A Review of the Evidence for the Use of Ventilation as a Surrogate Measure of Energy Expenditure

Precise measurement of sedentary behavior and physical activity is necessary to characterize the dose‐response relationship between these variables and health outcomes. The most frequently used methods employ portable devices to measure mechanical or physiological parameters (eg, pedometers, heart rate monitors, accelerometers). There is considerable variability in the accuracy of total energy expenditure (TEE) estimates from these devices. This review examines the potential of measurement of ventilation () to provide an estimate of free‐living TEE. The existence of a linear relationship between and energy expenditure (EE) was demonstrated in the mid‐20th century. However, few studies have investigated this parameter as an estimate of EE due to the cumbersome equipment required to measure . Portable systems that measure without the use of a mouthpiece have existed for about 20 years (respiratory inductive plethysmography). However, these devices are adapted for clinical monitoring and are too cumbersome to be used in conditions of daily life. Technological innovations of recent years (small electromagnetic coils glued on the chest/back) suggest that could be estimated from variations in rib cage and abdominal distances. This method of TEE estimation is based on the development of individual/group calibration curves to predict the relationship between ventilation and oxygen consumption. The new method provides a reasonably accurate estimate of TEE in different free‐living conditions such as sitting, standing, and walking. Further work is required to integrate these electromagnetic coils into a jacket or T‐shirt to create a wearable device suitable for long‐term use in free‐living conditions.     

Approximation of Resting Energy Expenditure in Intensive Care Unit Patients Using the SenseWear Bracelet: A Comparison With Indirect Calorimetry

Background and Aims: Indirect calorimetry (IC) is the gold standard for determining energy expenditure in patients requiring mechanical ventilation. Metabolic armbands using data derived from dermal measurements have been proposed as an alternative to IC in healthy subjects, but their utility during critical illness is unclear. The aim of this study was to determine the level of agreement between the SenseWear armband and the Deltatrac Metabolic Monitor in mechanically ventilated intensive care unit (ICU) patients. Methods: Adult ICU patients requiring invasive ventilator therapy were eligible for inclusion. Simultaneous measurements were performed with the SenseWear Armband and Deltatrac under stable conditions. Resting energy expenditure (REE) values were registered for both instruments and compared with Bland‐Altman plots. Results: Forty‐two measurements were performed in 30 patients. The SenseWear Armband measured significantly higher REE values as compared with IC (mean bias, 85 kcal/24 h; P = .027). Less variability was noted between individual SenseWear measurements and REE as predicted by the Harris‐Benedict equation (2 SD, ±327 kcal/24 h) than when IC was compared with SenseWear and Harris‐Benedict (2 SD, ±473 and ±543 kcal/24 h, respectively). Conclusions: The systematic bias and large variability of the SenseWear armband when compared with gas exchange measurements confer limited benefits over the Harris Benedict equation in determining caloric requirements of ICU patients.     

Energy Expenditure in Critically Ill Elderly Patients: Indirect Calorimetry vs Predictive Equations

Background: Predictive equations (PEs) are used for estimating resting energy expenditure (REE) when the measurements obtained from indirect calorimetry (IC) are not available. This study evaluated the degree of agreement and the accuracy between the REE measured by IC (REE‐IC) and REE estimated by PE (REE‐PE) in mechanically ventilated elderly patients admitted to the intensive care unit (ICU). Methods: REE‐IC of 97 critically ill elderly patients was compared with REE‐PE by 6 PEs: Harris and Benedict (HB) multiplied by the correction factor of 1.2; European Society for Clinical Nutrition and Metabolism (ESPEN) using the minimum (ESPENmi), average (ESPENme), and maximum (ESPENma) values; Mifflin–St Jeor; Ireton‐Jones (IJ); Fredrix; and Lührmann. Degree of agreement between REE‐PE and REE‐IC was analyzed by the interclass correlation coefficient and the Bland‐Altman test. The accuracy was calculated by the percentage of male and/or female patients whose REE‐PE values differ by up to ±10% in relation to REE‐IC. Results: For both sexes, there was no difference for average REE‐IC in kcal/kg when the values obtained with REE‐PE by corrected HB and ESPENme were compared. A high level of agreement was demonstrated by corrected HB for both sexes, with greater accuracy for women. The best accuracy in the male group was obtained with the IJ equation but with a low level of agreement. Conclusions: The effectiveness of PEs is limited for estimating REE of critically ill elderly patients. Nonetheless, HB multiplied by a correction factor of 1.2 can be used until a specific PE for this group of patients is developed.     

