Hill, Rebecca J.; Davies, Peter S.W.
Growth and nutritional status are important issues in paediatric inflammatory bowel disease (IBD); however, research in both adults and children suggests that patients with ulcerative colitis (UC) are not as compromised, if at all, compared with those with Crohn disease (1–5). However, research in this area is not extensive, as detailed by Valentini et al, (6) and is often compromised by the use of simple measures, such as the body mass index (BMI) and skinfold thickness, to assess nutritional status rather than more appropriate and sophisticated techniques to measure body composition.
BMI is frequently used both clinically and in research as an indicator of nutritional status because of its ease of calculation; however, it is a proxy measure at best. BMI is a means of adjusting weight for height, and as such, gives information about a patient's body size. Although it can give an indication if a patient is overly large or small for their height, it does not provide information on body composition, and it is body composition that is a much more accurate measure of nutritional status.
Jahnsen et al (3) reported BMIs in adults with UC to be significantly greater than both patients with CD and healthy controls, with no differences found between the latter 2 groups; however, when body composition was assessed via the measurement of lean body mass (dual-energy x-ray absorptiometry), this parameter was not significantly different in the patients with UC compared with controls. A higher percentage of fat mass in the patients with UC contributed to increased weight, and hence, BMI, which in itself has detrimental implications for the course of UC, as well as general health (7). Elevated fat mass was also found by Valentini et al (6) in apparently well-nourished patients with UC (defined by subjective global assessment, BMI and albumin measures) who were matched for BMI with healthy controls. Although BMI was not different and fat mass was significantly increased, lean body mass and body cell mass (BCM), assessed via bioelectrical impedance, were significantly decreased compared with controls. These changes, suggestive of malnutrition, were not related to disease location, duration, or activity.
Changes in BCM are clinically important because it is a good indicator of functional nutritional status and it is sensitive to early changes in nutrition, medication, or disease progression (8). BCM is the metabolically active component of fat-free mass. In patients with IBD, the measurement of BCM is superior to other available methods that effectively derive a value for fat-free mass because fat-free mass contains extracellular water, which is known to be expanded in volume in cases of anorexia and CD during active treatment (9,10). Using BCM therefore negates the effect of an expanded extracellular water space on estimates of body composition, and hence, nutritional status. Furthermore, we believe measurements of nutritional status in conditions where malnutrition is potentially evident should involve the measurement of fat-free mass, or a component thereof, as opposed to the measurement of fat mass because it is changes in fat-free mass that are more indicative of disease progression and outcomes such as morbidity and mortality (11).
The present study detailed nutritional status in patients with UC in relation to their BCM, and illustrated the problems of using BMI to assess nutritional status in this cohort.
Children with UC were recruited as part of ongoing research on the effect of IBD on growth, development, and nutritional status. Measurements of nutritional status were available in 18 children with UC (7 boys; 11 girls) ages 8.6 to 17.1 years. Repeated measurements of nutritional status were performed on this cohort at 6-monthly intervals, yielding a total number of measurements available for analysis of 77. Eighty-three percent of the cohort had measurements through to 1 year, with 50% and 28% to 2 and 3 years, respectively.
Body weight was recorded to the nearest 100 g using a digital scale, and height was measured to the last completed millimetre using a wall-mounted stadiometer. BMI was calculated as weight/height2 (kg/m2). z scores were calculated using CDC 2000 data (12). The 5th percentile was used as a cutoff for underweight (as defined by the Centers for Disease Control and Prevention (13)), which is equal to a z score of −1.65.
Nutritional Status: BCM
In the present study, we report total body potassium40 (TBK) measurements of BCM and calculated BCM (adjusted for height) z scores (14). The measurement of BCM was calculated via the assessment of TBK using a shadow shield whole body counter (Accuscan, Canberra Industries, Meridien, CT). This equipment uses 3 sodium iodide crystal scintillation detectors placed above a movable bed. TBK was measured during two 18-minute scans, during which all of the subjects’ body passes under the detectors. An average was taken of the 2 scans. In our laboratory, TBK is measured in grams; however, this can be easily converted to mmol and BCM was then calculated using the equation described by Wang et al (15), which has recently been validated for use in children (16).
BCM (kg) = 0.00918 × TBK (mmol),
where 0.00918 is a constant that represents the relation between the TBK in mmol found in 1 kg of BCM.
As with BMI, the 5th percentile was defined as a z score of −1.65, below which indicated compromised nutritional status.
Disease activity was determined using the Pediatric Ulcerative Colitis Activity Index (17).
Blood samples were collected at each time point by experienced phlebotomists and analysed using standard procedures for C-reactive protein (CRP), erythrocyte sedimentation rate (ESR), and serum albumin.
