Advertisement

Meal Skipping and Shorter Meal Intervals Are Associated with Increased Risk of All-Cause and Cardiovascular Disease Mortality among US Adults

Published:August 10, 2022DOI:https://doi.org/10.1016/j.jand.2022.08.119

      Abstract

      Background

      Previous dietary studies and current dietary guidelines have mainly focused on dietary intake and food patterns. Little is known about the association between eating behaviors such as meal frequency, skipping and intervals, and mortality.

      Objective

      The objective was to examine the associations of meal frequency, skipping, and intervals with all-cause and cardiovascular disease (CVD) mortality.

      Design

      This was a prospective study.

      Participants/setting

      A total of 24,011 adults (aged ≥40 years) who participated in the National Health and Nutrition Examination Survey 1999-2014 were included in this study. Eating behaviors were assessed using 24-hour recall. Death and underlying causes of death were ascertained by linkage to death records through December 31, 2015.

      Main outcome measures

      The outcomes were all-cause and CVD mortality.

      Statistical analyses performed

      Multivariable Cox proportional hazards models were used to estimate adjusted hazard ratios (HRs) of all-cause and CVD mortality.

      Results

      During 185,398 person-years of follow-up period, 4,175 deaths occurred, including 878 cardiovascular deaths. Most participants ate three meals per day. Compared with participants eating three meals per day, the multivariable-adjusted HRs for participants eating one meal per day were 1.30 (95% CI 1.03 to 1.64) for all-cause mortality, and 1.83 (95% CI 1.26 to 2.65) for CVD mortality. Participants who skipped breakfast have multivariable-adjusted HRs 1.40 (95% CI 1.09 to 1.78) for CVD mortality compared with those who did not. The multivariable-adjusted HRs for all-cause mortality were 1.12 (95% CI 1.01 to 1.24) for skipping lunch and 1.16 (95% CI 1.02 to 1.32) for skipping dinner compared with those who did not. Among participants eating three meals per day, the multivariable-adjusted HR for participants with an average interval of ≤4.5 hours in two adjacent meals was 1.17 (95% CI 1.04 to 1.32) for all-cause mortality, comparing with those having a meal interval of 4.6 to 5.5 hours.

      Conclusions

      In this large, prospective study of US adults aged 40 years or older, eating one meal per day was associated with an increased risk of all-cause and CVD mortality. Skipping breakfast was associated with increased risk of CVD mortality, whereas skipping lunch or dinner was associated with increased risk of all-cause mortality. Among participant with three meals per day, a meal interval of ≤4.5 hours in two adjacent meals was associated with higher all-cause mortality.

      Keywords

      Research Question: Are eating behaviors such as meal frequency, skipping, and intervals associated with all-cause and cardiovascular disease (CVD) mortality?
      Key Findings: Eating one meal per day was associated with an increased risk of all-cause and CVD mortality. Skipping breakfast was associated with higher risk of CVD mortality, whereas skipping lunch or dinner was associated with higher risk of all-cause mortality. In addition, among participants with three meals per day, a shorter meal interval of ≤4.5 hours in two adjacent meals was associated with higher risk of all-cause mortality.
      Currently, the influence of meal frequency, skipping, and intervals on health and disease is a topic of universal interest.
      • St-Onge M.P.
      • Ard J.
      • Baskin M.L.
      • et al.
      Meal timing and frequency: implications for cardiovascular disease prevention: a Scientific Statement From the American Heart Association.
      According to the US National Health and Nutrition Examination Survey (NHANES) in 2009-2010,
      • Kant A.K.
      • Graubard B.I.
      40-year trends in meal and snack eating behaviors of American adults.
      the proportion of adults who reported consuming all three standard meals (breakfast, lunch, and dinner) was 59% in men and 63% in women. The prevalence of skipping breakfast, lunch, and dinner was 19%, 22%, and 8% for men, and 19%, 21%, and 7% for women. This indicated that about 40% of Americans skipped meals, and at least one in five skipped breakfast or lunch among populations aged 20 to 74 years. As for meal intervals, the average meal interval between breakfast and lunch was about 5 hours, and the average interval between lunch and dinner was about 6 hours.
      • Kant A.K.
      • Graubard B.I.
      40-year trends in meal and snack eating behaviors of American adults.
      When compared with 1971-1974, the length of interval between the first and the last meal had decreased by about 30 minutes.
      • Kant A.K.
      • Graubard B.I.
      40-year trends in meal and snack eating behaviors of American adults.
      However, the meal frequency, skipping, and intervals have not been addressed by the 2020-2025 Dietary Guidelines for Americans
      US Depts of Agriculture and Health and Human Services
      Dietary Guidelines for Americans, 2020-2025.
      because the Dietary Guidelines Advisory Committee “was unable to find sufficient evidence on which to summarize the evidence between frequency of eating and health.”
      Dietary Guidelines Advisory Committee
      Scientific Report of the 2020 Dietary Guidelines Advisory Committee: Advisory Report to the Secretary of Agriculture and the Secretary of Health and Human Services.
      Prior observational and interventional evidence on the association between meal frequency and risk of obesity, diabetes, and cardiovascular disease (CVD) has been mixed and inconsistent.
      • van der Heijden A.A.
      • Hu F.B.
      • Rimm E.B.
      • van Dam R.M.
      A prospective study of breakfast consumption and weight gain among U.S. men.
      • Mekary R.A.
      • Giovannucci E.
      • Willett W.C.
      • van Dam R.M.
      • Hu F.B.
      Eating patterns and type 2 diabetes risk in men: breakfast omission, eating frequency, and snacking.
      • Ma Y.S.
      • Bertone E.R.
      • Stanek E.J.
      • et al.
      Association between eating patterns and obesity in a free-living US adult population.
      • Cahill L.E.
      • Chiuve S.E.
      • Mekary R.A.
      • et al.
      Prospective study of breakfast eating and incident coronary heart disease in a cohort of male US health professionals.
      Skipping breakfast has been associated with higher CVD risk and mortality in previous studies.
      • Cahill L.E.
      • Chiuve S.E.
      • Mekary R.A.
      • et al.
      Prospective study of breakfast eating and incident coronary heart disease in a cohort of male US health professionals.
      • Rong S.
      • Snetselaar L.G.
      • Xu G.F.
      • et al.
      Association of skipping breakfast with cardiovascular and all-cause mortality.
      • Kubota Y.
      • Iso H.
      • Sawada N.
      • Tsugane S.
      • Group J.S.
      Association of breakfast intake with incident stroke and coronary heart disease: the Japan Public Health Center-Based Study.
      However, the association of meal frequency, skipping lunch or dinner, and meal intervals with mortality remains unknown. In this study, data from a nationally representative cohort was used to examine the associations of specific eating behaviors (meal frequency, skipping, and intervals) with all-cause and CVD mortality among US adults.

      Materials And Methods

      Study Population

      The study population consisted of participants from the NHANES 1999-2014. Briefly, NHANES is a large-scale, ongoing, nationally representative health survey of the noninstitutionalized US population. It is conducted by the National Center for Health Statistics (NCHS) of the Centers for Disease Control and Prevention. Since 1999-2000, NHANES surveys have been organized in 2-year cycles; each cycle consists of approximately 10,000 participants.
      • Ahluwalia N.
      • Dwyer J.
      • Terry A.
      • Moshfegh A.
      • Johnson C.
      Update on NHANES dietary data: focus on collection, release, analytical considerations, and uses to inform public policy.
      Through in-house and telephone interviews, NHANES collects a wide range of health-related data to assess diet, nutritional status, general health, disease history, and health behaviors.
      • Ahluwalia N.
      • Dwyer J.
      • Terry A.
      • Moshfegh A.
      • Johnson C.
      Update on NHANES dietary data: focus on collection, release, analytical considerations, and uses to inform public policy.
      The surveys use multistage, probability clusters to develop a population sample that is nationally representative of the United States based on age, gender, and race and ethnicity. NHANES data along with documents on the survey methods and other information are publicly available on the NHANES website.
      Centers for Disease Control and Prevention, National Center for Health Statistics
      National Health and Nutrition Examination Survey.
      All participants gave written informed consent. The study protocol was approved by the NCHS Research Ethics Review Board.
      For this analysis included 27,759 adult participants aged 40 years or older who had complete information on meal frequency, skipping, and intervals, and had linked mortality information, including underlying causes of death. After further exclusion of 534 women who are currently pregnant or breastfeeding, 3,136 participants with implausible total energy intake (<600 or >5000 kcal/day), and 78 participants who reported no meal consumption, 24,011 participants were left in the final analysis.

