Dietary Fiber,C-Reactive Protein,and Leisure-Time Physical Activity Among U.S. Adults

Abstract

Background:

Some evidence suggests an inverse association between increased fiber intake and C-reactive protein (CRP). However, few studies have examined the associations among CRP, dietary fiber, and leisure-time physical activity (LTPA) in a nationally representative sample of U.S. adults.

Methods:

Sample (n = 8372) included adults (≥20 years of age) who participated in the 2007–2010 National Health and Nutrition Examination Survey. Tertiles of reported fiber intake were created. The dependent variable was elevated CRP (>3–10 mg/L). Logistic regression models were stratified by LTPA participation and adjusted for age, gender, race, waist circumference (WC), and standing height.

Results:

In adults reporting any volume of LTPA participation, increased fiber intake was significantly (P < 0.05 for the upper tertile of fiber intake) associated with lower odds of having an elevated CRP concentration when compared with the lowest tertile. Similar associations were not revealed in analyses limited to adults reporting no LTPA participation. After additional adjustment for WC and standing height, this protective association was no longer statistically significant.

Conclusions:

Results suggest that WC and standing height may mediate the beneficial association between increased fiber intake and lower odds of elevated CRP in adults reporting LTPA participation.

Introduction

C-reactive protein (CRP) is a nonspecific marker of inflammation that has proven useful in predicting cardiometabolic risk.^1,2 A 2010 review by Butcher and Beckstrand³ reported a favorable inverse association between increased fiber intake and CRP concentrations. These authors stated that the association between fiber intake and decreased risk for cardiovascular disease (CVD) was evidenced by the decreases in CRP concentration as fiber intake increased. Recent studies,^4,5 including a 2015 meta-analysis of 14 randomized control trials,⁶ have continued to support the evidence suggesting a favorable inverse association between increased fiber intake and CRP concentrations.^7

–12

The inverse association between CRP concentrations and leisure-time physical activity (LTPA) participation has been reported in cross-sectional studies conducted with various populations, including adult Americans,¹³ Japanese elderly,¹⁴ and adult Greeks.¹⁵ Furthermore, a 2016 meta-analysis examining the effects of exercise intervention trials on CRP concentrations in both healthy individuals and in individuals with heart disease suggests that exercise may significantly reduce CRP concentrations, independent of the presence of heart disease.¹⁶

However, the evidence examining the associations among fiber intake, elevated CRP concentrations, LTPA participation, and the possible mediating effects of waist circumference (WC) and standing height, using nationally representative samples of U.S. adults, remains limited. The purpose of this study was to examine the associations among fiber intake, elevated CRP concentrations, LTPA participation, and the possible mediating effects of WC and standing height in a nationally representative sample of U.S. adults.

Materials and Methods

This study utilized 4 years of data from the 2007 to 2010 National Health and Nutrition Examination Survey (NHANES), a continuous survey conducted by the National Center for Health Statistics.¹⁷ The NHANES is a stratified multistage probability sample of the civilian noninstitutionalized population in the United States. The unweighted response rates were 75.4% and 77.3% for the examined sample from the 2007–2008 and 2009–2010 cycles of the continuous NHANES, respectively. The study sample included U.S. adult (≥20 years of age) participants (n = 8372) in the 2007–2010 NHANES. The testing and examination procedures were approved by the institutional review board of the National Center for Health Statistics and all participants provided informed consent.¹⁸ All of the analyses utilized in this study use deidentified NHANES data, which was approved by the University of North Florida Institutional Review Board.

The primary independent variable in this study was fiber intake (grams/day). Fiber intake data were recorded during two separate 24-hour dietary recall interviews. The first interview was conducted in-person and the second interview was conducted through telephone ∼3 to 10 days later. Tertiles of fiber intake were calculated using data from the 2007–2008 and 2009–2010 cycles of the NHANES (total nutrient intake files 2007–2008, DR1TOT_E and DR2TOT_E; total nutrient intake files 2009–2010 DR1TOT_F and DR2TOT_F).¹⁹ The total nutrient intake files provide a summary record of total nutrient intakes for each individual. The in-person interview was conducted in a private room in the NHANES mobile examination center (MEC). A set of measuring guides (various glasses, bowls, mugs, drink boxes and bottles, household spoons, measuring cups and spoons, a ruler, thickness sticks, bean bags, and circles) was made available in the MEC dietary interview room for the participant to use for reporting amounts of foods. Data were collected using the United States Department of Agriculture's dietary data collection instrument [the Automated Multiple Pass Method (AMPM)]. The AMPM has been previously validated and shown to be an effective method for collecting accurate energy intake of adults.²⁰