A feasibility study comparing two metabolic monitors in hospitalised patients following an ischaemic stroke

Background: Data on the energy requirements of patients following acute ischaemic stroke are scarce. A recent draft report highlighted the lack of data on physical activity levels during and following acute illness (SACN, 2009). The aims of this study were to establish if two metabolic monitors (CCM Express? and the Sensewear? armband) were feasible for use in hospitalised stroke patients and to determine the relative contributions of resting energy expenditure (REE) and physical activity to total energy expenditure (TEE). Methods: Eleven medically stable patients (seven male; four female) were recruited within 7 days of ischaemic stroke. Exclusion criteria: unable to give informed consent, receiving renal replacement therapy, body mass index (BMI) ≥ 50 kg m^?2, known nickel allergy or receiving end‐of‐life care. All subjects were fasted from midnight and REE was measured early in the morning using the CCM Express? for a period of up to 1 h (including establishment of steady‐state (i.e. <10% difference in minute to minute VO₂ and VCO₂ measurements over 5 min). Concurrently, TEE was measured using the Sensewear? armband for a period of 24 h. Assessments of patient acceptability and tolerance of both metabolic monitors were conducted by direct observation, completion of a checklist and, where clinically appropriate, a brief patient questionnaire. REE was compared with predicted basal metabolic rate (BMR) (Henry, 2005) and physical activity was estimated using the Metabolic Equivalent Task (MET) method, where 1.0 MET is equivalent to the energy expended at rest. Results: Mean age was 69.8 years (range 42–84 years) and mean (SD) BMI was 25.4 (5.2) kg m^?2. All subjects were able to tolerate measurement of REE using the CCM Express?, although the facemask caused some discomfort to one subject with facial abrasions. Mean (SD) REE was 1257 (357) kcal day^?1 and, perhaps unexpectedly, was lower than predicted BMR [1503 (226) kcal day^?1; t‐test, P = 0.07]. It was, however, difficult to achieve steady‐state in four patients; thus, these REE measurements were unreliable. All subjects were able to tolerate measurement of TEE using the Sensewear? armband. Mean (SD) TEE was 1663 (303) kcal day^?1. Physical activity on the ward was very low, with subjects expending very little more energy than would be expected at rest [METS = 1.01 (SD 0.15)]. Discussion: Both metabolic monitors were well tolerated by the subjects; however, the unreliable REE measurements in some patients made it impossible to determine the relative contribution of REE to TEE. The results obtained regarding TEE and the low activity level in this study were comparable to results reported in other metabolic studies of patients who have had a stroke (Weekes & Elia, 1992; Finestone et al., 2003; Leone & Pencharz, 2010). Conclusions: Both metabolic monitors were feasible for use in patients following ischaemic stroke; however, some measurements of REE using the CCM Express? were unreliable because of difficulties in establishing steady‐state and the reasons for this merit further investigation. In this group of patients, physical activity on the ward was very low following a stroke. References: Finestone, H.M., Greene‐Finestone, L.S., Foley, N.C. & Woodbury, M.G. (2003) Measuring longitudinally the metabolic demands of stroke patients. Stroke 34 , 502–507. Leone, A. & Pencharz, P.B. (2010) Resting energy expenditure in stroke patients who are dependent on tube feeding: a pilot study. Clin. Nutr. 29 , 370–372. Henry, C.J.K. (2005) Basal metabolic rate studies in humans: measurement and development of new equations. Public Health Nutr. 8 , 1133–1152. Scientific Advisory Committee on Nutrition (SACN) (2009) Energy Requirements Working Group Draft Report. London: SACN. Weekes, C.E. & Elia, M. (1992) Resting energy expenditure and body composition following cerebro‐vascular accident. Clin. Nutr. 11 , 18–22.     

