Abstract
Objective
The purpose of this study was to examine the relationship between usual sugar-sweetened beverage (SSB) consumption and prevalence of abnormal metabolic health across body mass index (BMI) categories.
Design and Methods
The metabolic health of 6,842 non-diabetic adults was classified using cross-sectional data from the Framingham Heart Study Offspring (1998–2001) and Third Generation (2002–2005) cohorts. Adults were classified as normal weight, overweight or obese and, within these categories, metabolic health was defined based on five criteria – hypertension, elevated fasting glucose, elevated triglycerides, low HDL cholesterol, and insulin resistance. Individuals without metabolic abnormalities were considered metabolically healthy. Logistic regression was used to examine the associations between categories of SSB consumption and risk of metabolic health after stratification by BMI.
Results
Comparing the highest category of SSB consumers (median of 7 SSB per week) to the lowest category (non-consumers), odds ratios (95% confidence intervals) for metabolically abnormal phenotypes, compared to the metabolically normal, were 1.9 (1.1–3.4) among the obese, 2.0 (1.4–2.9) among the overweight, and 1.9 (1.4–2.6) among the normal weight individuals.
Conclusions
In this cross-sectional analysis, it is observed that, irrespective of weight status, consumers of SSB were more likely to display metabolic abnormalities compared to non-consumers in a dose-dependent manner.
Keywords: Obesity, metabolically healthy obese phenotype, sugar-sweetened beverages, chronic disease
Introduction
Several large prospective cohort studies have linked sugar-sweetened beverage (SSB) intake with a 20–30% higher incidence of type 2 diabetes (1) and cardiovascular diseases (2, 3). A recent population-based study of 8,157 adults followed for six years observed that an increase in SSB consumption was associated with a greater risk of developing metabolic syndrome (4). We previously observed that soda drinking was associated with greater insulin resistance (5) and increased risk of hypertriglyceridemia and low high-density lipoprotein cholesterol (6) in the Framingham Heart Study.
Obesity continues to be a major public health concern in the US, with over 30% of the population classified as obese (7, 8). While metabolic abnormalities typically accompany excess adiposity, not all obese adults display a “metabolically abnormal” phenotype. These “metabolically healthy obese (MHO)” individuals appear to be protected from developing metabolic abnormalities associated with obesity in some studies (9–16) and may be at lower risk of developing chronic diseases (17). In contrast, metabolically obese normal weight (MONW) adults appear to be at comparable risk as metabolically abnormal obese adults for CVD events (18).
To date, only a few observational studies have examined the role of diet in distinguishing metabolically healthy obese or overweight adults from metabolically unhealthy adults (19–21). One population-based study in a diverse sample of 775 obese, middle-aged adult Americans observed no difference in the diet composition between obesity phenotypes. Consistent with this, other cross-sectional studies have not observed any differences in the macronutrient composition of the diet according to metabolic health status in obese adults (20, 21), although diet quality was reportedly better in MHO individuals in one study (20). Thus, the purpose of the present study was to examine the relationship between usual SSB consumption and metabolic health across weight categories in disease-free, non-diabetic adults.
Methods and Procedures
Study Participants
Study participants included members of the Framingham Heart Study Offspring and Generation 3 cohorts. The Framingham Heart Study was initiated in 1948 as a longitudinal, population-based study of cardiovascular disease in 5,209 adult residents of Framingham, Massachusetts aged 28–62 (22). The study has continued over the past 60 years, with the survivors returning every two years for a physical examination and to complete a series of questionnaires, laboratory and cardiovascular tests. By 1971, the original cohort included 1,644 husband-wife pairs and 378 individuals who had developed CVD. The offspring of these subjects and the offspring’s spouses were invited to form a new cohort, and 5,135 of the 6,838 eligible individuals participated in the first Framingham Offspring Study examination (23). The Offspring cohort undergoes repeat examination approximately every 3–4 years. The 7th examination of the Offspring cohort began in October 1998 and was completed in October 2001. A total of 3,539 adults were examined. To provide greater resources for phenotypic and genotypic information, recruitment and examination of a third generation cohort (Gen3) was started in 2002. By 2005, the Gen3 study had enrolled 4,095 adults (ages 19 to 72 years). For the present study, we included all participants from the 7th Offspring cohort and the 1st Gen3 examinations who were free of diabetes and had a BMI≥18.5 at the time of the examination (n= 6,842).
The Boston University Medical Campus Institutional Review Board reviewed and approved all Framingham Heart Study protocols and procedures, and the current analysis of the existing Framingham data was reviewed and approved by the Tufts Medical Center Institutional Review Board.
Dietary Assessment
Usual dietary intake for the previous year was assessed using a 126-item semi-quantitative FFQ concurrently at the seventh and first examination cycles of the Framingham Offspring and Gen3 studies, respectively (24). The questionnaires were mailed to participants before the examination, and the participants were asked to bring the completed questionnaire with them to their examination at the Framingham Heart Study. The FFQ consists of a list of foods with a standardized serving size and a selection of nine frequency categories ranging from ‘never or less than one serving per month’ to ‘more than six servings per day’. Separate questions about the use of vitamin and mineral supplements and the type of breakfast cereal most commonly consumed were also included in the FFQ. FFQs with reported energy intakes of less than 600 kcal/d for men and women, or more than 4,000 kcal/d for women and 4,200 kcal/d for men, or with more than twelve food items left blank were considered invalid. Nutrient intakes were calculated by multiplying the frequency of consumption of each unit of food from the FFQ by the nutrient content of the specified portion. The relative validity of the FFQ has been reported for nutrients (24, 25), foods (26, 27), and dietary patterns (28). SSB consumption included total reported intake of 1) cola (e.g., Coke or Pepsi) with sugar, 2) caffeine free cola with sugar, 3) other carbonated beverage with sugar (e.g., 7-Up, ginger ale), and 4) non-carbonated fruit drinks (e.g., Hawaiian Punch, lemonade). One serving of SSB or diet soda was recorded as 360 mL (12 fl oz.). We created four categories of SSB intake for the sample: Non consumers, < 1 serving /wk, 1–3.5 servings/wk and greater than 3.5 servings/wk.. The median SSB intake in each of the 4 categories was 0, 0.5, 1.9 and 7.0 servings per week, respectively.
