If you don't remember your password, you can reset it by entering your email address and clicking the Reset Password button. You will then receive an email that contains a secure link for resetting your password
If the address matches a valid account an email will be sent to __email__ with instructions for resetting your password
There are no data on the effects of fat-free mass (FFM) and body fat (BF) on prognosis in patients with myocardial infarction (MI). We investigated the effects of FFM and BF (which were estimated using formulas rather than direct measurements) on 30-day and long-term all-cause mortality in patients with MI who underwent percutaneous coronary intervention. We analyzed data from 6,453 patients with MI. The patients were divided into 2 categories (high/low) according to the fat-free mass index (FFMI) and 2 categories (low/high) according to the BF. The resultant 4 patient groups: HighFFMI-LowBF, HighFFMI-HighBF, LowFFMI-LowBF, and LowFFMI-HighBF, were compared. The lowest crude mortality after 30 days and in the long term was observed in the HighFFMI-LowBF group (3.0%,9.8%, respectively), followed by the HighFFMI-HighBF group (6.6%, 27.0%, respectively), the LowFFMI-LowBF group (10.4%, 36.0%, respectively), and the LowFFMI-HighBF group (14.7%, 56.8%, respectively). The difference was significant (p <0.0001), as was the difference between groups. After adjustment, the FFMI-BF groups independently predicted 30-day mortality (p = 0.003), but the risk was similar in all groups. Compared with the HighFFMI-LowBF group, the long-term mortality risk was similar in the HighFFMI-HighBF group (hazard ratio [HR] 1.11, 95% confidence interval [CI] 0.84 to 1.47, p = 0.47), but the LowFFMI-LowBF and LowFFMI-HighBF patients had a higher risk (HR 1.59, 95% CI 1.20 to 2.11, p = 0.001, HR 1.40, 95% CI 1.03 to 1.91, p = 0.033, respectively). Body composition predicted mortality better than body mass index in patients with MI. Mortality appeared to be inversely related to FFM, with patients with low FFM and low BF having a particularly high mortality risk. The body composition groups also confirmed the obesity paradox.
Obesity is an independent risk factor for a worse prognosis.
However, in patients with established coronary artery disease (CAD), patients who are overweight and patients with obesity have a more favorable prognosis, which is referred to as the “obesity paradox.”
The body mass index (BMI) may differ considerably from true fatness in various age groups and between the Sex, and it cannot discriminate between an excess of body fat (BF) and an increase in fat-free mass (FFM), which may result in nutritional misclassification.
but there are no data on patients with myocardial infarction (MI). We aimed to evaluate the effects of FFM and BF on 30-day and long-term all-cause mortality in patients with MI who underwent percutaneous coronary intervention (PCI).
The cohort of the present single-center retrospective study was recruited from patients with MI who underwent PCI between 2007 and 2017 at the University Medical Center Maribor, a tertiary referral hospital, with 24 hours a day, 7 days a week PCI service. Of 7,343 consecutively screened patients, 846 patients without BMI data and 43 underweight patients (BMI <18.5 kg/m2) were excluded. The final patient cohort comprised 6,453 patients eligible for the analyses. Underweight patients were eliminated because of the possibility of reverse causation. BMI was calculated as the weight (kg) divided by the square of the height in meters. The Jackson-Pollock equation, which has previously been found to correlate well with the sum of skinfold measurements, was used to calculate the BF percentage (BF = [1.61 × BMI] + [0.13 × age] – [12.1 × gender] – 13.9), where Sex = 1 for men and 0 for women.
Patients were divided according to the FFMI into 2 groups based on the previously defined 50th percentile for FFMI analysis in Caucasians (the cutoff for the low FFMI group was ≤19.8 kg/m2 for men and ≤15.9 kg/m2 for women).
We then combined the FFMI and BF groups resulting in 4 groups of patients: HighFFMI-LowBF (633 patients), HighFFMI-HighBF (3,736 patients), LowFFMI-LowBF (1574 patients), and LowFFMI-HighBF (511 patients). These groups were compared. In addition, BMI groups were created using established cutoffs; normal weight 18.5 to 24.9 kg/m2, overweight 25.0 to 29.9 kg/m2, class I obesity 30.0 to 34.9 kg/m2, class II obesity 35.0 to 39.9 kg/m2, and class III obesity ≥ 40 kg/m2. The diagnosis of MI was established following published guidelines, as were the treatment protocols.
ESC Scientific Document Group 2017 ESC Guidelines for the management of acute myocardial infarction in patients presenting with ST-segment elevation: The Task Force for the management of acute myocardial infarction in patients presenting with ST-segment elevation of the European Society of Cardiology (ESC).