The mean height, weight, BMI, and BCM z scores were compared with zero using a single-sample t test, and BMI and BCM z scores were compared using a paired-sample t test and Bland-Altman analysis (18). Sample size was calculated based on finding a significant difference of 0.50 of a z score between BMI and BCM. At 80% power and 5% significance, a minimum of 64 measurements were required. Relations were determined through correlation analysis, and differences in frequencies between groups were determined using χ2. Differences in z scores between disease activity groups were determined using analysis of variance.
The present study was conducted according to the guidelines laid down in the Declaration of Helsinki, and all the procedures involving human subjects/patients were approved by the human research ethics committee of the University of Queensland, Brisbane, and the Royal Children's Hospital and Health Services District ethics committee, Brisbane. Informed written consent was obtained from the parents or guardian of the children involved who, in turn, gave assent to be involved in the study.
BMI and BCM Comparisons
For the 77 measurements, mean (±SD) height, weight, BMI, and BCM z scores were 0.39 (0.96), 0.20 (1.08), −0.05 (1.18), and −0.74 (1.41), respectively. No differences were found for these parameters between each 6-month measurement. The mean z score at the population level should be zero, and so the z scores for height, weight, BMI, and BCM were compared against zero via t test. Height z scores were significantly above zero (t = 3.56, P < 0.01); however, neither weight nor BMI z scores were significantly different from zero. In contrast, BCM z scores were significantly below zero (t = −4.60, P < 0.01). Height z score was significantly higher in the boys (t = −4.03, P < 0.01); however, there were no significant differences among weight, BMI, or BCM z scores between the sexes.
As measures of nutritional status, BMI and BCM z scores were not related (r = 0.20, P = ns), and, in fact, BCM z scores were significantly lower than BMI z scores (t = 3.64, P < 0.01). According to Bland-Altman analysis (Fig. 1), BCM z scores were, on average, over half a z score below that of BMI z scores (bias −0.69 ± 1.65), and in 64.9% of measurements (n = 50), BCM z scores were lower than BMI z scores. There were also differences between BMI z scores and BCM z scores when they were categorised according to whether values were below or above the 5th percentile (z score −1.65) (Fig. 2). Nineteen measurements were <5th percentile for BM; however, only 5 measurements were lower for BMI. Only 3 measurements were categorised as <5th percentile according to both BMI and BCM z scores. According to χ2 analysis, these differences were not significant; however, they approached significance (P = 0.06), and it is possible that the lack of significance may simply have been a factor of the small numbers of measurements falling below this cutoff.
Markers of Inflammation
BMI z scores were significantly negatively correlated with CRP (r = −0.40, P < 0.01) and ESR (r = −0.29, P = 0.02), but not albumin; however, albumin was significantly negatively correlated with both CRP (r = −0.60, P < 0.01) and ESR (r = −0.71, P < 0.01) in this cohort. In contrast, BCM z scores were significantly correlated with CRP, ESR, and albumin. The relations with CRP (r = −0.52, P < 0.01) and ESR (r = −0.45, P < 0.01) were negative, whereas the relation with albumin (r = 0.36, P < 0.01) was positive. These relations were to be expected, as albumin would increase with improving nutritional status; however, nutritional status would decrease with increasing inflammation (CRP and ESR).
Finally, where BMI z scores showed no relation to disease activity (Pediatric Ulcerative Colitis Activity Index), BCM z scores were significantly negatively correlated (r = −0.32, P = 0.01), suggesting, as expected, a worsening of nutritional status with increased disease activity. Table 1 displays the number of measurements within each disease activity status and differences between and within these groups according to either BMI or BCM z scores. Of the 18 patients participating in repeated measurements, 13 patients had disease activity classifications that differed between measurement times. The number of measurements in each disease activity group, therefore, corresponds to 51 measurements in 17 patients, 11 measurements in 9 patients, and 15 measurements in 11 patients for remission, mild, and moderate/severe groups, respectively. BMI z scores were not different according to whether a patient was in remission, had mild, or had moderate/severe disease at the time of measurement. In contrast, BCM z scores were significantly lower in patients with moderate/severe disease compared with those in remission (F = 4.59; P = 0.01). BCM z scores in those with mild disease were not different from those patients in remission, or from those with moderate/severe disease. When comparing BMI z scores with BCM z scores within each disease activity category, there were trends to significantly lower z scores for BCM in the remission and mild groups, with significantly lower scores found in the moderate/severe group.
The present study compared BMI against BCM as a measure of nutritional status. Nutritional status measured by BMI was significantly better than that indicated by BCM and was not related to disease activity or serum albumin. Both BMI and BCM z scores were correlated with markers of inflammation (CRP and ESR), with stronger relations being found between these markers and BCM z scores. Furthermore, the variance in CRP and ESR accounted for by BCM (27% and 20%, respectively) was higher than that accounted for by BMI (16% and 8%, respectively).