Defining metabolically healthy and abnormal participants
We based our classification of metabolic health on the National Cholesterol Education Program Adult Treatment Panel III (ATP-III) criteria for the following metabolic syndrome risk factors: hypertension, elevated fasting glucose concentrations, elevated triglyceride concentration, and low HDL cholesterol concentrations (29). To improve our sensitivity to identify insulin resistant individuals, we also included elevated fasting insulin as a metabolic risk factor (29). Participants meeting the criteria for any of the metabolic abnormalities were considered to be metabolically abnormal. Participants with no metabolic abnormalities were considered metabolically healthy.
Glucose concentrations were measured in fresh plasma specimens with a hexokinase reagent kit (A-gent glucose test; Abbott Laboratories, Inc., South Pasadena, CA, USA). Fasting plasma insulin was measured using human-specific radioimmunoassay in the 7th exam of the Offspring and enzyme-linked immunosorbent assay in the Third Generation (Linco Research Inc., St. Charles, MO). The intra-assay coefficient for fasting insulin was below 4% for both assays.
Triglycerides were measured enzymatically, and the HDL fraction was measured after precipitation of low-density and very low-density lipoprotein cholesterol with dextran sulfate magnesium. The intra-assay coefficients of variation were less than 3% for glucose, triglycerides and HDL.
We used the homeostasis model of insulin resistance (HOMA-IR) to evaluate insulin resistance in our study population using the following formula:
(Fasting Plasma Glucose Level (mg/dL) × Fasting Serum Insulin Level (µU/L))/405
We defined insulin resistance as a HOMA-IR level in the top quartile of the distribution among subjects without diabetes (30, 31).
Blood pressure was measured to the nearest 2 mmHg with a mercury column sphygmomanometer on the left arm after the subject had been seated quietly for 5 minutes. Two readings obtained by a physician were averaged to calculate the systolic and diastolic blood pressures. The use of medications including insulin, oral-hypoglycemic, antihyperlipidemic or antihypertensive was determined during the physical examination.
We stratified our analyses on body mass index (BMI)-based weight categories. BMI was calculated as weight in kilograms divided by the square of height in meters using examiner measured weight and height. Obesity was defined as a BMI > 30 kg/m2, overweight as a BMI > 25 and ≤30 kg/m2, and normal weight was defined as a BMI between 18.5 and 25 kg/m2.
As elevated LDL (≥ 130 mg/dL) is a significant cardiometabolic risk factor, we conducted secondary analyses including LDL criteria in the definition of metabolic health.
Covariates
Covariates included age, sex, BMI (as a continuous variable within the BMI strata), current smoking dose (0, 1–15, 16–25, or >25 cigarettes/d), cohort, menopausal status (y/n), energy intake, and diet quality as assessed using the Dietary Guidelines for Americans Adherence Index (DGAI), a measure of adherence to the key dietary intake recommendations in the 2005 Dietary Guidelines for Americans. In brief, there are eleven index items that assess adherence to energy-specific food intake recommendations and 9 items that assess adherence to nutrient intake recommendations. Possible values of the DGAI range from 0–20, with a higher score indicating greater adherence to the guidelines.
Statistical analysis
To describe subject characteristics across the categories of SSB intake, we stratified based on BMI weight categories (obese, overweight, normal weight) and calculated age- and sex-adjusted means and prevalence and their 95 percent confidence intervals (95% CI) using analysis of covariance (PROC GLM; SAS Institute, Inc.). Sex was adjusted for age, and age was adjusted for sex. Dietary variables were also adjusted for energy intake (kcal/day). For positively skewed variables, the data were logarithmically transformed, and the geometric means and their 95% CIs presented.
We used logistic regression to examine the association between SSB intake and odds ratio of the metabolically abnormal phenotype after stratification by BMI weight categories. The covariates included in the logistic regression models were age, sex, BMI, current smoking dosage, cohort, menopausal status (y/n), energy intake, and DGAI. Analysis of covariance was also used to examine the estimated prevalence of metabolic health with adjustment for the same covariates included in the logistic regression.
All statistical analyses were performed using the Statistical Analysis Systems statistical software package (version 9.2; SAS Institute, Cary, NC, USA). Unless otherwise noted, statistical significance refers to p-values <0.05.
