Huang et al.

BMC Medicine (2015) 13:59

DOI 10.1186/s12916-015-0294-7

RESEARCH ARTICLE Open Access

Consumption of whole grains and cereal fiber

and total and cause-specific mortality:
prospective analysis of 367,442 individuals
Tao Huang1, Min Xu1, Albert Lee2, Susan Cho3 and Lu Qi1,4*

Abstract
Background: Intakes of whole grains and cereal fiber have been inversely associated with the risk of chronic
diseases; however, their relation with total and disease-specific mortality remain unclear. We aimed to prospectively
assess the association of whole grains and cereal fiber intake with all causes and cause-specific mortality.
Methods: The study included 367,442 participants from the prospective NIH-AARP Diet and Health Study (enrolled
in 1995 and followed through 2009). Participants with cancer, heart disease, stroke, diabetes, and self-reported
end-stage renal disease at baseline were excluded.
Results: Over an average of 14 years of follow-up, a total of 46,067 deaths were documented. Consumption of
whole grains were inversely associated with risk of all-cause mortality and death from cancer, cardiovascular disease
(CVD), diabetes, respiratory disease, infections, and other causes. In multivariable models, as compared with
individuals with the lowest intakes, those in the highest intake of whole grains had a 17% (95% CI, 14–19%) lower
risk of all-cause mortality and 11–48% lower risk of disease-specific mortality (all P for trend <0.023); those in the
highest intake of cereal fiber had a 19% (95% CI, 16–21%) lower risk of all-cause mortality and 15–34% lower risk of
disease-specific mortality (all P for trend <0.005). When cereal fiber was further adjusted, the associations of whole
grains with death from CVD, respiratory disease and infections became not significant; the associations with all-
cause mortality and death from cancer and diabetes were attenuated but remained significant (P for trend <0.029).
Conclusions: Consumption of whole grains and cereal fiber was inversely associated with reduced total and
cause-specific mortality. Our data suggest cereal fiber is one potentially protective component.
Keywords: Cereal fiber, Mortality, Whole grains

Background In epidemiology studies, evidence is accumulating in-

Grains, also called cereals, are the seeds of plants cul- dicating that consumption of whole grain products or
tivated for food. When whole, they include the germ, their effective components, especially dietary fiber found
bran, and endosperm [1]. Most whole grains are abun- in the grain, i.e., cereal fiber, may reduce the risk of
dant sources of dietary fiber and other nutrients, such chronic disease. Several recent meta-analyses taking in-
as minerals and antioxidants, which have shown benefi- to account a large number of subjects and prospective
cial effects on human health including improvement of studies showed significant and consistent protective ef-
weight loss, insulin sensitivity, and lipid profile, as well fects of high intake of whole grains and cereal fiber on
as inhibition of systemic inflammation [2-4]. type 2 diabetes [5], cardiovascular disease (CVD) [6],
and certain cancers (e.g., colorectal cancer) [7]. In our
earlier analysis in the Nurses’ Health Study, we ob-
* Correspondence: [email protected] served potential protective effects of whole grains on
1
Department of Nutrition, Harvard School of Public Health, 665 Huntington total or cardiovascular mortality in diabetic women [8].
Ave, Boston, MA 02115, USA
4
Channing Laboratory, Department of Medicine, Brigham and Women’s
Although the National Institutes of Health (NIH)-AARP
Hospital and Harvard Medical School, 75 Francis St, Boston, MA 02115, USA Diet and Health Study previously reported that dietary
Full list of author information is available at the end of the article

© 2015 Huang et al.; licensee BioMed Central. This is an Open Access article distributed under the terms of the Creative
Commons Attribution License (http://creativecommons.org/licenses/by/4.0), which permits unrestricted use, distribution, and
reproduction in any medium, provided the original work is properly credited. The Creative Commons Public Domain
Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article,
unless otherwise stated.
Huang et al. BMC Medicine (2015) 13:59 Page 2 of 9