Data on the date of death were provided by the Slovenian National Cause of Death Registry. All-cause mortality was assessed after 30 days and over a median period of 6.0 (25th, 75th percentile; 3,9) years. Group FFMI/BF data were provided for all patients, and data on all other essential patient and procedural characteristics were at least 99.6% complete. Ascertainment of end points was 100% complete. Ethical, governance, and waiver of consent approvals were granted by the University Medical Center Maribor Committee for Medical Ethics (Reference: UKC-MB-KME-59/19), and all methods were performed following relevant guidelines and regulations.
Continuous variables are presented as mean ± SD, or median and interquartile range, and categorical variables as frequencies and percentages, with 2-tailed p values <0.05 being statistically significant. Baseline characteristics of patients across the FFMI-BF groups were compared using chi-square tests for categorical variables and analysis of variance for continuous variables. Univariable associations between the FFMI-BF groups and other baseline clinical and procedural variables were evaluated with the analysis of variance and chi-square tests. Logistic regression models were used to compute odds ratios (OR) as estimates of 30-day mortality. Cox proportional hazards regression analysis was used to compute hazard ratios (HRs) as estimates of long-term mortality. The models were adjusted for age, gender, diabetes mellitus (DM), hyperlipidemia, hypertension, glomerular filtration rate (GFR) on admission, ST-elevation MI (STEMI), thrombolysis in MI 0/1 after PCI, dual antiplatelet therapy, BMI, and FFMI-BF groups. Model covariates were predefined within the study design based on their clinical and pathophysiologic relevance as a confounder, significance in the literature, and frequency of occurrence in this cohort of patients. ORs and HRs were calculated with stratified models according to the FFMI-BF group. The patient group with the lowest observed mortality (the HighFFMI-LowFM group) was used as a reference. SPSS Statistical Software for Windows version 25.0 (Armonk, New York) was used for statistical analyses. The end points were all-cause 30-day and long-term mortality in the different FFMI-BF groups.
During follow-up, 1,938 patients (30.0%) died. The survivors tended to be younger, male, with higher BMI and FFMI, higher serum cholesterol, triglyceride, low-density lipoprotein (LDL), high-density lipoprotein (HDL) cholesterol levels, and a higher GFR. In contrast, they were more likely to have lower BF and suffered less frequently from DM and hypertension but more frequently from hyperlipidemia. More radial PCIs were performed in these patients who had a better outcome. They were also more likely to have received dual antiplatelet therapy. Baseline data and procedural characteristics of the study population are listed in Table 1. We found 633 (9.8%), 3,735 (57.9%),1,574 (24.4%), and 511 (7.9%) patients in the HighFFMI-LowBF, HighFFMI-HighBF, LowFFMI-LowBF, and LowFFMI-HighBF patient groups, respectively. The baseline characteristics of the 4 different body composition groups based on both FFMI and BF are listed in Table 2. The groups differed in age, gender, BMI, cholesterol, LDL and HDL cholesterol levels, triglyceride levels, GFR, the incidence of DM, hypertension, hyperlipidemia, STEMI, PCI of the left main coronary artery, and the success rate of PCI.
Table 1Baseline and procedural characteristics of the study population
Survivors n = 4,515
Nonsurvivors n = 1,938
Population n = 6,453
24.3 (18.7, 31.7)
24.5 (18.9, 31.9)
24.3 (18.7, 31.8)
Fat-free mass (kg)
Fat-free mass index (kg/m2)
189 (155, 224)
170 (136, 201)
183 (150, 220)
131 (88, 195)
116 (83, 169)
124 (88, 186)
40 (34, 49)
40 (32, 49)
123 (93, 151)
104 (78, 135)
GFR (ml/min/1.73 m2)
86.5 (70.3, 102.9)
62.5 (43.7, 81.7)
80.8 (61.3, 98.6)
PCI radial access
TIMI 0/1 after PCI
Data are expressed as mean ± SD or as number (percentage) or as median (interquartile range).
BMI = body mass index; %BF = body fat percentage; DAPT = dual antiplatelet therapy; GFR = glomerular filtration rate; HDL = high-density lipoprotein cholesterol; LAD = left anterior descendant artery; LCX = circumflex artery; LDL = low-density lipoprotein cholesterol; LMCA = left main coronary artery; PCI = percutaneous coronary intervention; Right = right coronary artery; STEMI = ST-elevation myocardial infarction; TIMI = thrombolysis in myocardial infarction.
In the HighFFMI-LowBF group, 75.5% of patients were overweight, and 24.5% were normal weight according to BMI criteria, with a median BMI of 26.0 kg/m2 (25.1 to 26.7). HighFFMI-HighBF patients were almost evenly distributed between overweight (47.1%) and obesity grade I (40.4%) categories, whereas 10.0% of patients had obesity grade II, and 2.0% had obesity grade III. Their median BMI was 30.1 kg/m2 (28.4 to 32.7). In the LowFFMI-LowBF group, all patients were of normal weight (83.9%) or overweight (16.1%) with a median BMI of 23.7 kg/m2 (22.2 to 24.7). Most LowFFMI-HighBF patients were overweight (75.5%), 15.3% were normal weight, and around 10% were obese (1.2% patients with obesity grade II and 8.0% patients with obesity grade III). Their median BMI was 26.4 kg/m2 (25.7 to 27.3).
