Association Between Plant and Animal Protein Intake and Overall and Cause-Specific Mortality | Cardiology | JAMA Internal Medicine | JAMA Network
[Skip to Navigation]
Figure.  Risk of Overall Mortality Associated With Intake of Plant Protein According to Subgroups Among 237 036 Men and 179 068 Women
Risk of Overall Mortality Associated With Intake of Plant Protein According to Subgroups Among 237 036 Men and 179 068 Women

The illustrated analysis is detailed in the Results section at the first mention of the Figure. HR indicates hazard ratio; BMI, body mass index (calculated as weight in kilograms divided by height in meters squared).

aAchieved Bonferroni-corrected threshold (.05/21 = .0024).

Table 1.  Baseline Characteristics of Study Participants Stratified by Quintiles of Daily Dietary Plant Protein Intakea
Baseline Characteristics of Study Participants Stratified by Quintiles of Daily Dietary Plant Protein Intakea
Table 2.  Risk of Overall and Cause-Specific Mortality Associated With Daily Dietary Plant Protein Intake Among 237 036 Men
Risk of Overall and Cause-Specific Mortality Associated With Daily Dietary Plant Protein Intake Among 237 036 Men
Table 3.  Risk of Overall and Cause-Specific Mortality Associated With Daily Dietary Plant Protein Intake Among 179 068 Women
Risk of Overall and Cause-Specific Mortality Associated With Daily Dietary Plant Protein Intake Among 179 068 Women
Table 4.  Risk of Overall and Cause-Specific Mortality Associated With Substitution of 3% Dietary Energy Intake From Plant Protein for Various Animal Protein Sources Among 237 036 Men and 179 068 Womena
Risk of Overall and Cause-Specific Mortality Associated With Substitution of 3% Dietary Energy Intake From Plant Protein for Various Animal Protein Sources Among 237 036 Men and 179 068 Womena
1.
Rodriguez  NR.  Introduction to Protein Summit 2.0: continued exploration of the impact of high-quality protein on optimal health.   Am J Clin Nutr. 2015;101(6):1317S-1319S. doi:10.3945/ajcn.114.083980 PubMedGoogle Scholar
2.
Leidy  HJ, Clifton  PM, Astrup  A,  et al.  The role of protein in weight loss and maintenance.   Am J Clin Nutr. 2015;101(6):1320S-1329S. doi:10.3945/ajcn.114.084038 PubMedGoogle Scholar
3.
Westerterp-Plantenga  MS, Nieuwenhuizen  A, Tomé  D, Soenen  S, Westerterp  KR.  Dietary protein, weight loss, and weight maintenance.   Annu Rev Nutr. 2009;29:21-41. doi:10.1146/annurev-nutr-080508-141056 PubMedGoogle Scholar
4.
Wycherley  TP, Moran  LJ, Clifton  PM, Noakes  M, Brinkworth  GD.  Effects of energy-restricted high-protein, low-fat compared with standard-protein, low-fat diets: a meta-analysis of randomized controlled trials.   Am J Clin Nutr. 2012;96(6):1281-1298. doi:10.3945/ajcn.112.044321 PubMedGoogle Scholar
5.
Santesso  N, Akl  EA, Bianchi  M,  et al.  Effects of higher- versus lower-protein diets on health outcomes: a systematic review and meta-analysis.   Eur J Clin Nutr. 2012;66(7):780-788. doi:10.1038/ejcn.2012.37 PubMedGoogle Scholar
6.
Tielemans  SM, Altorf-van der Kuil  W, Engberink  MF,  et al.  Intake of total protein, plant protein and animal protein in relation to blood pressure: a meta-analysis of observational and intervention studies.   J Hum Hypertens. 2013;27(9):564-571. doi:10.1038/jhh.2013.16 PubMedGoogle Scholar
7.
Rebholz  CM, Friedman  EE, Powers  LJ, Arroyave  WD, He  J, Kelly  TN.  Dietary protein intake and blood pressure: a meta-analysis of randomized controlled trials.   Am J Epidemiol. 2012;176(suppl 7):S27-S43. doi:10.1093/aje/kws245 PubMedGoogle Scholar
8.
Dong  JY, Zhang  ZL, Wang  PY, Qin  LQ.  Effects of high-protein diets on body weight, glycaemic control, blood lipids and blood pressure in type 2 diabetes: meta-analysis of randomised controlled trials.   Br J Nutr. 2013;110(5):781-789. doi:10.1017/S0007114513002055 PubMedGoogle Scholar
9.
Wu  G.  Dietary protein intake and human health.   Food Funct. 2016;7(3):1251-1265. doi:10.1039/C5FO01530H PubMedGoogle Scholar
10.
Richter  CK, Skulas-Ray  AC, Champagne  CM, Kris-Etherton  PM.  Plant protein and animal proteins: do they differentially affect cardiovascular disease risk?   Adv Nutr. 2015;6(6):712-728. doi:10.3945/an.115.009654 PubMedGoogle Scholar
11.
Kelemen  LE, Kushi  LH, Jacobs  DR  Jr, Cerhan  JR.  Associations of dietary protein with disease and mortality in a prospective study of postmenopausal women.   Am J Epidemiol. 2005;161(3):239-249. doi:10.1093/aje/kwi038 PubMedGoogle Scholar
12.
Song  M, Fung  TT, Hu  FB,  et al.  Association of animal and plant protein intake with all-cause and cause-specific mortality.   JAMA Intern Med. 2016;176(10):1453-1463. doi:10.1001/jamainternmed.2016.4182 PubMedGoogle Scholar
13.
Budhathoki  S, Sawada  N, Iwasaki  M,  et al; Japan Public Health Center–based Prospective Study Group.  Association of animal and plant protein intake with all-cause and cause-specific mortality in a Japanese cohort.   JAMA Intern Med. 2019;179(11):1509-1518. doi:10.1001/jamainternmed.2019.2806 PubMedGoogle Scholar
14.
Schatzkin  A, Subar  AF, Thompson  FE,  et al.  Design and serendipity in establishing a large cohort with wide dietary intake distributions: the National Institutes of Health-American Association of Retired Persons Diet and Health Study.   Am J Epidemiol. 2001;154(12):1119-1125. doi:10.1093/aje/154.12.1119 PubMedGoogle Scholar
15.
Thompson  FE, Kipnis  V, Midthune  D,  et al.  Performance of a food-frequency questionnaire in the US NIH-AARP (National Institutes of Health-American Association of Retired Persons) Diet and Health Study.   Public Health Nutr. 2008;11(2):183-195. doi:10.1017/S1368980007000419 PubMedGoogle Scholar
16.
Willett  WC, Howe  GR, Kushi  LH.  Adjustment for total energy intake in epidemiologic studies.   Am J Clin Nutr. 1997;65(4)(suppl):1220S-1228S. doi:10.1093/ajcn/65.4.1220S PubMedGoogle Scholar
17.
Brown  CC, Kipnis  V, Freedman  LS, Hartman  AM, Schatzkin  A, Wacholder  S.  Energy adjustment methods for nutritional epidemiology: the effect of categorization.   Am J Epidemiol. 1994;139(3):323-338. doi:10.1093/oxfordjournals.aje.a117000 PubMedGoogle Scholar
18.
Willcox  BJ, Yano  K, Chen  R,  et al.  How much should we eat? the association between energy intake and mortality in a 36-year follow-up study of Japanese-American men.   J Gerontol A Biol Sci Med Sci. 2004;59(8):789-795. doi:10.1093/gerona/59.8.B789 PubMedGoogle Scholar
19.
D’Agostino  RB  Jr.  Propensity score methods for bias reduction in the comparison of a treatment to a non-randomized control group.   Stat Med. 1998;17(19):2265-2281. doi:10.1002/(SICI)1097-0258(19981015)17:19<2265::AID-SIM918>3.0.CO;2-B PubMedGoogle Scholar
20.
Chalvon-Demersay  T, Azzout-Marniche  D, Arfsten  J,  et al.  A systematic review of the effects of plant compared with animal protein sources on features of metabolic syndrome.   J Nutr. 2017;147(3):281-292. doi:10.3945/jn.116.239574 PubMedGoogle Scholar
21.
Holmes  MD, Pollak  MN, Willett  WC, Hankinson  SE.  Dietary correlates of plasma insulin-like growth factor I and insulin-like growth factor binding protein 3 concentrations.   Cancer Epidemiol Biomarkers Prev. 2002;11(9):852-861.PubMedGoogle Scholar
22.
Allen  NE, Appleby  PN, Davey  GK, Kaaks  R, Rinaldi  S, Key  TJ.  The associations of diet with serum insulin-like growth factor I and its main binding proteins in 292 women meat-eaters, vegetarians, and vegans.   Cancer Epidemiol Biomarkers Prev. 2002;11(11):1441-1448.PubMedGoogle Scholar
23.
Guasch-Ferré  M, Satija  A, Blondin  SA,  et al.  Meta-analysis of randomized controlled trials of red meat consumption in comparison with various comparison diets on cardiovascular risk factors.   Circulation. 2019;139(15):1828-1845. doi:10.1161/CIRCULATIONAHA.118.035225 PubMedGoogle Scholar
24.
Bernstein  AM, Sun  Q, Hu  FB, Stampfer  MJ, Manson  JE, Willett  WC.  Major dietary protein sources and risk of coronary heart disease in women.   Circulation. 2010;122(9):876-883. doi:10.1161/CIRCULATIONAHA.109.915165 PubMedGoogle Scholar
25.
Preis  SR, Stampfer  MJ, Spiegelman  D, Willett  WC, Rimm  EB.  Dietary protein and risk of ischemic heart disease in middle-aged men.   Am J Clin Nutr. 2010;92(5):1265-1272. doi:10.3945/ajcn.2010.29626 PubMedGoogle Scholar
26.
Bernstein  AM, Pan  A, Rexrode  KM,  et al.  Dietary protein sources and the risk of stroke in men and women.   Stroke. 2012;43(3):637-644. doi:10.1161/STROKEAHA.111.633404 PubMedGoogle Scholar
27.
Mariotti  F, Huneau  JF.  Plant and animal protein intakes are differentially associated with large clusters of nutrient intake that may explain part of their complex relation with CVD risk.   Adv Nutr. 2016;7(3):559-560. doi:10.3945/an.115.011932 PubMedGoogle Scholar
28.
Jennings  A, MacGregor  A, Welch  A, Chowienczyk  P, Spector  T, Cassidy  A.  Amino acid intakes are inversely associated with arterial stiffness and central blood pressure in women.   J Nutr. 2015;145(9):2130-2138. doi:10.3945/jn.115.214700 PubMedGoogle Scholar
29.
Magné  J, Huneau  JF, Tsikas  D,  et al.  Rapeseed protein in a high-fat mixed meal alleviates postprandial systemic and vascular oxidative stress and prevents vascular endothelial dysfunction in healthy rats.   J Nutr. 2009;139(9):1660-1666. doi:10.3945/jn.109.107441 PubMedGoogle Scholar
30.
Tang  WH, Wang  Z, Levison  BS,  et al.  Intestinal microbial metabolism of phosphatidylcholine and cardiovascular risk.   N Engl J Med. 2013;368(17):1575-1584. doi:10.1056/NEJMoa1109400 PubMedGoogle Scholar
31.
Wang  Z, Klipfell  E, Bennett  BJ,  et al.  Gut flora metabolism of phosphatidylcholine promotes cardiovascular disease.   Nature. 2011;472(7341):57-63. doi:10.1038/nature09922 PubMedGoogle Scholar
32.
Etemadi  A, Sinha  R, Ward  MH,  et al.  Mortality from different causes associated with meat, heme iron, nitrates, and nitrites in the NIH-AARP Diet and Health Study: population based cohort study.   BMJ. 2017;357:j1957. doi:10.1136/bmj.j1957 PubMedGoogle Scholar
33.
Bibbins-Domingo  K, Chertow  GM, Coxson  PG,  et al.  Projected effect of dietary salt reductions on future cardiovascular disease.   N Engl J Med. 2010;362(7):590-599. doi:10.1056/NEJMoa0907355 PubMedGoogle Scholar
34.
Lampe  JW.  Health effects of vegetables and fruit: assessing mechanisms of action in human experimental studies.   Am J Clin Nutr. 1999;70(3)(suppl):475S-490S. doi:10.1093/ajcn/70.3.475s PubMedGoogle Scholar
35.
Huang  J, Weinstein  SJ, Yu  K, Männistö  S, Albanes  D.  Serum beta carotene and overall and cause-specific mortality.   Circ Res. 2018;123(12):1339-1349. doi:10.1161/CIRCRESAHA.118.313409 PubMedGoogle Scholar
36.
van den Brandt  PA.  Red meat, processed meat, and other dietary protein sources and risk of overall and cause-specific mortality in The Netherlands Cohort Study.   Eur J Epidemiol. 2019;34(4):351-369. doi:10.1007/s10654-019-00483-9 PubMedGoogle Scholar
37.
Pan  A, Sun  Q, Bernstein  AM,  et al.  Red meat consumption and mortality: results from 2 prospective cohort studies.   Arch Intern Med. 2012;172(7):555-563. doi:10.1001/archinternmed.2011.2287 PubMedGoogle Scholar
38.
Wang  X, Lin  X, Ouyang  YY,  et al.  Red and processed meat consumption and mortality: dose-response meta-analysis of prospective cohort studies.   Public Health Nutr. 2016;19(5):893-905. doi:10.1017/S1368980015002062 PubMedGoogle Scholar
39.
Slavin  JL.  Mechanisms for the impact of whole grain foods on cancer risk.   J Am Coll Nutr. 2000;19(3)(suppl):300S-307S. doi:10.1080/07315724.2000.10718964 PubMedGoogle Scholar
40.
Flint  AJ, Hu  FB, Glynn  RJ,  et al.  Whole grains and incident hypertension in men.   Am J Clin Nutr. 2009;90(3):493-498. doi:10.3945/ajcn.2009.27460 PubMedGoogle Scholar
41.
McKeown  NM, Meigs  JB, Liu  S, Wilson  PW, Jacques  PF.  Whole-grain intake is favorably associated with metabolic risk factors for type 2 diabetes and cardiovascular disease in the Framingham Offspring Study.   Am J Clin Nutr. 2002;76(2):390-398. doi:10.1093/ajcn/76.2.390 PubMedGoogle Scholar
42.
Hollænder  PL, Ross  AB, Kristensen  M.  Whole-grain and blood lipid changes in apparently healthy adults: a systematic review and meta-analysis of randomized controlled studies.   Am J Clin Nutr. 2015;102(3):556-572. doi:10.3945/ajcn.115.109165 PubMedGoogle Scholar
43.
Aune  D, Keum  N, Giovannucci  E,  et al.  Whole grain consumption and risk of cardiovascular disease, cancer, and all cause and cause specific mortality: systematic review and dose-response meta-analysis of prospective studies.   BMJ. 2016;353:i2716. doi:10.1136/bmj.i2716 PubMedGoogle Scholar
44.
Huang  T, Xu  M, Lee  A, Cho  S, Qi  L.  Consumption of whole grains and cereal fiber and total and cause-specific mortality: prospective analysis of 367,442 individuals.   BMC Med. 2015;13:59. doi:10.1186/s12916-015-0294-7 PubMedGoogle Scholar
45.
Zhong  VW, Van Horn  L, Cornelis  MC,  et al.  Associations of dietary cholesterol or egg consumption with incident cardiovascular disease and mortality.   JAMA. 2019;321(11):1081-1095. doi:10.1001/jama.2019.1572 PubMedGoogle Scholar
46.
Pasiakos  SM, Agarwal  S, Lieberman  HR, Fulgoni  VL  III.  Sources and amounts of animal, dairy, and plant protein intake of US adults in 2007-2010.   Nutrients. 2015;7(8):7058-7069. doi:10.3390/nu7085322 PubMedGoogle Scholar
47.
Yancy  WS  Jr, Wang  CC, Maciejewski  ML.  Trends in energy and macronutrient intakes by weight status over four decades.   Public Health Nutr. 2014;17(2):256-265. doi:10.1017/S1368980012005423 PubMedGoogle Scholar
48.
Ford  ES, Dietz  WH.  Trends in energy intake among adults in the United States: findings from NHANES.   Am J Clin Nutr. 2013;97(4):848-853. doi:10.3945/ajcn.112.052662 PubMedGoogle Scholar
49.
Macdiarmid  J, Blundell  J.  Assessing dietary intake: who, what and why of under-reporting.   Nutr Res Rev. 1998;11(2):231-253. doi:10.1079/NRR19980017 PubMedGoogle Scholar
Limit 200 characters
Limit 25 characters
Conflicts of Interest Disclosure

