Highlights
-
•
Higher BMI linked to adverse cardiac remodeling independent of metabolic abnormalities.
-
•
Larger LV mass and LA dimension, and worse ventricular systolic/diastolic function independently linked to elevated BMI.
-
•
Weight control is crucial for preventing adverse remodeling regardless of metabolic comorbidities.
ABSTRACT
Background
Previous studies have reported that obesity-related metabolic abnormalities (eg, diabetes and hypertension) lead to myocardial dysfunction and adverse cardiac remodeling. However, it is unclear whether such cardiac remodeling is from obesity or obesity-related metabolic abnormalities. We hypothesize that overweight and obesity are associated with adverse cardiac remodeling independent of associated metabolic abnormalities.
Methods
We evaluated 6,639 participants from the Framingham Heart Study who underwent echocardiography and had no prevalent cardiovascular disease. Individuals were classified into 6 obesity sub-phenotypes based on metabolic health (metabolically healthy or metabolically unhealthy) and body mass index (normal weight, overweight, or obese). Obesity subphenotypes were related to echocardiographic measures using multivariable regression analyses.
Results
Mean age was 49 years and 55% were women. Overweight and obesity were consistently associated with adverse cardiac remodeling in both metabolic healthy and unhealthy participants. Among metabolically healthy participants, compared to the normal weight group (referent), overweight and obesity were significantly associated with increased left ventricular mass (11.6 and 21.4 gm), left atrium end-systolic dimension (0.27 and 0.48 cm), global longitudinal strain (0.82 and 1.06%), and the ratio of early diastolic trans-mitral flow velocity to early diastolic mitral annulus velocity (0.35 and 0.87) (all P <.001). Additionally, obesity was significantly associated with mitral annular plane systolic excursion (0.08 cm, P <.001) and relative wall thickness (0.01, P =.001) compared to the normal weight referent group.
Conclusions
Increasing body weight was associated with adverse cardiac remodeling regardless of metabolic health status, which suggests that obesity may directly increase the risk of adverse cardiac remodeling.
Background
Despite the decline in overall cardiovascular disease (CVD) mortality in the past several decades, CVD continues to be the leading cause of death in the US, and the prevalence and mortality of heart failure have increased. , Based on data from the National Health and Nutrition Examination Survey (NHANES), the prevalence of heart failure among Americans over 20 years of age increased from 6.0 million in 2015-2018 to 6.7 million in 2017-2020 and is projected to further increase to over 8 million in 2030. , In addition, obesity, a major risk factor for heart failure, has been increasing in the US population over the past several decades. , Overweight and obesity are associated with hemodynamic alterations characterized by increased cardiac output, increased central and total blood volume, higher central mean arterial pressure, and lower peripheral arterial resistance. The combined influences of these excess weight-related hemodynamic changes can lead to left ventricular dysfunction and hypertrophy. Other obesity-related metabolic abnormalities such as diabetes and hypertension may also lead to myocardial dysfunction and adverse cardiac remodeling. , As such, it can be challenging to disentangle the effects of obesity from those of its associated metabolic comorbidities.
Although overweight and obesity status are known precursors of cardiometabolic disorders, there has been increasing recognition of a subset of overweight or obese individuals labeled ‘metabolically healthy overweight or obese’ (MHOW or MHO, respectively), who appear unaffected by the metabolic abnormalities that often accompany overweight or obesity. , Previous studies have suggested that MHO is indicative of an early stage of obesity rather than denoting an inherent resistance to developing cardiometabolic disorders; most MHO individuals tend to develop metabolic abnormalities over time. , However, comparing metabolically healthy adults across different body mass index (BMI) categories can help delineate the independent effect of obesity on cardiac remodeling. While past studies have indicated an association between MHO and an increased risk of clinical CVD events, few studies have examined the independent effects of obesity on heart failure and its predecessor, subclinical cardiac remodeling. ,,,,
The present study investigated the relations between 6 distinct obesity sub-phenotypes (metabolically healthy and unhealthy normal weight, overweight, and obese) and measures of cardiac remodeling (as assessed by transthoracic echocardiography) to elucidate the independent effect of obesity on cardiac structure and function. We hypothesize that overweight or obese study participants will manifest echocardiographic evidence of adverse cardiac remodeling, regardless of their metabolic health category.