      Meal Frequency and Intervals

      In all surveys, a trained dietary interviewer used a standardized protocol to administer a 24-hour recall of foods and beverages. One in-person dietary recall was collected in the 1999-2000, and 2001-2002 surveys. For the later six surveys in 2003-2014, a second recall was collected, via telephone, 3 to 10 days after the mobile exam center visit. This study used the first recall for all surveys. During the 24-hour recall interview, participants were asked to report the clock time when a food or beverage was consumed. For every food or beverage item recalled, participants also chose a name of the eating event from a list. All items reported in one event were given the same clock time and event name in the recall. Participants could choose from “breakfast,” “brunch,” “lunch,” “dinner,” “supper,” “snack,” “drink,” or “extended consumption” or their equivalents in Spanish.
      Reports of breakfast, lunch, and dinner or their equivalents in Spanish were considered as meals in this study. Breakfast consumption was defined based on self-reported consumption of morning meal, including “breakfast” and the Spanish equivalents. Lunch consumption was defined based on self-reported consumption of noon meal, including “lunch,” “brunch,” and the Spanish equivalents. Dinner consumption was defined based on self-reported consumption of afternoon meal, including “dinner,” “supper,” and the Spanish equivalents. All other eating events were considered as snack. All foods and beverages reported at one discrete clock time were considered as part of one eating event. Reports of two similarly named meals (eg, two breakfasts) but at different clock times were considered as two meals. Meal frequency was calculated as the sum of frequency of breakfast, lunch, and dinner consumption, and further categorized as one, two, three and four or more meals per day. Skipping breakfast, lunch, or dinner was defined as no reporting of breakfast, lunch, or dinner consumption during the 24-hour recall. Average intervals between meals were calculated as interval between the reported clock time (in hours) of the first and the last meal divided by (meal frequency – 1). It was further categorized as ≤4.5, 4.6 to 5.5, and >5.5 hours, based on the average time required for complete emptying of the stomach.
      • Szarka L.A.
      • Camilleri M.
      Gastric emptying.

      Ascertainment of Mortality Outcomes

      Mortality status for each participant was determined using the NHANES Public-Use Linked Mortality File, which was created by the NCHS to permit a longitudinal study of participants in NHANES.
      Centers for Disease Control and Prevention, National Center for Health Statistics
      Office of Analysis and Epidemiology. Public-use linked mortality file.
      This file contains information based upon the results from a probabilistic match between NHANES and the National Death Index records to ascertain the vital status of each eligible NHANES subject through December 31, 2015. Underlying causes of death were classified according to the codes of the International Classification of Diseases 10th Revision.
      Centers for Disease Control and Prevention, National Center for Health Statistics
      International Classification of Diseases, Tenth Revision (ICD-10).
      The NCHS classified mortality from heart diseases, including acute rheumatic fever and chronic rheumatic heart diseases (codes I00-I09), hypertensive heart disease (codes I11), hypertensive heart and renal disease (codes I13), ischemic heart diseases (codes I20-I25) and other heart diseases (codes I26-I51), and mortality from cerebrovascular disease (ie, stroke) (codes I60-I69) according to International Classification of Diseases 10th Revision. Deaths from CVD were defined as death from either heart disease or cerebrovascular disease. Persons who survived were administratively censored on December 31, 2015. Follow-up time for each person was calculated as the difference between the NHANES survey date and the last known date alive or censored from the NHANES mortality study.

      Covariate Assessment

      Information on age, gender, race and ethnicity, education, annual household income, smoking status, and physical activity was collected during the in-house interviews which were administered by trained interviewers.
      Race and ethnicity was categorized as non-Hispanic White, non-Hispanic Black, Hispanic (Mexican and non-Mexican Hispanic), and other race and ethnicity. Educational attainment was grouped as less than high school diploma, high school graduate/GED or equivalent, and college graduate or above. Family income-to-poverty ratios were categorized as <1.0, 1.0 to 1.99, 2.0 to 3.99, and ≥4.0 based on the federal poverty level for the survey year.
      • Johnson C.L.
      • Paulose-Ram R.
      • Ogden C.L.
      • et al.
      National Health and Nutrition Examination Survey: analytic guidelines, 1999-2010.
      Individuals who smoked <100 cigarettes in their lifetime were defined as never smokers; those who had smoked >100 cigarettes but did not smoke at the time of survey were considered former smokers; those who had smoked >100 cigarettes and smoked cigarettes at the time of survey were current smokers.
      Centers for Disease Control and Prevention, National Center for Health Statistics
      Adult tobacco use information glossary.
      Physical activity was assessed using the Global Physical Activity Questionnaire, and was classified into three groups (<600, 600 to 1,199, and ≥1,200 metabolic equivalent of task minutes per week).
      World Health Organization
      Global recommendations on physical activity for health.
      Dietary intake was assessed through 24-hour dietary recalls. Total energy intake was calculated using the Food and Nutrient Database for Dietary Studies and the Food Patterns Equivalents Database.
      US Dept of Agriculture, Agricultural Research Service
      USDA Food and Nutrient Database for Dietary Studies.
      ,
      US Dept of Agriculture, Agricultural Research Service
      Food Patterns Equivalents Database.
      Alcohol intake was categorized as nondrinking (0 g/day), moderate drinking (0.1 to 27.9 g/day for men and 0.1 to 13.9 g/day for women), and heavy drinking (≥28 g/day for men and ≥14 g/day for women). Overall diet quality was assessed by the Healthy Eating Index (HEI) 2010, which is scored on the basis of the intake levels of 12 dietary components, including total fruit, whole fruit, total vegetables, greens and beans, whole grains, dairy, total protein foods, seafood and plant proteins, fatty acids, refined grains, sodium, and empty calories (ie, energy from solid fats, alcohol, and added sugars).
      • Guenther P.M.
      • Kirkpatrick S.I.
      • Reedy J.
      • et al.
      The Healthy Eating Index-2010 is a valid and reliable measure of diet quality according to the 2010 Dietary Guidelines for Americans.
      The Simple HEI Scoring Algorithm (Per Person) was used to calculate the individual HEI-2020 score.
      National Cancer Institute
      How to choose an analysis method dependent on purpose.
      Food insecurity was assessed by the Adult Food Security Survey Module, which is a standardized 10-item questionnaire developed by the US Department of Agriculture, validated and used in many national surveys.
      US Dept of Agriculture, Economic Research Service
      The U.S. Adult Food Security Survey Module.
      ,
      • Opsomer J.D.
      • Jensen H.H.
      • Pan S.
      An evaluation of the U.S. Department of Agriculture food security measure with generalized linear mixed models.
      Trained technicians measured weight and height. Body mass index (BMI) was calculated as weight in kilograms divided by height in meters squared, and was categorized as <25, 25 to 29.9, and ≥30. Baseline diabetes was defined based on a self-reported physician diagnosis, plasma fasting glucose levels of 126 mg/dL (to convert to mmol/L, multiply by 0.0555) or higher, glycated hemoglobin level 6.5% or higher, or 2-hour plasma glucose levels of 200 mg/dL or higher.
      • Menke A.
      • Casagrande S.
      • Geiss L.
      • Cowie C.C.
      Prevalence of and trends in diabetes among adults in the United States, 1988-2012.
      Baseline hypertension was defined based on a self-reported physician diagnosis, currently taking antihypertensive medications, mean systolic blood pressure ≥130 mm Hg, or mean diastolic blood pressure ≥80 mm Hg.
      • Whelton P.K.
      • Carey R.M.
      • Aronow W.S.
      • et al.
      2017 ACC/AHA/AAPA/ABC/ACPM/AGS/APhA/ASH/ASPC/NMA/PCNA Guideline for the Prevention, Detection, Evaluation, and Management of High Blood Pressure in Adults: Executive summary: a report of the American College of Cardiology/American Heart Association Task Force on Clinical Practice Guidelines.
      Baseline hypercholesterolemia was defined based on self-reported physician diagnosis, currently taking cholesterol-lowering medications, plasma levels of total cholesterol ≥240 mg/dL (to convert to mmol/L, multiply by 0.0259), or plasma levels of low-density lipoprotein-cholesterol ≥160 mg/dL.
      National Cholesterol Education Program
      Expert Panel on Detection, Evaluation, and Treatment of High Blood Cholesterol in Adults. Third Report of the National Cholesterol Education Program (NCEP) Expert Panel on Detection, Evaluation, and Treatment of High Blood Cholesterol in Adults (Adult Treatment Panel III) final report.
      Baseline CVD status was defined based on self-reported physician diagnosis of congestive health failure, coronary heart disease, angina/angina pectoris, heart attack, or stroke. Baseline cancer status was defined based on a self-reported physician diagnosis of cancer or malignancy.