fiber intake was inversely associated with the risk of total sizes, and nutrient database were constructed using the
death and death from CVD, infectious diseases, and US Department of Agriculture’s 1994–1996 Continuing
respiratory diseases [9], few studies have prospectively Survey of Food Intakes by Individuals. The FFQ used in
examined the associations of whole grains and its com- the study was calibrated using two non-consecutive
ponents, such as cereal fiber, with total or disease- 24-hour dietary recalls in 1953 NIH-AARP study par-
specific mortality. ticipants. The nutrient database for dietary fiber was
In the present study, we used data from 367,442 peo- based on AOAC methods.
ple who were at risk for a total of 12.3 million person- The whole grains were defined as the whole grain part
years. We aimed to provide reliable estimates of inde- of each product. The US Department of Agriculutre’s
pendent associations between baseline whole grains and Pyramid Servings Database enabled us to accurately esti-
cereal fiber intake and the risk of total or cause-specific mate whole-grain intake from all foods in the FFQ. The
death from CVD, cancers, diabetes, and other diseases. sources of whole-grain intake in the FFQ used in our
study were ready-to-eat cereals, high-fiber cereals, other
fiber cereals, whole-grain breads or dinner rolls, cooked
Methods
cereal, popcorn, pancakes, waffles, French toast or
Study population
crepes, rice or other cooked grains, bagels, English muf-
The NIH-AARP Diet and Health Study included
fins, tortillas, pasta, crackers, chips, cookies or brownies,
566,399 AARP members aged 50 to 71 from six US states
sweet pastries, and pies. In this Continuing Survey of Food
(California, Florida, Louisiana, New Jersey, North Carolina,
Intakes by Individuals dataset., whole grain foods were
and Pennsylvania) and two metropolitan areas (Atlanta,
defined as those containing at least 25% whole grains and/
Georgia, and Detroit, Michigan) [10]. Participants res-
or bran. Main fibers are from fruit, grains, vegetables, and
ponded to a questionnaire mailed in October 1995 and
beans in the present study. Cereal fiber was defined as
December 1997. Details of the NIH-AARP Study have
fiber from all cereals (e.g., ready-to-eat cereals, high-fiber
been previously described [11]. Among participants who
cereals, cooked cereal, and other fiber cereals) and grain-
returned the questionnaires with satisfactory dietary data,
based products.
we excluded individuals who indicated that they were
In specifying numbers of deaths by intake quintile,
proxies for the intended respondent (n = 15,760) as well
the number of deaths is determined by the energy-
as those who had cancer (n = 50,591), heart disease
adjusted intake quintile for the entire population; when
(n = 80,254), stroke (n = 12,812), diabetes (n = 52,647),
deaths are specific to a sex, we used the quintiles within
or self-reported end-stage renal disease at baseline (n =
the sex. We also collected demographic, anthropomet-
1,299). We also excluded those who reported extreme
ric, and lifestyle information, including history of smok-
consumption (>2 times the interquartile ranges of Box-
ing (the number of cigarettes smoked per day), time of
Cox transformed intake) of total energy (n = 3,771) and
smoking cessation (<1 years, 1 to 5 years, 5 to 10 years,
dietary fiber (n = 3,324). Exclusion of individuals report-
or ≥10 years before baseline), physical activity (never,
ing extreme energy intake is widely used in nutritional
rare, 1 to 2, 3 to 4, ≥5 hours/week), alcohol intake (g/day)
epidemiology studies since these participants are more
family history of cancers, menopausal hormone therapy
likely to over- or under-report their intake [12]. After
use in women, and some medical conditions at baseline.
exclusions (n = 198,957), the analytic cohort included
367,442 individuals. The NIH-AARP Diet and Health
Ascertaining mortality
study was approved by the Special Studies Institutional
The AARP dataset denotes date of death and cause of
Review Board of the US National Cancer Institute. All
death. There are 22 broad categories for cause of death.
participants provided written informed consent.
The modeling analysis for specific cause of death is per-
formed with the study end date of 2008. For total mor-
Assessment of dietary exposures tality, models for study end dates in 2008 and 2009 were
At baseline, dietary intake was assessed with a self- designed. Subjects with date of death after the study’s
administered 124-item food frequency questionnaire end date are treated as alive at the end of the study, with
(FFQ), which was an early version of the Diet History no death or cause of death in the model. When the
Questionnaire developed at the National Cancer Insti- study end date was 2008, and there was a date of death
tute [13,14]. Participants were asked to report their but no cause of death for 2008 or earlier, the subject was
usual frequency of intake and portion size over the past not included in the modeling for cause of death, but
12 months using 10 predefined frequency categories only total mortality. Thus, when specifying numbers of
ranging from never to 6+ times per day for beverages deaths it depends on both the study end date and whe-
and from never to 2+ times per day for solid foods with ther the cause of death field is missing. Vital status was
three portion size categories. The food items, portion determined through a periodic linkage of the cohort to
Huang et al. BMC Medicine (2015) 13:59 Page 3 of 9