Energy Expenditure in Patients With Esophageal,Gastric, and Colorectal Cancer

Background: Changes in resting energy expenditure (REE) appear to be one of the causes of nutritional depletion in cancer. Assessing REE may be an important tool for providing adequate nutritional therapy to these patients. The aims of this study were to evaluate REE of patients with gastrointestinal tract cancer and to compare it to that of healthy controls. Methods: A total of 20 patients, with esophageal (n = 3), gastric (n = 9), and colorectal (n = 8) cancers, and 20 healthy subjects were included. Indirect calorimetry (IC) was used to measure REE in both groups. The “pocket” equation (30 kcal/kg) and the Harris‐Benedict equation, with correction factors of 1.3 (activity) and 1.1 (injury), were employed for assessment of the estimated total energy expenditure (TEE). Statistics included Mann‐Whitney and paired t tests, Bland Altman analysis, and multivariate regression. Results: The REE of the patients (1,274.5 kcal [1,002.9–2,174.9]) was similar to that of the controls (1,445.5 kcal [1,114.5–1,762.6], not significant), even when corrected for the amount of metabolically active tissue. The pocket equation was effective in predicting the patients' TEE, with a 1.7% (32 kcal) difference being observed in comparison with the IC results corrected with the activity factor (not significant). Conclusions: The patients with digestive tract cancers showed a similar REE to that of the controls. The current formula of 30 kcal/kg is suitable for estimating the TEE of these patients.     

Validation of an equation for resting metabolic rate in older obese, critically ill patients

Background: In previous studies, the Penn State (PSU) equation was found to be a valid way to predict resting metabolic rate (RMR) in critically ill patients, with the exception of those who were aged 60 years or older and had a body mass index (BMI) ≥30 kg/m². A modification of the equation was proposed in this specific patient population. The current study was designed to test the validity of that equation and to retest the validity of the original equation. Methods: RMR was measured using a standardized evidence‐based protocol. Metabolic rates predicted with the PSU equation and a modification of it (PSU[m]) were compared with the measured values. Fifty patients were studied prospectively. Data were used from an additional 75 subjects from previous studies that had not been used to develop either of the equations being tested in the current study. This brought the final number of subjects for each equation to 74 for PSU(m) and 106 for PSU. Results: The PSU(m) equation was found to be biased by a narrow margin (95% confidence interval, ?87 to ?4 kcal/d), but both versions of the equation were precise. Accuracy rate for the PSU(m) equation was 74%, compared with 58% for the PSU equation (P < .04). Conclusions: The PSU(m) equation for predicting RMR in critically ill, mechanically ventilated patients is valid in patients aged cases where age is ≥60 years or older whose BMI is ≥30 kg/m².     

Day-to-day variance in measurement of resting metabolic rate using ventilated-hood and mouthpiece & nose-clip indirect calorimetry systems

BACKGROUND: To know if the magnitude of change in resting metabolic rate (RMR) observed during an intervention is meaningful, it is imperative to first identify the variability that occurs within individuals from day to day under normal conditions. The 2 most common systems used to measure RMR involve a ventilated hood or a mouthpiece & nose clip to collect expired gases. The variation in measurement using these 2 approaches has not been systematically compared. METHODS: RMR was measured in 10 healthy adults during 5 separate testing sessions within a 2-week period where usual diet and physical activity were maintained. Each testing session consisted of one measurement of RMR using a ventilated hood system, followed by another using a mouthpiece & nose-clip system. RESULTS: No significant difference in RMR was evident between measurement sessions using either indirect calorimeter. Oxygen consumption and RMR were significantly higher using the mouthpiece & nose-clip system. Average within-individual coefficient of variation for RMR was significantly lower for the ventilated-hood system. RMR measures were consistently lower using the ventilated-hood system by an average of 94.5 +/- 63.3 kcal. Day-to-day variance was between 2% and 4% for both systems. CONCLUSIONS: The use of either system is appropriate for assessing RMR in clinical and research settings, but alternating between systems should be undertaken with caution. A change in RMR must be greater than approximately 6% (96 kcal/d; 1.2 kcal/kg/d) or approximately 8% (135 kcal/d; 1.7 kcal/kg/d) when using a ventilated-hood system or a mouthpiece & nose-clip system, respectively, to observe any meaningful intervention-related differences within individuals.     