      Statistical Analysis

      This study followed the NHANES Analytic Guidelines developed by the NCHS when conducting and reporting this study.
      • Johnson C.L.
      • Paulose-Ram R.
      • Ogden C.L.
      • et al.
      National Health and Nutrition Examination Survey: analytic guidelines, 1999-2010.
      All statistical analyses accounted for the complex, multistage, stratified, cluster-sampling design of NHANES by using sample weights, strata, and primary sampling units embedded in the NHANES data. Comparisons of characteristics among participants across categories of meal frequency were performed using analysis of variance for continuous variables and chi-square test for categorical variables.
      Multivariable Cox proportional hazards regression models were used to estimate hazard ratios (HRs) and 95% CIs of mortality in relation to meal frequency and intervals. Model 1 adjusted for age, gender, and race and ethnicity. Model 2 additionally adjusted for education, income, smoking status, alcohol intake, physical activity levels, total energy intake, HEI-2010 score, household adult food insecurity status, and snacks frequency. Model 3 further adjusted for baseline diabetes, hypertension, hypercholesterolemia, CVD, cancer, and BMI status, because these variables may be mediators between meal frequency, intervals and timing, and mortality.
      • Ojo O.
      Nutrition and chronic conditions.
      ,
      • Raghupathi W.
      • Raghupathi V.
      An empirical study of chronic diseases in the United States: a visual analytics approach.
      For sensitivity analyses, the analyses were repeated by excluding participants with CVD or cancer at baseline. All analyses were performed using survey procedures in SAS 9.4.
      Two-sided P < 0.05 was considered statistically significant.