Identify all potential conflicts of interest that might be relevant to your comment.

Conflicts of interest comprise financial interests, activities, and relationships within the past 3 years including but not limited to employment, affiliation, grants or funding, consultancies, honoraria or payment, speaker's bureaus, stock ownership or options, expert testimony, royalties, donation of medical equipment, or patents planned, pending, or issued.

Err on the side of full disclosure.

If you have no conflicts of interest, check "No potential conflicts of interest" in the box below. The information will be posted with your response.

Not all submitted comments are published. Please see our commenting policy for details.

Limit 140 characters
Limit 3600 characters or approximately 600 words
    5 Comments for this article
    EXPAND ALL
    Plant protein intake and subsequent mortality: a risk assessment
    Tomoyuki Kawada, MD | Nippon Medical School
    Huang et al. conducted a long-term prospective study to examine the effect of plant protein intake on all-cause and cause-specific mortality (1). Higher plant protein intake was significantly associated with small reductions in the risk of all-cause and cardiovascular disease mortality. Risk reduction of all-cause mortality by the replacement of 3% energy from animal protein with plant protein was 10% in both men and women, and risk reduction of cardiovascular disease mortality by the same amount of replacement were 11% in men and 12% in women. The replacement from specific sources such as egg protein and red meat protein with plant protein contributed to larger reduction in mortality, and the authors recommended dietary modifications for keeping health and longevity. The authors cited the following two prospective studies for specifying the significance of their study, and I want to describe the content of two references precisely to improve the understanding the relationship.