Results
Of the 6,842 participants meeting the eligibility criteria for the present analyses, 1,735 were excluded from analyses because of missing or invalid data on dietary intake (n= 693), physical activity (n = 557), metabolic health markers (n= 456), or other covariates (n= 9). After exclusions, the sample size was 5,107 participants (2,349 men and 2,758 women). Participants who were excluded due to missing data were older (50.2 vs. 48.3 yrs), had a lower BMI (26.4 vs. 27.0 k/m2), and were more likely to be current smokers (17.6 vs. 13.0%) than those included in the analyses. Participants who were excluded (n = 1,042) did not differ from participants who were included with respect to SSB consumption, and there was no difference in metabolic health between those who were excluded (n= 1,178) and those who were included in the analyses. Among our study population, 25.2% were considered obese, 39.1% overweight, and 35.7% normal weight. The proportion of women classified as obese, overweight and normal weight were 50.8%, 58.7%, and 28.7%, respectively. Mean ages of the obese, overweight and normal weight adults were 50, 49 and 46 years, respectively. Prevalence of the metabolic healthy phenotype in the overall population was 31% and among the obese, overweight and normal weight participants was 10.8%, 25.6%, and 52.9%, respectively (Table 1). Of those displaying metabolic abnormalities in the total population, 24%, 17%, 14%, 8% and 4% of the sample had one, two, three, four and five metabolic risk factors, respectively.
Table 1.
Definition of Metabolically Healthy versus Metabolically Abnormal
| BMI Categories | ||||
|---|---|---|---|---|
| Metabolic Abnormalities Considered: | Normal 18.5– 25 kg/m2 (n=1825) |
Overweight >25–30 kg/m2 (n=1997) |
Obese >30 kg/m2 (n=1285) |
P trend |
| Insulin resistance (%): | ||||
| HOMA-IR ≥ 75thpercentile of non-diabetic population | 10.1 (8.2, 11.9) | 23.4 (21.7, 25.2) | 39.9 (37.7, 42.1) | < 0.001 |
| Elevated blood pressure (%): | ||||
| SBP ≥ 130 and/or DBP ≥ 85 or medication use | 21.2 (19.1, 23.3) | 40.6 (38.6, 42.6) | 57.1 (54.6, 59.7) | < 0.001 |
| Elevated triglycerides (%): | ||||
| Fasting triglycerides ≥ 150 mg/dL or medication use | 14.8 (12.8, 16.9) | 37.5 (35.5, 39.5) | 46.4 (43.9, 48.8) | < 0.001 |
| Elevated glucose (%): | ||||
| Fasting plasma glucose ≥ 100 mg/dL or medication use | 13.6 (11.6, 15.6) | 32.2 (30.3, 34.2) | 46.9 (44.5, 49.3) | < 0.001 |
| Low HDL-cholesterol level (%): | ||||
| HDL< 40 mg/dL in males;HDL< 50 mg/dL in females | 20.5 (18.4, 22.6) | 37.6 (35.6, 39.6) | 51.3 (48.8, 53.8) | < 0.001 |
| Metabolically Healthy (%): | ||||
| Free from any metabolic abnormality | 44.4 (42.6, 46.3) | 18.8 (17.0, 20.6) | 7 (4.8, 9.2) | < 0.001 |
We observed that individuals with greater SSB consumption were more likely to be men, more likely to be younger, more likely to be smokers, and had higher energy intakes and a lower overall diet quality relative to those with lower SSB intakes, irrespective of BMI category (Table 2). Within BMI weight categories, BMI was not related to SSB consumption.
Table 2.
Subject characteristics (Means and 95% CI’s) across categories of sugar sweetened beverage intake stratified by BMI category
| Categories of SSB intake (servings per week) | |||||
|---|---|---|---|---|---|
| 1 | 2 | 3 | 4 | P-trenda | |
| SSB intake median (min, max) | 0 (0,0) | 0.5 (0.5, 0.9) | 1.9 (1.0, 3.5) | 7.0 (3.9, 60.0) | |
| Number of participants | |||||
| Normal Weight (n) | 598 | 388 | 400 | 439 | |
| Overweight (n) | 582 | 369 | 508 | 538 | |
| Obese (n) | 449 | 239 | 278 | 319 | |
| Total | 1629 | 996 | 1186 | 1296 | |
| Male (%) b | |||||
| Normal Weight | 16.7 (13.1, 20.2) | 22.4 (18.1, 26.8) | 32.8 (28.5, 37.0) | 47.0 (42.9, 51.2) | < 0.001 |
| Overweight | 39.7 (35.8, 43.5) | 54.8 (50.0, 59.6) | 65.1 (61.0, 69.2) | 75.9 (71.8, 79.9) | < 0.001 |
| Obese | 36.2 (31.7, 40.7) | 46.0 (39.9, 52.1) | 56.2 (50.5, 61.9) | 70.3 (65.0, 75.7) | < 0.001 |
| Total | 29.9 (27.5, 32.2) | 40.0 (37.0, 42.9) | 52.2 (49.5, 55) | 65.2 (62.6, 67.9) | < 0.001 |
| Age (years) b | |||||
| Normal Weight | 50.5 (49.2, 51.7) | 46.6 (45.1, 48) | 45.1 (43.7, 46.4) | 42.1 (40.8, 43.4) | < 0.001 |
| Overweight | 52.7 (51.6, 53.7) | 50.8 (49.4, 52.1) | 48.8 (47.6, 49.9) | 45.8 (44.7, 46.9) | < 0.001 |