The BMI was initially proposed by Quetelet as a way to determine whether body weight was appropriate for height. Over the years it has morphed into an index with cutoffs for determining overweight or obesity, or alternatively, underweight. Hence, this cutoff for underweight is a proxy for nutritional adequacy, with a BMI falling below this cutoff being indicative of compromised nutritional status. The main problem of using BMI in this manner, however, is that it does not take into account differences in body composition, and it is changes in body composition, in particular the fat-free mass component, that more appropriately represent nutritional status with respect to health outcomes (11). Furthermore, as Thibault et al (11) discuss, weight recovery after disease exacerbation is not necessarily associated with a gain in metabolically active fat-free mass, and there is usually a preferential increase in fat mass before the restoration of fat-free mass (19,20); however, this does depend on the duration and degree of malnutrition.
The chronic malnutrition associated with IBD could be the instigator for the changes in body composition seen in the cohort of patients with UC described herein. Height, weight, and BMI z scores were comparative with the average scores for a healthy population of that age and sex; however, BCM z scores were significantly reduced. This suggests a reduction of fat-free mass in the face of increased fat mass. Hence, using subjective global assessment and/or BMI may falsely categorise some patients with UC as adequately nourished, when in fact, they are not. This was shown in Figure 2, with the misclassification of 74% of measurements as not being malnourished according to BMI, when in fact they were according to BCM.
Relations among BMI, disease activity, and traditional markers of nutritional status, such as albumin, were not found in this group of patients. Where BCM was related to both disease activity and albumin, BMI was not. Hence, although increased disease activity and decreased albumin were associated with a worsening of nutritional status, as indicated by lower BCM z scores, this was not reflected by changes in BMI. With respect to albumin, however, albumin has been criticised for not being a valid marker of nutritional status, but rather a marker of illness (21) that is influenced by the inflammatory response irrespective of malnutrition (22).
Furthermore, within each disease activity group, BCM z scores were consistently lower than BMI z scores. This was only a trend towards significance for both the remission and mild disease groups; however, the differences were significantly reflected in patients within the moderate/severe disease group. Interestingly, the mean BMI z score in this group was on the 40th percentile, whereas the mean BCM z score was on the 5th percentile and, therefore, would be considered malnourished (13). So, if BMI alone was calculated, the nutritional status of this group would not be considered clinically compromised, when, in fact, their BCM was significantly reduced and they were malnourished.
Recent data by Long et al (23) suggest an increase in overweight and obesity in children with UC, which was significantly higher with a history of any corticosteroid use (35.0% vs 27.1%); however, their study classified patients only according to BMI, which further highlights the problem of using this method in this cohort. Corticosteroid use remains a common treatment modality for children with UC and is known to be associated with fluid retention (24). In this instance, BMI would always be elevated because of increased extracellular water weight and it is not possible to separate the effects of corticosteroid use on body compartment changes versus actual changes in fat and fat-free mass. From our data, it appears there are body compartment changes in children with UC, with loss of functional tissue and an increase in fat mass. Similar findings were reported in adults with UC by Valentini et al (6), where decreased BCM and increased fat mass were found in patients who were not considered malnourished according to the standard measures of malnutrition, including BMI.
One of the problems associated with these changes in body composition is the effect of increased body fatness on long-term health, as well as its direct influence on disease progression. There are data to suggest that circulating adipokines play a role in modulating intestinal inflammation, and therefore, increased adiposity may directly influence disease prognosis in patients with UC (25–27); however, research in this area is inconclusive and often conflicting, and more studies are needed to clarify the role of adipokines not only in UC but also in IBD in general.
In conclusion, overfatness in the face of malnutrition can exist, and the combination of increased fat mass with decreased functional mass in patients with UC will have deleterious effects on disease progression and overall health. It is acknowledged that access to technology that accurately measures nutritional status, such as TBK counting, is not widespread; however, it must be highlighted that although BMI is a simple measure, its calculation in patients with UC does not give adequate information regarding nutritional status and its use is flawed. Nutritional status may be compromised even if patients appear to “look” appropriately nourished, and caution is recommended when making treatment decisions based on BMI alone, especially because it was not related to disease activity in this cohort. More accurate, yet easily accessible “bedside” techniques need to be determined to improve patient care in patients with UC.
1. Capristo E, Mingrone G, Addolorato G, et al. Metabolic features of inflammatory bowel disease in a remission phase of the disease activity. J Intern Med 1998; 243:339–347.