Assessment of resting energy expenditure in mechanically ventilated patients

BACKGROUND: Usual equations for predicting resting energy expenditure (REE) are not appropriate for critically ill patients, and indirect calorimetry criteria render its routine use difficult. OBJECTIVE: Variables that might influence the REE of mechanically ventilated patients were evaluated to establish a predictive relation between these variables and REE. DESIGN: The REE of 70 metabolically stable, mechanically ventilated patients was prospectively measured by indirect calorimetry and calculated with the use of standard predictive models (Harris and Benedict's equations corrected for hypermetabolism factors). Patient data that might influence REE were assessed, and multivariate analysis was conducted to determine the relations between measured REE and these data. Measured and calculated REE were compared by using the Bland-Altman method. RESULTS: Multivariate analysis retained 4 independent variables defining REE: body weight (r(2) = 0.14, P < 0.0001), height (r(2) = 0.11, P = 0.0002), minute ventilation (r(2) = 0.04, P = 0.01), and body temperature (r(2) = 0.07, P = 0.002): REE (kcal/d) = 8 x body weight + 14 x height + 32 x minute ventilation + 94 x body temperature - 4834. REE calculated with this equation was well correlated with measured REE (r(2) = 0.61, P < 0.0001). Bland-Altman plots showed a mean bias approaching zero, and the limits of agreement between measured and predicted REE were clinically acceptable. CONCLUSION: Our results suggest that REE estimated on the basis of body weight, height, minute ventilation, and body temperature is clinically more relevant than are the usual predictive equations for metabolically stable, mechanically ventilated patients.     

Energy Expenditure After Liver Resection: Validation of a Mobile Device for Estimating Resting Energy Expenditure and an Investigation of Energy Expenditure Change After Liver Resection

Background: Resting energy expenditure (REE) is the major component of total energy expenditure. REE is traditionally performed by indirect calorimetry (IC) and is not well investigated after liver surgery. A mobile device (SenseWear Armband [SWA]) has been validated when estimating REE in other clinical settings but not liver resection. The aims of this study are to validate SWA vs IC, quantify REE change following liver resection, and determine factors associated with REE change. Materials and Methods: Patients listed for open liver resection prospectively underwent IC and SWA REE recordings pre‐ and postoperatively. In addition, the SWA was worn continuously postoperatively to estimate daily REE for the first 5 postoperative days. To determine acceptability of the SWA, validation analysis was performed. To assess REE change, peak postoperative REE was compared with preoperative levels. Factors associated with REE change were also analyzed. Results: SWA showed satisfactory validity compared with IC when estimating REE, although postoperatively, the 95% levels of agreement (–5.56 to 3.18 kcal/kg/d) may introduce error. Postoperative REE (median, 23.5 kcal/kg/d; interquartile range [IQR], 22.6–25.7 kcal/kg/d) was significantly higher than predicted REE (median, 19.7 kcal/kg/d; IQR, 19.1–21.0 kcal/kg/d; P < .0001). Median REE rise was 11% (IQR, –1% to 25%). Factors associated with REE rise of >11% were age (P = .017) and length of operation (P = .03). Conclusions: SWA offers a suitable alternative to IC when estimating postoperative REE, but the magnitude of the error (8.74 kcal/kg/d) could hinder its accuracy. REE quantification after liver resection is important to identify patients who could be prone to energy imbalance and therefore malnutrition.     

Another step in the search for better oral rehydration solutions.

Objective: To compare resting metabolic rate (RMR) measured by indirect calorimetry versus RMR predicted by several published formulas in a sample of healthy young women.