    Song et al. reported that high animal protein intake was positively associated with cardiovascular disease mortality and high plant protein intake was inversely associated with all-cause and cardiovascular disease mortality (2). They also recommended substitution of plant protein for animal protein, especially that from processed red meat, for keeping lower mortality. Budhathoki et al. evaluated the association between animal and plant protein intake and all-cause and cause-specific mortality (3). Higher plant protein intake was inversely associated with all-cause and cardiovascular disease mortality, but animal protein intake was not significantly associated with mortality. They also clarified that hazard ratios (95% confidence intervals) of substituting processed meat protein by plant protein for all-cause and cancer mortalities were 0.54 (0.38-0.75) and 0.50 (0.30-0.85), respectively.

    Although small discrepancies of the results existed in these studies, I think that dietary habits of taking plant protein might be related to other desirable lifestyles for keeping health, which would lead to subsequent longevity.


    References

    1. Huang J, Liao LM, Weinstein SJ, Sinha R, Graubard BI, Albanes D. Association Between Plant and Animal Protein Intake and Overall and Cause-Specific Mortality. JAMA Intern Med. 2020 doi:10.1001/jamainternmed.2020.2790

    2. Song M, Fung TT, Hu FB, et al. Association of Animal and Plant Protein Intake With All-Cause and Cause-Specific Mortality. JAMA Intern Med. 2016;176(10):1453-1463. doi:10.1001/jamainternmed.2016.4182

    3. Budhathoki S, Sawada N, Iwasaki M, et al. Association of Animal and Plant Protein Intake With All-Cause and Cause-Specific Mortality in a Japanese Cohort. JAMA Intern Med. 2019;179(11):1509-1518. doi:10.1001/jamainternmed.2019.2806
    CONFLICT OF INTEREST: None Reported
    READ MORE
    Subgroup analysis-Query
    Suryakanta Acharya, MD Radiation Oncology | Swami Vivekananada Cancer Hospital
    The association between plant protein intake and overall mortality was similar across the subgroups of smoking status, diabetes, fruit consumption, vitamin supplement use, and self-reported health status.

    I have query on this, if I am understanding correctly, a smoker & non-smoker have similar risk of death if both are taking similar plant protein. Kindly correct i am wrong.
    CONFLICT OF INTEREST: None Reported
    Re: Subgroup analysis-Query
    Jiaqi Huang, PhD | Division of Cancer Epidemiology and Genetics, National Cancer Institute
    Thank you for your interest in our paper, Dr. Acharya. To answer your question, it does not mean “a smoker & non-smoker have similar risk of death if both are taking similar plant protein”. With the plant protein association being similar across the smoking subgroups, it means that smoking status did not modify the observed protective association for plant protein intake. That is, both smokers and nonsmokers had similar inverse associations between plant protein intake and risk of mortality, not that higher intake led to similar smoker-nonsmoker death rates (which are well-known to differ substantially).
    CONFLICT OF INTEREST: None Reported
    Why does plant protein prevent accidental death?
    George Henderson | Auckland University of Technology
    The authors of the current analysis of data from the NIH-AARP cohort are to be commended for making the hazard ratios for accidental death available, as did the authors of a similar 2009 NIH-AARP analysis.[1, 2]
    Accidental death (defined as death by accidents, adverse effects, suicide, self-inflicted injury, homicide, and legal intervention) was associated with meat-eating in Sinha et al 2009 (though not directly measured in the paper of Huang et al). There was an increased risk associated with higher consumption of red meat (highest vs lowest quintile) in men HR 1.26 (1.04-1.54) p=0.008, but not in women. This
    HR for accidental death was of a similar size to the HR for CVD mortality and several other common causes of death in Sinha 2009; in the current paper it is greater than most at 0.76 (0.61-0.93) p=0.009.
    Could it be that the consumption of plant protein strongly prevents accidents, perhaps because of an effect on cognitive ability? This latter seems unlikely, given that heme iron had a protective association with Alzheimer's disease mortality, 0.84 (0.72,0.97) p=<0.05, in the 2017 NIH-AARP analysis of Etemadi et al, for which accidental death HRs were not supplied.[3]
    Plant protein intake is associated with a healthy lifestyle, and may signify conscientiousness and an aversion to risk-taking, including potential routes of infection exposure and dangerous or polluted occupations. Residual confounding in this case could therefore lead to an inverse association between plant protein intake and accidental and other causes of death. In contrast, consumption of meat, milk products, and cheese contributes to animal protein intake and is related to an unhealthier lifestyle which may include greater risk-taking behavior; residual confounding in this case could therefore lead to a positive association between animal protein intake and accidental and other causes of death. For this reason, it is important that HRs for mortality from injuries and sudden deaths always be provided in studies analyzing dietary behaviours that may be associated with the longevity-promoting trait of contentiousness.[4]


    [1] Huang J, Liao LM, Weinstein SJ, Sinha R, Graubard BI, Albanes D. Association Between Plant and Animal Protein Intake and Overall and Cause-Specific Mortality. JAMA Intern Med. Published online July 13, 2020. doi:10.1001/jamainternmed.2020.2790

    [2] Sinha R, Cross AJ, Graubard BI, Leitzmann MF, Schatzkin A. Meat intake and mortality: a prospective study of over half a million people. Arch Intern Med. 2009;169(6):562-571. doi:10.1001/archinternmed.2009.6

    [3]E temadi A, Sinha R, Ward MH, et al. Mortality from different causes associated with meat, heme iron, nitrates, and nitrites in the NIH-AARP Diet and Health Study: population based cohort study. BMJ. 2017;357:j1957. Published 2017 May 9. doi:10.1136/bmj.j1957

    [4] Hill PL, Turiano NA, Hurd MD, Mroczek DK, Roberts BW. Conscientiousness and longevity: an examination of possible mediators. Health Psychol. 2011;30(5):536-541. doi:10.1037/a0023859
    CONFLICT OF INTEREST: None Reported
    READ MORE
    "Other" Plant Proteins
    Kavitha Shankar, MS MBA (pursuing PhD) | Corporatation

    I enjoyed reading this paper. Thank you.

    I have one question. In the methods, could the authors clarify what they mean by other plant protein (41.7%)? A few examples of the plant proteins in this category would be helpful.

    CONFLICT OF INTEREST: None Reported
    Original Investigation
    July 13, 2020

    Association Between Plant and Animal Protein Intake and Overall and Cause-Specific Mortality

    Author Affiliations
    • 1Division of Cancer Epidemiology and Genetics, National Cancer Institute, National Institutes of Health, Department of Health and Human Services, Bethesda, Maryland
    JAMA Intern Med. 2020;180(9):1173-1184. doi:10.1001/jamainternmed.2020.2790
    Key Points

    Question  Does an association exist between dietary protein choice, particularly from various food sources, and long-term overall mortality or cause-specific mortality in the US population?

    Findings  In this cohort of 237 036 men and 179 068 women with 16 years of observation and nearly 78 000 deaths, greater intake of plant protein was significantly associated with lower overall mortality and cardiovascular disease mortality independent of several other risk factors.

    Meaning  This study provides evidence for public health recommendations regarding dietary modifications in choice of protein sources that may promote health and longevity.

    Abstract

    Importance  Although emphasis has recently been placed on the importance of high-protein diets to overall health, a comprehensive analysis of long-term cause-specific mortality in association with the intake of plant protein and animal protein has not been reported.

    Objective  To examine the associations between overall mortality and cause-specific mortality and plant protein intake.

    Design, Setting, and Participants  This prospective cohort study analyzed data from 416 104 men and women in the US National Institutes of Health–AARP Diet and Health Study from 1995 to 2011. Data were analyzed from October 2018 through April 2020.

    Exposures  Validated baseline food frequency questionnaire dietary information, including intake of plant protein and animal protein.

    Main Outcomes and Measures  Hazard ratios and 16-year absolute risk differences for overall mortality and cause-specific mortality.

    Results  The final analytic cohort included 237 036 men (57%) and 179 068 women. Their overall median (SD) ages were 62.2 (5.4) years for men and 62.0 (5.4) years for women. Based on 6 009 748 person-years of observation, 77 614 deaths (18.7%; 49 297 men and 28 317 women) were analyzed. Adjusting for several important clinical and other risk factors, greater dietary plant protein intake was associated with reduced overall mortality in both sexes (hazard ratio per 1 SD was 0.95 [95% CI, 0.94-0.97] for men and 0.95 [95% CI, 0.93-0.96] for women; adjusted absolute risk difference per 1 SD was −0.36% [95% CI, −0.48% to −0.25%] for men and −0.33% [95% CI, −0.48% to −0.21%] for women; hazard ratio per 10 g/1000 kcal was 0.88 [95% CI, 0.84-0.91] for men and 0.86 [95% CI, 0.82-0.90] for women; adjusted absolute risk difference per 10 g/1000 kcal was −0.95% [95% CI, −1.3% to −0.68%] for men and −0.86% [95% CI, −1.3% to −0.55%] for women; all P < .001). The association between plant protein intake and overall mortality was similar across the subgroups of smoking status, diabetes, fruit consumption, vitamin supplement use, and self-reported health status. Replacement of 3% energy from animal protein with plant protein was inversely associated with overall mortality (risk decreased 10% in both men and women) and cardiovascular disease mortality (11% lower risk in men and 12% lower risk in women). In particular, the lower overall mortality was attributable primarily to substitution of plant protein for egg protein (24% lower risk in men and 21% lower risk in women) and red meat protein (13% lower risk in men and 15% lower risk in women).

