Hostname: page-component-586b7cd67f-t7fkt Total loading time: 0 Render date: 2024-11-22T09:08:27.699Z Has data issue: false hasContentIssue false

Assessing nutritional quality as a ‘vital sign’ of cardiometabolic health

Published online by Cambridge University Press:  25 June 2019

Dorothée Buteau-Poulin
Affiliation:
Centre de recherche de l’Institut universitaire de cardiologie et de pneumologie de Québec – Université Laval, Québec, QC, G1V 4G5, Canada Department of Kinesiology, Faculty of Medicine, Université Laval, Québec, QC, G1V 0A6, Canada
Paul Poirier
Affiliation:
Centre de recherche de l’Institut universitaire de cardiologie et de pneumologie de Québec – Université Laval, Québec, QC, G1V 4G5, Canada Faculty of Pharmacy, Université Laval, Québec, QC, G1V 0A6, Canada
Jean-Pierre Després
Affiliation:
Centre de recherche de l’Institut universitaire de cardiologie et de pneumologie de Québec – Université Laval, Québec, QC, G1V 4G5, Canada Department of Kinesiology, Faculty of Medicine, Université Laval, Québec, QC, G1V 0A6, Canada Centre de recherche sur les soins et les services de première ligne – Université Laval, Québec, QC, G1J 0A4, Canada
Natalie Alméras*
Affiliation:
Centre de recherche de l’Institut universitaire de cardiologie et de pneumologie de Québec – Université Laval, Québec, QC, G1V 4G5, Canada Department of Kinesiology, Faculty of Medicine, Université Laval, Québec, QC, G1V 0A6, Canada
*
*Corresponding author: Natalie Alméras, email [email protected]
Rights & Permissions [Opens in a new window]

Abstract

High overall nutritional quality (NQ) is an important component of ideal cardiovascular health, a concept introduced in 2010 by the American Heart Association. However, data on the independent contribution of overall NQ to the variation in the cardiometabolic risk (CMR) profile are limited. This observational study aimed to investigate the association between overall NQ and the CMR profile in 4785 participants (65⋅4 % of men, age 43⋅3 (sd 10⋅8) years) who underwent a cardiometabolic health evaluation, including lifestyle habits, anthropometric measurements, blood pressure, lipid profile and HbA1c concentrations. In addition, a submaximal exercise test was conducted to assess cardiorespiratory fitness (CRF). Using a standardised NQ questionnaire (twenty-five items food-based questionnaire), participants were classified into three subgroups: (1) low, (2) moderate or (3) high NQ and variance and multiple linear regression analyses were performed. Results showed that less than 15 % of participants presented a high NQ. A high NQ was associated with a healthier lifestyle habits and a more favourable CMR profile (lower values of waist circumference and cholesterol:HDL-cholesterol ratio, lower concentrations of non-HDL-cholesterol, TAG and HbA1c). Some of these associations were independent of age, physical activity level (PAL) and CRF. A better NQ was also associated with a lower proportion of participants presenting the hypertriacylglycerolaemic waist phenotype independently of both PAL and CRF. The present study suggests that overall NQ can be assessed with a short food-based questionnaire and should be considered in clinical practice as a new ‘vital sign’ associated with other health behaviours and cardiometabolic health.

Type
Full Papers
Copyright
© The Authors 2019 

Despite major improvements in its clinical management over the last decades, the prevalence of CVD and associated risk factors remains high, representing a major societal economic burden(Reference Benjamin, Blaha and Chiuve1). To address this issue, the concept of primordial prevention has been put forward, focusing on maintaining an optimal health rather than trying to prevent CVD in individuals with risk factors(Reference Strasser2,Reference Lloyd-Jones, Hong and Labarthe3) . In its strategic goals for 2020, the American Heart Association (AHA) introduced the concept of ideal cardiovascular (CV) health based on three biological risk factors (blood pressure (BP), lipids and glucose) and four health behaviours or markers (BMI, non-smoking, physical activity level (PAL) and nutritional quality (NQ))(Reference Lloyd-Jones, Hong and Labarthe3). Epidemiological studies have confirmed that each of these seven metrics are independently associated with CVD(Reference Folsom, Yatsuya and Nettleton4,Reference Yang, Cogswell and Flanders5) . For instance, at any level of biological risk factors, health behaviours remain strongly related with the incidence of CVD and mortality risk(Reference Yang, Cogswell and Flanders5,Reference Younus, Aneni and Spatz6) . The proportions of Americans (20–49 years) reaching the ideal targets for each health behaviour metrics are 32⋅6 % for BMI, 73⋅1 % for non-smoking, 42⋅0 % for PAL and only 0⋅2 % for NQ(Reference Benjamin, Blaha and Chiuve1). Therefore, improvements in eating habits are obviously needed to improve CV health at the population level.

To promote healthier eating habits, both the AHA and the 2015 Dietary Guidelines Advisory Committee have highlighted the relevance of targeting overall NQ(Reference Eckel, Jakicic and Ard7,8) . In this regard, although there is no consensual definition on how to assess overall NQ, many studies that have used various approaches and tools to assess food-based NQ have shown that some of these food-based dietary patterns are clearly associated with an overall better cardiometabolic risk (CMR) profile consistent with current nutritional guidelines or evidence-based recommendations(Reference Benjamin, Blaha and Chiuve1,Reference Alkerwi9Reference Gunther, Liese and Bell18) . Over the past decades, several methods have been developed to assess overall NQ. Approaches that target food items rather than isolated nutrients or distribution of macronutrients have been reported to be quite useful in the evaluation and the management of eating habits related to cardiometabolic health(Reference Mozaffarian, Appel and Van Horn12). CMR encompasses traditional CVD risk factors such as age, sex, LDL- and HDL-cholesterol, BP and genetics in addition to the dysmetabolic state of abdominal obesity and is used to assess the global risk of developing cardiometabolic disorders such as CVD and type 2 diabetes(Reference Després and Lemieux19).

The objective of the present study was to evaluate the ability of overall NQ assessed with a short food-based questionnaire to predict markers of the CMR profile in a workplace health programme targeting lifestyle habits (the Grand Défi Entreprise project).

