Correction: A novel scaling methodology to reduce the biases associated with missing data from commercial activity monitors.

[This corrects the article DOI: 10.1371/journal.pone.0235144.].

A novel scaling methodology to reduce the biases associated with missing data from commercial activity monitors.

BACKGROUND: Commercial physical activity monitors have wide utility in the assessment of physical activity in research and clinical settings, however, the removal of devices results in missing data and has the potential to bias study conclusions. This study aimed to evaluate methods to address missingness in data collected from commercial activity monitors. METHODS: This study utilised 1526 days of near complete data from 109 adults participating in a European weight loss maintenance study (NoHoW). We conducted simulation experiments to test a novel scaling methodology (NoHoW method) and alternative imputation strategies (overall/individual mean imputation, overall/individual multiple imputation, Kalman imputation and random forest imputation). Methods were compared for hourly, daily and 14-day physical activity estimates for steps, total daily energy expenditure (TDEE) and time in physical activity categories. In a second simulation study, individual multiple imputation, Kalman imputation and the NoHoW method were tested at different positions and quantities of missingness. Equivalence testing and root mean squared error (RMSE) were used to evaluate the ability of each of the strategies relative to the true data. RESULTS: The NoHoW method, Kalman imputation and multiple imputation methods remained statistically equivalent (p<0.05) for all physical activity metrics at the 14-day level. In the second simulation study, RMSE tended to increase with increased missingness. Multiple imputation showed the smallest RMSE for Steps and TDEE at lower levels of missingness (<19%) and the Kalman and NoHoW methods were generally superior for imputing time in physical activity categories. CONCLUSION: Individual centred imputation approaches (NoHoW method, Kalman imputation and individual Multiple imputation) offer an effective means to reduce the biases associated with missing data from activity monitors and maximise data retention.

Social, temporal and situational influences on meat consumption in the UK population.

The amount of meat consumed is having a negative impact on both health and the environment. This study investigated the probability of eating meat and the amount eaten at a meal within different social, temporal and situational contexts. Dietary intake data from 4-day diet diaries of adults (19 years and above) taken from the UK National Diet and Nutrition Survey (2008/9-2013/14) were used for the analysis. Individual eating occasions were identified and the effects of where the food was eaten, with whom, day of the week, age and gender on the probability of eating meat and amount of meat eaten were modelled using general linear mixed models. Each factor showed distinctive effects on the probability of eating meat and the amount consumed. The amount of meat eaten was greater when eating with family members compared to when alone or with other companions. Both the probability and amount of meat eaten in a single eating occasion were higher on Sundays compared to the rest of the week. Eating out (e.g. restaurants/cafes) increased the probability of consuming meat and the amount compared to other situations (e.g. home, work). When considering the factors influencing meat consumption, attention must be paid to the effects of social, temporal, and situational factors as they all work to shape consumption behaviour. This information should be used in the design of interventions and development of policies for the most effective way to reduce meat consumption.

Modelling the associations between fat-free mass, resting metabolic rate and energy intake in the context of total energy balance.

BACKGROUND: The relationship between body composition, energy expenditure and ad libitum energy intake (EI) has rarely been examined under conditions that allow any interplay between these variables to be disclosed. OBJECTIVE: The present study examined the relationships between body composition, energy expenditure and EI under controlled laboratory conditions in which the energy density and macronutrient content of the diet varied freely as a function of food choice. METHODS: Fifty-nine subjects (30 men: mean body mass index=26.7±4.0 kg m(-2); 29 women: mean body mass index=25.4±3.5 kg m(-)(2)) completed a 14-day stay in a residential feeding behaviour suite. During days 1 and 2, subjects consumed a fixed diet designed to maintain energy balance. On days 3-14, food intake was covertly measured in subjects who had ad libitum access to a wide variety of foods typical of their normal diets. Resting metabolic rate (RMR; respiratory exchange), total daily energy expenditure (doubly labelled water) and body composition (total body water estimated from deuterium dilution) were measured on days 3-14. RESULTS: Hierarchical multiple regression indicated that after controlling for age and sex, both fat-free mass (FFM; P<0.001) and RMR (P<0.001) predicted daily EI. However, a mediation model using path analysis indicated that the effect of FFM (and fat mass) on EI was fully mediated by RMR (P<0.001). CONCLUSIONS: These data indicate that RMR is a strong determinant of EI under controlled laboratory conditions where food choice is allowed to freely vary and subjects are close to energy balance. Therefore, the conventional adipocentric model of appetite control should be revised to reflect the influence of RMR.