      Results

      During 185,398 person-years of period, 4,175 deaths occurred, including 878 CVD deaths. Most participants (56.7%) ate three meals per day. As shown in Table 1, compared with participants with three meals per day, participants eating fewer than three meals per day were more likely to be younger, men, non-Hispanic Black, with less education and lower family income, current smokers, heavy alcohol drinkers, higher physical activity levels, lower total energy intake and lower diet quality, food insecure, and higher frequency of snacks. They were less likely to have diabetes, and cancer at baseline. The characteristics for participants who ate more than three meals per day were comparable to those for participants with three meals per day, except participants who ate more than three meals per day were more likely to be with higher physical activity levels, higher total energy intake, and higher diet quality, lower BMI, lower frequency of snacks, and less likely to have CVD at baseline.
      Table 1Characteristics of 24,011 US adults aged 40 years or older according to meal frequency, National Health and Nutrition Examination Survey, 1999-2014
      CharacteristicMeal frequencyP value
      P values indicated comparisons of characteristics among participants across categories of meal frequency. The comparisons were performed using analysis of variance for continuous variables and χ2 test for categorical variables.
      123≥4
      No. of participants1,1166,31513,6262,954
      mean (SE)
      Age (y)53.1 (0.3)55.8 (0.2)57.5 (0.2)55.7 (0.3)< 0.001
      Total energy intake (kcal/d)1,732 (33)1,945 (14)2,090 (10)2,229 (24)< 0.001
      Heealthy Eating Index-2010 score43.2 (0.6)47.2 (0.3)52.5 (0.2)53.0 (0.4)< 0.001
      Snacks frequency (times/d)2.8 (0.1)2.4 (0.03)2.1 (0.02)2.0 (0.04)< 0.001
      Body mass index28.9 (0.3)29.5 (0.1)29.0 (0.1)28.5 (0.2)< 0.001
      % (SE)
      The numbers in the table were unweighted and the percentages of values were weighted.
      Male gender (%)55.9 (1.9)49.9 (0.8)45.5 (0.5)49.6 (1.2)< 0.001
      Race and ethnicity
      Non-Hispanic White61.7 (2.7)69.3 (1.5)78.7 (1.0)74.9 (1.2)< 0.001
      Non-Hispanic Black24.6 (2.1)15.1 (1.0)7.4 (0.5)7.6 (0.6)
      Hispanic8.8 (1.2)11.3 (0.9)8.8 (0.7)10.9 (0.8)
      Other
      Other included other races and multiracial.
      4.8 (1.1)4.3 (0.4)5.1 (0.3)6.5 (0.6)
      Education
      Less than high school diploma30.6 (1.7)22.7 (0.8)15.9 (0.5)18.2 (1.1)< 0.001
      High school graduate/GED or equivalent28.1 (1.9)25.6 (0.7)23.9 (0.7)24.6 (1.2)
      College graduate or above41.4 (2.2)51.7 (1.0)60.3 (0.9)57.1 (1.5)
      Ratio of family income to poverty
      <1.022.2 (1.6)12.9 (0.6)8.0 (0.4)8.4 (0.6)< 0.001
      1.00-1.9922.7 (1.9)20.0 (0.8)16.5 (0.6)17.1 (1.0)
      2.00-3.9926.5 (2.1)27.5 (0.9)27.6 (0.7)26.7 (1.1)
      ≥422.1 (2.1)32.4 (1.0)41.3 (1.0)41.4 (1.4)
      Missing6.5 (0.9)7.2 (0.5)6.6 (0.4)6.3 (0.5)
      Smoking status
      Never smoker37.9 (2.0)44.8 (0.9)52.7 (0.6)48.8 (1.4)< 0.001
      Former smoker19.7 (1.7)29.2 (0.8)32.1 (0.5)34.4 (1.2)
      Current smoker42.4 (2.3)26.0 (0.9)15.3 (0.4)16.8 (1.0)
      Alcohol intake
      Nondrinker = 0 g/day alcohol; moderate drinking = 0.1 to 27.9 g/day alcohol for men and 0.1 to 13.9 g/day alcohol day for women; and heavy drinking: ≥28 g/day alcohol for men and ≥14 g/day alcohol for women.
      Non-drinking69.8 (2.2)72.1 (0.9)73.4 (0.8)64.2 (1.4)< 0.001
      Moderate drinking8.0 (1.3)9.9 (0.6)11.0 (0.4)13.7 (1.0)
      Heavy drinking22.1 (1.7)17.9 (0.7)15.7 (0.6)22.1 (1.3)
      Physical activity (MET-min/wk)
      <60051.0 (2.0)47.2 (0.9)41.3 (0.7)37.9 (1.4)< 0.001
      ≥600-1,1999.8 (1.5)11.8 (0.5)14.0 (0.4)15.2 (0.8)
      ≥1,20039.2 (1.8)41.0 (0.8)44.7 (0.7)46.9 (1.4)
      Food insecurity18.2 (1.4)11.9(0.5)7.1 (0.3)8.7 (0.7)< 0.001
      Baseline diabetes14.2 (1.2)16.7 (0.5)17.5 (0.5)15.3 (0.8)0.01
      Baseline hypertension59.7 (1.9)60.6 (0.9)59.7 (0.6)57.7 (1.2)0.18
      Baseline hypercholesterolemia46.9 (2.2)51.1 (0.8)50.2 (0.6)51.0 (1.2)0.20
      Baseline cardiovascular disease12.3 (1.1)13.8 (0.5)12.4 (0.4)10.0 (0.7)< 0.001
      Baseline cancer8.7 (1.1)11.1 (0.6)14.4 (0.4)12.7 (0.8)< 0.001
      Body mass index status
      Normal weight and underweight29.2 (1.9)24.5 (0.7)27.5 (0.6)28.5 (1.1)< 0.001
      Overweight36.5 (1.8)34.7 (0.9)35.2 (0.5)37.1 (1.3)
      Obesity33.2 (2.0)39.1 (0.9)35.9 (0.6)33.2 (1.2)
      Missing1.1 (0.2)1.7 (0.2)1.3 (0.1)1.2 (0.2)
      a P values indicated comparisons of characteristics among participants across categories of meal frequency. The comparisons were performed using analysis of variance for continuous variables and χ2 test for categorical variables.
      b The numbers in the table were unweighted and the percentages of values were weighted.
      c Other included other races and multiracial.
      d Nondrinker = 0 g/day alcohol; moderate drinking = 0.1 to 27.9 g/day alcohol for men and 0.1 to 13.9 g/day alcohol day for women; and heavy drinking: ≥28 g/day alcohol for men and ≥14 g/day alcohol for women.
      The association between meal frequency and all-cause and CVD mortality were shown in Table 2. After adjustment for age, gender, race and ethnicity, socioeconomic status, lifestyle factors, and baseline history of diseases, compared with participants with three meals per day, the multivariable-adjusted HRs all-cause mortality for participants eating one meal and two meals were 1.30 (95% CI 1.03 to 1.64), and 1.07 (95% CI 0.98 to 1.16), respectively. For CVD mortality, the multivariable-adjusted HRs for participants eating one meal and two meals were 1.83 (95% CI 1.26 to 2.65), and 1.10 (95% CI 0.90 to 1.35), respectively. Eating more than three meals per day was not associated with all-cause or CVD mortality compared with eating three meals per day. The multivariable-adjusted HRs for the participants who ate more than three meals per day were 0.91 (95% CI 0.79 to 1.04) for all-cause mortality, and 0.92 (95% CI 0.67 to 1.25) for CVD mortality.
      Table 2Association of meal frequency with mortality among 24,011 US adults aged 40 years or older with meals at least 1 per day, National Health and Nutrition Examination Survey, 1999-2014
      Meal Frequency
      123≥4
      Hazard ratio (95% CI)
      All-cause mortality
      No. of cases/participants224/1,1161,137/6,3152,424/13,626390/2,954
      Model 1
      Model 1 was adjusted for age, gender, and race and ethnicity.
      1.69 (1.34-2.13)1.24 (1.14-1.34)Ref
      Ref = reference category.
      0.87 (0.76-0.99)
      Model 2
      Model 2 was model 1 + education, income, smoking status, alcohol intake, physical activity levels, total energy intake, overall diet quality indicated by Healthy Eating Index 2010 score, food insecurity status, and snacks frequency.
      1.21 (0.95-1.55)1.06 (0.98-1.15)Ref0.89 (0.78-1.03)
      Model 3
      Model 3 was model 2 + baseline diabetes, baseline hypertension, baseline hypercholesterolemia, baseline cardiovascular disease, baseline cancer, and body mass index status.
      1.30 (1.03-1.64)1.07 (0.98-1.16)Ref0.91 (0.79-1.04)
      CVD mortality
      No. of cases/participants54/1,116242/6,315504/13,62678/2,954
      Model 1
      Model 1 was adjusted for age, gender, and race and ethnicity.
      2.21 (1.46-3.33)1.24 (1.01-1.53)Ref0.84 (0.62-1.14)
      Model 2
      Model 2 was model 1 + education, income, smoking status, alcohol intake, physical activity levels, total energy intake, overall diet quality indicated by Healthy Eating Index 2010 score, food insecurity status, and snacks frequency.
      1.67 (1.11-2.51)1.09 (0.89-1.34)Ref0.89 (0.65-1.21)
      Model 3
      Model 3 was model 2 + baseline diabetes, baseline hypertension, baseline hypercholesterolemia, baseline cardiovascular disease, baseline cancer, and body mass index status.
      1.83 (1.26-2.65)1.10 (0.90-1.35)Ref0.92 (0.67-1.25)
      a Model 1 was adjusted for age, gender, and race and ethnicity.
      b Ref = reference category.
      c Model 2 was model 1 + education, income, smoking status, alcohol intake, physical activity levels, total energy intake, overall diet quality indicated by Healthy Eating Index 2010 score, food insecurity status, and snacks frequency.
      d Model 3 was model 2 + baseline diabetes, baseline hypertension, baseline hypercholesterolemia, baseline cardiovascular disease, baseline cancer, and body mass index status.
      The associations of skipping a specific meal with mortality were shown in Table 3. In the fully adjusted model, skipping breakfast was associated with higher risk of CVD mortality, whereas skipping lunch or dinner was associated with higher risk of all-cause mortality. Compared with participant who did not skip breakfast, the multivariable-adjusted HRs for participants who skipped breakfast were 1.11 (95% CI 0.98 to 1.26) for all-cause mortality, and 1.40 (95% CI 1.09 to 1.78) for CVD mortality. Compared with participant who did not skip lunch, the multivariable-adjusted HRs for participants who skipped lunch were 1.12 (95% CI 1.01 to 1.24) for all-cause mortality, and 1.15 (95% CI 0.94 to 1.40) for CVD mortality. Compared with participant who did not skip dinner, the multivariable-adjusted HRs for participants who skipped dinner were 1.16 (95% CI 1.02 to 1.32) for all-cause mortality, and 1.19 (95% CI 0.89 to 1.59) for CVD mortality.
      Table 3Association of skipping meal with mortality among 24,011 US adults aged 40 years or older with meals at least one per day, National Health and Nutrition Examination Survey, 1999-2014
      Skipping Breakfast
      NoYes
      Hazard ratio (95% CI)
      All-cause mortality
      No. of cases/participants3,736/21,217439/2,794
      Model 1
      Model 1 was adjusted for age, gender, and race and ethnicity.
      Ref
      Ref = reference category.
      1.38 (1.22-1.57)
      Model 2
      Model 2 was model 1 + education, income, smoking status, alcohol intake, physical activity levels, total energy intake, overall diet quality indicated by Healthy Eating Index 2010 score, food insecurity status, snacks frequency, and mutual adjustment for each other (skipping breakfast, skipping lunch and slipping dinner).
      Ref1.10 (0.96-1.25)
      Model 3
      Model 3 was model 2 + baseline diabetes, baseline hypertension, baseline hypercholesterolemia, baseline cardiovascular disease, baseline cancer, and body mass index status.
      Ref1.11 (0.98-1.26)
      Cardiovascular disease mortality
      No. of cases/participants779/21,21799/2,794
      Model 1
      Model 1 was adjusted for age, gender, and race and ethnicity.
      Ref1.68 (1.32-2.15)
      Model 2
      Model 2 was model 1 + education, income, smoking status, alcohol intake, physical activity levels, total energy intake, overall diet quality indicated by Healthy Eating Index 2010 score, food insecurity status, snacks frequency, and mutual adjustment for each other (skipping breakfast, skipping lunch and slipping dinner).
      Ref1.36 (1.05-1.75)
      Model 3
      Model 3 was model 2 + baseline diabetes, baseline hypertension, baseline hypercholesterolemia, baseline cardiovascular disease, baseline cancer, and body mass index status.
      Ref1.40 (1.09-1.78)
      Skipping lunch
      All-cause mortality
      No. of cases/participants2,970/17,8301,205/6,181
      Model 1
      Model 1 was adjusted for age, gender, and race and ethnicity.
      Ref1.26 (1.16-1.38)
      Model 2
      Model 2 was model 1 + education, income, smoking status, alcohol intake, physical activity levels, total energy intake, overall diet quality indicated by Healthy Eating Index 2010 score, food insecurity status, snacks frequency, and mutual adjustment for each other (skipping breakfast, skipping lunch and slipping dinner).
      Ref1.07 (0.97-1.18)
      Model 3
      Model 3 was model 2 + baseline diabetes, baseline hypertension, baseline hypercholesterolemia, baseline cardiovascular disease, baseline cancer, and body mass index status.
      Ref1.12 (1.01-1.24)
      Cardiovascular disease mortality
      No. of cases/participants614/17,830264/6,181
      Model 1
      Model 1 was adjusted for age, gender, and race and ethnicity.
      Ref1.28 (1.06-1.55)
      Model 2
      Model 2 was model 1 + education, income, smoking status, alcohol intake, physical activity levels, total energy intake, overall diet quality indicated by Healthy Eating Index 2010 score, food insecurity status, snacks frequency, and mutual adjustment for each other (skipping breakfast, skipping lunch and slipping dinner).
      Ref1.09 (0.89-1.34)
      Model 3
      Model 3 was model 2 + baseline diabetes, baseline hypertension, baseline hypercholesterolemia, baseline cardiovascular disease, baseline cancer, and body mass index status.
      Ref1.15 (0.94-1.40)
      Skipping dinner
      All-cause mortality
      No. of cases/participants3726/21,722449/2,289
      Model 1
      Model 1 was adjusted for age, gender, and race and ethnicity.
      Ref1.22 (1.07-1.38)
      Model 2
      Model 2 was model 1 + education, income, smoking status, alcohol intake, physical activity levels, total energy intake, overall diet quality indicated by Healthy Eating Index 2010 score, food insecurity status, snacks frequency, and mutual adjustment for each other (skipping breakfast, skipping lunch and slipping dinner).
      Ref1.13 (0.99-1.29)
      Model 3
      Model 3 was model 2 + baseline diabetes, baseline hypertension, baseline hypercholesterolemia, baseline cardiovascular disease, baseline cancer, and body mass index status.
      Ref1.16 (1.02-1.32)
      Cardiovascular disease mortality
      No. of cases/participants779/21,72299/2,289
      Model 1
      Model 1 was adjusted for age, gender, and race and ethnicity.
      Ref1.22 (0.91-1.63)
      Model 2
      Model 2 was model 1 + education, income, smoking status, alcohol intake, physical activity levels, total energy intake, overall diet quality indicated by Healthy Eating Index 2010 score, food insecurity status, snacks frequency, and mutual adjustment for each other (skipping breakfast, skipping lunch and slipping dinner).
      Ref1.16 (0.87-1.56)
      Model 3
      Model 3 was model 2 + baseline diabetes, baseline hypertension, baseline hypercholesterolemia, baseline cardiovascular disease, baseline cancer, and body mass index status.
      Ref1.19 (0.89-1.59)
      a Model 1 was adjusted for age, gender, and race and ethnicity.
      b Ref = reference category.
      c Model 2 was model 1 + education, income, smoking status, alcohol intake, physical activity levels, total energy intake, overall diet quality indicated by Healthy Eating Index 2010 score, food insecurity status, snacks frequency, and mutual adjustment for each other (skipping breakfast, skipping lunch and slipping dinner).
      d Model 3 was model 2 + baseline diabetes, baseline hypertension, baseline hypercholesterolemia, baseline cardiovascular disease, baseline cancer, and body mass index status.
      The associations of meal intervals with mortality were shown in Table 4. Among those eating 3 meals/day, compared with an average interval between two daytime meals of 4.6 to 5.5 hours, the fully adjusted HRs associated with an average interval of ≤4.5 hours were 1.17 (95% CI 1.04 to 1.32) for all-cause mortality and 1.22 (95% CI 0.94 to 1.59) for CVD mortality. An interval >5.5 hours between two daytime meals was not associated with all-cause or CVD mortality, with the multivariable-adjusted HRs as 0.98 (95% CI 85 to 1.12), and 0.98 (95% 0.74 to 1.29), respectively.
      Table 4Association of average meal intervals with mortality among 13,626 US adults aged 40 years or older with three meals per day, National Health and Nutrition Examination Survey, 1999-2014
      Meal Intervals (h)
      ≤4.54.6-5.5>5.5
      Hazard ration (95% CI)
      All-cause mortality
      No. of cases/participants666/3,2701111/5,770647/4,586
      Model 1
      Model 1 was adjusted for age, gender, and race and ethnicity.
      1.28 (1.14-1.45)Ref
      Ref = reference category.
      0.94 (0.82-1.08)
      Model 2
      Model 2 was model 1 + education, income, smoking status, alcohol intake, physical activity levels, total energy intake, overall diet quality indicated by Healthy Eating Index 2010 score, and food insecurity status.
      1.19 (1.06-1.33)Ref0.98 (0.84-1.13)
      Model 3
      Model 3 was model 2 + baseline diabetes, baseline hypertension, baseline hypercholesterolemia, baseline cardiovascular disease, baseline cancer, and body mass index status.
      1.17 (1.04-1.32)Ref0.98 (0.85-1.12)
      Cardiovascular disease mortality
      No. of cases/participants136/3,270230/5,770138/4,586
      Model 1
      Model 1 was adjusted for age, gender, and race and ethnicity.
      1.40 (1.08-1.82)Ref0.96 (0.73-1.28)
      Model 2
      Model 2 was model 1 + education, income, smoking status, alcohol intake, physical activity levels, total energy intake, overall diet quality indicated by Healthy Eating Index 2010 score, and food insecurity status.
      1.27 (0.98-1.64)Ref0.99 (0.75-1.31)
      Model 3
      Model 3 was model 2 + baseline diabetes, baseline hypertension, baseline hypercholesterolemia, baseline cardiovascular disease, baseline cancer, and body mass index status.
      1.22 (0.94-1.59)Ref0.98 (0.74-1.29)
      a Model 1 was adjusted for age, gender, and race and ethnicity.
      b Ref = reference category.
      c Model 2 was model 1 + education, income, smoking status, alcohol intake, physical activity levels, total energy intake, overall diet quality indicated by Healthy Eating Index 2010 score, and food insecurity status.
      d Model 3 was model 2 + baseline diabetes, baseline hypertension, baseline hypercholesterolemia, baseline cardiovascular disease, baseline cancer, and body mass index status.
      Some of the results became statistically insignificant in the sensitivity analyses by excluding participants with CVD or cancer at baseline (Table 5, Table 6, Table 7, available at www.jandonline.org), probably due to reduced sample size because similar HRs were observed.