the Social Security Administration Death Master File follow-up (total person-years, 5,148,760), we documented
and follow-up searches of the National Death Index Plus 46,067 deaths, among them 11,283 from CVD, 19,043
for participants who matched the Social Security Ad- from cancer, 371 from diabetes, 3,796 from respiratory
ministration Death Master File, cancer registry linkage, disease, 922 from infection, and 5,223 from other causes.
questionnaire responses, and responses to other mailings. At baseline, intakes of whole grains and cereal fiber were
The International Classification of Diseases, Ninth Revi- inversely correlated with prevalence of overweight, obesity,
sion [15] and the International Statistical Classification of and current smoking, as well as intake of red meat. The
Diseases, 10th Revision [16] were used to define death as levels of moderate and vigorous physical activity were
follows: cancer (ICD-9, 140–239; ICD-10, C00–C97 and higher among participants with higher intakes of whole
D00–D48), CVD (ICD-9, 390–398, 401–404, 410–429, grains or cereal fiber than those with lower intakes.
and 440–448; ICD-10, I00–I13, I20–I51, and I70–I78),
diabetes (ICD-9, 250; ICD-10, E10–E14), respiratory dis- Whole grains and cereal fiber intake with total mortality
ease (ICD-9, 480–487 and 490–496; ICD-10, J10–J18 and In age- and gender-adjusted analysis (Model 1), we found
J40–J47), infections (ICD-9, 001–139; ICD-10, A00–B99), that intake of whole grains were inversely associated with
and all other/unknown causes. all-cause mortality (Table 2). As compared with the lowest
quintile, the HRs across increasing quintiles of whole
Statistical analysis grain intake were 0.78 (95% CI, 0.76–0.80), 0.70 (95% CI,
We used the Cox proportional hazards model to esti- 0.68–0.72), 0.63 (95% CI, 0.61–0.65), and 0.61 (95% CI,
mate hazard ratios (HRs) and two-sided 95% confidence 0.59–0.62) (P trend <0.0001). Further adjustment for smo-
intervals (CIs) using the SAS PROC PHREG procedure king status and time since smoking cessation (Model 2)
(Version 9.1; SAS Institute Inc., Cary, NC, USA). Person- did not appreciably change the associations. When the
years of follow-up were calculated from the date of the models further included race/ethnicity, education, marital
baseline questionnaire until the date of death or the end status, self-rated health status, obesity (underweight, over-
of follow-up (December 31, 2009), whichever occurred weight, and obesity), physical activity, use of menopau-
first. Intake of whole grains and cereal fiber were ad- sal hormone therapy, and intake of alcohol, red meat,
justed for total energy intake using the residual method fruits, vegetables, and total energy (Model 3), the high-
[17], and were categorized into quintiles. est quintile of whole grain intake was associated with
We presented the results from four analysis models. 17% (95% CI, 14–19%) lower risk of all-cause mortality
Model 1, adjusted for age and gender; Model 2, adjusted (P trend <0.0001). The associations between whole grain
for age, gender, the number of cigarettes smoked per day, intake and all-cause mortality was attenuated, the highest
and time of smoking cessation (<1 years, 1 to 5 years, 5 to quintile of whole grain intake was associated with 6%
10 years, or ≥10 years before baseline); Model 3, adjusted (95% CI, 3–10%) lower risk, but remained significant
for age, gender, the number of cigarettes smoked per day, when cereal fiber was additionally adjusted (Model 4;
time of smoking cessation (<1 years, 1 to 5 years, 5 to P trend = 0.002). These results suggested that the pro-
10 years, or ≥10 years before baseline), race or ethnicity tective effects of whole grain may be due, at least in the
group, alcohol intake, education level, marital status (yes, main part, to its cereal fiber component.
no), health status (poor, fair, good, very good), obesity Similarly, we found that cereal fiber intake was sig-
(underweight, overweight, obesity), physical activity, con- nificantly associated with all-cause mortality in age-
sumption of red meat (processed and fresh meat), total and gender-adjusted and multivariate-adjusted models
fruit and total vegetables, total energy intake, and hor- (Models 1, 2 and 3; all P trend <0.0001; Table 3). In
mone usage; and Model 4, based on Model 3 further ad- model 3, the highest quintile of cereal intake was asso-
justed for cereal fiber (whole grains analysis). ciated with 19% (16–21%) lower risk of all-cause mor-
For missing data in each covariate, we created indi- tality (P trend <0.0001).
cator variables. Overall, missing data was less than 5%.
The model results summary includes the results of stat- Whole grains and cereal fiber intake with cause-specific
istical tests for trend in the response for the risk variable. mortality
Quintiles Trend P denotes the P value when the median We next tested the associations for cause-specific mor-
value within the risk variable quintile is included in the talities. In age- and gender-adjusted and multivariate ad-
hazard model as linear. justed models (Models 1, 2 and 3), intakes of whole
grains or cereal fiber were inversely associated with risk
Results of death from CVD, cancer, diabetes, respiratory disease,
Table 1 shows baseline characteristics of study parti- infections, and other/unknown causes (all P trend <0.023).
cipants (n = 367,442), according to intake of whole In Model 3, as compared with the lowest quintiles, people
grains and cereal fiber. During an average of 14 years of in the highest quintile of whole grain intake had 11%
Huang et al. BMC Medicine (2015) 13:59 Page 4 of 9

Table 1 Baseline characteristics of the study participants according to intake of whole grains and cereal fiber
Total Whole grains Cereal fiber
participants
Q1 Q3 Q5 Q1 Q3 Q5
n 367,442 73,488 73,489 73,489 73,488 73,489 73,489
Age, mean years 61.7 61.1 61.7 62.1 61.0 61.7 62.2
Female, % 43.9 30.7 51.9 40.4 35.0 49.9 40.2
Whites, % 92.9 92.6 92.9 93.2 91.2 93.0 94.4
College graduate, % 41.0 35.2 41.4 44.9 36.2 40.4 46.4
Married, % 68.1 72.7 65.2 68.9 69.8 66.1 69.9
Moderate physical activity (3–4 times/week), % 27.3 23.0 27.8 30.6 23.6 27.2 30.9
Vigorous physical activity (≥5 times/week), % 19.2 17.5 18.0 23.6 18.1 17.6 24.1
Overweight, % 42.4 44.5 42.0 41.0 43.9 42.2 41.1
Obesity, % 19.6 22.7 19.7 16.6 23.6 20.0 15.0
Very good or excellent self-report health, % 61.4 56.5 62.2 64.3 58.8 61.4 65.0
Previous or current use of postmenopausal hormone therapy, % 55.0 47.9 55.6 58.1 45.5 54.4 59.2
Former smoker, % 48.7 48.5 48.2 49.5 47.6 48.4 50.2
Current smoker, % 12.8 21.4 11.3 8.0 21.4 11.5 7.0
Median total energy intake (kcal/d) 1,805 2,394 1,527 1,855 2,330 1,563 1,832
Median alcohol intake (g/d) 14.7 32.5 10.0 8.8 27.5 11.6 9.9
Median servings of food
Red meat (oz/d) 2.0 3.1 1.6 1.7 3.1 1.71 1.6
Fruits (cup eq/d) 2.0 2.1 1.8 2.2 2.3 1.8 2.2
Vegetables (cup eq/d) 1.9 2.2 1.7 2.0 2.3 1.7 2.0