Methods: RMR was measured using indirect calorimetry and predicted using 6 commonly used equations (Nelson, 1992; Mifflin, 1990; Owen, 1986; Schofield_Weight, 1985; Schofield_{Weight and Height}, 1985; Harris-Benedict, 1919) in 47 reportedly healthy young females (age = 22.8 ± 2.9 years; body mass index = 21.8 ± 2.1 kg/m²). Comparisons between measured versus predicted RMR were conducted using paired t tests, and agreement using Pearson's correlation coefficient, analysis of variance, and the method of Bland-Altman.

Results: All 6 equations overestimated measured RMR by 140–738 kcal/d (all p < 0.001). The proportion of subjects for whom measured versus predicted RMR differed by ±10% ranged from 74% (Nelson) to 100% (Harris-Benedict). The adjusted coefficients of determination (R²) between measured and predicted RMR ranged from 0.13 to 0.19 (all p < 0.05). Bland-Altman analysis R² values ranged from 0.03 (p = 0.233; Harris-Benedict) to 0.72 (p = 0.000; Owen). Given its continued popularity, we modified the Harris-Benedict equation (RMR_{modified Harris-Benedict} (kcal/d) = 738 / (RMR_{Harris-Benedict} ? 738)). Doing so reduced the mean difference between measured and predicted RMR from +738 kcal/d to ?0.53 kcal/d (p = 0.984).

Conclusion: No equation performed well, and none should be used interchangeably with measured RMR. We recommend that a new equation be validated for, and prospectively tested in, young women. In the interim, RMR should be measured in this population or predicted using the modified Harris-Benedict equation that we developed.     

Energy balance in predialysis patients on a low-protein diet.

OBJECTIVE: To determine if energy intake on a low-protein diet (0.6 g protein/kg ideal body weight (ibw)/d) with 70% animal protein (Diet A) or 30% animal protein (Diet B) meets energy expenditure derived from measured resting energy expenditure and activity levels. DESIGN: Patients already on a conventional low-protein diet with 70% animal protein kept a 5-day weighed dietary intake, with a 3-day activity diary, and had their resting metabolic rate (RMR) measured. Patients then switched to a diet with 30% animal protein for a minimum of 2 weeks (range, 2 to 16 weeks) and repeated the weighed intake and RMR measurement. SETTING: Predialysis hospital outpatients. PATIENTS: Seven patients were recruited, 5 male. Mean age, 56 years (range, 43 to 78 years); mean serum creatinine 300 micromol/L (range, 180 to 560 micromol/L). INTERVENTION: Indirect calorimetry used to measure RMR. MAIN OUTCOME MEASURE: RMR compared with standard formulae and total energy expenditure compared with dietary intake. RESULTS: Mean RMR was 5.76 MJ/d (1,385 kcal/d) or 84.9 kJ/kg ibw/d (20.3 kcal/kg ibw/d); which was 108% to 113% of that predicted by standard formulae. Total energy expenditure (RMR plus activity) was 8.35 MJ/d (1,996 kcal/d) or 123.3 kJ/kg/d (29.5 kcal/kg ibw/d). Mean energy intake was 116.3 (27.8 kcal/kg ibw/d) on Diet A and 131.2 (31.4 kcal/kg ibw/d) on Diet B (P = .096) with 3 of the 7 patients meeting their energy expenditure on Diet A and 4 on Diet B. CONCLUSION: RMR of patients with chronic renal failure is within expected range for healthy individuals, and the activity of these relatively fit patients similar to healthy individuals with light to moderate activity. Energy intake on the low-protein diets failed to meet energy expenditure in 4 patients on Diet A and 3 patients on Diet B. Low energy intake may contribute to the development of malnutrition in some patients.     