    Conclusions and Relevance  In this large prospective cohort, higher plant protein intake was associated with small reductions in risk of overall and cardiovascular disease mortality. Our findings provide evidence that dietary modification in choice of protein sources may influence health and longevity.

    Introduction

    Great emphasis has been placed on the human health effects of a high-quality protein diet owing to its proposed benefits regarding weight management, metabolism, and healthy aging.1 Previous studies, including meta-analyses of short-term clinical trials, have shown that high-protein diets lead to weight loss and fat mass loss, possibly from the negative energy balance that results from sustained satiety and energy expenditure.2-5 In addition, cardiovascular metrics, including blood pressure, circulating lipid and lipoprotein profiles, and glycemic modulation, have shown improvement with protein substitution for carbohydrate.4-8 Notwithstanding such evidence and interest in the beneficial health effects of dietary protein, recommendations regarding food sources of protein have yet to be made. A thorough analysis of disparate long-term health outcomes resulting from choices of dietary protein sources (ie, plant protein or animal protein) remains to be elucidated given the different dimensions of nutritional profiles, including macronutrients, micronutrients, polyphenols, and amino acid combinations from varied protein sources.9,10

    The association between dietary protein sources and mortality has been examined in only a few population-based studies.11-13 The Iowa Women’s Health Study showed that plant protein substitution for animal protein is associated with reduced coronary heart disease mortality but not with overall or cancer mortality.11 The Nurses’ Health Study (NHS) and Health Professionals Follow-up Study (HPFS) cohorts with 131 342 participants found lower overall and cardiovascular disease (CVD) mortality with greater plant protein intake,12 whereas reduced risk of mortality overall and mortality related to cancer or CVD was associated with a 3% energy substitution of plant protein for red meat protein in the Japan Public Health Center-Based Prospective Cohort Study of 70 696 participants.13 Additional evidence is needed regarding the association between the intake of plant protein and animal protein and mortality in a large cohort that examines specific plant and animal protein sources in detail as well as in population subgroups based on risk factors.

    To this end, we examined whether plant protein intake and animal protein intake from various sources are associated with overall and cause-specific mortality in the National Institutes of Health (NIH)-AARP Diet and Health Study, a large prospective cohort of more than 400 000 participants with more than 77 000 deaths.

    Methods
    Study Population

    Details of the NIH-AARP Diet and Health Study have been described.14 In brief, 617 119 participants 50 to 71 years of age who returned a baseline questionnaire were enrolled from 6 states (California, Florida, Louisiana, New Jersey, North Carolina, and Pennsylvania) and 2 metropolitan areas (Atlanta and Detroit) in 1995 and 1996. The study was approved by the Special Studies institutional review board of the US National Cancer Institute. Completion of the self-administered questionnaire was considered to imply informed consent to participate in the study and consistent with the Common Rule requirements. No one received compensation or was offered any incentive for participating in the study.

    Exposure Assessment

    At baseline, in addition to completing a questionnaire of demographic and lifestyle characteristics, participants completed the National Cancer Institute Diet History Questionnaire (DHQ) of 124 dietary items.14 The DHQ included 124 food items and portion sizes, and it was linked with the US Department of Agriculture 1994-1996 Continuing Survey of Food Intakes by Individuals food composition nutrient database to provide an accurate representation relative to the dietary habits and nutrition of the general US population. Validity and reliability of the DHQ has been confirmed and reported.15 Total dietary protein intake included plant protein and animal protein. Primary dietary sources for plant protein (mean percentages of total plant protein intake) included bread, cereal, and pasta (45.8%), nuts (4.5%), beans and legumes (8.0%), and other plant protein (41.7%). Primary dietary sources of animal protein (mean percentages of total animal protein intake) were red meat (eg, fresh and processed red meat) (30.6%), white meat (eg, poultry, fish, and processed white meat) (31.3%), and other animal protein (eg, dairy products [31.6%] and eggs [4.0%]). Dietary intake of protein, fiber, fruits, and vegetables were adjusted for total energy intake (per 1000 kcal intake) according to the nutrient-density approach.

    Mortality Assessment

    Vital status of participants was ascertained via linkage with the Social Security Administration Death Master File, and specific causes of death were identified through linkage with the National Death Index Plus, updated by the National Center for Health Statistics (for details, see eAppendix 1 in the Supplement).

    Statistical Analysis

    Applying the nutrient density method to adjust for caloric intake,16,17 we divided all nutritional variables by a standardized calorie intake (grams per 1000 kcal) in addition to including energy as a continuous covariate in the model because it can be independently associated with mortality.18 Person-time of observation was calculated for cohort participants from study entry in 1995 to 1996 until date of death or the end of follow-up (December 31, 2011), whichever occurred first. Cox proportional hazards regression models using person-time as the underlying time metric were used to estimate hazard ratios (HRs) and 95% CIs for the association between plant protein intake (per 1 SD or 10 g/1000 kcal increment) and mortality risk, including overall and cause-specific mortality. We used the nutrient-density model (ie, grams of protein per 1000 kcal dietary intake) and reported risk estimates for men and women separately. In the models of cause-specific mortality, mortality from causes other than the cause of interest was censored at the time of death. To test for proportional hazards, we included an interaction of intervals of follow-up time as categories with plant protein intake in the regression model, and P values for the significance of the interaction were determined from likelihood ratio tests. However, we did find several quantitative interactions with follow-up time (ie, significant risk estimates in the same direction, with only magnitudes differing by the mortality outcome and protein source), but not qualitative interactions (ie, risk estimates in opposite directions). Because interactions were only quantitative in nature, we report main analyses without stratifying by follow-up time intervals and separately present models stratified by follow-up time. For each HR from the primary analysis, the adjusted absolute risk difference (ARD) was estimated for each 1 SD or 10 g/1000 kcal increment of plant protein intake per day at the end of the observation follow-up of 16 years. The corresponding 95% CIs for ARD were calculated based on 100 bootstrap samples. To elucidate residual confounding status, propensity scores reflecting the association between plant protein intake and potential confounding covariables were generated and included in the models.19 To evaluate the possibility that the inverse mortality associations may be explained by other confounding factors, we further adjusted for consumption of sugar-sweetened beverages (ie, regular soft drinks and regular fruit drinks) and total carbohydrates, as well as median household income (as a census variable in group levels). To minimize the influence of potential bias from reverse causality, we also performed sensitivity analyses that excluded the first 5 years of follow-up.

    Sensitivity analyses were also performed stratified by age at entry (<60, 60 to <65, and ≥65 years), cigarette-smoking status (never, former, or current), diabetes (no or yes), body mass index (BMI [calculated as weight in kilograms divided by height in meters squared], <18.5, 18.5 to <25, 25 to <30, 30 to <35, and ≥35), alcohol consumption (0 to <1, 1 to <3, and ≥3 drinks per day), consumption of fruits and vegetables (median split low/high), supplemental vitamin use (no or yes), self-reported health status (poor to fair, good, and very good to excellent), postmenopausal hormone therapy use (for women; no or yes), and mortality follow-up period (0 to <5, 5 to <10, and ≥10 years). We used likelihood-ratio tests to evaluate P values for interactions by comparing Cox proportional hazards regression models with or without the cross-product terms for the preselected factors and dietary plant protein intake (per 1 SD).

    We examined mortality associations of substituting 3% of energy from plant protein with an equal decrement in animal protein from various sources, including red meat, white meat, dairy, and eggs. Plant, animal, and total protein were modeled as percentages of energy, using energy from the specific protein divided by total energy intake. The regression coefficient for plant protein in the model was the log hazard ratio for a 3% increment in energy from plant protein with an equal decrement in animal protein while total energy from protein was held constant. Missing values for any single covariable were less than 5%; we included a missing value indicator for each covariable in the models.

    All analyses were conducted from October 2018 through April 2020 using SAS software, version 9.4 (SAS Institute Inc). All P values are 2-sided at a type I error rate of 0.05. The Bonferroni correction threshold was used to account for multiple comparisons and define statistical significance (.05/18 = .0028 for primary and secondary analyses [9 tests each for men and for women], and .05/21 = .0024 for the subgroup interaction tests). Details of covariate adjustment and sensitivity analyses are described in eAppendix 1 in the Supplement.

    Results
    Participants

    After checking dietary questionnaire responses for data quality, 566 398 participants constituted the baseline cohort. After excluding those with proxy questionnaires (n = 15 760), previous diagnoses of cancer except nonmelanoma skin cancer (n = 51 334), self-reported end-stage kidney disease (n = 997), or heart disease or stroke (n = 73 549), and those with only a death record for cancer (n = 4253), extreme daily caloric consumption (n = 4372), or zero years of follow-up (n = 29), the final analytic cohort included 416 104 individuals comprising 237 036 men (57%) and 179 068 women. Their overall median (SD) ages were 62.2 (5.4) years for men and 62.0 (5.4) years for women.