Methods

Participants

The Grand Défi Entreprise is a workplace health and wellness programme with onsite comprehensive cardiometabolic and cardiorespiratory health evaluations of workers, including a 3-month lifestyle intervention programme. Participants were recruited on a voluntary basis through the workplace health programme from March 2011 to November 2017 among twenty-eight corporations of the Province of Québec with no inclusion or exclusion criteria. The present paper reports analyses conducted on the baseline CMR data obtained on a total sample of 4785 workers (3128 men and 1657 women) from a cohort of 4831. Only forty-six participants were excluded from the analyses since they did not complete the Dietary Screening Tool (DST) for the evaluation of the NQ. More details on this programme have been previously published(Reference Lévesque, Vallières and Poirier20). Briefly, participants completed standardised questionnaires on medical history and health behaviours such as NQ, PAL and smoking. Collected data included anthropometric measurements, body composition, waist circumference (WC), BP, lipid profile, HbA1c and cardiorespiratory fitness (CRF). The present study was conducted according to the guidelines laid down in the Declaration of Helsinki and all procedures involving human subjects were approved by the local Institutional Review Board (2011–1858, 20636). Written informed consent was obtained from all participants.

Overall nutritional quality

Among the different tools proposed in the literature to evaluate NQ, we have selected the DST(Reference Bailey, Miller and Mitchell21), as it allows to evaluate the NQ of the participants and to provide nutritional recommendations on the worksite. This NQ questionnaire was developed in a North American adult population. The DST is, to our knowledge, the only tool that does not require the use of an intermediate method such as a 24 h recall, a food diary or a quantitative and exhaustive food frequency questionnaire. In contrast to other frequently used NQ indices(Reference Gunther, Liese and Bell18,Reference Rumawas, Dwyer and McKeown22) , the DST does not measure adherence to nutrition guidelines or predefined dietary patterns. The DST is rather based on the consumption frequency of several food items (poultry, fruits, processed meats, etc.) and food groups (cereal products, dairy products, etc.) as well as the evaluation of certain habits such as the addition of sugar, the consumption of alcohol or the use of nutritional supplements. Participants answered the twenty-five items of the DST in approximately 10 min. A score is assigned to each question according to the subject’s response using a predefined scoring system. Frequent consumption of high NQ foods or food groups (whole fruits, vegetables, unfried fish, whole grains and milk) is given more points as well as low-frequency intakes of low NQ food or food groups (processed meat, snacks, sugary foods and added fats). The total NQ score ranged from 0 to 100 and participants were classified according to three predefined NQ subgroups: (1) low (score < 60), (2) moderate (score 60–75) or (3) high (score > 75)(Reference Bailey, Miller and Mitchell21).

Physical activity level

PAL was assessed with a self-administered, short and validated questionnaire which focused on the volume of aerobic physical activities (cycling, running, swimming, etc.) performed during leisure time and by season(Reference Khaw, Jakes and Bingham23). A mean PAL in min per week was calculated and then used to classify participants in four PAL subgroups: (1) sedentary (<30 min/week), (2) moderately inactive (30–149 min/week), (3) moderately active (150–299 min/week) or (4) active (≥300 min/week).

Anthropometric measurements and body composition

Height and weight were measured and BMI was calculated(Reference Gordon, Chumlea, Roche, Lohman, Roche and Martorell24). WC was obtained following standardised procedures(25). Percentage of body fat was estimated by bioelectrical impedance with a Tanita TBF-300A body composition analyser (Tanita Corporation).

Cardiometabolic risk profile

Resting BP was measured on both arms of participants after they had been seated for at least 5 min with an automated Suntech 247 sphygmomanometer (Suntech Medical). Blood samples in the non-fasting state were collected on the forearm vein and analysed with an Abaxis Piccolo Xpress Chemistry Analyzer to obtain different cholesterol fractions and TAG (Thermo Fisher). From 2011 to 2015, HbA1c concentrations were assessed by a turbidimetric inhibition immunoassay with a Cobas Integra 400/800 system (Roche)(Reference Abadie and Koelsch26). Since 2016, HbA1c concentrations are measured by monoclonal antibody agglutination reaction with an immunoassay DCA Vantage analyser (Siemens Healthcare).

Cardiorespiratory fitness

According to a previously described protocol, a submaximal treadmill exercise test was performed to assess CRF(Reference Lévesque, Vallières and Poirier20). Using both ACSM’s metabolic equations and the least square method, VO2max was estimated by extrapolating oxygen consumption to age-estimated maximal heart rate(27,Reference Astrand and Ryhming28) . Depending on their age, sex and estimated VO2max, participants were classified in four CRF subgroups: (1) very poor/poor, (2) fair, (3) good or (4) excellent/superior(27).

Hypertriacylglycerolaemic waist phenotype

The hypertriacylglycerolaemic (hyperTG) waist phenotype is a simple clinical marker useful for identifying viscerally obese individuals who are also likely to present metabolic abnormalities(Reference Lemieux, Pascot and Couillard29,Reference Arsenault, Lemieux and Després30) . Criteria are WC ≥ 90⋅0 cm and TAG ≥ 2⋅0 mmol/l for men or WC ≥ 85⋅0 cm and TAG ≥ 1⋅5 mmol/l for women(Reference Lemieux, Pascot and Couillard29,Reference Arsenault, Lemieux and Després30) .

Cardiovascular risk

Predicted 10-year CV risk of the participants and their vascular age were determined using standardised methods(Reference Wilson, D’Agostino and Levy31,Reference D’Agostino, Vasan and Pencina32) .