Resting Energy Expenditure in Critically Ill Patients With Spontaneous Intracranial Hemorrhage

Background: Data on energy requirements of patients with spontaneous intracranial hemorrhage (SICH) are scarce. The objective of this study was to determine the resting energy expenditure (REE) in critically ill patients with SICH and to compare it with the predicted basal metabolic rate (BMR). Methods: In 30 nonseptic patients with SICH, the REE was measured during the 10 first posthemorrhage days with the use of indirect calorimetry (IC). Predicted BMR was also evaluated by the Harris‐Benedict (HB) equation. Bland‐Altman analysis was used to evaluate the agreement between measured and predicted values. The possible effect of confounding factors (demographics, disease, and severity of illness score) on the evolution of continuous variables was also tested. Results: mean predicted BMR, calculated by the HB equation, was 1580.3 ± 262 kcal/d, while measured REE was 1878.9 ± 478 kcal/d (117.5% BMR). Compared with BMR, measured REE values showed a statistically significant increase at all studied points (P < .005). Measured and predicted values showed a good correlation (r = 0.73, P < .001), but the test of agreement between the 2 methods with the Bland‐Altman analysis showed a mean bias (294.6 ± 265.6 kcal/d) and limits of agreement (–226 to 815.29 kcal/d) that were beyond the clinically acceptable range. REE values presented a trend toward increase over time (P = .077), reaching significance (P < .005) after the seventh day. Significant correlation was found between REE and temperature (P = .002, r = 0.63), as well as between REE and cortisol level (P = .017, r = 0.62) on the 10th day. No correlation was identified between REE and depth of sedation, as well as Acute Physiology and Chronic Health Evaluation II, Glasgow Coma Scale, and Hunt and Hess scores. Conclusions: During the early posthemorrhagic stage, energy requirements of critically ill patients with SICH are increased, presenting a trend toward increase over time. Compared with IC, the HB equation underestimates energy requirements and is inefficient in detecting individual variability of REE in this group of patients.     

Can We Rely on Predicted Basal Metabolic Rate in Patients With Intestinal Failure on Home Parenteral Nutrition?

Background and Aims: Intestinal failure (IF) is a serious and common complication of short bowel syndrome with patients depending on parenteral nutrition (PN) support. Effective nutrition management requires an accurate estimation of the patient's basal metabolic rate (BMR) to avoid underfeeding or overfeeding. However, indirect calorimetry, considered the gold standard for BMR assessment, is a time‐ and resource‐consuming procedure. Consequently, several equations for prediction of BMR have been developed in different settings, but their accuracy in patients with IF are yet to be investigated. We evaluated the accuracy of predicted BMR in clinically stable patients with IF dependent on home parenteral nutrition (HPN). Methods: In total, 103 patients with IF were included. We used indirect calorimetry for assessment of BMR and calculated predicted BMR using different equations based on anthropometric and/or bioelectrical impedance parameters. The accuracy of predicted BMR was evaluated using Bland‐Altman analysis with measured BMR as the gold standard. Results: The average measured BMR was 1272 ± 245 kcal/d. The most accurate estimations of BMR were obtained using the Harris‐Benedict equation (mean bias, 14 kcal/d [P = .28]; limits of agreement [LoA], ?238 to 266 kcal/d) and the Johnstone equation (mean bias, ?16 kcal/d [P = .24]; LoA, ?285 to 253 kcal/d). For both equations, 67% of patients had a predicted BMR from 90%–110% All other equations demonstrated a statistically and clinically significant difference between measured and predicted BMR. Conclusions: The Harris‐Benedict and Johnstone equations reliably predict BMR in two‐thirds of clinically stable patients with IF on HPN.     

24-hour indirect calorimetry in mechanically ventilated critically ill patients.