    Dietary Plant and Animal Protein Intake, Diet, and Lifestyle Factors

    In our study, the median percentage of daily energy intake from total protein was 15.3%, and the median percentage of daily dietary protein intake was 40% from plant sources and 60% from animal protein sources (including 19% from dairy foods). Table 1 provides cohort baseline characteristics according to categories (quintiles) of plant protein intake. The median plant protein intake was 26.9 g/d (14.4 g/1000 kcal per day) for men and 21.6 g/d (14.9 g/1000 kcal per day) for women. Participants with higher plant protein intake were more likely to have diabetes, higher educational level, lower BMI and total energy intake, and higher fiber, fruit, and vegetable intake; be more physically active and use vitamin supplements; and be less likely to be current smokers or report their health as poor or fair (Table 1). Cohort characteristics by quintiles of animal protein intake are presented in eAppendix 2 and eTable 1 in the Supplement.

    Dietary Plant Protein Intake and Overall Mortality

    During 16 years of follow-up (median, 15.5 years; interquartile range, 15.5-15.8) with 6 009 748 accumulated person-years, there were 77 614 deaths (18.7%; 49 297 men and 28 317 women). These deaths included 28 099 from cancer (17 816 men and 10 283 women), 22 228 from CVD (14 647 men and 7581 women), of whom 17 901 were from heart disease and 3748 were from stroke; 5606 from respiratory disease (3135 men and 2471 women); 2923 from infectious disease (1796 men and/1127 women); 2614 from injuries and accidents (1878 men and 736 women); and 16 144 from all other causes combined (10 025 men and 6119 women).

    Plant protein intake was significantly inversely associated with age-adjusted mortality from all causes in both men (Table 2) and women (Table 3). After multivariable adjustment, inverse associations remained significant for overall mortality in both sexes. In men, the HRs were 0.95 (95% CI, 0.94-0.97) per 1 SD and 0.88 (95% CI, 0.84-0.91) per 10 g/1000 kcal increment, with adjusted ARDs (at 16 years of follow-up) of −0.36% (95% CI, −0.48% to −0.25%) per 1 SD and −0.95% (95% CI, −1.3% to −0.68%) per 10 g/1000 kcal increment (P < .001) (Table 2). Corresponding adjusted HRs in women were 0.95 (95% CI, 0.93-0.96) per 1 SD and 0.86 (95% CI, 0.82-0.90) per 10 g/1000 kcal increment, with adjusted ARDs of −0.33% (95% CI, −0.48% to −0.21%) per 1 SD and −0.86% (95% CI, −1.3% to −0.55%) per 10 g/1000 kcal increment (P < .001) (Table 3). The propensity score–adjusted risk estimates remained essentially unchanged compared with those from the multivariable-adjusted analyses for both men and women (for overall mortality, propensity score–adjusted HRs were 0.95 [95% CI, 0.94-0.97] per 1 SD for men and 0.92 [95% CI, 0.90-0.93] per 1 SD for women; both P < .001). Additional adjustment for consumption of sugar-sweetened beverages and total carbohydrates as well as for median household income did not materially change our association estimates (eTable 2 and eTable 3 in the Supplement).

    By contrast, animal protein intake was not associated with multivariable-adjusted mortality in men (HR, 0.99; 95% CI, 0.98-1.00; P = .18) or women (HR, 0.98; 95% CI, 0.97-1.00; P = .02) (eTable 4 in the Supplement).

    Dietary Plant Protein Intake and Cause-Specific Mortality

    Plant protein intake was inversely associated with mortality from all CVD, stroke, and “other causes combined,” representing risk reductions of 5% to 12% per 1-SD intake increment and 12% to 30% per 10 g/1000 kcal intake increment in men, and 7% to 12% per 1 SD and 17% to 29% per 10 g/1000 kcal in women (all P < .003). The adjusted ARDs at 16 years of follow-up for CVD mortality were −0.13% (95% CI, −0.22% to −0.05%) per 1-SD intake increment and −0.35% per (95% CI, −0.59% to −0.14%) per 10 g/1000 kcal increment in men, and −0.13% (95% CI, −0.22% to −0.053%) per 1 SD and −0.35% (95% CI, −0.57% to −0.15%) per 10 g/1000 kcal in women. The corresponding adjusted ARDs for stroke mortality were −0.07% (95% CI, −0.11% to −0.03%) per 1-SD intake increment and −0.17% (95% CI, −0.26% to −0.08%) per 10 g/1000 kcal increment in men, and −0.06% (95% CI, −0.1% to −0.019%) per 1-SD intake increment and −0.15% (95% CI, −0.25% to −0.052%) per 10 g/1000 kcal increment in women (Table 2 and Table 3). By contrast, plant protein intake was not significantly associated with mortality from cancer, heart disease, respiratory disease, infections or injuries/accidents in either men or women (all P > .003) (Table 2 and Table 3).

    Dietary Plant Protein Intake and Overall Mortality in Cohort Subgroups

    The Figure provides key findings from stratified multivariable adjusted analyses of cohort subgroups. As illustrated there, HRs of overall mortality are for 1-SD increment of plant protein intake. Multivariable analyses were adjusted for age at entry (continuous), BMI, alcohol consumption, smoking status (never, former with ≤20 cigarettes a day, former with >20 cigarettes a day, current with ≤20 cigarettes a day, current with >20 cigarettes a day, or missing), physical activity (never or rarely, 1-3 times per month, 1-2 times per week, 3-4 times per week, ≥5 times per week, or missing), race or ethnic group (non-Hispanic white, non-Hispanic black, other race/ethnicity), educational level (<high school, high school graduate, post–high school training or some college, college graduate or higher, or missing), marital status (yes vs no), diabetes, health status, vitamin supplement use, and daily dietary total energy, animal protein, saturated fat, polyunsaturated fat, monounsaturated fat, trans fat, fiber, vegetables, and fruits (all continuous). For women, risk estimates were additionally adjusted for postmenopausal hormone replacement therapy. For the end point of cancer mortality, the model was further adjusted for history of cancer in a first-degree relative (yes vs no). The value of P for interaction was assessed by the likelihood ratio test, entering the cross-product term of plant protein (continuous variable) and the stratification variables (ordinal variable) into the Cox proportional hazards regression model.

    The overall mortality association with a 1-SD increment of plant protein intake was largely similar across population subsets based on smoking status, diabetes, fruit consumption, vitamin supplement use, and self-reported health status in both men and women, and on use of postmenopausal hormone therapy for women. By contrast, significant differences were observed in subgroups of BMI and years of follow-up (in both sexes), alcohol consumption (men), and age and vegetable consumption (women) (Figure). We observed stronger inverse associations for those with BMI of 18.5 to 25 (men: HR, 0.94; 95% CI, 0.91-0.96; women: HR, 0.92; 95% CI, 0.89-0.95), follow-up time of less than 5 years (men: HR, 0.92; 95% CI, 0.90-0.95; women: HR, 0.90; 95% CI, 0.87-0.93), consumption of 3 drinks or more daily (HR, 0.89; 95% CI, 0.85-0.93) for men, and for women who were younger than 60 years of age (HR, 0.90; 95% CI, 0.87-0.94) or who reported lower dietary vegetable consumption (HR, 0.90; 95% CI, 0.88-0.93). Although we observed a stronger inverse plant protein intake–mortality association during the first 5 years of cohort follow-up, our primary findings were not materially altered by excluding the first 5 years of observation, with overall HRs among men of 0.95 (95% CI, 0.94-0.97) per 1-SD intake increment and 0.88 (95% CI, 0.84-0.95) per 10 g/1000 kcal and corresponding HRs of 0.96 (95% CI, 0.94-0.98) per 1 SD and 0.88 (95% CI, 0.84-0.93) per 10 g/1000 kcal for women (all P < .001).

    Substitution of 3% Energy From Plant Protein for Specific Animal Protein Sources

    Table 4 presents mortality data for substitution of 3% energy from plant protein for the equivalent amount of animal protein of various sources and shows a significant association with lower overall mortality (men: HR, 0.90; 95% CI, 0.88-0.93; women: HR, 0.90; 95% CI, 0.87-0.93; all P < .001), and CVD mortality (men: HR, 0.89; 95% CI, 0.85-0.94; women: HR, 0.88; 95% CI, 0.82-0.94; P < .001), including in men mortality from heart disease (HR, 0.91; 95% CI, 0.86-0.96; P < .001) and stroke (HR, 0.78; 95% CI, 0.68-0.90; P < .001).

    Examination of substitution of 3% energy from plant protein for specific sources of animal protein revealed that the mortality reduction appeared strongest with substitution for egg protein, with 15% to 39% lower risk in men and 17% to 34% lower risk in women for overall (24% lower risk in men and 21% lower risk in women), cancer, CVD, heart disease, and respiratory disease mortality (Table 4). Substitution of 3% energy from plant protein for red meat protein intake showed stronger mortality reductions of 11% to 21% in men and 11% to 25% in women for overall (13% lower risk in men and 15% lower risk in women), CVD, heart disease, and stroke mortality in both sexes, and cancer and respiratory disease mortality in women. A similar change from dairy protein to plant protein was associated with lower overall and CVD mortality in men and women, and heart disease and stroke mortality in men, with overall mortality HRs of 0.92 in both men and women (P < .001). However, substituting of 3% energy from plant protein for white meat protein was not associated with risk of mortality (Table 4).