Ideal cardiovascular health

The proportion of participants within each category (ideal, intermediate and poor) of the seven metrics defining the ideal CV health was determined mainly according to the AHA definition(Reference Lloyd-Jones, Hong and Labarthe3). Criteria used were (1) smoking: never smoke or quit smoking for at least 12 months (ideal) and current smoking (poor), (2) BMI: < 25 kg/m2 (ideal) and ≥ 30 kg/m2 (poor), (3) PAL: ≥ 150 min/week (ideal) and none (poor), (4) total cholesterol: < 5⋅2 mmol/l (ideal) and ≥ 6⋅2 mmol/l (poor) and (5) BP: systolic BP < 120 mmHg and diastolic BP < 80 mmHg (ideal) and systolic BP ≥ 140 mmHg or diastolic BP ≥ 90 mmHg (poor). Unlike the AHA definition, the healthy diet score was replaced by the NQ: > 75 (ideal) and < 60 (poor)(Reference Bailey, Miller and Mitchell21) and fasting plasma glucose was replaced by HbA1c concentrations: < 5⋅7 % (ideal) and ≥ 6⋅5 % (poor)(33). Metrics between ideal and poor levels were considered in the intermediate category. Participants taking medications for dyslipidaemia, hypertension or diabetes were classified in the intermediate category if treated to goal. Otherwise, they were classified in the intermediate or poor category depending on the level reached under treatment.

Statistical analyses

No sample size calculation was performed since data presented are from an exploratory analysis of an observational study of 4785 workers that provided us with an opportunity to examine the potential relationships between lifestyle variables and biological risk variables. Furthermore, a few studies have examined the relationship between NQ indices and CMR factors in different sample sizes ranging from 488 to 1493 participants(Reference Alessa, Malik and Yuan34Reference Saraf-Bank, Haghighatdoost and Esmaillzadeh36).

As a significant sex interaction term was found for many cardiometabolic variables, analyses have been performed by sex separately, with the exception of analyses leading to findings presented in Fig. 3 where men and women were pooled but using sex-dependent criteria to define CRF and hyperTG waist. Differences in the CMR profile across NQ subgroups were analysed using one-way ANOVA and adjusted for age on separate residual variances in each group, as effect that specifies heterogeneity in the covariance structure was significant (heteroscedasticity) compared with the same variance between groups. The Satterthwaite’s degree of freedom statement was added for variables analysed using unequal variance structures. Posteriori comparisons were performed using the Tukey–Kramer adjustment for multiple comparisons. Categorical variables were compared by χ2 tests.

To investigate the relationships between the CMR profile variables and a set of explanatory variables (NQ, PAL and VO2max), multiple linear regression models were performed. All statistical regression models were first adjusted for age (Table 2), and then for additional potential confounding factors such as smoking status, alcohol intake, educational level and use of lipid-lowering, hypotensive and hypoglycaemic drugs (Table 3). The univariate normality assumption was verified with the Shapiro–Wilk tests on the error distribution from the statistical models. The Brown and Forsythe’s variation of Levene’s test statistic was used to verify the homogeneity of variances. For variables which normality and variance assumptions were not fulfilled, the logarithm transformation was used. The results were considered significant with P values ≤ 0⋅05. All data were analysed using Statistical Analysis Software Studio 3.4 (SAS Institute Inc.).

Results

Mean age of the 4785 participants was 43⋅3 (sd 10⋅8) years; 94⋅2 % of them were Caucasians and 50⋅7 % blue-collar workers. The mean NQ score was 62 (sd 13) and the proportion of participants with a high NQ was 14⋅2 %. Mean BMI was 27⋅1 (sd 4⋅9) kg/m2; 40 % were overweight and 24 % met the criterion for obesity. Moreover, 59⋅2 % of all participants showed an elevated WC and 81⋅9 % of overweight/obese participants were abdominally obese. Use of lipid-lowering and antihypertensive drugs was reported by 11⋅4 and 11⋅0 % of the participants, respectively. Among all participants, 2⋅3 % had a history of CHD or stroke.

Fig. 1 presents the proportion of participants in the poor, intermediate and ideal categories of metrics of ideal CV health according to the AHA definition(Reference Lloyd-Jones, Hong and Labarthe3). Proportions of men and women classified in the ideal category were 81⋅5 and 83⋅4 % for non-smoking; 26⋅9 and 53⋅5 % had a BMI < 25 kg/m2; 63⋅0 and 60⋅3 % reached PAL recommendations; 10⋅4 and 21⋅5 % presented a high NQ; 56⋅5 and 70⋅7 % had total cholesterol concentrations < 5⋅17 mmol/l; 14⋅1 and 39⋅1 % had a systolic BP < 120 and a diastolic BP < 80 mmHg; and 65⋅3 and 82⋅1 % had HbA1c concentrations < 5⋅7 %, respectively.

Fig. 1. Proportion of participants in the poor (), intermediate () or ideal () category of metrics of ideal cardiovascular health in (a) men and (b) women. The analysis included 3128 men and 1657 women. The proportion of participants within each category was determined mainly according to the definition of the American Heart Association (AHA)(Reference Lloyd-Jones, Hong and Labarthe3). Criteria used were (1) smoking: never smoke or quit smoking for at least 12 months (ideal) and current smoking (poor), (2) BMI: < 25 kg/m2 (ideal) and ≥ 30 kg/m2 (poor), (3) physical activity level (PAL): ≥ 150 min/week (ideal) and none (poor), (4) total cholesterol (TC): < 5⋅2 mmol/l (ideal) and ≥ 6⋅2 mmol/l (poor) and (5) blood pressure (BP): systolic BP < 120 mmHg and diastolic BP < 80 mmHg (ideal) and systolic BP ≥ 140 mmHg or diastolic BP ≥ 90 mmHg (poor). Unlike the AHA definition, the healthy diet score was replaced by nutritional quality (NQ): > 75 (ideal) and < 60 (poor)(Reference Bailey, Miller and Mitchell21) and fasting plasma glucose was replaced by HbA1c concentrations: < 5⋅7 % (ideal) and ≥ 6⋅5 % (poor)(33). Metrics between ideal and poor levels were considered in the intermediate category. Participants taking medications for dyslipidaemia, hypertension or diabetes were classified in the intermediate category if treated to goal. Otherwise, they were classified in the intermediate or poor category depending on the level reached under treatment.