(respiratory disease) to 48% (diabetes) lower risk of cause- To the best of our knowledge, the present study is,
specific mortality, while people in the highest quintile of thus far, the largest in size regarding deaths in a prospec-
cereal fiber intake had 15% (cancer) to 34% (diabetes) tive setting. Our findings are concordant with previously
lower risk of cause-specific mortality. observed protective effects of whole grain intake on CVD,
When cereal fiber was further adjusted, the associations diabetes, and certain cancers [18,19]. Based on a meta-
of whole grains with death from CVD, respiratory disease, analysis of six cohort studies including 286,125 partici-
infections, and other causes became non-significant; how- pants and 10,944 cases, a two servings per day increment
ever, the associations with death from cancer and diabetes in whole grain consumption was associated with a 21%
remained significant (P trend <0.029). (95% CI, 13–28%) decrease in risk of type 2 diabetes after
adjustment for potential confounders and BMI [5]. These
Discussion findings were confirmed by Ye et al.’s meta-analysis [20],
In this large prospective cohort study of the US popula- in which it was also reported that compared with never/
tion, we found that high consumption of whole grains or rare consumers of whole grains, individuals consuming 48
cereal fiber was significantly associated with reduced risk to 80 g of whole grains per day (3 to 5 serving/day) had
of all-cause mortality and death from CVD, cancer, dia- a 21% lower risk of CVD (relative risk = 0.79; 95% CI,
betes, respiratory disease, infections, and other causes. 0.74–0.85). Inverse associations were also reported be-
As compared with individuals with the lowest intake of tween intake of whole grains and incident hypertension
whole grains, those in the highest intake group had a [21]. In a meta-analysis of 25 prospective studies, the
17% lower risk of all-cause mortality and 11 to 48% lower summary relative risk of developing colorectal cancer
risk of disease-specific mortality. As compared with indi- for 10 g daily of cereal fiber was 0.90 (95% CI, 0.83–0.97),
viduals with the lowest intake of cereal fiber, those in the pooled results from six studies showed the relative risk for
highest intake group had a 19% lower risk of all-cause an increment of three servings daily of whole grains was
mortality and 15 to 34% lower risk of disease-specific mor- 0.83 (95% CI, 0.78–0.89) [7]. High whole grain intakes
tality. Furthermore, our results suggested that the protect- have been related to reduced risk of other cancers, such as
ive effects of whole grains may due, at least in the main digestive cancer, in prospective studies, although the pro-
part, to its cereal fiber component. tective effects were not consistently observed [22,23].
Huang et al. BMC Medicine (2015) 13:59 Page 5 of 9

Table 2 Association of whole grain intake with total and cause-specific mortality
All Whole grains (oz eq/d) P trend
participants
Q1 (n = 41,248) Q2 (n = 41,248) Q3 (n = 41,249) Q4 (n = 41,248) Q5 (n = 41,249)
0.13 0.30 0.47 0.69 1.20
Causes of death
All cause
No. of deaths 46,067 11,845 9,450 8,694 8,054 8,024
Model 1 1.00 0.78 (0.76–0.80) 0.70 (0.68–0.72) 0.63(0.61–0.65) 0.61 (0.59–0.62) <0.0001
Model 2 1.00 0.88 (0.86–0.91) 0.83 (0.81–0.85) 0.78 (0.75–0.80) 0.77 (0.75–0.79) <0.0001
Model 3 1.00 0.93 (0.90–0.95) 0.89 (0.87–0.92) 0.85 (0.82–0.87) 0.83 (0.81–0.86) <0.0001
Model 4 1.00 0.96 (0.93–0.99) 0.95 (0.92–0.98) 0.92 (0.89–0.96) 0.94 (0.90–0.97) 0.002
Cardiovascular disease
No. of deaths 11,283 2,921 2,330 2,121 1,914 1,997 <0.0001
Model 1 1.00 0.78 (0.74–0.83) 0.69 (0.65–0.73) 0.60 (0.57–0.64) 0.60 (0.57–0.64) <0.0001
Model 2 1.00 0.88 (0.83–0.93) 0.81 (0.77–0.86) 0.73 (0.69–0.77) 0.75 (0.71–0.80) <0.0001
Model 3 1.00 0.93 (0.88–0.98) 0.88 (0.83–0.93) 0.81 (0.77–0.86) 0.83 (0.78–0.88) <0.0001
Model 4 1.00 0.96 (0.91–1.02) 0.95 (0.89–1.01) 0.90 (0.84–0.97) 0.95 (0.88–1.03) 0.188
Cancer
No. of deaths 19,043 4,836 3,912 3,616 3,388 3,291
Model 1 1.00 0.79 (0.76–0.83) 0.71 (0.68–0.75) 0.65 (0.62–0.68) 0.61 (0.59–0.64) <0.0001
Model 2 1.00 0.91 (0.87–0.95) 0.86 (0.83–0.90) 0.82 (0.79–0.86) 0.80 (0.76–0.84) <0.0001
Model 3 1.00 0.94 (0.90–0.98) 0.91 (0.87–0.95) 0.88 (0.84–0.92) 0.85 (0.81–0.89) <0.0001
Model 4 1.00 0.96 (0.92–1.00) 0.95 (0.90–0.99) 0.93 (0.88–0.98) 0.93 (0.88–0.99) 0.025
Diabetes
No. of deaths 371 113 72 73 66 47
Model 1 1.00 0.62 (0.46–0.84) 0.62 (0.46–0.83) 0.54 (0.40–0.73) 0.37 (0.27–0.52) <0.0001
Model 2 1.00 0.67 (0.49–0.90) 0.67 (0.50–0.91) 0.60 (0.44–0.82) 0.42 (0.30–0.60) <0.0001
Model 3 1.00 0.71 (0.53–0.96) 0.76 (0.56–1.03) 0.72 (0.53–0.99) 0.52 (0.37–0.75) 0.0009
Model 4 1.00 0.74 (0.54–1.00) 0.81 (0.59–1.13) 0.78 (0.55–1.13) 0.57 (0.37–0.89) 0.029
Respiratory disease
No. of deaths 3,796 1,123 802 673 606 592
Model 1 1.00 0.69 (0.63–0.75) 0.56 (0.50–0.61) 0.48 (0.43–0.53) 0.45 (0.41–0.50) <0.0001
Model 2 1.00 0.88 (0.80–0.96) 0.79 (0.72–0.87) 0.74 (0.67–0.82) 0.74 (0.67–0.82) <0.0001
Model 3 1.00 0.99 (0.90–1.09) 0.94 (0.85–1.03) 0.91 (0.82–1.01) 0.89 (0.80–0.98) 0.0099
Model 4 1.00 1.02 (0.93–1.12) 1.00 (0.90–1.11) 1.01 (0.90–1.14) 1.03 (0.91–1.18) 0.67
Infections
No. of deaths 922 251 184 163 161 163
Model 1 1.00 0.71 (0.59–0.86) 0.61 (0.50–0.75) 0.58 (0.48–0.71) 0.57 (0.47–0.70) <0.0001
Model 2 1.00 0.78 (0.64–0.94) 0.69 (0.57–0.84) 0.68 (0.55–0.83) 0.68 (0.55–0.83) 0.0002
Model 3 1.00 0.84 (0.70–1.02) 0.78 (0.64–0.96) 0.79 (0.65–0.97) 0.77 (0.62–0.95) 0.02
Model 4 1.00 0.87 (0.71–1.06) 0.83 (0.67–1.04) 0.87 (0.69–1.10) 0.89 (0.68–1.16) 0.54
Huang et al. BMC Medicine (2015) 13:59 Page 6 of 9