    Substitution of 3% Energy From Specific Dietary Plant Protein Sources for Egg Protein and Red Meat Protein Intake

    In this sensitivity analysis of specific plant protein sources substituting 3% energy for egg protein and red meat protein, we observed lower overall and cause-specific mortality (with the exception of deaths from infection, injuries, and accidents) when exchanging bread, cereal, and pasta plant protein for egg protein, with risk reductions of 18% to 45% in men and 20% to 39% in women (eTable 5 in the Supplement). Similarly, we found 12% to 30% lower mortality in men, and 13% to 28% lower mortality in women, when substituting for red meat protein (again, except for deaths from infection in men and women, and from injuries and accidents in women) (eTable 6 in the Supplement).

    Discussion

    In this analysis of a large prospective cohort of men and women in the US with 16 years of observation, we found higher plant protein intake was associated with reduced risk of overall mortality, with men and women experiencing (respectively) 12% and 14% lower mortality per 10 g/1000 kcal intake increment (5% lower mortality per 1-SD increment). The inverse association was apparent for CVD and stroke mortality in both sexes, was independent of several risk factors, and was evident in most other cohort subgroups. Replacement of 3% energy from various animal protein sources with plant protein was associated with 10% decreased overall mortality in both sexes. Of note, substitution analyses suggested that replacement of egg protein and red meat protein with plant protein resulted in the most prominent protective associations for overall mortality, representing 24% and 21% lower risk for men and women, respectively, for egg protein replacement, and 13% and 15% lower risk for men and women for red meat protein replacement. The effect sizes of these risk estimates were small.

    Relatively few prospective studies have examined the association between dietary protein sources and overall mortality. Concordant with the present findings, the NHS and HPFS cohorts found an inverse association for higher plant protein intake and overall and CVD mortality, and reduced overall mortality for greater substitution of plant protein for animal protein.12 Increased plant protein intake was also associated with lower risk of total and CVD mortality in a cohort in Japan,13 whereas a study of postmenopausal women in the US showed that replacement of animal protein with plant protein led to decreased coronary heart disease mortality.11 Collectively, the present and previous studies indicate the importance and diverse effects of primary dietary protein sources, including beneficial associations for higher plant protein intake in long-term health outcomes.

    Considerable evidence has accumulated supporting a beneficial role for plant-based diets in the prevention of cardiovascular disease. Dietary plant protein has been associated with reduced CVD risk factors, including lower systolic and diastolic blood pressure, improved lipid and lipoprotein profiles, and decreased circulating concentrations of insulin-like growth factor-1.20-22 Consistent with our findings, a recent meta-analysis of randomized clinical trials has shown that substitution of red meat with protein from high-quality plant sources, including legumes, soy, nuts, and other foods, was associated with benefits on lipid profiles, including favorable reductions in blood total cholesterol and low-density lipoprotein cholesterol.23 Replacement of 1 standard serving of red meat with nuts and seeds daily reduced coronary heart disease risk by 30% in the NHS,24 and greater animal protein intake was associated with 25% higher risk of ischemic heart disease among healthy men in the HPFS.25 In addition, lower risks of coronary heart disease and stroke were associated with substituting other protein sources (ie, fish and poultry) for red meat in the analysis of the NHS and HPFS.24,26 Underlying mechanisms for the CVD mortality associations remain unclear, although different amino acid constituents, protein source-related endogenous responses, and coexisting bioactive nutrients in the plant food matrix have been suggested as contributory. For example, plant-based proteins contain higher concentrations of nonessential amino acids (eg, arginine and glycine) and fewer essential amino acids (eg, methionine, lysine, and tryptophan), which could potentially influence cardiovascular health through lower arterial stiffness or decreased systemic and vascular generation of reactive oxygen species.10,27-29 In addition, increased concentrations in circulation of the gut microbiome-generated metabolite trimethylamine N-oxide, which results from higher intake of L-carnitine and choline from animal-based diets,10 has been proposed to mediate the effect of protein sources on cardiovascular health associated with increased risk of major adverse cardiac events.30,31 Beyond such potential direct effects of protein per se, other bioactive nutrients in foods may influence health, including heme iron, nitrates and nitrites, and sodium in red and processed meat32,33 and unsaturated fatty acids, antioxidants, vitamins, and trace minerals in plant-based foods.34,35 Our findings of more pronounced inverse plant protein associations in women who were younger than 60 years, and in both men and women with a favorable BMI, suggest that those health effects can be overshadowed by, for example, metabolic impairment-associated oxidative stress and systemic chronic inflammation.

    Our observation of stronger inverse mortality associations for plant protein from bread, cereal, and pasta substitution of red meat protein and egg protein do not stand alone in the literature. Higher red and processed meat consumption have been consistently associated with increased premature death, possibly owing to systemic oxidative stress, inflammation, heme iron, and endogenous formation of N-nitroso compounds.32,36-38 Grains and cereal foods have high contents of bioactive substances, including phenolic compounds, antioxidants, vitamins, minerals, and phytoestrogens,39 and increased whole grain consumption has been favorably associated with reduced cardiovascular disease risk factors, including hypertension, total cholesterol and low-density lipoprotein cholesterol.40-42 Moreover, a meta-analysis of 17 prospective cohorts that included 150 328 deaths found greater consumption of total red meat and processed meat was associated with excess risk of overall, CVD, and cancer mortality,38 and lower total mortality risk and cause-specific mortality risk in persons having greater consumption of plant protein–rich whole grains and cereal fiber.43,44 Consistent with our findings, it has been shown that replacement of 3% energy from egg protein with plant protein was associated with decreased risk of overall mortality, with mortality risk reductions of 19% and 18% in US cohorts12 and a Japanese cohort,13 respectively. In addition, an analysis of 6 prospective US cohorts, showed that higher egg consumption was associated with increased risk of incident CVD and all-cause mortality in a dose-response manner.45

    Strengths and Limitations

    There are several strengths of our study. The large cohort sample size and number of events as well as long-term follow-up afforded substantial power to detect moderate associations between substitution of plant protein for animal protein and mortality risk and to assess effect modification by other risk factors. The large number of deaths permitted a robust examination of cause-specific mortality. Of note, approximately 40% of daily dietary protein was from plant sources and 60% was animal protein (including 19% dairy protein), which are similar to the plant and animal protein intakes of the US population.46 In addition, 15.3% of total calories (median) from protein consumption in our study is very similar to that in the general US population (15%-16%),47,48 supporting the representativeness of our population. Study limitations include our use of a food frequency questionnaire to assess intake of plant protein and animal protein, with inherent measurement error (eg, underreporting of dietary intakes49) that would underestimate associations as a result of nondifferential misclassification and bias risk estimates toward the null. Although 24-hour dietary recalls are reasonable validation instruments but not criterion standards (eg, duplicate plate measurements), and they entail substantial measurement error, a previous validation study of the NIH-AARP DHQ and 24-hour dietary recalls reported correlations for total protein intake of 0.43 in men and 0.50 in women, reflecting moderate validity and reliability.15 Our analysis was based on a single dietary assessment at baseline, with possible unknown dietary changes during the follow-up period. Although the quantitative plant protein substitution interaction with follow-up time results were significant, the mortality risk estimates for the periods of 0 to fewer than 5 years, 5 to fewer than 10 years, and 10 or more years were essentially similar, suggesting consistency of the associations over time. The vast majority of our study participants were non-Hispanic white, which reduces the generalizability of the findings to other racial/ethnic populations. The latest linkage of the AARP-NIH cohort to the National Death Index contained cause-specific mortality through December 31, 2011, with the current analysis being based on 237 036 men, 179 068 women, and 77 614 deaths during 16 years of follow-up. The next linkage has been delayed logistically by the coronavirus disease 2019 pandemic, with no precise estimate of when we might have access to those data. In addition, because our findings are unlikely to be materially altered by the additional events, the present analysis was conducted with the currently available follow-up. Finally, we cannot preclude the possibility of residual confounding, including from unmeasured factors, which may have biased the observed associations. We did, however, control for a wide range of known potential confounding factors and used a propensity score to adjust for variation between comparison groups, and our main findings remained essentially unchanged.

    Conclusions

    In conclusion, this large cohort investigation showed small but significant associations between higher intake of plant protein and lower overall and CVD mortality, with prominent inverse associations observed for replacement of egg protein and red meat protein with plant protein, particularly for plant protein derived from bread, cereal, and pasta. Findings from this and previous studies provide evidence that dietary modifications in choice of protein sources may promote health and longevity.

    Back to top
    Article Information

    Accepted for Publication: May 22, 2020.

    Corresponding Author: Jiaqi Huang, PhD (jiaqi.huang@nih.gov), and Demetrius Albanes, MD (daa@nih.gov), Division of Cancer Epidemiology and Genetics, National Cancer Institute, 9609 Medical Center Dr, 6E316, Bethesda, MD 20892.

    Published Online: July 13, 2020. doi:10.1001/jamainternmed.2020.2790

    Author Contributions: Drs Huang and Albanes had full access to all the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis.