Characteristics of participants according to NQ subgroups adjusted for age are shown in Table 1. In both men and women, proportion of smokers as well as BMI and percentage of body fat values progressively decreased from low to high subgroups of NQ, while PAL and estimated VO2max increased. Furthermore, both 10-year CV risk and vascular age were lower in the moderate and high NQ subgroups compared with the low NQ subgroups in both sexes. Fig. 2 presents various CMR variables across NQ subgroups adjusted for age. In both men and women, non-HDL-cholesterol concentrations were lower in the moderate and high NQ subgroups compared with the low NQ subgroup (P < 0⋅05). Moreover, men and women with a high NQ had lower cholesterol:HDL-cholesterol ratio compared with participants with a low NQ (P < 0⋅001). In comparison with men with low NQ, the mean WC of men with moderate or high NQ was 2⋅5 and 5⋅9 cm smaller, respectively. A similar trend was observed in women, WC being 3⋅4 and 7⋅1 cm smaller in the moderate and the high NQ subgroups, respectively, compared with the low NQ subgroup (P < 0⋅001). Among men with a moderate or a high NQ, TAG concentrations were lower by 7⋅9 and 17⋅7 %, respectively, in comparison with men with a low NQ (P < 0⋅01). TAG concentrations were also 5⋅1 and 10⋅9 % lower among women in the moderate and the high NQ subgroups, respectively, compared with women with a low NQ (P < 0⋅01). Accordingly, there was an inverse association between the percentage of men and women with the hyperTG waist phenotype and NQ (P < 0⋅001). In both men and women, HbA1c concentrations were lower in the moderate and the high NQ subgroups in comparison with workers with a low NQ (P < 0⋅05). Resting systolic BP was 2 mmHg lower in men and in women with a high NQ compared with those with a low NQ (P < 0⋅05). Resting diastolic BP showed an inverse association with NQ in men only (P < 0⋅01).

Table 1. Characteristics of participants according to nutritional quality

(Numbers; percentages; mean values and standard deviations)

a,b,cMean values with unlike superscript letters were significantly different.

* P < 0⋅05, ** P < 0⋅01, *** P < 0⋅001.

Analyses are age-adjusted one-way ANOVA with Tukey–Kramer’s post hoc corrections for multiple comparisons and were performed separately in men and in women.

Tertiary education was defined as post-secondary education.

§ A high household income was defined as an income greater than CAN$ 60 000.

Analyses were performed on log-transformed data.

Fig. 2. Cardiometabolic risk profile across nutritional quality (NQ) subgroups: low NQ (; <60), moderate NQ (; 60–75) and high NQ (; >75). (a) Non-HDL-cholesterol (men, n 3035; women, n 1621); (b) cholesterol:HDL-cholesterol (men, n 3034; women, n 1621); (c) waist circumference (men, n 3126; women, n 1649); (d) TAG (men, n 3083; women, n 1645); (e) hypertriacylglycerolaemic (hyperTG ) waist carriers (men, n 2776; women, n 1357); (f) HbA1c (men, n 3083; women, n 1645); (g) systolic blood pressure (BP) (men, n 3127; women, n 1657); (h) diastolic BP (men, n 3127; women, n 1657). Values are means with their standard errors, except for panel (e) where values are expressed in percentages. Analyses are age-adjusted one-way ANOVA and were performed separately in men and in women. Posteriori comparisons were performed using the Tukey–Kramer adjustment for multiple comparisons. Categorical variables were compared by χ 2 tests. * P < 0⋅05, ** P < 0⋅01, *** P < 0⋅001. † Analyses were performed on log-transformed data. Criteria for hyperTG waist are waist circumference ≥ 90⋅0 cm and TAG ≥ 2⋅0 mmol/l in men and waist circumference ≥ 85⋅0 cm and TAG ≥ 1⋅5 mmol/l for women(Reference Lemieux, Pascot and Couillard29,Reference Arsenault, Lemieux and Després30) .

Fig. 3 shows that the percentage of participants with the hyperTG waist phenotype was progressively lower as a function of higher PAL as well as higher NQ. It was shown by χ 2 tests that there was (1) 2⋅5 times more hyperTG waist carriers in the sedentary/low NQ subgroup than in the active/high NQ subgroup (P < 0⋅001) and (2) 3⋅2 times more hyperTG waist carriers in the very poor/poor CRF/low NQ subgroup than in the excellent/superior CRF/high NQ subgroup (P < 0⋅001). Furthermore, within the moderately inactive, moderately active and the active subgroups of PAL, the percentage of men and women with the hyperTG waist phenotype decreased as a function of increased NQ. Moreover, when PAL subgroups were substituted by CRF categories, both CRF and NQ were inversely associated with the percentage of individuals with the hyperTG waist phenotype.

Fig. 3. Proportion of carriers of the hypertriacylglycerolaemic (hyperTG) waist phenotype according to nutritional quality (NQ) and (a) physical activity level or (b) cardiorespiratory fitness. NQ subgroups: low (<60), moderate (60–75) and high (>75). Physical activity level subgroups: sedentary (<30 min/week), moderately inactive (30–149 min/week), moderately active (150–299 min/week) and active (≥ 300 min/week). Cardiorespiratory fitness subgroups are defined as proposed by American College of Sports and Medicine guidelines(27). Criteria for hyperTG waist are waist circumference ≥ 90·0 cm and TAG ≥ 2·0 mmol/l in men and waist circumference ≥ 85·0 cm and TAG ≥ 1·5 mmol/l in women(Reference Lemieux, Pascot and Couillard29,Reference Arsenault, Lemieux and Després30) . * Statistical difference between the sedentary/low NQ subgroup and the active/high NQ subgroup. † Statistical difference between the very poor/poor cardiorespiratory fitness/low NQ subgroup and the excellent/superior cardiorespiratory fitness/high NQ subgroup.

