There are two major findings in the present study. First, compared with non-CHF participants, CHF participants exhibited higher BMI with lower energy and macronutrient intakes, lower physical activity level and longer rest time and hemodilution (lower hematocrit and hemoglobin). Second, higher water intake, sedentary time and hematocrit were observed in CHF patients with BMI ≥ 30kg/m2s, while energy, macronutrient intake and physical level were similar between CHF patients with BMI ≥ 30kg/m2 and < 30kg/m2.
Excess body weight and a sedentary lifestyle are major public health problems worldwide . Obesity (BMI ≥ 30 kg/m2) has been considered as a major independent risk factor for cardiovascular disease . A previous study showed that up to 40% of patients with CHF suffered from obesity . The interplay between obesity and CHF is complex. Despite obesity increases the risk of CHF, studies found that CHF patients who developed obesity or over-weight (25-29.9 kg/m2) are associated with better prognosis compared with those who did not [18–20], with reductions in cardiovascular disease mortality (19% and 40%, respectively) and all-cause mortality (16% and 33%, respectively) compared to heart failure patients with normal-weight (BMI 20-24.9 kg/m2), whereas heart failure patients with BMI < 20 kg/m2 increased total and cardiovascular mortality (27% and 20%, respectively) . This phenomenon is termed as “obesity paradox” and several potential reasons may account for it, such as earlier presentation, different etiology, greater metabolic reserves, protective cytokines, more tolerance of cardiac medications, different cardiorespiratory fitness level and less cachexia . Based on previous studies, our study termed “obesity” as “BMI ≥ 30 kg/m2” to investigate the differences for increased BMI in CHF participants.
In our analysis, we recognized CHF participants presented with higher BMI than non-CHF participants and a higher proportion of BMI ≥ 30 kg/m2 among CHF participants. Meanwhile, we also noticed an inadequate energy and macronutrient intakes in CHF participants, seen as significantly lower energy, protein, carbohydrate, total sugar, total fat, cholesterol, and water intakes. This implied that nutritional deficiency existed in CHF participants. Malnutrition is commonly prevalent in CHF, with a reported incidence of 7.5% and 57% in severe or moderate malnutrition, respectively . These patients are older and frailty, with negative calorie and nitrogen balance and energy availability  for physical activity and are triggered by multiple factors such as anorexia, malabsorption secondary to intestinal edema, high energy demand, and cytokine-induced hypercatabolism [7, 25] and worse outcomes [6, 26, 27]. Although when compared with CHF patients with BMI < 30kg/m2, those with BMI ≥ 30kg/m2 participants did not show a distinct increase in energy and macronutrient intakes. A constellation of researches revealed that failing myocardium utilized glucose switching from fatty acids as the primary energetic substrate to produce ATP , which may exert a protective role in preventing cardiomyocytes from oxidative radical excess and cell damage [29, 30], but a lower energy production . With more total fat intake and less glucose intake, cells may have the less energetic substrate to produce enough energy. This would affect metabolism in the body and cause adipose tissue accumulation, and subsequently increased body weight. This may imply that inadequate or unbalanced energy and macronutrient intakes are account for the increase in BMI or obesity among CHF participants. A daily caloric intake of about 29 kcal/kg and a daily protein intake of 1.2–1.4 g/kg was recommended for CHF patients at normal weights and a less energy intake was required (20–24 kcal/kg/day), and a reduction in dietary fat intake to about 25% of total caloric intake (0.6–0.8 g/kg/day) for overweight and obese CHF patients [32, 33].
Our study found lower hematocrit and hemoglobin levels in CHF participants, while CHF with BMI ≥ 30kg/m2 participants had higher hematocrit and hemoglobin concentration than those with BMI < 30kg/m2. These results demonstrated that hemodilution and fluid retention are common in CHF, especially in those with lower BMI level (BMI < 30kg/m2). This observation may be another potential reason to explain “obesity paradox”. Considerable evidence demonstrated that total blood volume and cardiac output are positively correlated with the degree of excess body weight [18, 22, 34] and further preserved or even increased skeletal muscle mass (i.e. lean mass) [35, 36], which may cause BMI higher vice versa. This may explain the phenomenon of elevated hematocrit and hemoglobin concentration in CHF with BMI ≥ 30kg/m2 participants and fluid retention may not be the reason for increased BMI or obesity in CHF participants. Hemodilution was common in chronic heart failure  and has a deleterious effect as it may impair peripheral oxygen delivery  and is often neglected as compensatory mechanisms may mask signs of volume . Our study found lower hematocrit and hemoglobin level in CHF participants, while CHF with BMI ≥ 30kg/m2 participants had higher hematocrit and hemoglobin concentration than BMI < 30kg/m2 ones. These results demonstrated that hemodilution and fluid retention are common in CHF, especially in those with lower BMI level (BMI < 30kg/m2). Previous studies indicated a higher mortality rate in patients with hemodilution than in those with hemoconcentration in acute heart failure patients , while fluid restriction could only improve signs and symptoms of chronic heart failure in patients in moderate to severe chronic heart failure  and aggressive fluid removal positively affected survival . In the present study, CHF with BMI ≥ 30kg/m2 participants had higher hematocrit and hemoglobin concentration than BMI < 30kg/m2 ones, indicating that obese CHF participants were less likely to develop hemodilution. This could be one explanation for “obesity paradox”. Furthermore, hemodynamic changes including increased stroke volume and increased arterial pressure may compensate the impaired peripheral oxygen delivery, leading CHF patients with increased BMI or obesity better prognoses. Still, we could not ignore that there was fluid retention in our CHF group which implied an inadequate usage of diuretics and fluid management should be enhanced.
Another finding in our analysis is that we demonstrated that CHF participants had lower physical activity levels and spent a longer time in rest, especially those BMI ≥ 30kg/m2 participants. It was revealed that sedentary time caused metabolic alterations at the muscle level and next influence gross metabolic disturbances in the whole body . It has been reported that prolonged sedentary time would impair mitochondrial function by elevating oxidative stress levels , which decreased the mitochondrial respiration level  and caused insufficient ATP production for daily activity and metabolism. Emerging evidences have demonstrated a significant dose-response association between sitting time and cardiovascular disease mortality [45–48] and the relative risks associated with sedentary time were higher among participants without regular physical activity [49, 50]. Meanwhile, there have been proven cardioprotective effects of regular physical activity on cardiovascular health, with improving cardiac compliance, reducing arterial stiffness and ventricular afterload, and finally reducing the risk of future cardiac dysfunction and improving cardiovascular outcomes . Moreover, researches showed an increasing lean mass with resistance exercise training could effectively improve muscular fitness in CHF [51, 52], which could prevent sarcopenia  or even cachexia in CHF patients and have better prognosis .
Our investigation has several limitations. We only included macronutrient in our analysis and lacked the data of micronutrient intake. Therefore, the present study may have underestimated the nutritional deficit and its effects on obesity among participants. Also, details regarding the etiology, subtype (HF with reduced ejection fraction versus HF with preserved ejection fraction), and severity of HF were not available. Moreover, CHF participants included in our analysis were self-reported in the NHANES survey. This may lead to possible reporting and recall bias. Finally, we lacked follow-up data of these participants, including BMI changes and relative outcomes, so we were unable to recognize the effects of nutrients intake and physical activity on BMI changes and prognoses in patients with CHF.