      Discussion

      In this large, prospective study of a nationally representative sample of US adults, eating 1 meal/day was associated with higher risk of all-cause and CVD mortality compared with eating 3 meals/day, whereas eating two meals per day or four or more meals per day was not associated with mortality in fully adjusted models. Specifically skipping breakfast was associated with higher risk of CVD mortality, whereas skipping lunch or dinner was associated with higher risk of all-cause mortality. Furthermore, meal intervals ≤4.5 hours was associated with higher risk of all-cause mortality, compared with meals intervals of 4.6 to 5.5 hours, whereas meal intervals >5.5 hours was not associated with mortality.
      The findings that eating one meal per day was associated with higher risk of all-cause and CVD mortality were consistent with the findings of the only previous study on eating frequency and mortality. It was found that a decreasing trend in mortality was observed to be associated with increasing eating frequency.
      • Chen H.J.
      • Wang Y.F.
      • Cheskin L.J.
      Relationship between frequency of eating and cardiovascular disease mortality in US adults: the NHANES III follow-up study.
      The findings of this study were also consistent with findings from previous studies that greater eating frequency was related to improved risk status for coronary heart disease.
      • Cahill L.E.
      • Chiuve S.E.
      • Mekary R.A.
      • et al.
      Prospective study of breakfast eating and incident coronary heart disease in a cohort of male US health professionals.
      ,
      • Titan S.M.O.
      • Bingham S.
      • Welch A.
      • et al.
      Frequency of eating and concentrations of serum cholesterol in the Norfolk population of the European prospective investigation into cancer (EPIC-Norfolk): cross sectional study.
      A meta-analysis of 21 randomized controlled trials found that higher meal frequency, not less meal frequency, could improve total cholesterol and low-density lipoprotein cholesterol levels in human.
      • Abdollahi S.
      • Kazemi A.
      • de Souza R.J.
      • Clark C.C.T.
      • Soltani S.
      The effect of meal frequency on biochemical cardiometabolic factors: a systematic review and meta-analysis of randomized controlled trials.
      Furthermore, evidence from animal experimental studies also showed that increased feeding frequency could also improve the postprandial glucose and insulin concentrations, which might decrease cardiometabolic risk of hyperglycemia, glucosuria, and excessive insulinemia.
      • Vicari T.
      • van den Borne J.J.
      • Gerrits W.J.
      • Zbinden Y.
      • Blum J.W.
      Postprandial blood hormone and metabolite concentrations influenced by feeding frequency and feeding level in veal calves.
      Consistent with previous studies, this study confirmed that skipping breakfast was associated with higher CVD mortality.
      • Rong S.
      • Snetselaar L.G.
      • Xu G.F.
      • et al.
      Association of skipping breakfast with cardiovascular and all-cause mortality.
      ,
      • Chen H.
      • Zhang B.
      • Ge Y.
      • et al.
      Association between skipping breakfast and risk of cardiovascular disease and all cause mortality: a meta-analysis.
      However, to our best knowledge, there has not been a previous study that has investigated the association of skipping lunch or dinner with mortality. There was only one study that found a significant association between skipping dinner and increased risk of obesity, which provides support for the findings of this current study that skipping dinner was related with increased risk of all-cause mortality.
      • Yamamoto R.
      • Tomi R.
      • Shinzawa M.
      • et al.
      Associations of skipping breakfast, lunch, and dinner with weight gain and overweight/obesity in university students: a retrospective cohort study.
      The association of skipping lunch with mortality disappeared with the adjustment for some lifestyle factors, which indicated that these lifestyle factors, such as diet quality, might have important role in the influence of skipping meals on mortality. Evidence from previous studies showed that people who skipped lunch had lower diet quality, less consumption of fruits and vegetables, and more consumption of sodium and refined grains, and these were also risk factors for mortality.
      • Zeballos E.
      • Todd J.E.
      The effects of skipping a meal on daily energy intake and diet quality.
      There are several possible reasons that skipping meals is a probable risk factor of disease and mortality. First, skipping meals was an indicator of the inability to acquire adequate food, which has been shown to be associated with higher CVD risk among adults.
      • Mendy V.L.
      • Vargas R.
      • Cannon-Smith G.
      • Payton M.
      • Enkhmaa B.
      • Zhang L.
      Food insecurity and cardiovascular disease risk factors among Mississippi adults.
      ,
      • Vercammen K.A.
      • Moran A.J.
      • McClain A.C.
      • Thorndike A.N.
      • Fulay A.P.
      • Rimm E.B.
      Food security and 10-year cardiovascular disease risk among U.S. adults.
      Second, skipping meals, in particular skipping breakfast, might also be a behavioral marker for unhealthy dietary and lifestyle habits.
      • Zeballos E.
      • Todd J.E.
      The effects of skipping a meal on daily energy intake and diet quality.
      However, the results remained significant after adjustment for food insecurity status, and a variety of dietary and lifestyle factors including smoking, alcohol use, physical activity levels, total energy intake and overall diet quality, making them less likely explanations for the observed association between skipping meals and mortality. Third, skipping meals usually means eating meals with a larger energy load, which could aggravate the burden of glucose metabolism regulation, and thus might lead to subsequent metabolic deterioration.
      • Chen H.J.
      • Wang Y.F.
      • Cheskin L.J.
      Relationship between frequency of eating and cardiovascular disease mortality in US adults: the NHANES III follow-up study.
      This could also explain the association between a shorter meal interval; for example, ≤4.5 hours, and mortality, as a shorter waiting time in between meals also means a larger energy load in a given period of time. In addition, skipping meals, in particular skipping breakfast, which is related to changes in appetite and decreased satiety, might lead to subsequent overeating later and impairment in insulin sensitivity.
      • Leidy H.J.
      • Campbell W.W.
      The effect of eating frequency on appetite control and food intake: brief synopsis of controlled feeding studies.
      By contrast, eating meals regularly has a beneficial effect on appetite regulation and also improves the glycemic response at the next eating occasion with increased sensitivity to insulin.
      • Alhussain M.H.
      • Macdonald I.A.
      • Taylor M.A.
      Irregular meal-pattern effects on energy expenditure, metabolism, and appetite regulation: a randomized controlled trial in healthy normal-weight women.
      ,
      • Jovanovic A.
      • Gerrard J.
      • Taylor R.
      The second-meal phenomenon in type 2 diabetes.
      Fourth, skipping meals may reflect psychosocial conditions which are associated with elevated CVD risk; for example, anorexia nervosa.
      • Liu C.
      • Xie B.
      • Chou C.P.
      • et al.
      Perceived stress, depression and food consumption frequency in the college students of China Seven Cities.
      • Mikolajczyk R.T.
      • El Ansari W.
      • Maxwell A.E.
      Food consumption frequency and perceived stress and depressive symptoms among students in three European countries.
      • Mattson M.P.
      Energy intake, meal frequency, and health: a neurobiological perspective.
      • Papadopoulos F.C.
      • Ekbom A.
      • Brandt L.
      • Ekselius L.
      Excess mortality, causes of death and prognostic factors in anorexia nervosa.
      To our best knowledge, there were no prior study that has examined the association of duration of the daily postprandial period with cardiometabolic health. This study is the first to show that among participants with three meals per day, an average meal interval between two daytime meals of ≤4.5 hours was associated with a higher risk of all-cause mortality.