Tertiles of fiber intake were created using the 2-day average total fiber intake data collected in the 2007–2010 NHANES adult (≥20 years of age) population. The upper, middle, and lower fiber intake tertiles for this study correspond to >18.5, 11.3–18.5, and <11.3 (grams/day), respectively. Three categories of age were created: 20–39, 40–59, and ≥60 years of age. The four categories of race included non-Hispanic white, non-Hispanic black, Mexican American, and other. WC, measured just above the uppermost lateral border of the right ilium,^21,22 was examined continuously in our logistic regression models to examine the odds of elevated CRP concentration per every centimeter of WC. The primary dependent variable in this study was elevated CRP concentration (>3–10 mg/L).²³ The CRP assays were performed using a Behring Nephelometer for quantitative CRP determination. Blood specimens were stored and shipped to the University of Washington, Seattle, WA, for processing.^24,25

The data in this study were initially managed using SAS 9.4.²⁶ SAS was used to conduct both complex variable recodes and data coding validation. SAS-callable SUDAAN²⁷ was then used to conduct the analysis, incorporating sampling weights within the context of the correlated multistage complex sampling design inherent to NHANES. Initial analysis (Table 2) revealed a significant interaction between LTPA participation and fiber intake (P = 0.014), thus mutually exclusive LTPA participation (active/no LTPA) models were created to better elucidate this relationship. The following questions were used to identify those who reported engaging in LTPA.

(1) The next questions exclude the work and transportation activities that you have already mentioned. Now I would like to ask you about sports, fitness, and recreational activities. Do you do any vigorous-intensity sports, fitness, or recreational activities that cause large increases in breathing or heart rate such as running or basketball for at least 10 min continuously?

(2) Do you do any moderate-intensity sports, fitness, or recreational activities that cause a small increase in breathing or heart rate such as brisk walking, bicycling, swimming, or golf for at least 10 min continuously?

Those responding no to both survey items would be categorized as performing no LTPA.

Subsequently, a forward selection process based on Wald F- test results was used to create the best-fit models used to examine the odds of elevated CRP by tertiles of total fiber intake. The resultant logistic regression models were adjusted for age, gender, race, WC, and standing height. As per the NHANES analytic recommendations, our analysis of dietary data was conducted using a separate set of adjusted weights. For our analyses, the dietary 2-day sample weight (WTDR2D) was selected due to the combined use of day 1 and day 2 dietary data. The 2-day weight adjusts for the potential nonresponse during the second recall and the combination of weekend and weekday recalls.

Results

Table 1 illustrates the prevalence estimates for elevated CRP for the total sample, for those reporting any volume of LTPA participation (active), and for those reporting no LTPA participation according to sample characteristics.

Table 1.

Prevalence of Elevated CRP According to Sample Characteristics: NHANES 2007–2010

	Total		Active		No LTPA
Covariates	n	Weighted % (95% CI)	n	Weighted % (95% CI)	n	Weighted % (95% CI)
Total	8372	24.1 (22.9–25.4)	3928	19.7 (18.4–21.2)	4444	29.2 (26.5–32.0)
Tertiles of fiber intake (grams/day)
1 (lower)	2452	27.8 (25.3–30.3)	924	22.3 (19.1–26.0)	1528	32.2 (27.7–37.1)
2	3128	25.1 (23.3–27.1)	1433	22.6 (20.1–25.4)	1695	28.1 (25.2–31.3)
3	2792	20.0 (17.9–22.3)	1571	15.8 (13.9–17.9)	1221	27.0 (22.5–31.7)
Age (years)
20–39	2664	20.6 (18.2–23.1)	1552	17.6 (14.6–21.2)	1112	25.8 (22.0–30.0)
40–59	2811	24.0 (21.8–26.1)	1260	19.4 (16.5–22.6)	1551	28.9 (25.9–32.1)
≥60	2897	30.4 (28.2–32.7)	1116	23.8 (20.6–27.4)	1781	35.5 (31.7–39.5)
Gender
Male	4145	19.9 (18.2–21.7)	2106	16.5 (14.5–18.8)	2039	24.5 (21.1–28.0)
Female	4227	28.4 (26.5–30.3)	1822	23.5 (21.1–26.1)	2405	33.1 (30.2–36.2)
Race
nH white	4200	22.7 (20.9–24.7)	2167	18.1 (16.6–19.8)	2033	29.3 (25.0–34.1)
nH black	1512	31.6 (28.6–34.8)	635	31.5 (27.3–36.0)	877	31.8 (27.5–36.5)
Mexican Americans	1429	28.3 (24.8–32.1)	573	25.0 (21.5–28.8)	856	30.7 (25.9–36.1)
Other	1231	21.8 (17.2–27.1)	553	18.2 (12.8–25.0)	678	24.3 (18.6–31.1)
Augmented WC (cm)
Yes	4669	36.0 (33.7–38.4)	1889	31.2 (28.5–34.0)	2780	40.3 (36.5–44.2)
No	3703	12.4 (11.0–14.0)	2039	11.1 (9.8–12.6)	1664	14.3 (11.7–17.4)