Table 2 Association of whole grain intake with total and cause-specific mortality (Continued)
Other/unknown causes
No. of deaths 5,223 1,206 1,058 1,038 940 981
Model 1 1.00 0.86 (0.79–0.93) 0.82 (0.76–0.90) 0.71 (0.66–0.78) 0.72 (0.66–0.78) <0.0001
Model 2 1.00 0.91 (0.84–0.99) 0.90 (0.83–0.98) 0.80 (0.73–0.87) 0.81 (0.74–0.88) <0.0001
Model 3 1.00 0.97 (0.88–1.06) 0.97 (0.89–1.06) 0.87 (0.79–0.96) 0.86 (0.78–0.94) 0.0001
Model 4 1.00 0.99 (0.91–1.09) 1.03 (0.93–1.13) 0.96 (0.86–1.06) 0.98 (0.87–1.09) 0.54
Data are hazard ratios (HR) and 95% confidence interval (CI). The numbers of deaths are for participants who died during follow-up. The co-variables are
baseline assessments.
Model 1, Adjusted for age and gender; Model 2, Adjusted for age, gender, the number of cigarettes smoked per day, and time of smoking cessation (<1 years,
1 to 5 years, 5 to 10 years, or ≥10 years before baseline); Model 3, Adjusted for age, gender, the number of cigarettes smoked per day, time of smoking cessation
(<1 years, 1 to 5 years, 5 to 10 years, or ≥10 years before baseline), race or ethnicity group, alcohol intake, education level, marital status (yes, no), health status
(poor, fair, good, very good), obesity (underweight, overweight, obesity), physical activity, consumption of red meat, total fruit and total vegetables, total energy
intake, and hormone usage. Model 4, Based on Model 3 further adjusted for cereal fiber.

Very few previous studies have examined the relation- tumor necrosis factor α receptor 2, which play key roles
ship between whole grains and their components with in chronic inflammatory conditions [29,30]. Whole grain
mortality in humans. Our findings are consistent with foods are rich in fiber. Therefore, the anti-inflammatory
the results reported in the Nurses’ Health Study, in effects of dietary fiber may help explain, at least in part,
which whole grain intake, especially bran, was associ- the inverse associations of whole grains and fiber con-
ated with lower all-cause and CVD mortality in diabetic sumption with chronic disease death. Moreover, whole
women [24]. Similarly, higher fiber intake was associa- grains and cereal fiber have a high content of antioxidants,
ted with lower total mortality, particularly mortality from vitamins, trace minerals, phenolic acids, lignans, and
circulatory, digestive, and non-CVD non-cancer inflam- phytoestrogens, which have been associated with a re-
matory diseases in a large European prospective study duced risk of colorectal cancer [31] and lower risk of
of 452,717 men and women [25]. In a previous analysis death from non-cardiovascular, non-cancer inflamma-
among our study samples, it was found that intake of tory diseases and respiratory system diseases [32]. In
fiber from grains but not from other sources was inversely addition, dietary fibers have specific and unique impacts
related to all-cause mortality and death from cancer, CVD, on intestinal microbiota composition and metabolism
infections, and respiratory disease [9]. In this updated ana- [33,34]. Additionally, recent studies have related gut mi-
lysis, we found cereal fiber intake was inversely associated crobiota with various chronic diseases such as obesity,
with death from diabetes. However, we did not report the CVD, diabetes, and cancer [34,35]. Further functional
associations of specific types of whole grain foods/pro- investigations are warranted to verify these potential
ducts with mortality and cause-specific mortality, since it mechanisms.
is hard to further differentiate such food groups; this pre-
sents a limitation of this observational study. Strengths and limitations of the study
In addition, we found that the associations of whole In our study cohort, both whole grains and cereal fiber
grains with death from CVD, respiratory disease, and in- were correlated with high levels of physical activity and
fections became non-significant after adjustment for better health status, as well as with low BMI, low levels
cereal fiber intake. The associations with total mortality of smoking, and low intakes of alcohol and red meat.
and death from cancer and diabetes were also largely at- However, our results were less likely due to the potential
tenuated, although they remained significant after adjust- confounding of these factors because careful adjustment
ment for cereal fiber intake. These observations suggest for these factors in our analyses did not significantly
that the protective effects of whole grains on mortality are change the results. Nevertheless, we acknowledge that
at least partly mediated by its cereal fiber component. the positive associations may still be related to residual
Such a postulation is supported by previous evidence that confounding of non-measured covariates. Reverse caus-
shows cereal fiber intake is related to an improvement of ality might also affect the associations, since people with
insulin sensitivity and lipid profile, an increase in protect- chronic disease might modify their eating habits by con-
ive molecules such as adiponectin, and a reduction in in- suming healthy foods including those rich in whole
flammation markers [26-28]. grains and cereal fiber. In our analyses, however, we have
The protective effect of whole grains and fiber con- excluded patients with cancer, heart disease, and dia-
sumption on risk of mortality is biologically plausible. betes at baseline and only analyzed the associations with
Dietary fiber intake is associated with lower levels of incident cases. Whole grains and cereal fiber intakes
inflammation markers, such as C-reactive protein, and were evaluated by self-report at a single time point. It is
Huang et al. BMC Medicine (2015) 13:59 Page 7 of 9