Multiple linear regression analyses were conducted by sex to examine the independent contributions of NQ, PAL and estimated VO2max to the variance of CMR variables after statistical adjustment for age (Table 2). In both men and women, estimated VO2max was the main independent variable explaining the largest proportion of the variance of BMI, WC, non-HDL-cholesterol, cholesterol:HDL-cholesterol, TAG and systolic and diastolic BP. In this multiple regression model, in men NQ remained independently associated with all the CMR variables with the exception of systolic BP, whereas in women NQ was only independently associated with BMI, WC, HbA1c and systolic BP. The second multiple regression model, including WC, showed that this crude index of visceral adipose tissue (VAT) contributed the most to the variance of CMR variables in both men and women after adjustment for age (Table 3). However, NQ remained significantly related to non-HDL-cholesterol, TAG and diastolic BP in men and only to HbA1c concentrations in women. Additional adjustments for potential confounding factors such as smoking status, alcohol intake, educational level and use of lipid-lowering hypotensive or hypoglycaemic drugs yielded similar results in men, with the exception of HbA1c that did not reach statistical significance (data not shown). In women, data remained unchanged after adjustment for these potential confounders.

Table 2. Association of nutritional quality (NQ), physical activity level (PAL) and VO2max with cardiometabolic risk profile variables*

(R 2; β; standard errors; t values; P values and standardised β)

* Analyses are multiple linear regression and age-adjusted and were performed separately in men and in women.

Analyses were performed on log-transformed data.

Discussion

Among participants (n 4785; 65⋅4 % of men) recruited through a workplace health programme, a higher NQ was associated with a more favourable CMR profile and with a lower estimated 10-year CV risk in both men and women. Participants with a high NQ were also more physically active and less likely to be smokers, a finding consistent with the literature(Reference Monfort-Pires, Salvador and Folchetti37,Reference Alkerwi, Baydarlioglu and Sauvageot38) . Hence, these findings suggest that overall NQ could be a simple marker of a global healthier lifestyle and of a more favourable CMR profile. However, overall NQ was the health behaviour with the lowest proportion of participants in the ideal category, a finding also reported in other population studies(Reference Benjamin, Blaha and Chiuve1). Therefore, assessment of lifestyle habits such as overall NQ appears relevant to target health behaviours contributing to a more deteriorated CMR profile.

The cross-sectional design of the present study has limitations. Although the workplace health programme was offered to all employees within each participating company, we cannot translate our results to the general population. Social desirability and response set biases are expected and both NQ and PAL were likely overestimated(Reference Hebert39). Consequently, the proportion of participants with non-ideal NQ and PAL may be greater than reported. Despite these limitations, NQ was strongly associated with variation in CMR profile as well as with the percentage of participants with the hyperTG waist phenotype (a simple marker of visceral obesity)(Reference Lemieux, Pascot and Couillard29,Reference Arsenault, Lemieux and Després30) independently of PAL and CRF.

Men and women with a high NQ showed a more favourable CMR profile and a decreased percentage of participants with visceral obesity or type 2 diabetes in comparison with participants with a low NQ. These findings are consistent with many studies where adherence to healthy dietary patterns such as the Mediterranean diet or the Dietary Approaches to Stop Hypertension (DASH) type eating plan has been inversely associated with the CMR profile and with a reduced risk for type 2 diabetes(Reference Estruch, Martinez-Gonzalez and Corella16,Reference Mozaffarian, Hao and Rimm40Reference Gadgil, Appel and Yeung43) . These studies assessed adherence to a priori defined dietary patterns known to be beneficial regarding CVD prevention, whereas the DST used in the present study targets various food items associated with overall NQ regardless of a specific dietary pattern. From a practical standpoint, a high NQ score implied, for instance, regular consumption of whole grains (both breads and cereals) ≥ 3 times a week, eating cakes and pies less than once a week, avoiding deli meats, having one fruit a day, eating unfried fish once a week and eating two daily servings of vegetables. Thereby, it is suggested that a diet of high overall NQ is achievable through various healthy dietary patterns by following very simple food-based nutritional advice consistent with current guidelines.

As in other studies, NQ, PAL and CRF were closely interrelated and seemed to show synergistic association with CMR profile(Reference Monfort-Pires, Salvador and Folchetti37,Reference Charreire, Kesse-Guyot and Bertrais44,Reference Cuenca-Garcia, Artero and Sui45) . To our knowledge, the present study is the first to report the association between health behaviours and the hyperTG waist phenotype, a simple clinical marker of visceral obesity related to metabolic abnormalities(Reference Lemieux, Pascot and Couillard29,Reference Arsenault, Lemieux and Després30) . NQ was inversely associated with the hyperTG waist phenotype independently of either PAL or CRF. These results reinforce the importance of the contribution of NQ to the most detrimental form of overweight/obesity, visceral obesity and emphasise the significance of targeting NQ even within active or fit individuals. A large longitudinal study has previously reported no association between NQ and CMR profile after adjustment for CRF(Reference Cuenca-Garcia, Artero and Sui45). In the present study, multiple regression analyses showed that association of NQ with most of the CMR variables examined remained significant after control for age, PAL and estimated VO2max as an indicator of CRF. Thus, the contribution of NQ to the CMR profile is partly independent from its contribution to traditional CV risk factors such as lipid profile, diastolic BP and HbA1c concentrations. Consequently, these results suggest that NQ may provide protection against CVD through mechanisms beyond traditional CV risk factors(Reference Mozaffarian14,Reference Mora, Cook and Buring46,Reference Alves, Viana and Cavalcante47) . A plausible hypothesis would be that VAT modulates the association of NQ with CMR variables(Reference Shah, Murthy and Allison48,Reference Fischer, Pick and Moewes49) . In this regard, our second model (Table 3) aimed at testing this hypothesis and revealed that WC was the main independent factor associated with most CMR profile variables in both men and women. Nevertheless, NQ remained significantly related to the lipid profile and diastolic BP in men as well as to HbA1c in women. It is therefore likely that physiological mechanisms related to visceral adiposity contribute to these associations as men are known to store more VAT than women(Reference Tchernof and Després50). Therefore, PAL, CRF and probably VAT modulated the association of NQ with CMR variables, but the present cross-sectional study cannot address this question. Nevertheless, the contribution of health behaviours to the CMR profile are probably synergistic and closely interrelated, the real issue being whether they can be assessed in clinical practice. For this purpose, NQ assessment appears as a promising indicator of lifestyle habits.