Methods
All Framingham Heart Study (FHS) cohort data and materials are publicly available at the Biologic Specimen and Data Repositories Information Coordinating Center of the National Heart, Lung, and Blood Institute at https://biolincc.nhlbi.nih.gov/studies/fhs/ .
Study participants
FHS design and participant recruitment have been described elsewhere. , The present investigation included participants from 5 separate FHS cohorts at 2 distinctive examination cycles. The final study sample ( n = 6639) included second-generation participants who attended their eighth (2005-2008) examination cycle ( n = 2184), third-generation participants who attended their first (2002-2005) examination cycle ( n = 3843), and Omni-1 and Omni-2 cohort participants who attended their third (2007-2008) ( n = 226) and first (2003-2005) ( n = 386) examination cycles, respectively. Attendees at these examinations underwent routine transthoracic echocardiography and a cardiovascular-focused physical examination. We excluded participants who had prevalent CVD or were on beta-blockers or cardiac glycosides. The Institutional Review Board of the Boston University Medical Center approved the study protocol, and all participants provided written informed consent.
Definition of obesity subphenotypes
All FHS participants underwent standardized physical examinations, including a medical interview, anthropometry and blood pressure (BP) assessments, and phlebotomy in a fasting state to assess standard CVD risk factors. We categorized participants into normal weight (BMI <25 kg/m 2), overweight (BMI ≥25 and <30 kg/m 2), and obese (BMI ≥30 kg/m 2) groups. We defined a metabolically unhealthy state as the presence of at least one of the following criteria: use of antihypertensive medications or elevated systolic or diastolic BP (≥140/90 mmHg); use of medications for diabetes or elevated fasting plasma glucose (≥126 mg/dL); use of lipid-lowering drugs or elevated serum triglycerides (≥150 mg/dL); and low high-density lipoprotein cholesterol (HDL <40 mg/dL in men and <50 mg/dL in women).
Based on the cross-classification of BMI (three groups) and metabolic health status (metabolically healthy vs metabolically unhealthy), we defined 6 obesity subphenotypes. Thus, participants were classified as metabolically healthy normal weight (MHNW), metabolically unhealthy normal weight (MUNW), metabolically healthy overweight (MHOW), metabolically unhealthy overweight (MUOW), metabolically healthy obese (MHO), and metabolically unhealthy obese (MUO).
Echocardiographic methods
All study participants underwent routine transthoracic echocardiography using a standardized protocol on an HP Sonos 5500 ultrasound machine (Philips Medical Systems, Andover, MA). In accordance with the recommendations of the American Society of Echocardiography, digitized images were obtained and measured offline using DigiView Software (Digisonics, Inc., Houston, TX). All echocardiograms were interpreted by a sonographer or cardiologist blinded to clinical information, following standardized protocols. Detailed descriptions of how these measures were obtained were published elsewhere. , Echocardiographic measures of interest for the present investigation included left ventricular mass (LVM), left atrium internal dimension (LAD), global longitudinal strain (GLS), the ratio between early mitral inflow velocity and mitral annular early diastolic velocity (E/e’ ratio), global circumferential strain (GCS), left ventricular fractional shortening (LVFS), mitral annular plane systolic excursion (MAPSE), and the ratio of left ventricular internal diastolic dimension to the sum of the diastolic thicknesses of the interventricular septum and left ventricular posterior wall (RWT). Binary outcomes of interest included left ventricular hypertrophy (LVH), defined as an LV mass/body surface area >95 g/m 2 in women and >115 g/m 2 in men, and left ventricular systolic dysfunction (LVSD), defined as LVFS <29%. An LVSD value of <29% corresponds to a left ventricular ejection fraction (LVEF) ≤50% and represents the first percentile of LVFS in a healthy reference sample.