2. Geerling BJ, Badart-Smook A, Stockbrügger RW, et al. Comprehensive nutritional status in recently diagnosed patients with inflammatory bowel disease compared with population controls. Eur J Clin Nutr 2000; 54:514–521.
3. Jahnsen J, Falch JA, Mowinckel P, et al. Body composition in patients with inflammatory bowel disease: a population based study. Am J Gastroenterol 2003; 98:1556–1562.
4. Paerregaard A, Uldall Urne F. Anthropometry at the time of diagnosis in Danish children with inflammatory bowel disease. Acta Paediatrica 2005; 94:1682–1683.
5. Rocha R, Santana GO, Almeida N, et al. Analysis of fat and muscle mass in patients with inflammatory bowel disease during remission and active phase. Br J Nutr 2009; 101:1–4.
6. Valentini L, Schaper L, Buning C, et al. Malnutrition and impaired muscle strength in patients with Crohn's disease and ulcerative colitis in remission. Nutrition 2008; 24:694–702.
7. Teixeira LG, Leonel AJ, Aguilar EC, et al. The combination of high-fat diet-induced obesity and chronic ulcerative colitis reciprocally exacerbates adipose tissue and colon inflammation. Lipids in Health and Disease 2011; 10:204.
8. Ellis KJ. Whole-body counting and neutron activation analysis. In: Heymsfield SB, Lohman TG, Wang Z, et al, eds. Human Body Composition. Champagne, IL: Human Kinetics; 2005:51–62.
9. Azcue M, Rashid M, Griffiths A, et al. Energy expenditure and body composition in children with Crohn's disease: effect of enteral nutrition and treatment with prednisolone. Gut 1997; 41:203–208.
10. Vaisman N, Rossi MF, Goldberg E, et al. Energy expenditure and body composition in patients with anorexia nervosa. J Pediatr 1988; 113:919–924.
11. Thibault R, Genton L, Pichard C. Body composition: why, when and for who? Clin Nutr 2012; 31:435–447.
12. Ogden CL, Kuczmarski RJ, Flegal KM, et al. Centers for Disease Control and Prevention 2000 growth charts for the United States: improvements to the 1977 National Center for Health Statistics version. Pediatrics 2002; 109:45–60.
14. Murphy AJ, Davies PS. Body cell mass index in children: interpretation of total body potassium results. Br J Nutr 2008; 100:666–668.
15. Wang Z, St-Onge MP, Lecumberri B, et al. Body cell mass: model development and validation at the cellular level of body composition. Am J Physiol Endocrinol Metab 2004; 286:123–128.
16. Wang Z, Heshka S, Wang J, et al. Body cell mass: validation of total body potassium prediction model in children and adolescents. IJBCR 2005; 3:153–158.
17. Turner D, Otley AR, Mack D, et al. Development, validation, and evaluation of a pediatric ulcerative colitis activity index: a prospective multicenter study. Gastroenterology 2007; 133:423–432.
18. Bland JM, Altman DG. Statistical methods for assessing agreement between two methods of clinical measurement. Lancet 1986; 8:307–310.
19. Pagliato E, Corradi E, Gentile MG, et al. Changes in body composition and resting energy expenditure in anorectic patients after a weight gain of fifteen percent. Ann N Y Acad Sci 2000; 904:617–620.
20. Polito A, Cuzzolaro M, Raguzzini A, et al. Body composition changes in anorexia nervosa. Eur J Clin Nutr 1998; 52:655–662.
21. Friedman AN, Fadem SZ. Reassessment of albumin as a nutritional marker in kidney disease. J Am Soc Nephrol 2010; 21:223–230.
22. Valentini L, Schulzke J-D. Mundane, yet challenging: the assessment of malnutrition in inflammatory bowel disease. Eur J Int Med 2011; 22:13–15.
23. Long MD, Crandall WV, Leibowitz IH, et al. Prevalence and epidemiology of overweight and obesity in children with inflammatory bowel disease. Inflamm Bowel Dis 2011; 17:2162–2168.
24. Cronin E. Prednisolone in the management of patients with Crohn's disease. Brit J Nurs 2010; 19:1333–1336.
25. Karmiris K, Koutroubakis IE, Xidakis C, et al. Circulating levels of leptin, adiponectin, resistin, and ghrelin in inflammatory bowel disease. Inflamm Bowel Dis 2006; 12:100–105.
26. Valentini L, Wirth EK, Schweizer U, et al. Circulating adipokines and the protective effects of hyperinsulinemia in inflammatory bowel disease. Nutrition 2009; 25:172–181.
27. Weigert J, Obermeier F, Neumeier M, et al. Circulating levels of chemerin and adiponectin are higher in ulcerative colitis and chemerin is elevated in Crohn's disease. Inflamm Bowel Dis 2010; 16:630–637.