Table 3 Association of cereal fiber intake with total and cause-specific mortality
All Cereal fiber (g/d) P trend
participants
Q1 (n = 73,488) Q2 (n = 73,488) Q3 (n = 73,489) Q4 (n = 73,488) Q5 (n = 73,489)
2.02 4.15 5.27 6.65 10.22
Causes of death
All cause
No. of deaths 46,067 11,700 9,652 8,664 8,133 7,918
Model 1 1.00 0.80 (0.78–0.83) 0.70 (0.68–0.72) 0.64 (0.62–0.66) 0.59 (0.58–0.61) <0.0001
Model 2 1.00 0.89 (0.87–0.92) 0.83 (0.81–0.85) 0.78 (0.76–0.81) 0.76 (0.73–0.78) <0.0001
Model 3 1.00 0.93 (0.90–0.95) 0.87 (0.85–0.90) 0.84 (0.81–0.86) 0.81 (0.79–0.84) <0.0001
Cardiovascular disease 11,283
No. of deaths 2,901 2,368 2,094 1,986 1,934
Model 1 1.00 0.80 (0.76–0.85) 0.69 (0.65–0.73) 0.63 (0.59–0.66) 0.57 (0.54–0.61) <0.0001
Model 2 1.00 0.88 (0.83–0.93) 0.80 (0.76–0.85) 0.76 (0.71–0.80) 0.72 (0.68–0.76) <0.0001
Model 3 1.00 0.93 (0.87–0.98) 0.86 (0.81–0.91) 0.83 (0.78–0.88) 0.80 (0.75–0.85) <0.0001
Cancer
No. of deaths 19,043 4,772 3,974 3,616 3,391 3,290
Model 1 1.00 0.81 (0.78–0.85) 0.72 (0.69–0.76) 0.66 (0.63–0.69) 0.61 (0.59–0.64) <0.0001
Model 2 1.00 0.91 (0.87–0.95) 0.87 (0.83–0.91) 0.83 (0.79–0.86) 0.80 (0.77–0.84) <0.0001
Model 3 1.00 0.94 (0.90–0.98) 0.90 (0.86–0.95) 0.87 (0.83–0.91) 0.85 (0.81–0.89) <0.0001
Diabetes
No. of deaths 371 92 92 72 57 51
Model 1 1.00 0.92 (0.69–1.22) 0.70 (0.52–0.95) 0.54 (0.39–0.74) 0.45 (0.32–0.64) <0.0001
Model 2 1.00 0.97 (0.73–1.29) 0.77 (0.56–1.04) 0.60 (0.43–0.83) 0.52 (0.37–0.73) <0.0001
Model 3 1.00 1.00 (0.74–1.35) 0.83 (0.60–1.15) 0.70 (0.50–0.99) 0.66 (0.46–0.94) 0.005
Respiratory disease
No. of deaths 3,796 1,082 866 652 649 547
Model 1 1.00 0.75 (0.68–0.82) 0.54 (0.49–0.59) 0.52 (0.47–0.57) 0.42 (0.38–0.46) <0.0001
Model 2 1.00 0.91(0.83–1.00) 0.76 (0.69–0.84) 0.80 (0.72–0.88) 0.70 (0.63–0.78) <0.0001
Model 3 1.00 0.98 (0.89–1.08) 0.82 (0.74–0.91) 0.89 (0.80–0.99) 0.79 (0.71–0.88) <0.0001
Infections
No. of deaths 922 230 210 165 157 160
Model 1 1.00 0.87 (0.72–1.05) 0.66 (0.54–0.81) 0.61 (0.50–0.75) 0.60 (0.49–0.74) <0.0001
Model 2 1.00 0.94 (0.78–1.13) 0.75 (0.61–0.91) 0.71 (0.58–0.87) 0.71 (0.58–0.88) 0.0002
Model 3 1.00 1.04 (0.85–1.27) 0.84 (0.68–1.05) 0.82 (0.66–1.02) 0.83 (0.67–1.03) 0.023
Other/unknown causes
No. of deaths 5,223 1,215 1,050 1,044 963 951
Model 1 1.00 0.84 (0.77–0.91) 0.81 (0.74–0.88) 0.72 (0.66–0.78) 0.67 (0.62–0.73) <0.0001
Model 2 1.00 0.88 (0.81–0.96) 0.88 (0.81–0.95) 0.79 (0.73–0.87) 0.76 (0.70–0.83) <0.0001
Model 3 1.00 0.90 (0.83–0.99) 0.91 (0.83–0.99) 0.83 (0.76–0.91) 0.79 (0.72–0.87) <0.0001
Data are hazard ratios (HR) and 95% confidence interval (CI). The numbers of deaths are for participants who died during follow-up. The co-variables are
baseline assessments.
Model 1, Adjusted for age and gender; Model 2, Adjusted for age, gender, the number of cigarettes smoked per day, and time of smoking cessation (<1 years,
1 to 5 years, 5 to 10 years, or ≥10 years before baseline); Model 3, Adjusted for age, gender, the number of cigarettes smoked per day, time of smoking cessation
(<1 years, 1 to 5 years, 5 to 10 years, or ≥10 years before baseline), race or ethnicity group, alcohol intake, education level, marital status (yes, no), health status
(poor, fair, good, very good), obesity (underweight, overweight, obesity), physical activity, consumption of red meat, total fruit and total vegetables, total energy
intake, and hormone usage.
Huang et al. BMC Medicine (2015) 13:59 Page 8 of 9