| Obese | 52.7 (51.5, 53.9) | 50.6 (49.0, 52.2) | 48.2 (46.7, 49.7) | 46.5 (45.2, 47.9) | < 0.001 |
| Total | 51.9 (51.3, 52.6) | 49.0 (48.2, 49.8) | 47.2 (46.5, 48) | 44.5 (43.7, 45.2) | < 0.001 |
| Current Smoker (%) b | |||||
| Normal Weight | 11.5 (8.5, 14.6) | 11.3 (7.7, 14.8) | 12.0 (8.6, 15.5) | 22.4 (19.2, 25.6) | < 0.001 |
| Overweight | 10.8 (8.0, 13.5) | 11.7 (8.3, 15.1) | 13.2 (10.3, 16.2) | 16.4 (13.4, 19.3) | 0.006 |
| Obese | 11.4 (8.3, 14.5) | 11.1 (6.9, 15.2) | 9.3 (5.5, 13.2) | 17.1 (13.4, 20.8) | 0.008 |
| Total | 11.0 (9.3, 12.6) | 11.0 (8.9, 13.1) | 11.6 (9.7, 13.5) | 18.4 (16.5, 20.2) | < 0.001 |
| BMI (kg/m2)b,c | |||||
| Normal Weight | 22.8 (22.6, 22.9) | 22.6 (22.5, 22.8) | 22.8 (22.6, 22.9) | 22.7 (22.5, 22.8) | 0.629 |
| Overweight | 27.1 (27.0, 27.3) | 27.3 (27.2, 27.4) | 27.3 (27.2, 27.4) | 27.4 (27.2, 27.5) | 0.054 |
| Obese | 34.4 (34.0, 34.8) | 33.9 (33.5, 34.4) | 33.9 (33.4, 34.3) | 34.1 (33.6, 34.5) | 0.561 |
| Total | 27.3 (27.1, 27.5) | 26.8 (26.5, 27.1) | 26.9 (26.7, 27.2) | 26.9 (26.7, 27.2) | 0.260 |
| Calorie intake (kcal/day) b | |||||
| Normal Weight | 1828 (1777, 1880) | 1864 (1804, 1925) | 2010 (1952, 2068) | 2346 (2291, 2401) | < 0.001 |
| Overweight | 1728 (1677, 1778) | 1843 (1780, 1906) | 1959 (1905, 2014) | 2280 (2225, 2335) | < 0.001 |
| Obese | 1819 (1761, 1878) | 1952 (1874, 2031) | 1997 (1924, 2070) | 2306 (2236, 2377) | < 0.001 |
| Total | 1785 (1755, 1816) | 1871 (1833, 1909) | 1981 (1947, 2016) | 2306 (2272, 2340) | < 0.001 |
| DGAI b | |||||
| Normal Weight | 10.0 (9.8, 10.2) | 9.6 (9.3, 9.8) | 9.5 (9.3, 9.8) | 8.3 (8.1, 8.6) | < 0.001 |
| Overweight | 9.6 (9.4, 9.8) | 9.3 (9.0, 9.5) | 8.8 (8.6, 9.1) | 8.3 (8.1, 8.5) | < 0.001 |
| Obese | 8.7 (8.5, 9.0) | 8.9 (8.6, 9.2) | 8.5 (8.2, 8.8) | 7.8 (7.5, 8.0) | < 0.001 |
| Total | 9.5 (9.3, 9.6) | 9.3 (9.1, 9.4) | 9.0 (8.8, 9.1) | 8.2 (8.0, 8.3) | < 0.001 |
P-trend: Linear trend across the categories of SSB intake was tested by using the median SSB intake in each group as a continuous variable.
Lifestyle variable adjusted for age (years) and sex. Sex was adjusted for age and age was adjusted for sex.
Geometric means and their 95% CIs are presented.
After multivariate adjustment, we observed a positive dose-response association between consumption of SSB and a metabolically abnormal phenotype (Table 3). The prevalence of the metabolically abnormal phenotype was almost double among the highest SSB consumers (highest category, median consumption of 7 SSB per week) relative to the lowest SSB consumers (lowest category, median consumption of 0 SSB per week). The odds ratio (95% confidence interval; P-trend) in the daily SSB consumers was 1.9 (1.1–3.4; P=0.04) for obese participants, 2.0 (1.4–2.9; P<0.001) for overweight participants, and 1.9 (1.4–2.6; P<0.001) for normal weight participants.
Table 3.
Odds ratios (OR) and 95% confidence intervals (CI) of the metabolically abnormal phenotype across categories of sugar sweetened beverage intake (servings per week) by stratum of weight status
| Categories of SSB intake (servings per week) | ||||||
|---|---|---|---|---|---|---|
| N | 1 | 2 | 3 | 4 | P-trenda | |
| Median SSB intake (servings/week) | 0 | 0.5 | 1.9 | 7.0 | ||
| Obese | 1285 | |||||
| Age and sex adjusted | 1 | 1.07 (0.64, 1.78) | 1.46 (0.88, 2.42) | 1.92 (1.11, 3.32) | 0.02 | |
| Multivariable adjustedb | 1 | 1.17 (0.69, 1.99) | 1.66 (0.98, 2.81) | 1.88 (1.05, 3.39) | 0.04 | |
| Overweight | 1997 | |||||
| Age and sex adjusted | 1 | 1.32 (0.96, 1.83) | 1.56 (1.15, 2.12) | 1.92 (1.40, 2.62) | < 0.001 | |
| Multivariable adjusted | 1 | 1.34 (0.96, 1.86) | 1.59 (1.17, 2.18) | 2.03 (1.44, 2.86) | < 0.001 | |
| Normal weight | 1825 | |||||
| Age and sex adjusted | 1 | 1.10 (0.83, 1.46) | 1.23 (0.93, 1.64) | 1.82 (1.37, 2.43) | < 0.001 | |
| Multivariable adjusted | 1 | 1.18 (0.88, 1.57) | 1.27 (0.95, 1.69) | 1.87 (1.36, 2.56) | < 0.001 | |
| Total | 5107 | |||||
| Age and sex adjusted | 1 | 1.09 (0.91, 1.31) | 1.30 (1.09, 1.56) | 1.65 (1.37, 1.98) | <0.001 | |
| Multivariable adjusted | 1 | 1.25 (1.02, 1.53) | 1.44 (1.18, 1.75) | 1.93 (1.56, 2.40) | <0.001 | |
P-trend: Linear trend across the categories of SSB intake was tested by using the median SSB intake in each group as a continuous variable.