BACKGROUND: Energy imbalance in critically ill, mechanically ventilated patients may lead to medical complications. The nutrition care team needs accurate, noninvasive, rapid methods to estimate energy requirements. We investigated whether brief measurements of indirect calorimetry at any time of the day would give valid estimates of 24-hour energy expenditure (EE). METHODS: EE of 12 mechanically ventilated critically ill patients (6 men, 6 women, mean +/- SD age 67 +/- 18 years, weight 70.2 +/- 8.8 kg) was recorded every minute during 24 hours by indirect calorimetry. All patients were continuously fed enteral nutrition. RESULTS: Mean +/- SD EE was 1658 +/- 279 kcal/d (6941 +/- 1167 kJ/d). Within patients, EE during the day fluctuated by 234 kcal in the most constant patient to 1190 kcal in the least constant patient, with a mean fluctuation of 521 kcal (12 patients). No statistically significant difference (p = .53) in mean EE between morning (6-12 hours, 1676 kcal), afternoon (12-18 hours, 1642 kcal), evening (18-24 hours, 1658 kcal), and night (0-6 hours, 1655 kcal) was found. A 2-hour instead of a 24-hour measurement resulted in a maximal error of 128 kcal (536 kJ), which was <10% of the average EE. The maximal error decreased with longer time intervals. CONCLUSIONS: In mechanically ventilated critically ill patients, 24-hour indirect calorimetry measurements can be replaced by shorter (>/=2 hours) measurements. Time of day did not affect EE.     

A new measure of between‐studies heterogeneity in meta‐analysis

Assessing the magnitude of heterogeneity in a meta‐analysis is important for determining the appropriateness of combining results. The most popular measure of heterogeneity, I², was derived under an assumption of homogeneity of the within‐study variances, which is almost never true, and the alternative estimator, , uses the harmonic mean to estimate the average of the within‐study variances, which may also lead to bias. This paper thus presents a new measure for quantifying the extent to which the variance of the pooled random‐effects estimator is due to between‐studies variation, , that overcomes the limitations of the previous approach. We show that this measure estimates the expected value of the proportion of total variance due to between‐studies variation and we present its point and interval estimators. The performance of all three heterogeneity measures is evaluated in an extensive simulation study. A negative bias for was observed when the number of studies was very small and became negligible as the number of studies increased, while and I² showed a tendency to overestimate the impact of heterogeneity. The coverage of confidence intervals based upon was good across different simulation scenarios but was substantially lower for and I², especially for high values of heterogeneity and when a large number of studies were included in the meta‐analysis. The proposed measure is implemented in a user‐friendly function available for routine use in r and sas . will be useful in quantifying the magnitude of heterogeneity in meta‐analysis and should supplement the p‐value for the test of heterogeneity obtained from the Q test. Copyright © 2016 John Wiley & Sons, Ltd.     

A Method for Gene‐Based Pathway Analysis Using Genomewide Association Study Summary Statistics Reveals Nine New Type 1 Diabetes Associations

Pathway analysis can complement point‐wise single nucleotide polymorphism (SNP) analysis in exploring genomewide association study (GWAS) data to identify specific disease‐associated genes that can be candidate causal genes. We propose a straightforward methodology that can be used for conducting a gene‐based pathway analysis using summary GWAS statistics in combination with widely available reference genotype data. We used this method to perform a gene‐based pathway analysis of a type 1 diabetes (T1D) meta‐analysis GWAS (of 7,514 cases and 9,045 controls). An important feature of the conducted analysis is the removal of the major histocompatibility complex gene region, the major genetic risk factor for T1D. Thirty‐one of the 1,583 (2%) tested pathways were identified to be enriched for association with T1D at a 5% false discovery rate. We analyzed these 31 pathways and their genes to identify SNPs in or near these pathway genes that showed potentially novel association with T1D and attempted to replicate the association of 22 SNPs in additional samples. Replication P‐values were skewed () with 12 of the 22 SNPs showing . Support, including replication evidence, was obtained for nine T1D associated variants in genes ITGB7 (rs11170466, ), NRP1 (rs722988, ), BAD (rs694739, ), CTSB (rs1296023, ), FYN (rs11964650, ), UBE2G1 (rs9906760, ), MAP3K14 (rs17759555, ), ITGB1 (rs1557150, ), and IL7R (rs1445898, ). The proposed methodology can be applied to other GWAS datasets for which only summary level data are available.     