Table 3. Association of nutritional quality (NQ), physical activity level (PAL), VO2max and waist circumference (WC) with cardiometabolic risk profile variables*

(R 2; β; standard errors; t values; P values and standardised β)

* Analyses are multiple linear regression and age-adjusted and were performed separately in men and in women.

Analyses were performed on log-transformed data.

Conclusions

The independent contribution of NQ to the CMR profile and its inverse association with the hyperTG waist phenotype, regardless of PAL or CRF, suggests that targeting physical activity or fitness alone is not enough to optimise the CV health of men and women. Although causal relationships cannot be inferred from these cross-sectional analyses, results of the present study suggest that targeting overall NQ assessed with a short food-based questionnaire is feasible in clinical practice and should be considered as a new ‘vital sign’ associated with other health behaviours and cardiometabolic health.

Acknowledgements

The present study was achieved through a collaboration between the Grand Défi Entreprise and the researchers of the Centre de recherche de l’Institut universitaire de cardiologie et de pneumologie de Québec – Université Laval. The authors thank all the companies and their employees for their involvement in the Grand Défi Entreprise workplace health programme.

The present study was partly funded by an unrestricted grant from Pfizer as well as by the Fondation de l’Institut universitaire de cardiologie et de pneumologie de Québec. These two organisations and the Grand Défi Entreprise were not involved in any way in the study design, the conduction of the present study and did not participate in data analysis/interpretation or the writing of the present paper.

The authors’ contributions were: P. P., J. P. D. and N. A. designed the research (project conception, development of overall research plan and study oversight) and conducted the research (hands-on conduct of the experiments and data collection); J. P. D. and N. A. provided the database; D. B. P., J. P. D. and N. A. analysed data and performed statistical analyses; D. B. P. drafted the manuscript; and P. P., J. P. D. and N. A. reviewed and edited it. All authors read and approved the final manuscript.

The authors declare that there are no conflicts of interest.