Statistical methods
The baseline characteristics and echocardiographic measures of study participants were compared according to obesity subphenotypes. The associations between obesity subphenotypes (independent variables; with MHNW serving as the referent) and echocardiographic measures (dependent variable; separate model for each) were assessed using multivariable linear (continuous measures) and logistic (binary responses) regression. The models were adjusted for age, sex, smoking, resting systolic and diastolic BP, and resting heart rate. Covariates were selected based on prior knowledge of risk factors for cardiac remodeling.
Next, the associations of continuous BMI (independent variable, with effects measured per-unit increase) and echocardiographic measures (dependent variable; separate model for each) were assessed using multivariable linear (continuous measures) and logistic (binary responses) regression. These models were also adjusted for age, sex, smoking, resting SBP and DBP, resting heart rate, hypertension medications, and diabetes. Additionally, the associations of BMI and echocardiographic measures were explored in subgroup analyses by sex (men vs women) and age (<40 vs ≥40 years). The median age among metabolically healthy participants was approximately 40 years. Using the same approach, the associations between BMI and echocardiographic measures were examined in a subgroup analysis comprising solely metabolically healthy individuals from our study sample (including MHNW, MHOW, and MHO obesity subphenotypes).
All analyses were conducted using SAS version 9.4. The P -values of all models were based on 2-sided tests, with values <.05 indicating statistical significance. It’s important to note that multiple comparisons were not adjusted, so these results should be considered as hypothesis generation.
Results
Baseline characteristics of study participants
Table 1 shows the baseline characteristics of all 6,639 FHS participants included in this analysis, stratified by metabolic health and BMI categories. Metabolically healthy participants in a given BMI category were younger and more likely to be women relative to their metabolically unhealthy counterparts. As expected, metabolically healthy individuals had lower mean SBP and DBP, higher mean HDL-cholesterol, and lower mean blood triglycerides and glucose concentrations.
Table 1
Baseline characteristics of 6,339 Framingham heart study participants according to obesity subphenotypes.
| Characteristics | Metabolically healthy | Metabolically unhealthy | ||||
|---|---|---|---|---|---|---|
| Normal weight ( n = 1,350) | Overweight ( n = 651) |
Obesity
( n = 200) |
Normal weight ( n = 1,166) |
Overweight
( n = 1,799) |
Obesity
( n = 1,473) |
|
| Age, years | 41 ± 12 | 41 ± 12 | 44 ± 11 | 52 ± 16 | 54 ± 15 | 53 ± 14 |
| Women, n (%) | 1027 (76) | 341 (52) | 127 (64) | 735 (63) | 692 (38) | 724 (49) |
| Body mass index, kg/m 2 | 22 ± 2 | 27 ± 1 | 33 ± 3 | 23 ± 2 | 27 ± 1 | 35 ± 5 |