Models adjusted for: age (years), sex, BMI (kg/m2), cohort, smoking dosage dose (0, 1–15, 16–25, 26+ cigarettes/day), menopausal status (y/n), energy intake (kcal/day) and DGAI.
In all BMI categories, the prevalence of metabolically healthy individuals decreased in a dose-response manner, across increasing categories of SSB (Table 4). In obese adults, the prevalence of the metabolically healthy phenotype decreased from 13.2% to 7.6% (P-trend = 0.03), in overweight individuals from 32.5% to 22.0% (P-trend=0.001) and in normal weight individuals from 53.9% to 41.2% (P-trend≤0.001) across SSB intake categories.
Table 4.
Prevalence (%) of metabolically healthy phenotype across categories of sugar sweetened beverage intake (servings/week) by stratum of weight status
| Categories of SSB intake (servings per week) | |||||||
|---|---|---|---|---|---|---|---|
| N | 1 | 2 | 3 | 4 | P-trenda | ||
| Median SSB intake (servings/week) | 0.0 | 0.5 | 1.9 | 7 | |||
| Obese | 1285 | ||||||
| Age and sex adjusted | 12.7 (9.7, 15.6) | 12.1 (8.3, 15.9) | 9.8 (6.2, 13.3) | 7.1 (3.7, 10.5) | 0.014 | ||
| Multivariable adjustedb | 13.2 (10.3, 16.1) | 11.8 (8.0, 15.5) | 9.1 (5.6, 12.6) | 7.6 (4.1, 11.0) | 0.026 | ||
| Overweight | 1997 | ||||||
| Age and sex adjusted | 31.8 (28.5, 35.2) | 27.5 (23.3, 31.6) | 24.9 (21.4, 28.5) | 21.6 (18.1, 25.2) | < 0.001 | ||
| Multivariable adjusted | 32.5 (29.1, 36.0) | 28.2 (24.1, 32.3) | 25.5 (21.9, 29.0) | 22.0 (18.3, 25.7) | 0.001 | ||
| Normal Weight | 1825 | ||||||
| Age and sex adjusted | 52.8 (48.5, 57.0) | 51.0 (46, 55.9) | 48.5 (43.8, 53.2) | 40.1 (35.7, 44.6) | < 0.001 | ||
| Multivariable adjusted | 53.9 (49.6, 58.1) | 50.6 (45.7, 55.5) | 49.0 (44.4, 53.7) | 41.2 (36.5, 46.0) | < 0.001 | ||
| Total | 5107 | ||||||
| Age and sex adjusted | 33.9 (31.7, 36.1) | 32.8 (30.1, 35.5) | 29.6 (27.2, 32.1) | 25.1 (22.8, 27.5) | < 0.001 | ||
| Multivariable adjusted | 35.4 (33.3, 37.5) | 32.0 (29.5, 34.5) | 29.3 (27, 31.6) | 25.1 (22.8, 27.5) | < 0.001 | ||
P-trend: Linear trend across the categories of SSB intake was tested by using the median SSB intake in each group as a continuous variable.
Models adjusted for: age (years), sex, BMI (kg/m2), cohort, smoking dosage (0, 1–15, 16–25, 26+ cigarettes/day), menopausal status (y/n), energy intake (kcal/day) and DGAI.
Inclusion of elevated LDL (>130 mg/dL) reduced the prevalence of metabolically healthy phenotype and attenuated the odds ratio across increasing categories of SSB in the obese and overweight statum, while strengthening the association in the normal weight adults (Supplemental table 1a and 1b). However, all associations observed in Table 3 and Table 4 remained statistically significant after inclusion of elevated LDL-cholesterol in the criteria.
Discussion
In this cross-sectional study, we observed a significant inverse association between the consumption of SSB and prevalence of the metabolically healthy phenotype, irrespective of body weight status. Adults who reportedly consumed one SSB daily were almost twice as likely to be metabolically unhealthy compared with those who did not consume any SSB. This strong positive association was observed between SSB consumption and prevalence of metabolic abnormalities among obese, overweight, and normal weight subgroups.
The findings of this study suggest that diet, in particular daily SSB consumption, is associated with metabolic health regardless of body size. This finding is supported by an earlier study that showed that higher SSB consumption was associated with a greater risk of developing metabolic syndrome over a 6 year follow-up (4). Hankinson and colleagues published the only prior study examining the influence of dietary factors on prevalence of the MHO phenotype. They found no association between 14 food groups, including sugar-sweetened beverages, and metabolic health among a diverse population of 775 obese American adults (19). Our study differs in many features from that of Hankinson and colleagues which may explain, in part, the inconsistent findings. For instance, we used a previously validated FFQ to capture long-term intake, rather than repeat 24-hour recalls to assess dietary intake (19). Furthermore, our study was larger with greater power to capture associations between diet and metabolic health. Finally, our characterization of metabolic health was based on objective biomarkers of dyslipidemia, insulin resistance, and blood pressure in conjunction with reported medication, physician diagnosis and prevalence of CVD. The difference in characterizing metabolic health is likely the reason that the prevalence of the MHO phenotype in the above-mentioned study was nearly double (i.e., 19%) the prevalence we observed in the Framingham Heart Study (i.e., 10.9%).