    Concept and design: Huang, Albanes.

    Acquisition, analysis, or interpretation of data: All authors.

    Drafting of the manuscript: Huang, Graubard, Albanes.

    Critical revision of the manuscript for important intellectual content: Huang, Liao, Weinstein, Sinha, Albanes.

    Statistical analysis: Huang, Liao, Graubard.

    Obtained funding: Huang, Albanes.

    Administrative, technical, or material support: Huang, Liao, Weinstein, Sinha, Albanes.

    Supervision: Huang, Sinha, Albanes.

    Conflict of Interest Disclosures: None reported.

    Funding/Support: This research was supported in part by the Intramural Research Program of the US National Institutes of Health (NIH) National Cancer Institute.

    Role of the Funder/Sponsor: The funder had no role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation, review, or approval of the manuscript; and decision to submit the manuscript for publication.

    Disclaimer: The views expressed herein are solely those of the authors and do not necessarily reflect those of the Florida Cancer Data System or Florida Department of Health. The Pennsylvania Department of Health specifically disclaims responsibility for any analyses, interpretations, or conclusions.

    Additional Contributions: We are indebted to the participants in the NIH-AARP Diet and Health Study for their outstanding cooperation. Sigurd Hermansen, MA, and Kerry Grace Morrissey, MPH, from Westat provided study outcomes ascertainment and management. Leslie Carroll, BA, at Information Management Services assisted with data support and analysis. Participants were not financially compensated for their contribution.

    Additional Information: Cancer incidence data from the Atlanta metropolitan area were collected by the Georgia Center for Cancer Statistics, Department of Epidemiology, Rollins School of Public Health, Emory University, Atlanta, Georgia. Cancer incidence data from California were collected by the California Cancer Registry, California Department of Public Health’s Cancer Surveillance and Research Branch, Sacramento, California. Cancer incidence data from the Detroit metropolitan area were collected by the Michigan Cancer Surveillance Program, Community Health Administration, Lansing, Michigan. The Florida cancer incidence data used in this report were collected by the Florida Cancer Data System (Miami, Florida) under contract with the Florida Department of Health, Tallahassee, Florida. Cancer incidence data from Louisiana were collected by the Louisiana Tumor Registry, Louisiana State University Health Sciences Center School of Public Health, New Orleans, Louisiana. Cancer incidence data from New Jersey were collected by the New Jersey State Cancer Registry, The Rutgers Cancer Institute of New Jersey, New Brunswick, New Jersey. Cancer incidence data from North Carolina were collected by the North Carolina Central Cancer Registry, Raleigh, North Carolina. Cancer incidence data from Pennsylvania were supplied by the Division of Health Statistics and Research, Pennsylvania Department of Health, Harrisburg, Pennsylvania. Cancer incidence data from Arizona were collected by the Arizona Cancer Registry, Division of Public Health Services, Arizona Department of Health Services, Phoenix, Arizona. Cancer incidence data from Texas were collected by the Texas Cancer Registry, Cancer Epidemiology and Surveillance Branch, Texas Department of State Health Services, Austin, Texas. Cancer incidence data from Nevada were collected by the Nevada Central Cancer Registry, Division of Public and Behavioral Health, State of Nevada Department of Health and Human Services, Carson City, Nevada.