| Current Smoking, n (%) | 177 (13) | 86 (13) | 16 (8) | 202 (17) | 216 (12) | 193 (13) |
| Resting heart rate, beats/min | 60 ± 9 | 60 ± 9 | 62 ± 9 | 63 ± 10 | 62 ± 10 | 66 ± 10 |
| Systolic blood pressure, mmHg | 109 ± 10 | 113 ± 8 | 115 ± 8 | 123 ± 18 | 127 ± 16 | 128 ± 15 |
| Diastolic blood pressure, mmHg | 70 ± 7 | 72 ± 7 | 74 ± 7 | 74 ± 10 | 77 ± 10 | 79 ± 10 |
| Hypertension, n (%) | 0 | 0 | 0 | 404 (35) | 779 (43) | 804 (55) |
| Use of antihypertensive medications, n (%) | 0 | 0 | 0 | 271 (23) | 555 (31) | 581 (39) |
| HDL-cholesterol, mg/dL | 66 ± 15 | 59 ± 13 | 60 ± 13 | 59 ± 20 | 51 ± 15 | 48 ± 14 |
| Low HDL-cholesterol, n (%) | 0 | 0 | 0 | 534 (46) | 990 (55) | 924 (63) |
| Triglycerides, mg/dL | 65 (53, 87) | 77 (62, 104) | 82 (63, 107) | 90 (68, 131) | 115 (81, 167) | 130 (93, 181) |
| High triglycerides no medications, n (%) | 0 | 0 | 0 | 218 (19) | 591 (33) | 575 (39) |
| High triglycerides, n (%) | 0 | 0 | 0 | 408 (35) | 947 (53) | 842 (57) |
| Use of lipid-lowering medications, n (%) | 0 | 0 | 0 | 227 (19) | 489 (27) | 427 (29) |
| Diabetes, n (%) | 0 | 0 | 0 | 49 (5) | 104 (6) | 211 (15) |
| Fasting plasma glucose, mg/dL | 89 (84, 92) | 91 (88, 95) | 93 (89, 96) | 96 (89, 103) | 100 (93, 106) | 102 (95, 111) |
| Use of hypoglycemic agents, n (%) | 0 | 0 | 0 | 43 (4) | 97 (5) | 164 (11) |
| Echocardiographic measurements | Metabolically healthy | Metabolically unhealthy | ||||
| Normal weight ( n = 1,350) | Overweight ( n = 651) |
Obesity
( n = 200) |
Normal weight ( n = 1,166) |
Overweight
( n = 1,799) |
Obesity
( n = 1,473) |
|
| Left ventricular mass, gm | 121 (107, 141) | 147 (123, 171) | 146 (130, 178) | 133 (114, 157) | 161 (136, 185) | 170 (144, 200) |
| Left ventricular hypertrophy, n (%) | 23 (2) | 15 (2) | 7 (4) | 90 (8) | 147 (8) | 125 (8) |
| Left ventricular systolic dysfunction, n (%) | 2 (<1) | 4 (1) | 1 (1) | 6 (1) | 37 (2) | 24 (2) |
| Left atrium internal dimension, cm | 3.39 ± 0.38 | 3.75 ± 0.38 | 3.92 ± 0.38 | 3.54 ± 0.44 | 3.88 ± 0.46 | 4.09 ± 0.46 |
| Global longitudinal strain, % | −21 ± 3 | −20 ± 3 | −20 ± 3 | −21 ± 3 | −20 ± 3 | −19 ± 3 |
| E/e’ ratio | 5.3 (4.6, 6.2) | 5.5 (4.8, 6.4) | 6.2 (5.5, 7.3) | 5.8 (4.9, 7.0) | 6.0 (5.1, 7.3) | 6.6 (5.6, 8.0) |
| Global circumferential strain, % | −29 ± 5 | −29 ± 4 | −29 ± 4 | −30 ± 6 | −30 ± 5 | −30 ± 5 |
| Left ventricular fractional shortening, % | 36 ± 4 | 36 ± 4 | 36 ± 4 | 37 ± 4 | 37 ± 5 | 37 ± 5 |
| Mitral annular plane systolic excursion, cm | 1.6 ± 0.2 | 1.6 ± 0.2 | 1.7 ± 0.2 | 1.5 ± 0.2 | 1.5 ± 0.2 | 1.6 ± 0.2 |
| Relative wall thickness | 0.35 (0.33, 0.37) | 0.36 (0.33, 0.38) | 0.37 (0.34, 0.40) | 0.37 (0.34, 0.40) | 0.38 (0.35, 0.42) | 0.39 (0.36, 0.43) |
Values are mean ± standard deviation, median (inter-quarter range), or number (percentage).
Abbreviations: E/e’ ratio, the ratio between early mitral inflow velocity and mitral annular early diastolic velocity; HDL, high-density lipoprotein.