likely that dietary habits might change during the long 5. de Munter JS, Hu FB, Spiegelman D, Franz M, van Dam RM. Whole grain,
(14 years on average) follow-up period, and such temporal bran, and germ intake and risk of type 2 diabetes: a prospective cohort
study and systematic review. PLoS Med. 2007;4:e261.
patterns were not reflected in our analysis. Moreover, 6. Mellen PB, Walsh TF, Herrington DM. Whole grain intake and cardiovascular
the observational nature of our study limits causality disease: a meta-analysis. Nutr Metab Cardiovasc Dis. 2008;18:283–90.
inference between intakes of whole grains or cereal fiber 7. Aune D, Chan DS, Lau R, Vieira R, Greenwood DC, Kampman E, et al. Dietary
fibre, whole grains, and risk of colorectal cancer: systematic review and
and mortality. dose–response meta-analysis of prospective studies. BMJ. 2011;343:d6617.
8. Pajvani UB, Du X, Combs TP, Berg AH, Rajala MW, Schulthess T, et al.
Conclusions Structure-function studies of the adipocyte-secreted hormone Acrp30/
adiponectin: Implications fpr metabolic regulation and bioactivity. J Biol
Data from our study indicate that intake of whole grains Chem. 2003;278:9073–85.
and cereal fiber may reduce the risk of all-cause mor- 9. Park Y, Subar AF, Hollenbeck A, Schatzkin A. Dietary fiber intake and mortality
tality and death from chronic diseases such as cancer, in the NIH-AARP diet and health study. Arch Intern Med. 2011;171:1061–8.
10. Adams KF, Schatzkin A, Harris TB, Kipnis V, Mouw T, Ballard-Barbash R, et al.
CVD, diabetes, respiratory disease, infections, and other Overweight, obesity, and mortality in a large prospective cohort of persons
causes. Disappearance or attenuation of whole grain as- 50 to 71 years old. N Engl J Med. 2006;355:763–78.
sociations with total mortality and death from chronic 11. Lacey Jr JV, Brinton LA, Leitzmann MF, Mouw T, Hollenbeck A, Schatzkin A,
et al. Menopausal hormone therapy and ovarian cancer risk in the National
diseases after adjustment for cereal fiber intake suggests Institutes of Health-AARP Diet and Health Study Cohort. J Natl Cancer Inst.
that cereal fiber partly accounts for the protective effects 2006;98:1397–405.
of whole grains on mortality. 12. Willett W. Nutritional epidemiology: issues and challenges. Int J Epidemiol.
1987;16:312–7.
Abbreviations 13. Schatzkin A1, Mouw T, Park Y, Subar AF,Kipnis V, Hollenbeck A, et al. Dietary
CI: Confidence intervals; CVD: Cardiovascular diseases; FFQ: Food frequency fiber and whole-grain consumption in relation to colorectal cancer in the
questionnaire; HR: Hazard ratio; NIH: National Institutes of Health. NIH-AARP Diet and Health Study. Am J Clin Nutr. 2007;85:1353–60.
14. Thompson FE, Kipnis V, Subar AF, Krebs-Smith SM, Kahle LL, Midthune D,
Competing interests et al. Evaluation of 2 brief instruments and a food-frequency questionnaire
The authors declare that they have no competing interests. to estimate daily number of servings of fruit and vegetables. Am J Clin Nutr.
2000;71:1503–10.
Authors’ contributions 15. International Classification of Diseases, Ninth Revision (ICD-9). World Health
TH and LQ conceived the study. TH, MX, and LQ analyzed the data and Organization. http://www.who.int/classifications/icd/en/
wrote the draft of the paper. MX, AL, SC, and LQ contributed to writing, 16. International Classification of Diseases (ICD). World Health Organization.
reviewing, and revising of the paper. LQ is the guarantor. All authors read Retrieved 23 November 2010. http://apps.who.int/classifications/icd10/
and approved the final manuscript. browse/2010/en
17. Willett W. Nutritional epidemiology. 2nd ed. New York: Oxford University
Acknowledgments Press; 1998.
We thank all the participants in the NIH-AARP Diet and Health Study. We also 18. Borneo R, Leon AE. Whole grain cereals: functional components and health
thank Dr. David Hasza for statistical assistance. benefits. Food Funct. 2012;3:110–9.
19. Slavin J. Why whole grains are protective: biological mechanisms. Proc Nutr
Sources of support Soc. 2003;62:129–34.
This study is funded by an unrestricted research fund from NutraSource. 20. Ye EQ, Chacko SA, Chou EL, Kugizaki M, Liu S. Greater whole-grain intake is
Dr. Qi was supported by grants from the National Heart, Lung, and Blood associated with lower risk of type 2 diabetes, cardiovascular disease, and
Institute (HL071981), the National Institute of Diabetes and Digestive and weight gain. J Nutr. 2012;142:1304–13.
Kidney Diseases (DK091718), the Boston Obesity Nutrition Research Center 21. Flint AJ, Hu FB, Glynn RJ, Jensen MK, Franz M, Sampson L, et al. Whole
(DK46200), and United States–Israel Binational Science Foundation Grant grains and incident hypertension in men. Am J Clin Nutr. 2009;90:493–8.
2011036. Dr. Qi was a recipient of the American Heart Association Scientist 22. Egeberg R, Olsen A, Christensen J, Johnsen NF, Loft S, Overvad K, et al.
Development Award (0730094 N). Funding from NutraSource. There were no Intake of whole-grain products and risk of prostate cancer among men in
other relationships or activities that could appear to have influenced the the Danish Diet, Cancer and Health cohort study. Cancer Causes Control.
submitted work. 2011;22:1133–9.
23. Schatzkin A, Park Y, Leitzmann MF, Hollenbeck AR, Cross AJ. Prospective
Author details study of dietary fiber, whole grain foods, and small intestinal cancer.
1 Gastroenterology. 2008;135:1163–7.
Department of Nutrition, Harvard School of Public Health, 665 Huntington
Ave, Boston, MA 02115, USA. 2NutraSource (AWL), Royal Oak, MI 48073, USA. 24. He M, van Dam RM, Rimm E, Hu FB, Qi L. Whole-grain, cereal fiber, bran,
3
NutraSource (SSC), Clarksville, MD 21029, USA. 4Channing Laboratory, and germ intake and the risks of all-cause and cardiovascular disease-
Department of Medicine, Brigham and Women’s Hospital and Harvard specific mortality among women with type 2 diabetes mellitus. Circulation.
Medical School, 75 Francis St, Boston, MA 02115, USA. 2010;121:2162–8.
25. Chuang SC, Norat T, Murphy N, Olsen A, Tjonneland A, Overvad K, et al. Fiber
Received: 29 October 2014 Accepted: 13 February 2015 intake and total and cause-specific mortality in the European Prospective
Investigation into Cancer and Nutrition cohort. Am J Clin Nutr. 2012;96:164–74.
26. Qi L, Meigs JB, Liu S, Manson JE, Mantzoros C, Hu FB. Dietary fibers and
References glycemic load, obesity, and plasma adiponectin levels in women with type
1. Slavin J. Whole grains and human health. Nutr Res Rev. 2004;17:99–110. 2 diabetes. Diabetes Care. 2006;29:1501–5.
2. Qi L, Hu FB. Dietary glycemic load, whole grains, and systemic 27. Esposito K, Giugliano D. Whole-grain intake cools down inflammation.
inflammation in diabetes: the epidemiological evidence. Curr Opin Am J Clin Nutr. 2006;83:1440–1. Author reply 1441–2.
Lipidol. 2007;18:3–8. 28. Weickert MO, Mohlig M, Schofl C, Arafat AM, Otto B, Viehoff H, et al. Cereal
3. Qi L, van Dam RM, Liu S, Franz M, Mantzoros C, Hu FB. Whole-grain, bran, fiber improves whole-body insulin sensitivity in overweight and obese
and cereal fiber intakes and markers of systemic inflammation in diabetic women. Diabetes Care. 2006;29:775–80.
women. Diabetes Care. 2006;29:207–11. 29. Ma Y, Griffith JA, Chasan-Taber L, Olendzki BC, Jackson E, Stanek 3rd EJ,
4. Harland JI, Garton LE. Whole-grain intake as a marker of healthy body et al. Association between dietary fiber and serum C-reactive protein.
weight and adiposity. Public Health Nutr. 2008;11:554–63. Am J Clin Nutr. 2006;83:760–6.
Huang et al. BMC Medicine (2015) 13:59 Page 9 of 9