References

Benjamin, EJ, Blaha, MJ, Chiuve, SE, et al. (2017) Heart disease and stroke statistics-2017 update: a report from the American Heart Association. Circulation 135, e146e603.CrossRefGoogle ScholarPubMed
Strasser, T (1978) Reflections on cardiovascular diseases. Interdiscip Sci Rev 3, 225230.CrossRefGoogle Scholar
Lloyd-Jones, DM, Hong, Y, Labarthe, D, et al. (2010) Defining and setting national goals for cardiovascular health promotion and disease reduction: the American Heart Association’s strategic impact goal through 2020 and beyond. Circulation 121, 586613.CrossRefGoogle ScholarPubMed
Folsom, AR, Yatsuya, H, Nettleton, JA, et al. (2011) Community prevalence of ideal cardiovascular health, by the American Heart Association definition, and relationship with cardiovascular disease incidence. J Am Coll Cardiol 57, 16901696.CrossRefGoogle ScholarPubMed
Yang, Q, Cogswell, ME, Flanders, WD, et al. (2012) Trends in cardiovascular health metrics and associations with all-cause and CVD mortality among US adults. JAMA 307, 12731283.CrossRefGoogle ScholarPubMed
Younus, A, Aneni, EC, Spatz, ES, et al. (2016) A systematic review of the prevalence and outcomes of ideal cardiovascular health in US and non-US populations. Mayo Clin Proc 91, 649670.CrossRefGoogle ScholarPubMed
Eckel, RH, Jakicic, JM, Ard, JD, et al. (2014) 2013 AHA/ACC guideline on lifestyle management to reduce cardiovascular risk: a report of the American College of Cardiology/American Heart Association Task Force on Practice Guidelines. J Am Coll Cardiol 63, 29602984.CrossRefGoogle Scholar
Dietary Guidelines Advisory Committee (2015) Scientific Report of the 2015 Dietary Guidelines Advisory Committee: Advisory Report to the Secretary of Health and Human Services and the Secretary of Agriculture. Washingon, DC: U.S. Department of Agriculture, Agricultural Research Service.Google Scholar
Alkerwi, A (2014) Diet quality concept. Nutrition 30, 613618.CrossRefGoogle ScholarPubMed
Waijers, PM, Feskens, EJ & Ocke, MC (2007) A critical review of predefined diet quality scores. Br J Nutr 97, 219231.CrossRefGoogle ScholarPubMed
Wirt, A & Collins, CE (2009) Diet quality: what is it and does it matter? Public Health Nutr 12, 24732492.CrossRefGoogle ScholarPubMed
Mozaffarian, D, Appel, LJ & Van Horn, L (2011) Components of a cardioprotective diet: new insights. Circulation 123, 28702891.CrossRefGoogle ScholarPubMed
Micha, R, Penalvo, JL, Cudhea, F, et al. (2017) Association between dietary factors and mortality from heart disease, stroke, and type 2 diabetes in the United States. JAMA 317, 912924.CrossRefGoogle ScholarPubMed
Mozaffarian, D (2016) Dietary and policy priorities for cardiovascular disease, diabetes, and obesity: a comprehensive review. Circulation 133, 187225.CrossRefGoogle ScholarPubMed
Sala-Vila, A, Estruch, R & Ros, E (2015) New insights into the role of nutrition in CVD prevention. Curr Cardiol Rep 17, 26.CrossRefGoogle ScholarPubMed
Estruch, R, Martinez-Gonzalez, MA, Corella, D, et al. (2006) Effects of a Mediterranean-style diet on cardiovascular risk factors: a randomized trial. Ann Intern Med 145, 111.CrossRefGoogle ScholarPubMed
Rumawas, ME, Meigs, JB, Dwyer, JT, et al. (2009) Mediterranean-style dietary pattern, reduced risk of metabolic syndrome traits, and incidence in the Framingham Offspring Cohort. Am J Clin Nutr 90, 16081614.CrossRefGoogle ScholarPubMed
Gunther, AL, Liese, AD, Bell, RA, et al. (2009) Association between the dietary approaches to stop hypertension diet and hypertension in youth with diabetes mellitus. Hypertension 53, 612.CrossRefGoogle ScholarPubMed
Després, JP & Lemieux, I (2006) Abdominal obesity and metabolic syndrome. Nature 444, 881887.CrossRefGoogle ScholarPubMed
Lévesque, V, Vallières, M, Poirier, P, et al. (2015) Targeting abdominal adiposity and cardiorespiratory fitness in the workplace. Med Sci Sports Exerc 47, 13421350.CrossRefGoogle ScholarPubMed
Bailey, RL, Miller, PE, Mitchell, DC, et al. (2009) Dietary screening tool identifies nutritional risk in older adults. Am J Clin Nutr 90, 177183.CrossRefGoogle ScholarPubMed
Rumawas, ME, Dwyer, JT, McKeown, NM, et al. (2009) The development of the Mediterranean-style dietary pattern score and its application to the American diet in the Framingham Offspring Cohort. J Nutr 139, 11501156.CrossRefGoogle ScholarPubMed
Khaw, KT, Jakes, R, Bingham, S, et al. (2006) Work and leisure time physical activity assessed using a simple, pragmatic, validated questionnaire and incident cardiovascular disease and all-cause mortality in men and women: The European Prospective Investigation into Cancer in Norfolk prospective population study. Int J Epidemiol 35, 10341043.CrossRefGoogle ScholarPubMed
Gordon, CC, Chumlea, WC & Roche, AF (1988) Stature, recumbent length, and weight. In Anthropometric Standardization Reference Manual, pp. 38 [Lohman, TG, Roche, AF, Martorell, R, editors]. Champaign, IL: Human Kinetics Books.Google Scholar
National Heart Lung and Blood Institute (1998) Clinical guidelines on the identification, evaluation, and treatment of overweight and obesity in adults: the evidence report. Obes Res 6, Suppl. 2, 51S209S.Google Scholar
Abadie, JM & Koelsch, AA (2008) Performance of the Roche second generation hemoglobin A1c immunoassay in the presence of HB-S or HB-C traits. Ann Clin Lab Sci 38, 3136.Google ScholarPubMed
American College of Sports and Medicine (2017) ACSM’s Guidelines for Exercise Testing and Prescription, 10th ed. Philadelphia, PA: Wolters Kluwers/Lippincott Williams & Wilkins.Google Scholar
Astrand, PO & Ryhming, I (1954) A nomogram for calculation of aerobic capacity (physical fitness) from pulse rate during submaximal work. J Appl Physiol 7, 218221.CrossRefGoogle Scholar
Lemieux, I, Pascot, A, Couillard, C, et al. (2000) Hypertriglyceridemic waist: a marker of the atherogenic metabolic triad (hyperinsulinemia; hyperapolipoprotein B; small, dense LDL) in men? Circulation 102, 179184.CrossRefGoogle ScholarPubMed
Arsenault, BJ, Lemieux, I, Després, JP, et al. (2010) The hypertriglyceridemic-waist phenotype and the risk of coronary artery disease: results from the EPIC-Norfolk prospective population study. CMAJ 182, 14271432.CrossRefGoogle ScholarPubMed
Wilson, PW, D’Agostino, RB, Levy, D, et al. (1998) Prediction of coronary heart disease using risk factor categories. Circulation 97, 18371847.CrossRefGoogle ScholarPubMed
D’Agostino, RB Sr., Vasan, RS, Pencina, MJ, et al. (2008) General cardiovascular risk profile for use in primary care: the Framingham Heart Study. Circulation 117, 743753.CrossRefGoogle ScholarPubMed
American Diabetes Association (2014) Diagnosis and classification of diabetes mellitus. Diabetes Care 37, Suppl. 1, S81S90.CrossRefGoogle Scholar
Alessa, HB, Malik, VS, Yuan, C, et al. (2017) Dietary patterns and cardiometabolic and endocrine plasma biomarkers in US women. Am J Clin Nutr 105, 432441.CrossRefGoogle ScholarPubMed
Phillips, CM, Harrington, JM & Perry, IJ (2019) Relationship between dietary quality, determined by DASH score, and cardiometabolic health biomarkers: a cross-sectional analysis in adults. Clin Nutr 38, 16201628.CrossRefGoogle ScholarPubMed
Saraf-Bank, S, Haghighatdoost, F, Esmaillzadeh, A, et al. (2017) Adherence to Healthy Eating Index-2010 is inversely associated with metabolic syndrome and its features among Iranian adult women. Eur J Clin Nutr 71, 425430.CrossRefGoogle ScholarPubMed
Monfort-Pires, M, Salvador, EP, Folchetti, LD, et al. (2014) Diet quality is associated with leisure-time physical activity in individuals at cardiometabolic risk. J Am Coll Nutr 33, 297305.CrossRefGoogle ScholarPubMed
Alkerwi, A, Baydarlioglu, B, Sauvageot, N, et al. (2017) Smoking status is inversely associated with overall diet quality: findings from the ORISCAV-LUX study. Clin Nutr 36, 12751282.CrossRefGoogle ScholarPubMed
Hebert, JR (2016) Social desirability trait: biaser or driver of self-reported dietary intake? J Acad Nutr Diet 116, 18951898.CrossRefGoogle ScholarPubMed
Mozaffarian, D, Hao, T, Rimm, EB, et al. (2011) Changes in diet and lifestyle and long-term weight gain in women and men. N Eng J Med 364, 23922404.CrossRefGoogle ScholarPubMed
Estruch, R, Ros, E, Salas-Salvado, J, et al. (2013) Primary prevention of cardiovascular disease with a Mediterranean diet. N Eng J Med 368, 12791290.CrossRefGoogle ScholarPubMed
Appel, LJ, Sacks, FM, Carey, VJ, et al. (2005) Effects of protein, monounsaturated fat, and carbohydrate intake on blood pressure and serum lipids: results of the OmniHeart randomized trial. JAMA 294, 24552464.CrossRefGoogle ScholarPubMed
Gadgil, MD, Appel, LJ, Yeung, E, et al. (2013) The effects of carbohydrate, unsaturated fat, and protein intake on measures of insulin sensitivity: results from the OmniHeart trial. Diabetes Care 36, 11321137.CrossRefGoogle ScholarPubMed
Charreire, H, Kesse-Guyot, E, Bertrais, S, et al. (2011) Associations between dietary patterns, physical activity (leisure-time and occupational) and television viewing in middle-aged French adults. Br J Nutr 105, 902910.CrossRefGoogle ScholarPubMed
Cuenca-Garcia, M, Artero, EG, Sui, X, et al. (2014) Dietary indices, cardiovascular risk factors and mortality in middle-aged adults: findings from the Aerobics Center Longitudinal Study. Ann Epidemiol 24, 297303.e2.CrossRefGoogle ScholarPubMed
Mora, S, Cook, N, Buring, JE, et al. (2007) Physical activity and reduced risk of cardiovascular events: potential mediating mechanisms. Circulation 116, 21102118.CrossRefGoogle ScholarPubMed
Alves, AJ, Viana, JL, Cavalcante, SL, et al. (2016) Physical activity in primary and secondary prevention of cardiovascular disease: overview updated. World J Cardiol 8, 575583.CrossRefGoogle ScholarPubMed
Shah, RV, Murthy, VL, Allison, MA, et al. (2016) Diet and adipose tissue distributions: the multi-ethnic study of atherosclerosis. Nutr Metab Cardiovasc Dis 26, 185193.CrossRefGoogle ScholarPubMed
Fischer, K, Pick, JA, Moewes, D, et al. (2015) Qualitative aspects of diet affecting visceral and subcutaneous abdominal adipose tissue: a systematic review of observational and controlled intervention studies. Nutr Rev 73, 191215.CrossRefGoogle ScholarPubMed
Tchernof, A & Després, JP (2013) Pathophysiology of human visceral obesity: an update. Physiol Rev 93, 359404.CrossRefGoogle ScholarPubMed
Figure 0