Associations between echocardiographic measures and obesity sub-phenotypes among all participants
The unadjusted mean, median, or percentage of echocardiographic measures by obesity sub-phenotypes are presented in Table 1 . When compared to MHNW participants, those in higher BMI categories had greater LVM, LAD, and E/e’ ratio, as well as more positive (worse) GLS, regardless of their metabolic health category ( Table 2 ). These differences were greater in obese individuals than overweight individuals in both metabolic healthy and unhealthy categories, and greater in MUO than in MHO groups. This general trend of increased adverse cardiac remodeling among overweight and obese individuals across both metabolic health categories, as well as the more pronounced remodeling in MUO compared to MHO, is exemplified by the adjusted log-transformed mean LVM values. Specifically, the log-transformed mean differences were 0.092, 0.087, 0.158, and 0.201 grams higher in the MHOW, MUOW, MHO, and MUO groups, respectively, compared to the MHNW group (all P-values < 0.001). Likewise, the adjusted untransformed mean differences in LVM were 11.6, 10.7, 21.4, and 29.1 grams, respectively, higher in the MHOW, MUOW, MHO, and MUO groups compared to the MHNW group. Additionally, MAPSE and RWT were higher among obese participants compared to normal-weight participants in both metabolically healthy and unhealthy groups.
Table 2
Differences in echocardiogram measurements according to obesity subphenotypes in all 6,339 participants.
| Continuous echocardiogram measurements | Metabolically healthy * | Metabolically unhealthy | |||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|
| Normal weight ( n = 1,350) | Overweight ( n = 651) |
Obesity
( n = 200) |
Normal weight ( n = 1,166) |
Overweight
( n = 1,799) |
Obesity
( n = 1,473) |
||||||
| Beta (SE) † | P value ‡ | Beta (SE) † | P value ‡ | Beta (SE) † | P value ‡ | Beta (SE) † | P value ‡ | Beta (SE) † | P value ‡ | ||
| Log-transformed left ventricular mass, gm | Reference | 0.092 (0.009) | <.001 | 0.158 (0.014) | <.001 | <0.001 (0.008) | .986 | 0.087 (0.007) | <.001 | 0.201 (0.008) | <.001 |
| Left atrium internal dimension, cm | Reference | 0.269 (0.019) | <.001 | 0.480 (0.029) | <.001 | 0.029 (0.017) | .084 | 0.277 (0.016) | <.001 | 0.544 (0.017) | <.001 |
| Global longitudinal strain, % | Reference | 0.819 (0.138) | <.001 | 1.049 (0.216) | <.001 | 0.449 (0.123) | <.001 | 0.956 (0.119) | <.001 | 1.301 (0.124) | <.001 |
| Log-transformed E/e’ ratio | Reference | 0.059 (0.011) | <.001 | 0.151 (0.017) | <.001 | 0.017 (0.009) | .084 | 0.063 (0.009) | <.001 | 0.144 (0.009) | <.001 |
| Global circumferential strain, % | Reference | −0.118 (0.234) | .613 | −0.513 (0.369) | .166 | −0.607 (0.209) | .004 | −0.642 (0.202) | .002 | −0.602 (0.211) | .004 |
| Left ventricular fractional shortening, % | Reference | −0.022 (0.194) | .909 | 0.433 (0.308) | .161 | 0.488 (0.172) | .005 | 0.354 (0.167) | .034 | 0.281 (0.175) | .108 |
| Mitral annular plane systolic excursion, cm | Reference | 0.016 (0.010) | .129 | 0.078 (0.016) | <.001 | −0.022 (0.009) | .018 | −0.004 (0.009) | .614 | 0.028 (0.009) | .002 |
| Log-transformed relative wall thickness | Reference | 0.009 (0.006) | .138 | 0.035 (0.009) | <.001 | 0.014 (0.005) | .007 | 0.028 (0.005) | <.001 | 0.055 (0.005) | <.001 |
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
