In the bivariate analysis results, there was a significant association between anemia occurrence and total serum protein levels < 64 g/L (p = 0.009). The associations between BMI < 18.5 kg/m2, low body fat percentage, and anemia in crude analysis were non-significant.
The results of the binary logistic regression analysis are presented in Table 3. Two binary logistic models were used in this study. Model 1 included BMI-level binary variables and total serum protein level binary variables as the main independent variables and others as covariates. In Model 2, BMI was replaced by a fat mass level binary variable (low body fat percentage).In Model 1, variation in anemia across the total serum protein background remained at a significant level after adjustment for covariates. The variation in Model 2 across low body fat percentage shifted from a non-significant to a significant level after adjustment for covariates.
Model 2 fitted better than Model 1 (e.g., Nagelkerke R2 was 0.238 vs. 0.182, respectively), so patients with low body fat percentage had 8.5 times greater odds for belonging to the patients with anemia compared to the patients with normal or increased body fat. Those who had total serum protein level Figure 2. A direct correlation between hemoglobin level and total serum protein level was observed (ρ = 0.479, p 4. DiscussionThe main finding of the present study is the relationship between poor nutritional status and anemia in patients with UC. The phenotypic undernutrition indicators associated with anemia were decreased total serum protein and total body fat mass deficit. Based on the available literature, the pathogenetic pathways connecting these two factors with anemia were considered.
Hypoproteinemia in IBD may result from either systemic inflammation or decreased absorption of amino acids in the gastrointestinal tract. The main mechanism of hypoproteinemia in inflammation is elevated blood concentrations of proinflammatory cytokines, such as tumor necrosis factor-α, interleukin-1β, and interleukin-6 [19,20]. These cytokines inhibit the biosynthesis of negative acute phase proteins such as albumin and transferrin in hepatocytes, and increase the fractional catabolic rate of albumin. In addition, increased leakage of albumin into the interstitial space due to increased microvascular permeability occurs during acute phase or exacerbated chronic inflammation [21], which all result in hypoalbuminemia. Because many globulins are positive acute-phase proteins demonstrating elevated levels during inflammation, it is hypothesized that hypoproteinemia was due to hypoalbuminemia in the cohort of patients in this study. Therefore, if hypoalbuminemia is regarded as a marker of inflammation intensity, it would be logical to anticipate that patients with hypoalbuminemia will develop AI [12]. In this respect, it would be of interest to correlate the levels of albumin with disease activity and the extent of colonic involvement. The second potential cause of hypoproteinemia in the current patient sample is the decreased rate of blood protein synthesis in hepatocytes due to decreased available amino acids [22]. This can be attributed to a decreased protein intake, impaired digestion of dietary protein, and decreased amino acid absorption. Lambert et al. [23] demonstrated normal protein intake in patients with IBD. The decreased amino acid absorption rate secondary to mucosal injury is a likely plausible explanation mainly in patients with Crohn’s disease, because only 5% of dietary protein is absorbed in the colon. Nonetheless, decreased amino acid absorption can contribute to hypoproteinemia in patients with UC because of systemic hypercytokinemia effects on enterocyte apoptosis [24]. Irrespective of the exact mechanism, if one considers hypoproteinemia as a phenotypic criterion of undernutrition, it will lead to the appreciation of factors such as depletion of the free amino acid pool for globin synthesis within maturing erythroid cells in the pathogenesis of anemia in IBD.Low body fat percentage was the second nutritional parameter that demonstrated a significant association with anemia in the adjusted binary logistic regression model. Three principal mechanisms of decreased body fat mass in IBD have been described in the literature [25,26]. First, enterocellular malabsorption may be accompanied by impaired absorption and resynthesis of triglycerides due to mucosal injury by leukocyte infiltration [27], enterocyte apoptosis [28], and loss of tight junctions between enterocytes [29]. Second, fat deficit might be accounted for by decreased food intake due to decreased appetite and postprandial abdominal pain [23]. In this case, the caloric requirements of the body are not fully met, resulting in the partial utilization of endogenous substrates and triglycerides. Third, hypercytokinemia is associated with increased catabolism of triglycerides and an accelerated basal metabolic rate [30]. The changes in protein-related compartments were not analyzed in the present study; however, it is known that approximately 42% of patients with IBD develop sarcopenia, which persists during disease remission and is found in patients with normal and even increased BMI [31,32]. The latter phenomenon, termed sarcopenic obesity, may explain the lack of association between BMI and anemia in the current study, as BMI in such patients cannot be considered a reliable criterion of undernutrition [33]. Meanwhile, sarcopenia might have demonstrated a strong association with anemia, considering that the free amino acid pool for globin synthesis is derived from muscle protein catabolism or intestinal absorption.This research confirmed the association between anemia and fat tissue deficit. This finding could be explained by existing data on the stimulatory effects of the adipose tissue-derived hormone leptin on in vitro and in vivo erythropoiesis [34,35] in conjunction with the well-established direct correlation between plasma leptin concentration and body fat mass [36]. Thus, patients with body fat deficit might have relative leptin deficiency which contributes to the development of complex anemia in this patient cohort.The data obtained previously on cancer patients lend support to this view. In particular, Maccio et al. [37] investigated 888 patients with solid cancer and showed that leptin concentration, albumin level, and BMI were positively correlated with hemoglobin level. However, the association between hemoglobin concentration and total body fat percentage demonstrates a bell-shaped curve and seems to have a more complex pattern than direct correlation. It is well established that the majority of patients with morbid obesity have anemia, a mixture of IDA and AI according to its origin. Markedly elevated plasma leptin stimulates hepcidin expression in hepatocytes, thereby causing sequestration of iron in macrophages and decreasing iron absorption in the small intestine [38,39,40]. In conclusion, the results obtained improved the understanding of anemia pathogenesis in IBD patients and added two potentially new etiological factors: deficiency of amino acids for globin synthesis and leptin deficiency-mediated decreased erythropoiesis.The present study has several limitations. First, important parameters relevant to disease activity (e.g., C-reactive protein and proinflammatory cytokines) and iron metabolism (transferrin, ferritin, hepcidin, etc.) were not presented and analyzed because of the study’s retrospective design. Second, alterations in protein-related body compartments highly relevant to the mechanisms of anemia were not analyzed because of the lack of data in the records. Third, the number of patients with anemia in our cohort was relatively small (n = 32). When performing statistical analysis and modeling, we relied on the total numbers of patients in the sample rather than on the prevalence of parameters under study in the cohort. Moreover, it was impossible to envisage the real prevalence of anemia, hypoproteinemia, etc. in the population studied at the moment of study initiation. According to the generally held principles of logistic regression modeling, it is sufficient to have at least 10 observations per each studied independent variable. Fourth, the cross-sectional study design provides only limited evidence for the causal role of undernutrition in anemia. It should be noted, however, that the majority of analyzed factors with putative effects on iron metabolism and/or erythropoiesis, such as age, sex, and disease behavior, have already influenced the end-point at the time of hemoglobin measurement. This minimizes the probability of dynamic changes, although UC severity may be both the cause and consequence of malnutrition. Therefore, both patients with initial undernutrition and those with undernutrition secondary to severe UC might have been included in the study. It is presumed that such secondary interactions had minimal effect on the results of binary logistic regression modeling. Further studies will provide additional evidence of causality for the relationship between undernutrition and anemia in IBD patients using such criteria as “dose–response” effect, consistency of relationship, and experimental support.
留言 (0)