      Strengths and Limitations

      The strengths of this study include the use of detailed dietary information obtained from a 24-hour dietary recall in a large representative sample of US adults.
      • Kant A.K.
      • Graubard B.I.
      40-year trends in meal and snack eating behaviors of American adults.
      It was based on a nationally representative sample of US adults, the findings of which could be more directly translated to the general population. Furthermore, the reported clock time of each meal made it possible to calculate the intervals between meals. There are also several limitations. First, the measures of eating behaviors were self-reported, which might lead to recall bias. This may bias the results towards the null. Second, the lack of information on sleep such as sleep duration or bedtime makes it impossible to evaluate the role of sleep as a potential confounder for the association between meal patterns and mortality. Third, because only one in-person dietary recall was collected in the 1999-2000, and 2001-2002 cycles, this study used the first recall for all surveys. Fourth, despite the adjustment for a wide range of potential factors that were both related to meal consumption and mortality, there is still a possibility of residual confounding by unmeasured factors, such as neighborhood characteristics, detailed housing condition, or occupational environment. Fifth, this study focused on meal skipping regardless of snack consumption, thus the evidence provided for healthy effects of intermittent fasting is limited.
      This study has significant public health implications. Previous dietary studies and Dietary Guidelines for Americans have focused mainly on dietary components and food combinations. Little is known about the association of eating behaviors such as meal skipping and meal intervals with health. The findings have contributed to the current evidence regarding the association between eating behaviors and mortality by examining the association between meal skipping and mortality, in the context of meal timing and duration of the daily prandial period.

      Conclusions

      In this large, prospective study of US adults aged 40 years or older, eating one meal per day was significantly associated with an increased risk of all-cause and CVD mortality compared with eating three meals per day. Skipping breakfast was associated with increased risk of CVD mortality, whereas skipping lunch or dinner was associated with higher risk of all-cause mortality. Moreover, among participant with 3 meals/day, a shorter interval between meals (≤4.5 hours) was associated with higher risk of all-cause mortality. The results support the role of eating at least three meals per day and a >4.5 hours of waiting time in between two daytime meals in relation to better cardiovascular health. These findings need replication and assessment in other populations.

