Association of Estimated Cardiorespiratory Fitness in Midlife With Cardiometabolic Outcomes and Mortality

Key Points Question Is estimated cardiorespiratory fitness (eCRF) in midlife associated with subclinical atherosclerosis, vascular stiffness, and risk of cardiometabolic disease and mortality? Findings In this cohort study of 2962 Framingham Offspring Study participants, higher midlife eCRF was associated with lower burdens of subclinical atherosclerosis and vascular stiffness, and with a lower risk of hypertension, diabetes, chronic kidney disease, cardiovascular disease, and mortality over a mean follow-up of 15 years. Meaning These findings suggest that prognostic ability of midlife eCRF may extend to a wide range of cardiometabolic diseases.


eCRF assessment
During each Heart Study examination, data for eCRF calculations were obtained. Data on age and sex were collected using a medical history questionnaire. Height and weight were measured via a standardized scale, and BMI was calculated as the weight in kilograms divided by the square of the height in meters. Waist circumference was measured by a trained professional applying anthropometric tape at the level of the umbilicus and recording the reading at midrespiration and rounding to the nearest 0.25 of an inch. Resting heart rate was measured using standard supine 12-lead electrocardiography performed after approximately 5 minutes of resting quietly. Physical activity was measured using the Physical Activity Index (PAI), which measures self-reported intensity, frequency, and duration of physical activity during leisure time and is categorized as active (yes) or insufficiently active (no) based on the median values of PAI as previously described. 1 Participants who self-reported smoking ≥1 cigarette/day during the year before each examination cycle were classified as current smokers.

Indicators of arterial stiffness and subclinical atherosclerosis
To quantify CFPWV, applanation tonometry was performed after participants rested for 5 minutes in the supine position using a commercially available tonometer (SPT-301, Millar Instruments, Houston, TX). 2 For each participant, the transit distances from the suprasternal notch to carotid and femoral sites were measured, and the difference between the two lengths was divided by the delay in time between the bottom of the carotid and femoral waveforms, as previously described. 3 All measures were stored digitally in a core (Cardiovascular Engineering, Inc, Norwood, MA) and blinded to clinical information for subsequent analyses. 2 Standard coronary computed tomography angiography was performed to calculate the standardized CAC score in which a calcified lesion was defined per the following standards: an area ≥3 connected pixels, an attenuation >130 Hounsfield units. An Agatston score was calculated as previously described. 4 Results were evaluated by a certified clinician.
Lastly, ultrasound of the carotid arteries was performed during the eighth examination cycle to quantify CIMT. The common carotid arteries (CCA) were imaged with a 7.5-MHz transducer while the carotid bulb and internal carotid arteries (ICA) were imaged with a 5-MHz transducer (3-dB point: 6.2 MHz). Color Doppler and Doppler spectral analyzer software (model SSH140A; Toshiba America Medical Systems) were applied. 5,6 Gated diastolic images were obtained of the left and right carotid arteries at the level of the distal CCA, the carotid artery bulb, and the proximal 2 cm of the ICA. Assessment by two independent interpreters (replicate readings, n=25) for the mean ICA and CCA IMT showed intraclass correlation coefficients of 0.74 and 0.90, respectively. 7

Covariates
Baseline blood pressure was measured twice at rest in a seated position by a physician using a mercury column sphygmomanometer and measurements were averaged. Fasting plasma glucose, total cholesterol (TC), and high-density lipoprotein cholesterol (HDL-C) concentrations were measured using standard enzymatic methods. Serum creatinine was measured using the modified Jaffé method and adjusted by a two-step serum creatinine calibration process as previously described. 8 The CKD-Epidemiology Collaboration (CKD-EPI) equation was used to estimate the glomerular filtration rate (eGFR). 9 Information on cardiovascular disease was collected by a medical history questionnaire, physical examination, and hospitalization records.