Independent variables included age (years), gender, race, augmented WC men (yes: ≥102 cm, no: <102 cm), women (yes: ≥88 cm, no: <88 cm), and tertiles of fiber intake (grams/day).

CI, confidence interval; elevated CRP, elevated C-reactive protein (>3–10 mg/L); LTPA, leisure-time physical activity; nH, non-Hispanic; NHANES, National Health and Nutrition Examination Survey; WC, waist circumference.

The age-adjusted prevalence of elevated CRP concentration among U.S. adults was 24.1% (95% CI 22.9–25.4). The age-adjusted prevalence of elevated CRP concentration was 19.7% (95% CI 18.4%–21.2%) and 29.2% (95% CI 26.5%–32.0%) for those who reported any volume of LTPA participation and for those reporting no activity, respectively (Table 1). Prevalence estimates of elevated CRP increased with age independent of LTPA participation. Similarly, the prevalence of elevated CRP was found to be greater among women, non-Hispanic blacks, Mexican Americans, and those with augmented WC independent of reported LTPA participation. Table 2 illustrates the results of the logistic regression analyses examining the associations between elevated CRP and independent variables for the total sample and stratified by LTPA participation.

Table 2.

Odds Ratios for Fiber Intake as a Predictor of Elevated CRP Concentration in a Sample of U.S. Adults: NHANES 2007–2010

	Total (n = 8372)			Active (n = 3928)			No LTPA (n = 4444)
Variable	Model 1 OR (95% CI)	Model 2 OR (95% CI)	Model 3 OR (95% CI)	Model 1 OR (95% CI)	Model 2 OR (95% CI)	Model 3 OR (95% CI)	Model 1 OR (95% CI)	Model 2 OR (95% CI)	Model 3 OR (95% CI)
Tertiles of fiber intake (grams/day)
1 (lower)	1.00	1.00	1.00	1.00	1.00	1.00	1.00	1.00	1.00
2	0.87 (0.73–1.04)	0.91 (0.76–1.09)	0.99 (0.82–1.20)	1.01 (0.78–1.29)	1.06 (0.82–1.38)	1.19 (0.92–1.55)	0.84 (0.65–1.08)	0.86 (0.67–1.10)	0.83 (0.64–1.07)
3	0.65 (0.54–0.78)^a	0.72 (0.60–0.87)^a	0.84 (0.67–1.06)	0.64 (0.50–0.84)^a	0.72 (0.55–0.94)^a	0.84 (0.64–1.11)	0.83 (0.61–1.13)	0.89 (0.65–1.22)	0.89 (0.64–1.25)
Age (years)
20–39	1.00	1.00	1.00	1.00	1.00	1.00	1.00	1.00	1.00
40–59	1.23 (1.00–1.51)^a	1.24 (1.021.51)^a	0.91 (0.75–1.11)	1.15 (0.81–1.63)	1.17 (0.82–1.66)	0.89 (0.62–1.29)	1.18 (0.96–1.45)	1.20 (0.95–1.50)	0.89 (0.72–1.11)
≥60	1.71 (1.42–2.06)^a	1.74 (1.43–2.11)^a	0.98 (0.80–1.21)	1.51 (1.11–2.05)^a	1.59 (1.16–2.17)^a	1.01 (0.74–1.37)	1.59 (1.29–1.96)^a	1.57 (1.24–1.99)^a	0.89 (0.68–1.18)
Gender
Male		1.00	1.00		1.00	1.00		1.00	1.00
Female		1.53 (1.32–1.78)^a	1.70 (1.33–2.16)^a		1.55 (1.21–1.97)^a	1.67 (1.15–2.41)^a		1.48 (1.24–1.76)^a	1.73 (1.30–2.30)^a
Race
nH white		1.00	1.00		1.00	1.00		1.00	1.00
nH black		1.51 (1.27–1.80)^a	1.40 (1.13–1.72)^a		1.96 (1.50–2.56)^a	1.63 (1.18–2.24)^a		1.13 (0.85–1.49)	1.19 (0.86–1.65)
Mexican American		1.44 (1.15–1.81)^a	1.13 (0.89–1.43)		1.62 (1.25–2.10)^a	1.24 (0.95–1.63)		1.13 (0.81–1.60)	1.23 (0.87–1.75)
Other		0.94 (0.65–1.37)	1.02 (0.70–1.49)		1.07 (0.68–1.69)	1.14 (0.75–1.73)		0.77 (0.48–1.24)	1.00 (0.67–1.50)
Waist circumference (cm)			1.06 (1.05–1.07)^a			1.06 (1.05–1.07)^a			1.06 (1.05–1.07)^a
Standing height (cm)			0.97 (0.96–0.99)^a			0.97 (0.95–1.00)^a			0.97 (0.95–0.99)^a