LowFFMI patients were older and more often women who more frequently had non-STEMI. They were less likely to have DM, hypertension, and hyperlipidemia; however, they were more likely to have a poorer outcome of PCI and to have undergone PCI of the left main coronary artery. GFR on admission, serum cholesterol, triglycerides, and LDL cholesterol were more likely to be lower, but HDL cholesterol was higher. Patients with LowBF were predominantly men who were less likely to have DM, hypertension, hyperlipidemia, and renal dysfunction. They were more often admitted with STEMI, but the PCI was more likely to be successful. Their LDL and HDL cholesterols were more often lower but triglycerides higher.
After 30 days, 504 patients (7.8%) died. The lowest mortality was observed in the HighFFMI-LowBF group (3.0%), followed by the HighFFMI-HighBF group (6.6%). Mortality continued to increase in the LowFFMI-LowBF and LowFFMI-HighBF groups (10.4% and 14.7%, respectively). The relation between FFMI-BF groups and mortality was almost linear (p <0.0001) (Figure 1, Table 2). The same pattern was found for unadjusted long-term all-cause mortality. By the end of the follow-up period, 1,938 patients (30%) had died. The HighFFMI-LowBF group had the lowest long-term mortality (62 [9.8%] patients died), followed by the HighFFMI-HighBF group (1,009 [27.0%] patients died), the LowFFMI-LowBF group (577 [36.0%] patients died) and the LowFFMI-HighBF group (290 [56.8%] patients died), which had the highest observed long-term mortality. The difference in mortality rate was significant (p <0.0001), as was the difference between each group (p <0.0001) (Supplementary Figure 1, Table 2). After adjusting for confounders, the FFMI-BF groups independently predicted 30-day mortality (p = 0.031) (Table 3). The 30-day mortality risk was similar across the FFMI-BF groups (OR 0.74; 95% CI 0.41 to 1.32, p = 0.31, OR 1.42; 95% CI 0.80 to 2.51, p = 0.23 and OR 1.08; 95% CI 0.56 to 2.08, p = 0.82, respectively) for the HighFFMI-LowBF, LowFFMI-LowBF, and LowFFMI-HighBF patient groups compared with the HighFFMI-LowBF group (Table 3).
Table 3Predictors of 30-day and long-term mortality (HighFFMI-LowBF group as reference)
OR (95% CI)
HR (95% CI)
0.82 (0.65 to 1.03)
0.86 (0.78 to 0.96)
1.03 (1.02 to 1.05)
1.05 (1.04 to 1.06)
1.04 (1.01 to 1.07)
1.01 (0.99 to 1.03)
1.22 (0.96 to 1.55)
1.65 (1.50 to 1.82)
2.14 (1.72 to 2.67)
1.24 (1.13 to 1.37)
2.83 (2.18 to 3.68)
1.54 (1.40 to 1.70)
HighFFMI - HighBF
0.74 (0.41 to 1.32)
1.11 (0.94 to 1.47)
LowFFMI - LowBF
1.42 (0.80 to 2.51)
1.59 (1.20 to 2.11)
LowFFMI - HighBF
1.46 (1.03 to 1.91)
4.49 (3.49 to 5.78)
1.35 (1.23 to 1.48)
0.972 (0.968 to 0.976)
0.984 (0.982 to 0.986)
0.18 (0.14 to 0.25)
2.39 (1.75 to 3.27)
1.52 (1.31 to 1.76)
BF = body fat; BMI = body mass index; DAPT = dual antiplatelet therapy; FFMI = fat-free mass index; GFR = glomerular filtration rate; HR = hazard ratio; OR = odd ratio; STEMI = ST-elevation myocardial infarction; TIMI = thrombolysis in myocardial infarction.
The FFMI-BF groups independently predicted the long-term mortality risk (p <0.0001) (Table 3). We compared the groups using the HighFFMI-LowBF group as a reference. The multivariable-adjusted long-term mortality risk of the HighFFMI-HighBF group was similar (adjusted HR 1.11, 95% CI 0.84 to 1.47, p = 0.47) (Figure 2, Table 3). However, both the LowFFMI groups, LowFFMI-LowBF, and LowFFMI-HighBF, had a higher risk (HR 1.59, 95% CI 1.20 to 2.11, p = 0.001 and HR 1.40, 95% CI 1.03 to 1.91. p = 0.033, respectively) (Figure 2, Table 3). Patients with LowFFMI-LowBF had the highest long-term risk of death. Other independent predictors of 30-day and long-term mortality are listed in Table 3.