Indirect Calorimetry Reveals That Better Monitoring of Nutrition Therapy in Pediatric Intensive Care Is Needed

Background: Monitoring nutrition therapy is essential in the care of critically ill children, but the risk of nutrition failure seems to remain. The aims of the present study were to examine the prevalence of underfeeding, adequate feeding, and overfeeding in mechanically ventilated children and to identify barriers to the delivery of nutrition support. Materials and Methods: Children aged 0–14 years who fulfilled the criteria for indirect calorimetry were enrolled in this prospective, observational study and were studied for up to 5 consecutive days. Actual energy intake was recorded and compared with the required energy intake (measured energy expenditure plus 10%); energy intake was classified as underfeeding (<90% of required energy intake), adequate feeding (90%?110%), or overfeeding (>110%). The reasons for interruptions to enteral and parenteral nutrition were recorded. Results: In total, 104 calorimetric measurements for 140 total days were recorded for 30 mechanically ventilated children. Underfeeding, adequate feeding, and overfeeding occurred on 21.2%, 18.3%, and 60.5% of the 104 measurement days, respectively. There was considerable variability in the measured energy expenditure between children (median, 37.2 kcal/kg/d; range, 16.81?66.38 kcal/kg/d), but the variation within each child was small. Respiratory quotient had low sensitivity of 21% and 27% for detecting underfeeding and overfeeding, respectively. Fasting for procedures was the most frequent barrier that led to interrupted nutrition support. Conclusion: The high percentage of children (~61%) who were overfed emphasizes the need to measure energy needs by using indirect calorimetry.     

A practical approach to estimate resting energy expenditure in frail elderly people

Objectives: Some prediction equations of resting energy expenditure (REE) are available and can be used in clinical wards to determine energy requirements of patients. The aim of the present study was to assess the accuracy of those equations in sick elderly patients, using the Bland & Altman methods with our database of 187 REE measurements.Design: The 3 equations tested were Harris & Benedict equation of 1919, WHO/FAO/UNU equation of 1985 and Fredrix et al. equation of 1990. In addition, three models developed from the present data were tested.Results: The present study shows that the Fredrix et al equation gave an accurate prediction of REE without significant bias along the whole range of REE. It also shows that under-weight sick elderly patients (BMI ≤ 21 kg/m²) had a greater weight-adjusted REE than their normal weight counterparts.Conclusion: A simple formula using a factor multiplying body weight, i.e. 22 kcal/kg/d in under-weight and 19 kcal/kg/d in normal weight sick elderly was accurate to predicting REE and bias was not influenced by the level of REE. This model included half of the group in the range of ±10% of the difference between predicted REE and measured REE, but the confidence interval of the bias was ±400 kcal/d. Conversely, the Harris & Benedict and WHO formulae did accurately predict REE.     

Meta‐Analysis of Genome‐Wide Association Studies with Correlated Individuals: Application to the Hispanic Community Health Study/Study of Latinos (HCHS/SOL)

Investigators often meta‐analyze multiple genome‐wide association studies (GWASs) to increase the power to detect associations of single nucleotide polymorphisms (SNPs) with a trait. Meta‐analysis is also performed within a single cohort that is stratified by, e.g., sex or ancestry group. Having correlated individuals among the strata may complicate meta‐analyses, limit power, and inflate Type 1 error. For example, in the Hispanic Community Health Study/Study of Latinos (HCHS/SOL), sources of correlation include genetic relatedness, shared household, and shared community. We propose a novel mixed‐effect model for meta‐analysis, “MetaCor,” which accounts for correlation between stratum‐specific effect estimates. Simulations show that MetaCor controls inflation better than alternatives such as ignoring the correlation between the strata or analyzing all strata together in a “pooled” GWAS, especially with different minor allele frequencies (MAFs) between strata. We illustrate the benefits of MetaCor on two GWASs in the HCHS/SOL. Analysis of dental caries (tooth decay) stratified by ancestry group detected a genome‐wide significant SNP (rs7791001, P‐value = , compared to in pooled), with different MAFs between strata. Stratified analysis of body mass index (BMI) by ancestry group and sex reduced overall inflation from (pooled) to (MetaCor). Furthermore, even after removing close relatives to obtain nearly uncorrelated strata, a naïve stratified analysis resulted in compared to for MetaCor.