30. Wannamethee SG, Whincup PH, Thomas MC, Sattar N. Associations

between dietary fiber and inflammation, hepatic function, and risk of type 2
diabetes in older men: potential mechanisms for the benefits of fiber on
diabetes risk. Diabetes Care. 2009;32:1823–5.
31. Slavin JL. Mechanisms for the impact of whole grain foods on cancer risk. J
Am Coll Nutr. 2000;19:300S–7.
32. Jacobs Jr DR, Andersen LF, Blomhoff R. Whole-grain consumption is
associated with a reduced risk of noncardiovascular, noncancer death
attributed to inflammatory diseases in the Iowa Women’s Health Study.
Am J Clin Nutr. 2007;85:1606–14.
33. Yang J, Martinez I, Walter J, Keshavarzian A, Rose DJ. In vitro characterization
of the impact of selected dietary fibers on fecal microbiota composition
and short chain fatty acid production. Anaerobe. 2013;23:74–81.
34. Parnell JA, Reimer RA. Prebiotic fibres dose-dependently increase satiety
hormones and alter Bacteroidetes and Firmicutes in lean and obese
JCR:LA-cp rats. Br J Nutr. 2012;107:601–13.
35. Saura-Calixto F. Dietary fiber as a carrier of dietary antioxidants: an essential
physiological function. J Agric Food Chem. 2011;59:43–9.

Submit your next manuscript to BioMed Central

and take full advantage of:

• Convenient online submission

• Thorough peer review
• No space constraints or color figure charges
• Immediate publication on acceptance
• Inclusion in PubMed, CAS, Scopus and Google Scholar
• Research which is freely available for redistribution

Submit your manuscript at

www.biomedcentral.com/submit