The underlying mechanism whereby higher SSB may be linked to greater metabolic risk may be due to the proportion of visceral fat, rather than absolute body weight. A study by Neeland et al. found an association between insulin resistance and visceral adiposity with incidence of type 2 diabetes among 732 multi-ethnic obese adults; however, there was no significant association with total or subcutaneous body fat (32). In a subsample of the Framingham cohort, we observed that SSB consumption was associated with a higher ratio of visceral to subcutaneous fat (33), which may contribute to this cardiometabolic risk. Consistent with the current cross-sectional study, no positive association was observed between SSB consumption and BMI, which may be explained by a number of other factors such as variation in other aspects of diet, increased energy expenditure or genetics. In fact, a recent observational analysis in a large cohort of men and women observed that the genetic association with adiposity appeared to be more pronounced with greater intake of SSB (34).
A recent population-based study among more than 40,000 participants found that metabolically healthy individuals, compared to metabolically abnormal obese individuals, were more likely to be physically fit. Even after adjusting for fitness and other confounders, metabolically healthy obese individuals had lower risk of all-cause mortality, non-fatal and fatal cardiovascular disease, and cancer mortality than their metabolically unhealthy obese peers (35). Interestingly, we observed that our findings regarding SSB consumption and metabolic health were not unique to the obese. Although the prevalence of metabolic risk factors was much higher in the obese than in normal weight individuals in our study population, the relationship between SSB consumption and metabolic risk appeared to be independent of weight status. Increased SSB consumption was associated with elevated metabolic risk in obese, overweight and normal weight individuals. This suggests that diet itself may not be a unique or important correlate of the MHO phenotype per se, but rather only a correlate of metabolic health. Alternatively, other lifestyle factors, genetic factors, or the interaction of both, may be associated with a metabolically healthy phenotype.
There are some limitations to our findings. First, the concept of MHO is still not well-defined, and the use of varying definitions complicates the consideration of this potential phenotype in research studies. We classified a relatively small proportion of obese individuals (10.9%) as MHO in this cross-sectional study of adults without diabetes. This prevalence, however, is consistent with previous research that used a combination of insulin resistance and other metabolic factors to define metabolically healthy (9, 14). On the other hand, it is important to note that while the definition of metabolically healthy is very stringent, the classification of metabolically abnormal is very heterogeneous, with variation in the number of metabolic risk factors and the combination of risk factors. A universal definition of metabolic health would further enable comparisons and conclusions among research studies.
Another limitation is the cross-sectional nature of these analyses. We cannot distinguish dietary behaviors that may be a result of the presence of certain metabolic abnormalities from metabolic abnormalities that result from long-term dietary behaviors. Furthermore, although we adjusted for physical activity in our analysis, cardiorespiratory fitness may be an independent, modifiable, risk factor that may protect cardiometabolic risk factor clustering (35). Longitudinal analyses in cohorts would provide more insight into the risk factors for development of metabolic abnormalities among MHO adults, including SSB consumption (36).
These findings suggest that lower consumption of sugar-sweetened beverages is associated with healthier metabolic profiles even in obese and overweight individuals. Based on our results, those who frequently consume SSB are more likely to display metabolic abnormalities compared to those individuals who consume no SSB, irrespective of body weight. It is essential that these observations be explored in the future, given the potential impact of a relatively simple dietary modification on metabolic health.
Supplementary Material
What is already known about this subject.
Little is known about why some obese individuals, known as the “metabolically healthy obese (MHO),” appear to be protected from developing metabolic abnormalities typically associated with obesity.
Recent evidence suggest that normal weight individuals who display metabolic abnormalities typically seen in obese populations (i.e. “metabolically obese normal weight (MONW)” individuals) are at increased risk of CVD events and mortality.
Only two studies have examined the differences in aspects of diet, in particular macronutrient composition, between obese metabolic phenotypes, and no differences have been reported.
What this study adds
As far as we are aware, this is the first cross-sectional study to focus on examining sugar-sweetened beverage (SSB) consumption and the prevalence of being classified as metabolically healthy, metabolically healthy obese (MHO), metabolically healthy overweight (MHOW), or metabolically obese normal weight (MONW) across BMI categories.
Our findings suggest that individuals who frequently consume SSB are more likely to display metabolic risk than those individuals who do not consume SSB, irrespective of body weight status.
Acknowledgements
The authors' responsibilities were as follows: PFJ and NMM contributed to the study concept and analysis design; CSF, JBM, PFJ contributed to data acquisition; AKG and GR performed the statistical analysis; PFJ, NMM, AKG, JBM, CF contributed to the interpretation of data; AKG drafted the manuscript, and all authors were involved in critical revision and had final approval of the manuscript.
We thank Kara A. Livingston for her assistance in the preparation of this manuscript.
This study was supported by grants from the National Heart Lung and Blood Institute (contract N01-HC-25195) and the USDA Agricultural Research Service (agreement 58-1950-0-014). The funding providers did not play a role in any aspect of this study.
Footnotes
Competing interests
The authors have no competing interests.