Independent variables included in Model 1: tertiles of fiber and age; Model 2: tertiles of fiber, age, gender, and race; Model 3 included all variables from Model 2, WC, and standing height.

Significant predictors (P < 0.05).

OR, odds ratio.

An ANOVA table was used to test each model term and create activity-specific best-fit models. The independent variables included in our two mutually exclusive activity-stratified final analyses were age, gender, race, WC, and standing height (Table 2). In analysis limited to those reporting any LTPA participation, after adjustment for demographics alone, adults reporting increased fiber intake (upper tertile) were significantly (P < 0.05) less likely to have an elevated CRP concentration when compared with the lowest tertile referent group. After additional adjustment for WC and standing height, this protective association was no longer statistically significant. In contrast, a statistically significant protective association was not observed in analyses limited to adults reporting no LTPA.

Discussion

In this nationally representative sample of U.S. adults, increased fiber intake was associated with significantly lower odds of having an elevated CRP concentration only in adults reporting LTPA participation. However, this inverse association between increased fiber intake and CRP concentration did not remain statistically significant after additional adjustment for central adiposity, a well-recognized strong mediating factor when analyzing the relationship between CRP concentration and LTPA,^28

–31 or standing height, which has been associated with increased cardiometabolic risk.^32,33 Our findings add to the current evidence revealing a favorable inverse relationship between increased fiber intake and CRP concentration.^10,11 Furthermore, our findings also suggest that these associations may have important implications for U.S. adults reporting no LTPA.

Our findings contrast those from previous studies using nationally representative data from the NHANES. These studies have reported statistically significant inverse associations between increased fiber intake and CRP concentrations independent of LTPA,^10,11 whereas in this study, these associations were significant only in individuals reporting LTPA participation. King et al.¹⁰ analyzed the data collected from 4900 adult participants in the 1999–2000 NHANES and compared the CRP concentrations of adults in the highest quartile of fiber intake with a lowest quartile referent group. It was found that increased fiber intake was associated with lower CRP concentrations. Moreover, these associations were independent of any participation in moderate or vigorous intensity LTPA or body mass index (BMI). Similar findings were reported by Ajani et al.¹¹ in a 2004 study using data collected from 3920 adult participants of the 1999–2000 NHANES. These authors compared the CRP concentration of adults in the highest quintile of fiber intake with a lowest quintile referent group. After adjustment for several covariates, including LTPA and BMI, it was reported that fiber intake remained independently associated with CRP concentrations. In this study, significant inverse associations between fiber intake and CRP levels were present before adjusting for central adiposity and height in individuals reporting LTPA participation.

One important methodological change that occurred in NHANES dietary data collection since 2002 was the inclusion of a second dietary recall to account for day-to-day variation in dietary intake of the population.³⁴ Thus, although the studies by Ajani et al.¹¹ and King et al.¹² were based on dietary data collected from one 24-hour recall, this study included dietary data averaged over two 24-hour recalls, which reduced the within-person variation in fiber intake and allowed us to obtain a more accurate estimate of the population's true fiber intake. This may explain the contrast between our findings and those of studies using NHANES data collected before 2002.