Our analysis showed that body composition is important for both short-term and long-term outcomes in patients with MI. Both HighFFMI groups had better survival rates than the LowFFMI groups. Mortality was found to be inversely related to FFMI, as previously seen in patients with stable CAD.
Unfortunately, we lack data on muscle strength and cardiorespiratory fitness. After adjusting for confounders, body composition was important for 30-day mortality (p = 0.003), although the risk was similar across the FFMI-BF groups (Table 3). When the long-term mortality risk was compared with the reference group (HighFFMI-LowBF), the HighFFMI-HighBF group had a similar mortality risk (Figure 2, Table 3). High FFMI enables good mobility and largely neutralizes the negative effects of excess BF.
In contrast, patients with LowFFMI-LowBF and LowFFMI-HighBF had an almost 1.6-fold and 1.4-fold higher risk of dying in the long-term compared with the HighFFMI-LowBF group. It is worth noting that patients with LowFFMI-LowBF had the highest long-term mortality risk (Figure 2, Table 3). This differed from the observed crude long-term mortality, where the LowFFMI-HighBF group had the worst outcome (Supplementary Figure 1). At the time of MI, the LowFFMI-HighBF patients were sicker and older than the LowFFMI-LowBF patients (almost 11 years older, with a lower GFR and more DM and hypertension—all factors known to be associated with a worse outcome), which explains the higher unadjusted mortality.
Our finding confirms the previous data that the combination of low FFMI and low BF denotes patients with the highest risk of dying,
Unfortunately, we lacked data on waist circumference, and we were unable to distinguish between subcutaneous and visceral fat. Agreeing with previous findings, we found that BMI does not take into account which body compartment is responsible for excess weight, nor does it correctly estimate body composition, leading to misclassification of nutritional status.
Comparisons with previous studies must consider differences in the number of patients, patient selection, covariates in multivariable adjustment, observation time, and treatment options.
The body composition groups confirmed the obesity paradox in patients with MI. Patients with normal weight (LowFFMI-LowBF- median BMI 23.7 kg/m2 [22.2 to 24.7]) had the highest long-term mortality risk. All the other groups had a higher median BMI and lower long-term mortality risk.
Our findings may have potential implications for patients with MI. Although body composition assessment would be desirable for each patient with MI, the methods are time-consuming, not always accurate, and costly.
Patients with similar BMIs may have different body compositions and, more importantly, prognoses. Body composition groups also confirmed the obesity paradox. Nonetheless, because of its simplicity, BMI will probably remain the cornerstone of obesity assessment.
Several potential limitations of the study should be emphasized. BF was assessed using a calculation that correlates well with the sum of the skinfold measurements, as opposed to more sophisticated methods such as dual-energy x-ray absorptiometry, bioelectrical impedance, or hydrostatic weighing. We lack data on waist circumference and the waist:hip ratio, and we could not distinguish between visceral and subcutaneous fat. In addition, we could not focus on the possible differential association between subcutaneous and visceral fat and cardiometabolic risk and outcome, nor did we measure the distribution of FFM. Only all-cause mortality was assessed, and other cardiovascular events were not recorded. Because our study was retrospective in nature, it provides only associative, not causative, evidence. We could not exclude selection bias as only the healthiest patients with obesity survived long enough to be included in the analysis. The data on possible changes in BMI during the observation period were not available. We lacked data on the physical activity before and after MI, cardiorespiratory fitness, and inflammatory biomarkers known to be associated with mortality. We also lacked any assessment of muscle strength. Only data on the use of P2Y12 receptor inhibitors during hospitalization were available. Hence we were unable to adjust for potential differences in medication known to influence mortality. Only Caucasians were included in the analysis, so the generalizability of our findings is questionable. The data on smoking, previous MI or heart failure, previous revascularization, and socioeconomic status, known to be strong predictors of mortality, were not available.
In conclusion, our data indicate that body composition is important in patients with MI, suggesting that mortality is inversely related to FFMI. Our results suggest that patients with low FFMI and low BF are at particularly high risk of dying during follow-up, confirming the obesity paradox. Further studies with better methods to assess FFMI and BF are needed to determine the optimal body composition for patients with MI.
The authors have no conflicts of interest to declare.
The datasets used and/or analyzed during the present study are available from the corresponding author on reasonable request.
The study was approved by the Hospital Ethics Committee (UKC-MB-KME-24/19).
The authors thank Mario Gorenjak, PhD, for assistance with forest plot statistics.
2017 ESC Guidelines for the management of acute myocardial infarction in patients presenting with ST-segment elevation: The Task Force for the management of acute myocardial infarction in patients presenting with ST-segment elevation of the European Society of Cardiology (ESC).