at least once per weak that can reduce the incidence of CVD and prevent ASCVD. Funding: This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.