Some of the strengths of this study include the analysis of a nationally representative sample of U.S. adults, which allows for generalization of our findings to the general population of U.S. adults, the large sample size, which minimizes the chance that our findings are due to chance, and the use of dietary data estimated from two 24-hour recalls, a methodological change that was implemented starting in 2002. Some limitations of this study are its cross-sectional nature, which prevents us from establishing causality, the possibility of misclassification of dietary fiber intake, which may lead to inaccurate representation of the relationship between fiber intake and CRP concentrations, and the possibility of not including important confounders in the analyses.

Conclusion

Our findings add to the current evidence suggesting that increased fiber intake is associated with lower CRP concentrations. In a representative sample of U.S. adults, increased fiber intake was independently associated with lower odds of elevated CRP only in adults reporting LTPA. When adjusting for WC and standing height, the significant inverse associations between increasing tertiles of dietary fiber intake and odds of elevated CRP disappeared, suggesting that WC and standing height may mediate the beneficial association between increased fiber intake and lower odds of elevated CRP in adults reporting LTPA participation. Our findings do not support an association between dietary fiber intake and lower odds of elevated CRP in adults reporting no LTPA.

Footnotes

Acknowledgments

The findings and conclusions in this article are those of the authors and do not necessarily represent the official position of the Centers for Disease Control and Prevention. Furthermore, the results of this study do not constitute endorsement by the American College of Sports Medicine.

Author Disclosure Statement

No conflicting financial interests exist.

References

Black

, Kushner

, Samols

. C-reactive protein. J Biol Chem, 2004; 279:48487–48490.

Ridker

. High-sensitivity C-reactive protein: Potential adjunct for global risk assessment in the primary prevention of cardiovascular disease. Circulation, 2001; 103:1813–1818.

Butcher

, Beckstrand

. Fiber's impact on high-sensitivity C-reactive protein levels in cardiovascular disease. J Am Acad Nurse Pract, 2010; 22:566–572.

Johansson-Persson

, Ulmius

, Cloetens

, et al. A high intake of dietary fiber influences C-reactive protein and fibrinogen, but not glucose and lipid metabolism, in mildly hypercholesterolemic subjects. Eur J Nutr, 2014; 53:39–48.

Ning

, Van Horn

, Shay

, et al. Associations of dietary fiber intake with long-term predicted cardiovascular disease risk and C-reactive protein levels (from the National Health and Nutrition Examination Survey Data [2005–2010]). Am J Cardiol, 2014; 113:287–291.

Jiao

, Xu

, Zhang

, et al. Effect of dietary fiber on circulating C-reactive protein in overweight and obese adults: A meta-analysis of randomized controlled trials. Int J Food Sci Nutr, 2015; 66:114–119.

Esposito

, Pontillo

, Di Palo

, et al. Effect of weight loss and lifestyle changes on vascular inflammatory markers in obese women: A randomized trial. JAMA, 2003; 289:1799–1804.

Esposito

, Marfella

, Ciotola

, et al. Effect of a mediterranean-style diet on endothelial dysfunction and markers of vascular inflammation in the metabolic syndrome: A randomized trial. JAMA, 2004; 292:1440–1446.

, Griffith

, Chasan-Taber

, et al. Association between dietary fiber and serum C-reactive protein. Am J Clin Nutr, 2006; 83:760–766.

10.

King

, Egan

, Geesey

. Relation of dietary fat and fiber to elevation of C-reactive protein. Am J Cardiol, 2003; 92:1335–1339.

11.

Ajani

, Ford

, Mokdad

. Dietary fiber and C-reactive protein: Findings from national health and nutrition examination survey data. J Nutr, 2004; 134:1181–1185.

12.

King

, Egan

, Woolson

, et al. Effect of a high-fiber diet vs a fiber-supplemented diet on C-reactive protein level. Arch Intern Med, 2007; 167:502–506.

13.

Pischon

, Hankinson

, Hotamisligil

, et al. Leisure-time physical activity and reduced plasma levels of obesity-related inflammatory markers. Obes Res, 2003; 11:1055–1064.

14.

Niu

, Hozawa

, Fujita

, et al. Influence of leisure-time physical activity on the relationship between C-reactive protein and hypertension in a community-based elderly population of Japan: The Tsurugaya project. Hypertens Res, 2005; 28:747–754.