Statistical analysis
For analyses of eCRF trajectories, we limited the number of trajectory groups to ≤5, using the Bayesian Information Criterion to assess the best model fit. We accounted for the uncertainty of eCRF group membership by simulating 50 imputed datasets of trajectory group assignment reclassifying participants using the participant's posterior group probability. We then combined estimates from the 50 datasets using PROC MIANALYZE.

6.
Polak Abbreviations: eCRF, non-exercise estimated cardiorespiratory fitness; BMI, body mass index; WC, waist circumference; HR, heart rate; PAI, physical activity index. Note: Sex-specific eCRF at the seventh examinations were converted into z-score with mean of 0 and standard deviation of 1 (eCRFexam7); Sex-specific eCRF trajectories between second and seventh examinations were determined using group based modeling strategy (SAS PROC TRAJ, eCRFtrajectories); Sex-specific average of eCRF between second and seventh examinations were converted into z-score with mean of 0 and standard deviation of 1 (eCRFaverage .11 Abbreviations: eCRF, non-exercise estimated cardiorespiratory fitness; HR, hazard ratio; CI, confidence interval; SD, standard deviation; CKD, chronic kidney disease; CVD, cardiovascular disease. Note: Models were adjusted for age, sex, systolic blood pressure, diastolic blood pressure, antihypertensive medication, diabetes, total cholesterol/high-density lipoprotein cholesterol, lipid-lowering medication, and prevalence of CVD; Antihypertensive medication was not adjusted in the model using hypertension as an outcome; Fasting blood glucose was adjusted instead of diabetes in the model using diabetes as an outcome; estimated glomerular filtration rate was further adjusted in the model using CKD as an outcome; Prevalence of CVD was excluded in the model using CVD as an outcome; SDs are equal to 1.9 MET for hypertension, diabetes, and CKD samples and 2.0 MET for CVD and all-cause mortality sample. eTable 5. Associations of midlife eCRF with the incidence of cardiometabolic diseases and mortality among participants not on antihypertensive treatment.  .11 Abbreviations: eCRF, non-exercise estimated cardiorespiratory fitness; HR, hazard ratio; CI, confidence interval; SD, standard deviation; CKD, chronic kidney disease; CVD, cardiovascular disease. Note: Models were adjusted for age, sex, systolic blood pressure, diastolic blood pressure, antihypertensive medication, diabetes, total cholesterol/high-density lipoprotein cholesterol, lipid-lowering medication, and prevalence of CVD; No results were reported regarding the relation between eCRF and incident hypertension because the definition of hypertension include patients on antihypertensive treatment; Fasting blood glucose was adjusted instead of diabetes in the model using diabetes as an outcome; Estimated glomerular filtration rate was further adjusted in the model using CKD as an outcome; Prevalence of CVD was excluded in the model using CVD as an outcome; SDs are equal to 1.9 MET for diabetes and CKD samples and 2.0 MET for CVD and all-cause mortality sample. <.001 Abbreviations: eCRF, non-exercise estimated cardiorespiratory fitness; HR, hazard ratio; CI, confidence interval; SD, standard deviation; CIMT, carotid intima-media thickness; CAC, coronary artery calcium; AU, Agatston units; CKD, chronic kidney disease; CVD, cardiovascular disease. Note: Models were adjusted for age, sex, systolic blood pressure, diastolic blood pressure, antihypertensive medication, diabetes, total cholesterol/high-density lipoprotein cholesterol, lipid-lowering medication, and prevalence of CVD; Antihypertensive medication was not adjusted in the model using hypertension as an outcome; Fasting blood glucose was adjusted instead of diabetes in the model using diabetes as an outcome; estimated glomerular filtration rate was further adjusted in the model using CKD as an outcome; Prevalence of CVD was excluded in the model using CVD as an outcome; SDs are equal to 1.9 MET for hypertension, diabetes, and CKD samples and 2.0 MET for CVD and all-cause mortality sample.