      Supplementary Materials

      Table 5Association of meal frequency with mortality among 17,901 US adults aged 40 years or older with meals at least 1 per day (baseline cardiovascular disease [CVD] and cancer excluded), National Health and Nutrition Examination Survey, 1999-2014
      Meal Frequency
      123≥4
      Hazard ratio (95% CI)
      All-cause mortality
      No. of cases/participants136/884607/4,7761,192/9,961212/2,280
      Model 1
      Model 1 was adjusted for age, gender, and race and ethnicity.
      1.66 (1.25-2.19)1.22 (1.08-1.38)
      Ref = reference category.
      Ref
      0.90 (0.75-1.08)
      Model 2
      Model 2 was model 1 + education, income, smoking status, alcohol intake, physical activity levels, total energy intake, overall diet quality indicated by Healthy Eating Index 2010 score, food insecurity status, and snacks frequency.
      1.19 (0.89-1.59)1.08 (0.96-1.21)Ref0.88 (0.73-1.07)
      Model 3
      Model 3 was model 2 + baseline diabetes, baseline hypercholesterolemia, baseline hypertension, and body mass index status.
      1.26 (0.94-1.68)1.10 (0.98-1.23)Ref0.88 (0.73-1.07)
      CVD mortality
      No. of cases/participants27/884125/4,776226/9,96135/2,280
      Model 1
      Model 1 was adjusted for age, gender, and race and ethnicity.
      1.57 (0.96-2.58)1.23 (0.92-1.65)Ref0.80 (0.54-1.17)
      Model 2
      Model 2 was model 1 + education, income, smoking status, alcohol intake, physical activity levels, total energy intake, overall diet quality indicated by Healthy Eating Index 2010 score, food insecurity status, and snacks frequency.
      1.20 (0.77-1.86)1.12 (0.83-1.50)Ref0.80 (0.53-1.19)
      Model 3
      Model 3 was model 2 + baseline diabetes, baseline hypercholesterolemia, baseline hypertension, and body mass index status.
      1.31 (0.85-2.02)1.15 (0.85-1.55)Ref0.80 (0.54-1.19)
      a Model 1 was adjusted for age, gender, and race and ethnicity.
      b Ref = reference category.
      c Model 2 was model 1 + education, income, smoking status, alcohol intake, physical activity levels, total energy intake, overall diet quality indicated by Healthy Eating Index 2010 score, food insecurity status, and snacks frequency.
      d Model 3 was model 2 + baseline diabetes, baseline hypercholesterolemia, baseline hypertension, and body mass index status.
      Table 6Association of skipping meal with mortality among 17,901 US adults aged 40 years or older with meals at least 1 per day (baseline cardiovascular disease [CVD] and cancer excluded), National Health and Nutrition Examination Survey, 1999-2014
      Skipping Breakfast
      NoYes
      Hazard ratio (95% CI)
      All-cause mortality
      No. of cases/participants1882/15,657265/2,244
      Model 1
      Model 1 was adjusted for age, gender, and race and ethnicity.
      Ref
      Ref = reference category.
      1.42 (1.22-1.66)
      Model 2
      Model 2 was model 1 + education, income, smoking status, alcohol intake, physical activity levels, total energy intake, overall diet quality indicated by Healthy Eating Index 2010 score, food insecurity status, snacks frequency, and mutual adjustment for each other (skipping breakfast, skipping lunch and skipping dinner).
      Ref1.15 (0.98-1.35)
      Model 3
      Model 3 was model 2 + baseline diabetes, baseline hypertension, baseline hypercholesterolemia, baseline CVD, and body mass index status.
      Ref1.18 (0.99-1.39)
      CVD-mortality
      No. of cases/participants352/15,65761/2,244
      Model 1
      Model 1 was adjusted for age, gender, and race and ethnicity.
      Ref1.99 (1.44-2.74)
      Model 2
      Model 2 was model 1 + education, income, smoking status, alcohol intake, physical activity levels, total energy intake, overall diet quality indicated by Healthy Eating Index 2010 score, food insecurity status, snacks frequency, and mutual adjustment for each other (skipping breakfast, skipping lunch and skipping dinner).
      Ref1.70 (1.22-2.36)
      Model 3
      Model 3 was model 2 + baseline diabetes, baseline hypertension, baseline hypercholesterolemia, baseline CVD, and body mass index status.
      Ref1.77 (1.28-2.45)
      Skipping lunch
      All-cause mortality
      No. of cases/participants1512/13,303635/4,598
      Model 1
      Model 1 was adjusted for age, gender, and race and ethnicity.
      Ref1.25 (1.10-1.41)
      Model 2
      Model 2 was model 1 + education, income, smoking status, alcohol intake, physical activity levels, total energy intake, overall diet quality indicated by Healthy Eating Index 2010 score, food insecurity status, snacks frequency, and mutual adjustment for each other (skipping breakfast, skipping lunch and skipping dinner).
      Ref1.07 (0.94-1.21)
      Model 3
      Model 3 was model 2 + baseline diabetes, baseline hypertension, baseline hypercholesterolemia, baseline CVD, and body mass index status.
      Ref1.10 (0.96-1.25)
      CVD mortality
      No. of cases/participants286/13,303127/4,598
      Model 1
      Model 1 was adjusted for age, gender, and race and ethnicity.
      Ref1.18 (0.87-1.61)
      Model 2
      Model 2 was model 1 + education, income, smoking status, alcohol intake, physical activity levels, total energy intake, overall diet quality indicated by Healthy Eating Index 2010 score, food insecurity status, snacks frequency, and mutual adjustment for each other (skipping breakfast, skipping lunch and skipping dinner).
      Ref1.01 (0.73-1.40)
      Model 3
      Model 3 was model 2 + baseline diabetes, baseline hypertension, baseline hypercholesterolemia, baseline CVD, and body mass index status.
      Ref1.05 (0.76-1.45)
      Skipping dinner
      All-cause mortality
      No. of cases/participants1884/16,137263/1,764
      Model 1
      Model 1 was adjusted for age, gender, and race and ethnicity.
      Ref1.27 (1.04-1.54)
      Model 2
      Model 2 was model 1 + education, income, smoking status, alcohol intake, physical activity levels, total energy intake, overall diet quality indicated by Healthy Eating Index 2010 score, food insecurity status, snacks frequency, and mutual adjustment for each other (skipping breakfast, skipping lunch and skipping dinner).
      Ref1.18 (0.97-1.43)
      Model 3
      Model 3 was model 2 + baseline diabetes, baseline hypertension, baseline hypercholesterolemia, baseline CVD, and body mass index status.
      Ref1.20 (0.99-1.46)
      CVD mortality
      No. of cases/participants362/16,13751/1,764
      Model 1
      Model 1 was adjusted for age, gender, and race and ethnicity.
      Ref1.09 (0.69-1.70)
      Model 2
      Model 2 was model 1 + education, income, smoking status, alcohol intake, physical activity levels, total energy intake, overall diet quality indicated by Healthy Eating Index 2010 score, food insecurity status, snacks frequency, and mutual adjustment for each other (skipping breakfast, skipping lunch and skipping dinner).
      Ref1.01 (0.64-1.59)
      Model 3
      Model 3 was model 2 + baseline diabetes, baseline hypertension, baseline hypercholesterolemia, baseline CVD, and body mass index status.
      Ref1.02 (0.65-1.60)
      a Model 1 was adjusted for age, gender, and race and ethnicity.
      b Ref = reference category.
      c Model 2 was model 1 + education, income, smoking status, alcohol intake, physical activity levels, total energy intake, overall diet quality indicated by Healthy Eating Index 2010 score, food insecurity status, snacks frequency, and mutual adjustment for each other (skipping breakfast, skipping lunch and skipping dinner).
      d Model 3 was model 2 + baseline diabetes, baseline hypertension, baseline hypercholesterolemia, baseline CVD, and body mass index status.
      Table 7Association of average meal intervals with mortality among 9,961 US adults aged 40 years or older with 3 meals per day (baseline cardiovascular disease [CVD] and cancer excluded), National Health and Nutrition Examination Survey, 1999-2014
      VariableMeal Interval (h)
      ≤4.54.6-5.5>5.5
      Hazard ratio (95% CI)
      All-cause mortality
      No. of cases/participants317/2,303532/4,096343/3,562
      Model 1
      Model 1 was adjusted for age, gender, and race and ethnicity.
      1.30 (1.09-1.57)Ref
      Ref = reference category.
      0.95 (0.78-1.17)
      Model 2
      Model 2 was model 1 + education, income, smoking status, alcohol intake, physical activity levels, total energy intake, overall diet quality indicated by Healthy Eating Index 2010 score, and food insecurity status.
      1.21 (1.02-1.45)Ref1.00 (0.81-1.22)
      Model 3
      Model 3 indicates model 2 + baseline diabetes, baseline hypertension, baseline hypercholesterolemia, baseline CVD, and body mass index status.
      1.19 (0.99-1.43)Ref0.97 (0.79-1.19)
      CVD mortality
      No. of cases/participants58/2,30399/4,09669/3,562
      Model 1
      Model 1 was adjusted for age, gender, and race and ethnicity.
      1.45 (0.93-2.28)Ref1.10 (0.71-1.68)
      Model 2
      Model 2 was model 1 + education, income, smoking status, alcohol intake, physical activity levels, total energy intake, overall diet quality indicated by Healthy Eating Index 2010 score, and food insecurity status.
      1.38 (0.88-2.15)Ref1.16 (0.76-1.78)
      Model 3d
      Model 2 was model 1 + education, income, smoking status, alcohol intake, physical activity levels, total energy intake, overall diet quality indicated by Healthy Eating Index 2010 score, and food insecurity status.
      1.35 (0.85-2.14)Ref1.12 (0.74-1.71)
      a Model 1 was adjusted for age, gender, and race and ethnicity.
      b Ref = reference category.
      c Model 2 was model 1 + education, income, smoking status, alcohol intake, physical activity levels, total energy intake, overall diet quality indicated by Healthy Eating Index 2010 score, and food insecurity status.
      d Model 3 indicates model 2 + baseline diabetes, baseline hypertension, baseline hypercholesterolemia, baseline CVD, and body mass index status.

