Search in the Abstract Database

Abstracts Search 2014

* = Presenting author

P409. More targeted evaluation of nutrition in inflammatory bowel disease (IBD) reveals opportunities to optimise care: body mass index, body composition and iron deficiency

S.-Y. Ooi1, R. Bryant1, C. Schultz2, C. Goess1, R. Grafton1, J. Hughes1, J. McMahon1, D. Bartholomeusz1, J. Andrews1, 1Royal Adelaide Hospital, Gastroenterology and Hepatology, Adelaide, Australia, 2Royal Adelaide Hospital, Department of Nuclear Medicine, Adelaide, Australia


Addressing nutrition is important in the management of IBD [1]. Clinical evaluation of IBD patients routinely includes measurement of body mass index (BMI), iron studies and albumin. BMI has limitations as a normal range value may mask low lean body mass due to unrecognised excess fat mass. Body composition, a combination of fat-mass and fat-free mass (lean body mass), is poorly studied in IBD yet may be more accurate for assessing nutritional status. We evaluated BMI as a predictor of body composition and examined the prevalence of iron deficiency and hypoalbuminaemia in IBD.


Cross-sectional data were prospectively gathered on 18–50 year old IBD patients. Whole body composition analysis, height and weight were assessed at the time of dual energy X-ray absorptiometry (DXA) (GE - Lunar Prodigy) yielding body mass index (BMI) [weight (kg)/height (m2)], fat mass index (FMI) [fat mass (kg)/height (m2)] and lean mass index (LMI) [lean mass (kg)/height (m)2]. Baseline data included demographics, diagnosis, haemoglobin (Hb), iron studies and albumin. FMI and LMI Z-scores were calculated using standard deviation values from age-specific normative data [2]. Iron deficiency (ID) was defined as transferrin saturation ≤16%. Hypoalbuminaemia was defined as serum albumin <34 g/L. Relationships amongst variables were examined with correlation coefficients.


137 patients, 95 (69%) with Crohn's disease (CD), 76 (55%) male, had a mean age of 32.2 (median 31, range 20–49). 57% (78/137) had active disease at enrolment. Mean BMI was 26.6 (median 25.1, range 16.7–43.9), males 26.9 and females 26.3. BMI correlated well with FMI (r2=0.89, 95% CI 0.85, 0.92) (Figure 1) while it was a poor predictor of LMI (r2=0.45, 95% CI 0.30, 0.57), regardless of gender or IBD type. ID was highly prevalent at 27.7% (38/137). Hypoalbuminaemia was uncommon with a prevalence of 9.5% (13/137). Of the patients with ID 76% (29/38) had a normal range Hb and thus was only detected by specific testing, while 76% (29/38) had active disease on enrolment.

Figure 1. BMI vs FMI z-score: Pearson correlation = 0.89 (95% CI 0.85, 0.92).


Despite 57%% having clinically active disease, BMI was normal in this IBD cohort. BMI is a good predictor of FMI however correlated poorly with LMI. Clinical assessment of nutritional status using BMI alone would miss deficits in LMI and under recognise ID. ID is more common than hypoalbuminaemia in IBD patients and occurs despite normal Hb levels. Thus it is recommended that routine iron studies be included in the nutritional assessment of IBD patients.

1. Jahnsen J, Falch JA, Mowinckel P, Aadland E., (2003), Body composition in patients with inflammatory bowel disease: a population-based study., American Journal of Gastroenterology, 1556–62.

2. Kelly TL, Wilson KE, Heymsfield SB, (2009), Dual Energy X-Ray Absorptiometry Body Composition Reference Values from NHANES. Plo5 ONE,