Fig. 1. Proportion of participants in the poor (), intermediate () or ideal () category of metrics of ideal cardiovascular health in (a) men and (b) women. The analysis included 3128 men and 1657 women. The proportion of participants within each category was determined mainly according to the definition of the American Heart Association (AHA)(3). Criteria used were (1) smoking: never smoke or quit smoking for at least 12 months (ideal) and current smoking (poor), (2) BMI: < 25 kg/m2 (ideal) and ≥ 30 kg/m2 (poor), (3) physical activity level (PAL): ≥ 150 min/week (ideal) and none (poor), (4) total cholesterol (TC): < 5⋅2 mmol/l (ideal) and ≥ 6⋅2 mmol/l (poor) and (5) blood pressure (BP): systolic BP < 120 mmHg and diastolic BP < 80 mmHg (ideal) and systolic BP ≥ 140 mmHg or diastolic BP ≥ 90 mmHg (poor). Unlike the AHA definition, the healthy diet score was replaced by nutritional quality (NQ): > 75 (ideal) and < 60 (poor)(21) and fasting plasma glucose was replaced by HbA1c concentrations: < 5⋅7 % (ideal) and ≥ 6⋅5 % (poor)(33). Metrics between ideal and poor levels were considered in the intermediate category. Participants taking medications for dyslipidaemia, hypertension or diabetes were classified in the intermediate category if treated to goal. Otherwise, they were classified in the intermediate or poor category depending on the level reached under treatment.

Figure 1

Table 1. Characteristics of participants according to nutritional quality†(Numbers; percentages; mean values and standard deviations)

Figure 2

Fig. 2. Cardiometabolic risk profile across nutritional quality (NQ) subgroups: low NQ (; <60), moderate NQ (; 60–75) and high NQ (; >75). (a) Non-HDL-cholesterol (men, n 3035; women, n 1621); (b) cholesterol:HDL-cholesterol (men, n 3034; women, n 1621); (c) waist circumference (men, n 3126; women, n 1649); (d) TAG (men, n 3083; women, n 1645); (e) hypertriacylglycerolaemic (hyperTG ) waist carriers (men, n 2776; women, n 1357); (f) HbA1c (men, n 3083; women, n 1645); (g) systolic blood pressure (BP) (men, n 3127; women, n 1657); (h) diastolic BP (men, n 3127; women, n 1657). Values are means with their standard errors, except for panel (e) where values are expressed in percentages. Analyses are age-adjusted one-way ANOVA and were performed separately in men and in women. Posteriori comparisons were performed using the Tukey–Kramer adjustment for multiple comparisons. Categorical variables were compared by χ2 tests. * P < 0⋅05, ** P < 0⋅01, *** P < 0⋅001. † Analyses were performed on log-transformed data. Criteria for hyperTG waist are waist circumference ≥ 90⋅0 cm and TAG ≥ 2⋅0 mmol/l in men and waist circumference ≥ 85⋅0 cm and TAG ≥ 1⋅5 mmol/l for women(29,30).

Figure 3

Fig. 3. Proportion of carriers of the hypertriacylglycerolaemic (hyperTG) waist phenotype according to nutritional quality (NQ) and (a) physical activity level or (b) cardiorespiratory fitness. NQ subgroups: low (<60), moderate (60–75) and high (>75). Physical activity level subgroups: sedentary (<30 min/week), moderately inactive (30–149 min/week), moderately active (150–299 min/week) and active (≥ 300 min/week). Cardiorespiratory fitness subgroups are defined as proposed by American College of Sports and Medicine guidelines(27). Criteria for hyperTG waist are waist circumference ≥ 90·0 cm and TAG ≥ 2·0 mmol/l in men and waist circumference ≥ 85·0 cm and TAG ≥ 1·5 mmol/l in women(29,30). * Statistical difference between the sedentary/low NQ subgroup and the active/high NQ subgroup. † Statistical difference between the very poor/poor cardiorespiratory fitness/low NQ subgroup and the excellent/superior cardiorespiratory fitness/high NQ subgroup.

Figure 4

Table 2. Association of nutritional quality (NQ), physical activity level (PAL) and VO2max with cardiometabolic risk profile variables*(R2; β; standard errors; t values; P values and standardised β)

Figure 5

Table 3. Association of nutritional quality (NQ), physical activity level (PAL), VO2max and waist circumference (WC) with cardiometabolic risk profile variables*(R2; β; standard errors; t values; P values and standardised β)