Plausible self-reported dietary intakes in a residential facility are not necessarily reliable.

BACKGROUND/OBJECTIVES: Comparing reported energy intakes with estimated energy requirements as multiples of basal metabolic rate (Ein:BMR) is an established method of identifying implausible food intake records. The present study aimed to examine the validity of self-reported food intakes believed to be plausible. SUBJECTS/METHODS: One hundred and eighty men and women were provided with all food and beverages for two consecutive days in a residential laboratory setting. Subjects self-reported their food and beverage intakes using the weighed food diary method (WDR). Investigators covertly measured subjects' actual consumption over the same period. Subjects also reported intakes over four consecutive days at home. BMR was measured by indirect calorimetry. RESULTS: Average reported energy intakes were significantly lower than actual intakes (11.2 and 11.8 MJ/d, respectively, P<0.001). Two-thirds (121) of the WDR were under-reported to varying degrees. Only five of these were considered as implausible using an Ein:BMR cut-off value of 1.03*BMR. Under-reporting of food and beverage intakes, as measured by the difference between reported and actual intake, was evident at all levels of Ein;BMR. Reported energy intakes were lower still (10.2 MJ/d) while subjects were at home. CONCLUSIONS: Under-recording of self-reported food intake records was extensive but very few under-reported food intake records were identified as implausible using energy intake to BMR ratios. Under-recording was evident at all levels of energy intake.

Experience and acceptability of diets of varying protein content and glycemic index in an obese cohort: results from the Diogenes trial.

BACKGROUND/OBJECTIVES: To investigate acceptability and tolerability of diets of different protein and glycemic index (GI) content aimed at weight maintenance following a phase of rapid weight loss, as part of a large pan-European dietary intervention trial. SUBJECTS/METHODS: The Diogenes study (www.diogenes-eu.org) consisted of an initial 8-week rapid weight-loss phase (800-1000 kcal/day), followed by a 6-month weight maintenance intervention with five different diets varying in protein and GI content. Measurement of a range of outcomes relating to experience of the Diogenes diets in terms of acceptability, experience and mood were recorded via end of day questionnaires throughout the study. RESULTS: Weight change during the initial weight loss phase weakly, but positively correlated with acceptability of the programme (r range=-0.08 to 0.2, P 0.05, n=685 on four of five dimensions). Success at weight maintenance positively correlated with acceptance of the programme (r range=-0.21 to -0.34, P<0.001, n=540 for all five dimensions). The diets with higher protein content were more acceptable than the low protein (LP) diets, however, no differences between the high vs low GI diets were found concerning acceptability and tolerability. CONCLUSIONS: Results suggest that moderately high protein diets, compared with LP diets, are more acceptable diets for weight control in overweight individuals.

Gestational age, gender and parity specific centile charts for placental weight for singleton deliveries in Aberdeen, UK.

INTRODUCTION: The weight of the placenta is a crude but useful proxy for its function in vivo. Accordingly extremes of placental weight are associated with adverse pregnancy outcomes while even normal variations in placental size may impact lifelong health. Centile charts of placental weight for gestational age and gender are used to identify placental weight extremes but none report the effect of parity. Thus the objective was to produce gender and gestational age specific centile charts for placental weight in nulliparous and multiparous women. METHODS: Data was extracted from the Aberdeen Maternity and Neonatal Databank for all women delivering singleton babies in Aberdeen city and district after 24 weeks gestation. Gestational age specific centile charts for placental weight by gender and parity grouping (n = 88,649 deliveries over a 30 year period) were constructed using the LMS method after exclusion of outliers (0.63% of deliveries meeting study inclusion criteria). RESULTS: Tables and figures are presented for placental weight centiles according to gestational age, gender and parity grouping. Tables are additionally presented for the birth weight to placental weight ratio by gender. Placental weight and the fetal:placental weight ratio were higher in male versus female deliveries. Placental weight was greater in multiparous compared with nulliparous women. DISCUSSION: We present strong evidence that both gender and parity grouping influence placental weight centiles. The differences at any given gestational age are small and the effects of parity are greater overall than those of gender. In contrast the birth weight to placental weight ratio differs by gender only. CONCLUSION: These UK population specific centile charts may be useful in studies investigating the role of the placenta in mediating pregnancy outcome and lifelong health.