      References

        • St-Onge M.P.
        • Ard J.
        • Baskin M.L.
        • et al.
        Meal timing and frequency: implications for cardiovascular disease prevention: a Scientific Statement From the American Heart Association.
        Circulation. 2017; 135: E96-E121
        • Kant A.K.
        • Graubard B.I.
        40-year trends in meal and snack eating behaviors of American adults.
        J Acad Nutr Diet. 2015; 115: 50-63
        • US Depts of Agriculture and Health and Human Services
        Dietary Guidelines for Americans, 2020-2025.
        • Dietary Guidelines Advisory Committee
        Scientific Report of the 2020 Dietary Guidelines Advisory Committee: Advisory Report to the Secretary of Agriculture and the Secretary of Health and Human Services.
        • van der Heijden A.A.
        • Hu F.B.
        • Rimm E.B.
        • van Dam R.M.
        A prospective study of breakfast consumption and weight gain among U.S. men.
        Obesity (Silver Spring). 2007; 15: 2463-2469
        • Mekary R.A.
        • Giovannucci E.
        • Willett W.C.
        • van Dam R.M.
        • Hu F.B.
        Eating patterns and type 2 diabetes risk in men: breakfast omission, eating frequency, and snacking.
        Am J Clin Nutr. 2012; 95: 1182-1189
        • Ma Y.S.
        • Bertone E.R.
        • Stanek E.J.
        • et al.
        Association between eating patterns and obesity in a free-living US adult population.
        Am J Epidemiol. 2003; 158: 85-92
        • Cahill L.E.
        • Chiuve S.E.
        • Mekary R.A.
        • et al.
        Prospective study of breakfast eating and incident coronary heart disease in a cohort of male US health professionals.
        Circulation. 2013; 128: 337-343
        • Rong S.
        • Snetselaar L.G.
        • Xu G.F.
        • et al.
        Association of skipping breakfast with cardiovascular and all-cause mortality.
        J Am Coll Cardiol. 2019; 73: 2025-2032
        • Kubota Y.
        • Iso H.
        • Sawada N.
        • Tsugane S.
        • Group J.S.
        Association of breakfast intake with incident stroke and coronary heart disease: the Japan Public Health Center-Based Study.
        Stroke. 2016; 47: 477-481
        • Ahluwalia N.
        • Dwyer J.
        • Terry A.
        • Moshfegh A.
        • Johnson C.
        Update on NHANES dietary data: focus on collection, release, analytical considerations, and uses to inform public policy.
        Adv Nutr. 2016; 7: 121-134
        • Centers for Disease Control and Prevention, National Center for Health Statistics
        National Health and Nutrition Examination Survey.
        https://www.cdc.gov/nchs/nhanes/index.htm
        Date accessed: June 28, 2022
        • Szarka L.A.
        • Camilleri M.
        Gastric emptying.
        Clin Gastroenterol Hepatol. 2009; 7: 823-827
        • Centers for Disease Control and Prevention, National Center for Health Statistics
        Office of Analysis and Epidemiology. Public-use linked mortality file.
        2015
        • Centers for Disease Control and Prevention, National Center for Health Statistics
        International Classification of Diseases, Tenth Revision (ICD-10).
        https://www.cdc.gov/nchs/icd/icd10.htm
        Date accessed: June 22, 2022
      1. NHANES III Reference Manuals and Reports. Centers for Disease Control and Prevention, 1996
        • Johnson C.L.
        • Paulose-Ram R.
        • Ogden C.L.
        • et al.
        National Health and Nutrition Examination Survey: analytic guidelines, 1999-2010.
        Vital Health Stat 2. 2013; : 1-24
        • Centers for Disease Control and Prevention, National Center for Health Statistics
        Adult tobacco use information glossary.
        • World Health Organization
        Global recommendations on physical activity for health.
        (Accessed on June 21, 2022)
        • US Dept of Agriculture, Agricultural Research Service
        USDA Food and Nutrient Database for Dietary Studies.
        • US Dept of Agriculture, Agricultural Research Service
        Food Patterns Equivalents Database.
        • Guenther P.M.
        • Kirkpatrick S.I.
        • Reedy J.
        • et al.
        The Healthy Eating Index-2010 is a valid and reliable measure of diet quality according to the 2010 Dietary Guidelines for Americans.
        J Nutr. 2014; 144: 399-407
        • National Cancer Institute
        How to choose an analysis method dependent on purpose.
        • US Dept of Agriculture, Economic Research Service
        The U.S. Adult Food Security Survey Module.
        • Opsomer J.D.
        • Jensen H.H.
        • Pan S.
        An evaluation of the U.S. Department of Agriculture food security measure with generalized linear mixed models.
        J Nutr. 2003; 133: 421-427
        • CDC
        Defining adult overweight & obesity.
        • Menke A.
        • Casagrande S.
        • Geiss L.
        • Cowie C.C.
        Prevalence of and trends in diabetes among adults in the United States, 1988-2012.
        JAMA. 2015; 314: 1021-1029
        • Whelton P.K.
        • Carey R.M.
        • Aronow W.S.
        • et al.
        2017 ACC/AHA/AAPA/ABC/ACPM/AGS/APhA/ASH/ASPC/NMA/PCNA Guideline for the Prevention, Detection, Evaluation, and Management of High Blood Pressure in Adults: Executive summary: a report of the American College of Cardiology/American Heart Association Task Force on Clinical Practice Guidelines.
        J Am Coll Cardiol. 2018; 71: 2273-2275
        • National Cholesterol Education Program
        Expert Panel on Detection, Evaluation, and Treatment of High Blood Cholesterol in Adults. Third Report of the National Cholesterol Education Program (NCEP) Expert Panel on Detection, Evaluation, and Treatment of High Blood Cholesterol in Adults (Adult Treatment Panel III) final report.
        Circulation. 2002; 106: 3143-3421
        • Ojo O.
        Nutrition and chronic conditions.
        Nutrients. 2019; 11: 459
        • Raghupathi W.
        • Raghupathi V.
        An empirical study of chronic diseases in the United States: a visual analytics approach.
        Int J Environ Res Public Health. 2018; 15: 431
      2. SAS/IML. Version 14.1. SAS Institute Inc, 2015
        • Chen H.J.
        • Wang Y.F.
        • Cheskin L.J.
        Relationship between frequency of eating and cardiovascular disease mortality in US adults: the NHANES III follow-up study.
        Ann Epidemiol. 2016; 26: 527-533
        • Titan S.M.O.
        • Bingham S.
        • Welch A.
        • et al.
        Frequency of eating and concentrations of serum cholesterol in the Norfolk population of the European prospective investigation into cancer (EPIC-Norfolk): cross sectional study.
        BMJ. 2001; 323: 1286-1288a
        • Abdollahi S.
        • Kazemi A.
        • de Souza R.J.
        • Clark C.C.T.
        • Soltani S.
        The effect of meal frequency on biochemical cardiometabolic factors: a systematic review and meta-analysis of randomized controlled trials.
        Clin Nutr (Edinburgh, Scotland). 2021; 40: 3170-3181
        • Vicari T.
        • van den Borne J.J.
        • Gerrits W.J.
        • Zbinden Y.
        • Blum J.W.
        Postprandial blood hormone and metabolite concentrations influenced by feeding frequency and feeding level in veal calves.
        Domest Anim Endocrinol. 2008; 34: 74-88
        • Chen H.
        • Zhang B.
        • Ge Y.
        • et al.
        Association between skipping breakfast and risk of cardiovascular disease and all cause mortality: a meta-analysis.
        Clin Nutr. 2020; 39: 2982-2988
        • Yamamoto R.
        • Tomi R.
        • Shinzawa M.
        • et al.
        Associations of skipping breakfast, lunch, and dinner with weight gain and overweight/obesity in university students: a retrospective cohort study.
        Nutrients. 2021; 13: 271
        • Zeballos E.
        • Todd J.E.
        The effects of skipping a meal on daily energy intake and diet quality.
        Public Health Nutr. 2020; 23: 3346-3355
        • Mendy V.L.
        • Vargas R.
        • Cannon-Smith G.
        • Payton M.
        • Enkhmaa B.
        • Zhang L.
        Food insecurity and cardiovascular disease risk factors among Mississippi adults.
        Int J Env Res Pub He. 2018; 15: 2016
        • Vercammen K.A.
        • Moran A.J.
        • McClain A.C.
        • Thorndike A.N.
        • Fulay A.P.
        • Rimm E.B.
        Food security and 10-year cardiovascular disease risk among U.S. adults.
        Am J Prev Med. 2019; 56: 689-697
        • Leidy H.J.
        • Campbell W.W.
        The effect of eating frequency on appetite control and food intake: brief synopsis of controlled feeding studies.
        J Nutr. 2011; 141: 154-157
        • Alhussain M.H.
        • Macdonald I.A.
        • Taylor M.A.
        Irregular meal-pattern effects on energy expenditure, metabolism, and appetite regulation: a randomized controlled trial in healthy normal-weight women.
        Am J Clin Nutr. 2016; 104: 21-32
        • Jovanovic A.
        • Gerrard J.
        • Taylor R.
        The second-meal phenomenon in type 2 diabetes.
        Diabetes Care. 2009; 32: 1199-1201
        • Liu C.
        • Xie B.
        • Chou C.P.
        • et al.
        Perceived stress, depression and food consumption frequency in the college students of China Seven Cities.
        Physiol Behav. 2007; 92: 748-754
        • Mikolajczyk R.T.
        • El Ansari W.
        • Maxwell A.E.
        Food consumption frequency and perceived stress and depressive symptoms among students in three European countries.
        Nutr J. 2009; 8: 31
        • Mattson M.P.
        Energy intake, meal frequency, and health: a neurobiological perspective.
        Annu Rev Nutr. 2005; 25: 237-260
        • Papadopoulos F.C.
        • Ekbom A.
        • Brandt L.
        • Ekselius L.
        Excess mortality, causes of death and prognostic factors in anorexia nervosa.
        Br J Psychiat. 2009; 194: 10-17

      Biography

      Y. Sun is an assistant professor, Department of Preventive Medicine, The University of Tennessee Health Science Center, Memphis; at the time of the study, she was a postdoctoral research scholar, Department of Epidemiology, College of Public Health, University of Iowa, Iowa City.
      S. Rong is an associate professor, Department of Nutrition and Food Hygiene, School of Public Health, Medical College, Wuhan University of Science and Technology, Wuhan, China.
      B. Liu is a postdoctoral research scholar, Department of Epidemiology, College of Public Health, University of Iowa, Iowa City.
      Y. Du is a PhD candidate, Department of Epidemiology, College of Public Health, University of Iowa, Iowa City.
      Y. Wu isa PhD candidate, Department of Epidemiology, College of Public Health, University of Iowa, Iowa City.
      L. Snetselaar is a professor and chair, Section of Preventive Nutrition Education, Department of Epidemiology, College of Public Health, University of Iowa, Iowa City.
      R. Wallace is a professor and the Irene Ensminger Stecher Professorship in Cancer Research, Department of Epidemiology, College of Public Health, University of Iowa, Iowa City.
      Q. Xiao is an assistant professor, Department of Health and Human Physiology, College of Liberal Arts and Sciences, University of Iowa, Iowa City.
      L. Chen is a postdoctoral research scholar, Department of Nutrition and Food Hygiene, School of Public Health, Tongji Medical College, Huazhong University of Science and Technology, Wuhan, China.
      W. Bao is an assistant professor, Department of Epidemiology, College of Public Health, and a member, Obesity Research and Education Initiative and Fraternal Order of Eagles Diabetes Research Center, University of Iowa, Iowa City.