References
- 1.Malik VS, Popkin BM, Bray GA, Despres JP, Willett WC, Hu FB. Sugar-sweetened beverages and risk of metabolic syndrome and type 2 diabetes: a meta-analysis. Diabetes care. 2010;33:2477–2483. doi: 10.2337/dc10-1079. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Fung TT, Malik V, Rexrode KM, Manson JE, Willett WC, Hu FB. Sweetened beverage consumption and risk of coronary heart disease in women. The American journal of clinical nutrition. 2009;89:1037–1042. doi: 10.3945/ajcn.2008.27140. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.de Koning L, Malik VS, Kellogg MD, Rimm EB, Willett WC, Hu FB. Sweetened beverage consumption, incident coronary heart disease, and biomarkers of risk in men. Circulation. 2012;125:1735–1741. S1. doi: 10.1161/CIRCULATIONAHA.111.067017. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Barrio-Lopez MT, Martinez-Gonzalez MA, Fernandez-Montero A, Beunza JJ, Zazpe I, Bes-Rastrollo M. Prospective study of changes in sugar-sweetened beverage consumption and the incidence of the metabolic syndrome and its components: the SUN cohort. Br J Nutr. 2013:1–10. doi: 10.1017/S0007114513000822. [DOI] [PubMed] [Google Scholar]
- 5.Yoshida M, McKeown NM, Rogers G, et al. Surrogate markers of insulin resistance are associated with consumption of sugar-sweetened drinks and fruit juice in middle and older-aged adults. The Journal of nutrition. 2007;137:2121–2127. doi: 10.1093/jn/137.9.2121. [DOI] [PubMed] [Google Scholar]
- 6.Dhingra R, Sullivan L, Jacques PF, et al. Soft drink consumption and risk of developing cardiometabolic risk factors and the metabolic syndrome in middle-aged adults in the community. Circulation. 2007;116:480–488. doi: 10.1161/CIRCULATIONAHA.107.689935. [DOI] [PubMed] [Google Scholar]
- 7.Flegal KM, Carroll MD, Kit BK, Ogden CL. Prevalence of obesity and trends in the distribution of body mass index among US adults, 1999–2010. JAMA. 2012;307:491–497. doi: 10.1001/jama.2012.39. [DOI] [PubMed] [Google Scholar]
- 8.Flegal KM, Carroll MD, Ogden CL, Curtin LR. Prevalence and trends in obesity among US adults, 1999–2008. JAMA. 2010;303:235–241. doi: 10.1001/jama.2009.2014. [DOI] [PubMed] [Google Scholar]
- 9.Karelis AD, Brochu M, Rabasa-Lhoret R. Can we identify metabolically healthy but obese individuals (MHO)? Diabetes Metab. 2004;30:569–572. doi: 10.1016/s1262-3636(07)70156-8. [DOI] [PubMed] [Google Scholar]
- 10.Bonora E, Kiechl S, Willeit J, et al. Prevalence of insulin resistance in metabolic disorders: the Bruneck Study. Diabetes. 1998;47:1643–1649. doi: 10.2337/diabetes.47.10.1643. [DOI] [PubMed] [Google Scholar]
- 11.Brochu M, Tchernof A, Dionne IJ, et al. What are the physical characteristics associated with a normal metabolic profile despite a high level of obesity in postmenopausal women? J Clin Endocrinol Metab. 2001;86:1020–1025. doi: 10.1210/jcem.86.3.7365. [DOI] [PubMed] [Google Scholar]
- 12.Ferrannini E, Haffner SM, Mitchell BD, Stern MP. Hyperinsulinaemia: the key feature of a cardiovascular and metabolic syndrome. Diabetologia. 1991;34:416–422. doi: 10.1007/BF00403180. [DOI] [PubMed] [Google Scholar]
- 13.Bluher M. The distinction of metabolically 'healthy' from 'unhealthy' obese individuals. Curr Opin Lipidol. 2010;21:38–43. doi: 10.1097/MOL.0b013e3283346ccc. [DOI] [PubMed] [Google Scholar]
- 14.Wildman RP, Muntner P, Reynolds K, et al. The obese without cardiometabolic risk factor clustering and the normal weight with cardiometabolic risk factor clustering: prevalence and correlates of 2 phenotypes among the US population (NHANES 1999–2004) Arch Intern Med. 2008;168:1617–1624. doi: 10.1001/archinte.168.15.1617. [DOI] [PubMed] [Google Scholar]
- 15.Aguilar-Salinas CA, Garcia EG, Robles L, et al. High adiponectin concentrations are associated with the metabolically healthy obese phenotype. J Clin Endocrinol Metab. 2008;93:4075–4079. doi: 10.1210/jc.2007-2724. [DOI] [PubMed] [Google Scholar]
- 16.Karelis AD, Faraj M, Bastard JP, et al. The metabolically healthy but obese individual presents a favorable inflammation profile. J Clin Endocrinol Metab. 2005;90:4145–4150. doi: 10.1210/jc.2005-0482. [DOI] [PubMed] [Google Scholar]
- 17.Hamer M, Stamatakis E. Metabolically healthy obesity and risk of all-cause and cardiovascular disease mortality. The Journal of clinical endocrinology and metabolism. 2012;97:2482–2488. doi: 10.1210/jc.2011-3475. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Kramer CK, Zinman B, Retnakaran R. Are metabolically healthy overweight and obesity benign conditions?: A systematic review and meta-analysis. Ann Intern Med. 2013;159:758–769. doi: 10.7326/0003-4819-159-11-201312030-00008. [DOI] [PubMed] [Google Scholar]