15.

Panagiotakos

, Pitsavos

, Chrysohoou

, et al. The associations between leisure-time physical activity and inflammatory and coagulation markers related to cardiovascular disease: The ATTICA Study. Prev Med, 2005; 40:432–437.

16.

Hammonds

, Gathright

, Goldstein

, et al. Effects of exercise on c-reactive protein in healthy patients and in patients with heart disease: A meta-analysis. Heart Lung, 2016; 45:273–282.

17.

Centers for Disease Control and Prevention (CDC). National Center for Health Statistics (NCHS). National Health and Nutrtion Examination Survey Analytic and Reporting Guidelines. Hyattsville, MD: U.S. Department of Health and Human Services, Centers for Disease Control and Prevention; 2006.

18.

Centers for Disease Control and Prevention (CDC). National Center for Health Statistics (NCHS). National Health and Nutrition Examination Survey Examination Protocol. Hyattsville, MD: U.S. Department of Health and Human Services, Centers for Disease Control and Prevention; 2004.

19.

Centers for Disease Control and Prevention, National Center for Health Statistics. National Health and Nutrition Examination Survey MEC In-Person Dietary Interviewers Procedure Manual. Hyattsville, MD: U.S. Department of Health and Human Services, Centers for Disease Control and Prevention; 2009.

20.

Agricultural Research Service, Beltsville Human Nutrition Research Center, Food Surveys Research Group, et al. USDA Automated Multiple-Pass Method for Dietary Recalls. Accessed at www.ars.usda.gov/ba/bhnrc/fsrg on May 13 , 15.

21.

Centers for Disease Control and Prevention, National Center for Health Statistics. National Health and Nutrition Examination Survey Anthropometry Procedures Manual. Hyattsville, MD: U.S. Department of Health and Human Services, Centers for Disease Control and Prevention; 2007.

22.

23.

Pearson

, Mensah

, Alexander

, et al. Markers of inflammation and cardiovascular disease: Application to clinical and public health practice: A statement for healthcare professionals from the Centers for Disease Control and Prevention and the American Heart Association. Circulation, 2003; 107:499–511.

24.

Centers for Disease Control and Prevention, National Center for Health Statistics. National Health and Nutrition Examination Survey Laboratory Procedure Manual. C-reactive protein in Serum NHANES 2007–2008. Hyattsville, MD: U.S. Department of Health and Human Services, Centers for Disease Control and Prevention; 2007.

25.

Centers for Disease Control and Prevention, National Center for Health Statistics. National Health and Nutrition Examination Survey Laboratory Procedure Manual. C-reactive protein in Serum NHANES 2009–2010. Hyattsville, MD: U.S. Department of Health and Human Services, Centers for Disease Control and Prevention; 2011.

26.

Statistical Analysis System (SAS) Institute, Inc. Base SAS 9.4 Procedures Guide: Statistical Procedures, Second Edition. Cary (NC): SAS Institute Inc; 2013.

27.

Research Triangle Institute. SUDAAN User's Manual, Release 11.0. Durham, NC: Research Triangle Institute; 2012.

28.

Church

, Earnest

, Thompson

, et al. Exercise without weight loss does not reduce C-reactive protein: The INFLAME study. Med Sci Sports Exerc, 2010; 42:708–716.

29.

Visser

, Bouter

, McQuillan

, et al. Elevated C-reactive protein levels in overweight and obese adults. JAMA, 1999; 282:2131–2135.

30.

Geffken

, Cushman

, Burke

, et al. Association between physical activity and markers of inflammation in a healthy elderly population. Am J Epidemiol, 2001; 153:242–250.

31.

Thorand

, Baumert

, Doring

, et al. Sex differences in the relation of body composition to markers of inflammation. Atherosclerosis, 2006; 184:216–224.

32.

Nelson

, Hamby

, Saleheen

, et al. Genetically determined height and coronary artery disease. N Engl J Med, 2015; 372:1608–1618.

33.

Stefan

, Häring

, Hu

, et al. Divergent associations of height with cardiometabolic disease and cancer: Epidemiology, pathophysiology, and global implications. Lancet Diabetes Endocrinol, 2016; 4:457–467.

34.

Ahluwalia

, Dwyer

, Terry

, et al. Update on NHANES dietary data: Focus on collection, release, analytical considerations, and uses to inform public policy. Adv Nutr, 2016; 7:121–134.