    References
    1.
    Rodriguez  NR.  Introduction to Protein Summit 2.0: continued exploration of the impact of high-quality protein on optimal health.   Am J Clin Nutr. 2015;101(6):1317S-1319S. doi:10.3945/ajcn.114.083980 PubMedGoogle Scholar
    2.
    Leidy  HJ, Clifton  PM, Astrup  A,  et al.  The role of protein in weight loss and maintenance.   Am J Clin Nutr. 2015;101(6):1320S-1329S. doi:10.3945/ajcn.114.084038 PubMedGoogle Scholar
    3.
    Westerterp-Plantenga  MS, Nieuwenhuizen  A, Tomé  D, Soenen  S, Westerterp  KR.  Dietary protein, weight loss, and weight maintenance.   Annu Rev Nutr. 2009;29:21-41. doi:10.1146/annurev-nutr-080508-141056 PubMedGoogle Scholar
    4.
    Wycherley  TP, Moran  LJ, Clifton  PM, Noakes  M, Brinkworth  GD.  Effects of energy-restricted high-protein, low-fat compared with standard-protein, low-fat diets: a meta-analysis of randomized controlled trials.   Am J Clin Nutr. 2012;96(6):1281-1298. doi:10.3945/ajcn.112.044321 PubMedGoogle Scholar
    5.
    Santesso  N, Akl  EA, Bianchi  M,  et al.  Effects of higher- versus lower-protein diets on health outcomes: a systematic review and meta-analysis.   Eur J Clin Nutr. 2012;66(7):780-788. doi:10.1038/ejcn.2012.37 PubMedGoogle Scholar
    6.
    Tielemans  SM, Altorf-van der Kuil  W, Engberink  MF,  et al.  Intake of total protein, plant protein and animal protein in relation to blood pressure: a meta-analysis of observational and intervention studies.   J Hum Hypertens. 2013;27(9):564-571. doi:10.1038/jhh.2013.16 PubMedGoogle Scholar
    7.
    Rebholz  CM, Friedman  EE, Powers  LJ, Arroyave  WD, He  J, Kelly  TN.  Dietary protein intake and blood pressure: a meta-analysis of randomized controlled trials.   Am J Epidemiol. 2012;176(suppl 7):S27-S43. doi:10.1093/aje/kws245 PubMedGoogle Scholar
    8.
    Dong  JY, Zhang  ZL, Wang  PY, Qin  LQ.  Effects of high-protein diets on body weight, glycaemic control, blood lipids and blood pressure in type 2 diabetes: meta-analysis of randomised controlled trials.   Br J Nutr. 2013;110(5):781-789. doi:10.1017/S0007114513002055 PubMedGoogle Scholar
    9.
    Wu  G.  Dietary protein intake and human health.   Food Funct. 2016;7(3):1251-1265. doi:10.1039/C5FO01530H PubMedGoogle Scholar
    10.
    Richter  CK, Skulas-Ray  AC, Champagne  CM, Kris-Etherton  PM.  Plant protein and animal proteins: do they differentially affect cardiovascular disease risk?   Adv Nutr. 2015;6(6):712-728. doi:10.3945/an.115.009654 PubMedGoogle Scholar
    11.
    Kelemen  LE, Kushi  LH, Jacobs  DR  Jr, Cerhan  JR.  Associations of dietary protein with disease and mortality in a prospective study of postmenopausal women.   Am J Epidemiol. 2005;161(3):239-249. doi:10.1093/aje/kwi038 PubMedGoogle Scholar
    12.
    Song  M, Fung  TT, Hu  FB,  et al.  Association of animal and plant protein intake with all-cause and cause-specific mortality.   JAMA Intern Med. 2016;176(10):1453-1463. doi:10.1001/jamainternmed.2016.4182 PubMedGoogle Scholar
    13.
    Budhathoki  S, Sawada  N, Iwasaki  M,  et al; Japan Public Health Center–based Prospective Study Group.  Association of animal and plant protein intake with all-cause and cause-specific mortality in a Japanese cohort.   JAMA Intern Med. 2019;179(11):1509-1518. doi:10.1001/jamainternmed.2019.2806 PubMedGoogle Scholar
    14.
    Schatzkin  A, Subar  AF, Thompson  FE,  et al.  Design and serendipity in establishing a large cohort with wide dietary intake distributions: the National Institutes of Health-American Association of Retired Persons Diet and Health Study.   Am J Epidemiol. 2001;154(12):1119-1125. doi:10.1093/aje/154.12.1119 PubMedGoogle Scholar
    15.
    Thompson  FE, Kipnis  V, Midthune  D,  et al.  Performance of a food-frequency questionnaire in the US NIH-AARP (National Institutes of Health-American Association of Retired Persons) Diet and Health Study.   Public Health Nutr. 2008;11(2):183-195. doi:10.1017/S1368980007000419 PubMedGoogle Scholar
    16.
    Willett  WC, Howe  GR, Kushi  LH.  Adjustment for total energy intake in epidemiologic studies.   Am J Clin Nutr. 1997;65(4)(suppl):1220S-1228S. doi:10.1093/ajcn/65.4.1220S PubMedGoogle Scholar
    17.
    Brown  CC, Kipnis  V, Freedman  LS, Hartman  AM, Schatzkin  A, Wacholder  S.  Energy adjustment methods for nutritional epidemiology: the effect of categorization.   Am J Epidemiol. 1994;139(3):323-338. doi:10.1093/oxfordjournals.aje.a117000 PubMedGoogle Scholar
    18.
    Willcox  BJ, Yano  K, Chen  R,  et al.  How much should we eat? the association between energy intake and mortality in a 36-year follow-up study of Japanese-American men.   J Gerontol A Biol Sci Med Sci. 2004;59(8):789-795. doi:10.1093/gerona/59.8.B789 PubMedGoogle Scholar
    19.
    D’Agostino  RB  Jr.  Propensity score methods for bias reduction in the comparison of a treatment to a non-randomized control group.   Stat Med. 1998;17(19):2265-2281. doi:10.1002/(SICI)1097-0258(19981015)17:19<2265::AID-SIM918>3.0.CO;2-B PubMedGoogle Scholar
    20.
    Chalvon-Demersay  T, Azzout-Marniche  D, Arfsten  J,  et al.  A systematic review of the effects of plant compared with animal protein sources on features of metabolic syndrome.   J Nutr. 2017;147(3):281-292. doi:10.3945/jn.116.239574 PubMedGoogle Scholar
    21.
    Holmes  MD, Pollak  MN, Willett  WC, Hankinson  SE.  Dietary correlates of plasma insulin-like growth factor I and insulin-like growth factor binding protein 3 concentrations.   Cancer Epidemiol Biomarkers Prev. 2002;11(9):852-861.PubMedGoogle Scholar
    22.
    Allen  NE, Appleby  PN, Davey  GK, Kaaks  R, Rinaldi  S, Key  TJ.  The associations of diet with serum insulin-like growth factor I and its main binding proteins in 292 women meat-eaters, vegetarians, and vegans.   Cancer Epidemiol Biomarkers Prev. 2002;11(11):1441-1448.PubMedGoogle Scholar
    23.
    Guasch-Ferré  M, Satija  A, Blondin  SA,  et al.  Meta-analysis of randomized controlled trials of red meat consumption in comparison with various comparison diets on cardiovascular risk factors.   Circulation. 2019;139(15):1828-1845. doi:10.1161/CIRCULATIONAHA.118.035225 PubMedGoogle Scholar
    24.
    Bernstein  AM, Sun  Q, Hu  FB, Stampfer  MJ, Manson  JE, Willett  WC.  Major dietary protein sources and risk of coronary heart disease in women.   Circulation. 2010;122(9):876-883. doi:10.1161/CIRCULATIONAHA.109.915165 PubMedGoogle Scholar
    25.
    Preis  SR, Stampfer  MJ, Spiegelman  D, Willett  WC, Rimm  EB.  Dietary protein and risk of ischemic heart disease in middle-aged men.   Am J Clin Nutr. 2010;92(5):1265-1272. doi:10.3945/ajcn.2010.29626 PubMedGoogle Scholar
    26.
    Bernstein  AM, Pan  A, Rexrode  KM,  et al.  Dietary protein sources and the risk of stroke in men and women.   Stroke. 2012;43(3):637-644. doi:10.1161/STROKEAHA.111.633404 PubMedGoogle Scholar
    27.
    Mariotti  F, Huneau  JF.  Plant and animal protein intakes are differentially associated with large clusters of nutrient intake that may explain part of their complex relation with CVD risk.   Adv Nutr. 2016;7(3):559-560. doi:10.3945/an.115.011932 PubMedGoogle Scholar
    28.
    Jennings  A, MacGregor  A, Welch  A, Chowienczyk  P, Spector  T, Cassidy  A.  Amino acid intakes are inversely associated with arterial stiffness and central blood pressure in women.   J Nutr. 2015;145(9):2130-2138. doi:10.3945/jn.115.214700 PubMedGoogle Scholar
    29.
    Magné  J, Huneau  JF, Tsikas  D,  et al.  Rapeseed protein in a high-fat mixed meal alleviates postprandial systemic and vascular oxidative stress and prevents vascular endothelial dysfunction in healthy rats.   J Nutr. 2009;139(9):1660-1666. doi:10.3945/jn.109.107441 PubMedGoogle Scholar
    30.
    Tang  WH, Wang  Z, Levison  BS,  et al.  Intestinal microbial metabolism of phosphatidylcholine and cardiovascular risk.   N Engl J Med. 2013;368(17):1575-1584. doi:10.1056/NEJMoa1109400 PubMedGoogle Scholar
    31.
    Wang  Z, Klipfell  E, Bennett  BJ,  et al.  Gut flora metabolism of phosphatidylcholine promotes cardiovascular disease.   Nature. 2011;472(7341):57-63. doi:10.1038/nature09922 PubMedGoogle Scholar
    32.
    Etemadi  A, Sinha  R, Ward  MH,  et al.  Mortality from different causes associated with meat, heme iron, nitrates, and nitrites in the NIH-AARP Diet and Health Study: population based cohort study.   BMJ. 2017;357:j1957. doi:10.1136/bmj.j1957 PubMedGoogle Scholar
    33.
    Bibbins-Domingo  K, Chertow  GM, Coxson  PG,  et al.  Projected effect of dietary salt reductions on future cardiovascular disease.   N Engl J Med. 2010;362(7):590-599. doi:10.1056/NEJMoa0907355 PubMedGoogle Scholar
    34.
    Lampe  JW.  Health effects of vegetables and fruit: assessing mechanisms of action in human experimental studies.   Am J Clin Nutr. 1999;70(3)(suppl):475S-490S. doi:10.1093/ajcn/70.3.475s PubMedGoogle Scholar
    35.
    Huang  J, Weinstein  SJ, Yu  K, Männistö  S, Albanes  D.  Serum beta carotene and overall and cause-specific mortality.   Circ Res. 2018;123(12):1339-1349. doi:10.1161/CIRCRESAHA.118.313409 PubMedGoogle Scholar
    36.
    van den Brandt  PA.  Red meat, processed meat, and other dietary protein sources and risk of overall and cause-specific mortality in The Netherlands Cohort Study.   Eur J Epidemiol. 2019;34(4):351-369. doi:10.1007/s10654-019-00483-9 PubMedGoogle Scholar
    37.
    Pan  A, Sun  Q, Bernstein  AM,  et al.  Red meat consumption and mortality: results from 2 prospective cohort studies.   Arch Intern Med. 2012;172(7):555-563. doi:10.1001/archinternmed.2011.2287 PubMedGoogle Scholar
    38.
    Wang  X, Lin  X, Ouyang  YY,  et al.  Red and processed meat consumption and mortality: dose-response meta-analysis of prospective cohort studies.   Public Health Nutr. 2016;19(5):893-905. doi:10.1017/S1368980015002062 PubMedGoogle Scholar
    39.
    Slavin  JL.  Mechanisms for the impact of whole grain foods on cancer risk.   J Am Coll Nutr. 2000;19(3)(suppl):300S-307S. doi:10.1080/07315724.2000.10718964 PubMedGoogle Scholar
    40.
    Flint  AJ, Hu  FB, Glynn  RJ,  et al.  Whole grains and incident hypertension in men.   Am J Clin Nutr. 2009;90(3):493-498. doi:10.3945/ajcn.2009.27460 PubMedGoogle Scholar
    41.
    McKeown  NM, Meigs  JB, Liu  S, Wilson  PW, Jacques  PF.  Whole-grain intake is favorably associated with metabolic risk factors for type 2 diabetes and cardiovascular disease in the Framingham Offspring Study.   Am J Clin Nutr. 2002;76(2):390-398. doi:10.1093/ajcn/76.2.390 PubMedGoogle Scholar
    42.
    Hollænder  PL, Ross  AB, Kristensen  M.  Whole-grain and blood lipid changes in apparently healthy adults: a systematic review and meta-analysis of randomized controlled studies.   Am J Clin Nutr. 2015;102(3):556-572. doi:10.3945/ajcn.115.109165 PubMedGoogle Scholar
    43.
    Aune  D, Keum  N, Giovannucci  E,  et al.  Whole grain consumption and risk of cardiovascular disease, cancer, and all cause and cause specific mortality: systematic review and dose-response meta-analysis of prospective studies.   BMJ. 2016;353:i2716. doi:10.1136/bmj.i2716 PubMedGoogle Scholar
    44.
    Huang  T, Xu  M, Lee  A, Cho  S, Qi  L.  Consumption of whole grains and cereal fiber and total and cause-specific mortality: prospective analysis of 367,442 individuals.   BMC Med. 2015;13:59. doi:10.1186/s12916-015-0294-7 PubMedGoogle Scholar
    45.
    Zhong  VW, Van Horn  L, Cornelis  MC,  et al.  Associations of dietary cholesterol or egg consumption with incident cardiovascular disease and mortality.   JAMA. 2019;321(11):1081-1095. doi:10.1001/jama.2019.1572 PubMedGoogle Scholar
    46.
    Pasiakos  SM, Agarwal  S, Lieberman  HR, Fulgoni  VL  III.  Sources and amounts of animal, dairy, and plant protein intake of US adults in 2007-2010.   Nutrients. 2015;7(8):7058-7069. doi:10.3390/nu7085322 PubMedGoogle Scholar
    47.
    Yancy  WS  Jr, Wang  CC, Maciejewski  ML.  Trends in energy and macronutrient intakes by weight status over four decades.   Public Health Nutr. 2014;17(2):256-265. doi:10.1017/S1368980012005423 PubMedGoogle Scholar
    48.
    Ford  ES, Dietz  WH.  Trends in energy intake among adults in the United States: findings from NHANES.   Am J Clin Nutr. 2013;97(4):848-853. doi:10.3945/ajcn.112.052662 PubMedGoogle Scholar
    49.
    Macdiarmid  J, Blundell  J.  Assessing dietary intake: who, what and why of under-reporting.   Nutr Res Rev. 1998;11(2):231-253. doi:10.1079/NRR19980017 PubMedGoogle Scholar
    ×