- 19.Hankinson AL, Daviglus ML, Van Horn L, et al. Diet composition and activity level of at risk and metabolically healthy obese American adults. Obesity (Silver Spring) 2012 doi: 10.1002/oby.20257. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Phillips CM, Dillon C, Harrington JM, et al. Defining metabolically healthy obesity: role of dietary and lifestyle factors. PloS one. 2013;8:e76188. doi: 10.1371/journal.pone.0076188. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Bradshaw PT, Monda KL, Stevens J. Metabolic syndrome in healthy obese, overweight, and normal weight individuals: the Atherosclerosis Risk in Communities Study. Obesity (Silver Spring) 2013;21:203–209. doi: 10.1002/oby.20248. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Dawber TR, Meadors GF, Moore FE., Jr Epidemiological approaches to heart disease: the Framingham Study. Am J Public Health Nations Health. 1951;41:279–281. doi: 10.2105/ajph.41.3.279. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Feinleib M, Kannel WB, Garrison RJ, McNamara PM, Castelli WP. The Framingham Offspring Study. Design and preliminary data. Prev Med. 1975;4:518–525. doi: 10.1016/0091-7435(75)90037-7. [DOI] [PubMed] [Google Scholar]
- 24.Rimm EB, Giovannucci EL, Stampfer MJ, Colditz GA, Litin LB, Willett WC. Reproducibility and validity of an expanded self-administered semiquantitative food frequency questionnaire among male health professionals. Am J Epidemiol. 1992;135:1114–1126. doi: 10.1093/oxfordjournals.aje.a116211. discussion 27–36. [DOI] [PubMed] [Google Scholar]
- 25.Willett WC, Sampson L, Stampfer MJ, et al. Reproducibility and validity of a semiquantitative food frequency questionnaire. Am J Epidemiol. 1985;122:51–65. doi: 10.1093/oxfordjournals.aje.a114086. [DOI] [PubMed] [Google Scholar]
- 26.Feskanich D, Rimm EB, Giovannucci EL, et al. Reproducibility and validity of food intake measurements from a semiquantitative food frequency questionnaire. J Am Diet Assoc. 1993;93:790–796. doi: 10.1016/0002-8223(93)91754-e. [DOI] [PubMed] [Google Scholar]
- 27.Salvini S, Hunter DJ, Sampson L, et al. Food-based validation of a dietary questionnaire: the effects of week-to-week variation in food consumption. Int J Epidemiol. 1989;18:858–867. doi: 10.1093/ije/18.4.858. [DOI] [PubMed] [Google Scholar]
- 28.Hu FB, Rimm E, Smith-Warner SA, et al. Reproducibility and validity of dietary patterns assessed with a food-frequency questionnaire. Am J Clin Nutr. 1999;69:243–249. doi: 10.1093/ajcn/69.2.243. [DOI] [PubMed] [Google Scholar]
- 29.Expert Panel on Detection, Evaluation, and Treatment of High Blood Cholesterol in Adults: Executive summary of the Third Report of the National Cholesterol Education Program (NCEP) Expert Panel on Detection, Evaluation, and Treatment of High Blood Cholesterol in Adults (Adult Treatment Panel III) JAMA. 2001;285:2486–2497. doi: 10.1001/jama.285.19.2486. [DOI] [PubMed] [Google Scholar]
- 30.Matthews DR, Hosker JP, Rudenski AS, Naylor BA, Treacher DF, Turner RC. Homeostasis model assessment: insulin resistance and beta-cell function from fasting plasma glucose and insulin concentrations in man. Diabetologia. 1985;28:412–419. doi: 10.1007/BF00280883. [DOI] [PubMed] [Google Scholar]
- 31.Rutter MK, Wilson PW, Sullivan LM, Fox CS, D'Agostino RB, Sr, Meigs JB. Use of alternative thresholds defining insulin resistance to predict incident type 2 diabetes mellitus and cardiovascular disease. Circulation. 2008;117:1003–1009. doi: 10.1161/CIRCULATIONAHA.107.727727. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 32.Neeland IJ, Turer AT, Ayers CR, et al. Dysfunctional adiposity and the risk of prediabetes and type 2 diabetes in obese adults. JAMA : the journal of the American Medical Association. 2012;308:1150–1159. doi: 10.1001/2012.jama.11132. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 33.Ma J, Sloan M, Fox CS, Hoffman U, Jacques PF, McKeown NM. Sugar-sweetened beverage consumption is associated with relative distribution of abdominal adipose tissue in the Framingham Heart Study; Experimental Biology Conference; Boston, MA. 2013. [Google Scholar]
- 34.Qi Q, Qi L. Sugar-sweetened beverages, genetic risk, and obesity. The New England journal of medicine. 2013;368:286–287. doi: 10.1056/NEJMc1213563. [DOI] [PubMed] [Google Scholar]
- 35.Ortega FB, Lee DC, Katzmarzyk PT, et al. The intriguing metabolically healthy but obese phenotype: cardiovascular prognosis and role of fitness. Eur Heart J. 2013;34:389–397. doi: 10.1093/eurheartj/ehs174. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 36.Pataky Z, Bobbioni-Harsch E, Golay A. Open questions about metabolically normal obesity. Int J Obes (Lond) 2010;34(Suppl 2):S18–S23. doi: 10.1038/ijo.2010.235. [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.