Placental weight and efficiency in relation to maternal body mass index and the risk of pregnancy complications in women delivering singleton babies.

Herein we report placental weight and efficiency in relation to maternal BMI and the risk of pregnancy complications in 55,105 pregnancies. Adjusted placental weight increased with increasing BMI through underweight, normal, overweight, obese and morbidly obese categories and accordingly underweight women were more likely to experience placental growth restriction [OR 1.69 (95% CI 1.46-1.95)], while placental hypertrophy was more common in overweight, obese and morbidly obese groups [OR 1.59 (95% CI 1.50-1.69), OR 1.97 (95% CI 1.81-2.15) and OR 2.34 (95% CI 2.08-2.63), respectively]. In contrast the ratio of fetal to placental weight (a proxy for placental efficiency) was lower (P < 0.001) in overweight, obese and morbidly obese than in both normal and underweight women which were equivalent. Relative to the middle tertile reference group (mean 622 g), placental weight in the lower tertile (mean 484 g) was associated with a higher risk of pre-eclampsia, induced labour, spontaneous preterm delivery, stillbirth and low birth weight (P < 0.001). Conversely placental weight in the upper tertile (mean 788 g) was associated with a higher risk of caesarean section, post-term delivery and high birth weight (P < 0.001). With respect to assumed placental efficiency a ratio in the lower tertile was associated with an increased risk of pre-eclampsia, induced labour, caesarean section and spontaneous preterm delivery (P < 0.001) and a ratio in both the lower and higher tertiles was associated with an increased risk of low birth weight (P < 0.001). Placental efficiency was not related to the risk of stillbirth or high birth weight. No interactions between maternal BMI and placental weight tertile were detected suggesting that both abnormal BMI and placental growth are independent risk factors for a range of pregnancy complications.

The behaviour of a neutral model of weight regulated only by body mass.

While there are many mechanisms that may be involved in the regulation of body mass in humans and other animals, it is not so clear how much regulation is needed beyond the negative feedback effect of body mass itself. Here we model weight changes as a stochastic process, and show that it behaves approximately as an autoregressive process. Using published estimates of the energy cost of weight gain, the effect of weight on resting metabolic rate and the daily variation in intake and activity, we show that fluctuations in weight will be small. The effect of excess intake is also examined, and the assumptions and limitations of the model are discussed.

The relationship between BMI and percent body fat, measured by bioelectrical impedance, in a large adult sample is curvilinear and influenced by age and sex.

OBJECTIVE: The study aimed to establish the effects of age, gender and age-gender interactions on BMI-% fat relationships over a wide range of BMI and age. It also aimed to examine controversies regarding linear or curvilinear BMI-% fat relationships. METHODS: Body composition was measured using validated bio-impedance equipment (Bodystat) in a large self-selected sample of 23,627 UK adults aged 18-99 (99% ≤70) years, of which 11,582 were males with a mean BMI of 26.3±4.7 (sd) kg/m(2), and 12,044 females, with a mean BMI of 25.7±5.1 kg/m(2). Multiple regression analysis was used. RESULTS: BMI progressively increased with age in women and plateaued between 40 and 70 years in men. At a fixed BMI, body fat mass increased with age (1.9 kg/decade), as did % fat (1.1-1.4% per decade). The relationship between BMI and % fat was found to be curvilinear (quadratic) rather than linear, with a weaker association at lower BMI. There was also a small but significant age-gender interaction. CONCLUSION: The association between BMI and % body fat is not strong, particularly in the desirable BMI range, is curvilinear rather than linear, and is affected by age.

Optimising two-dye microarray designs for estimating associations with a quantitative trait.

We discuss how the samples should be arranged in two-dye microarray studies when the objective is to investigate associations between gene expression and quantitative traits measured on each sample. Because there is typically large between array variation, information about the association will come from the differences in traits and expression measurements between the two values hybridised to the two dyes on the same array. It is shown that within-slide correlation of trait values should be minimised. The arrangement of samples for which this occurs will depend on the trait values in question, and is a computationally demanding problem. An alternative is to minimise the rank correlation. We discuss this and related issues for different combinations of numbers of samples and arrays. Data analysis, including estimation of the variance components, is also described.

Interpretation of two-stage experiments in animal studies.

Two-stage experiments allow results to be analysed at the end of the first stage and the second stage to be omitted if the preliminary conclusions are clear-cut. They thus offer the potential to reduce the number of experimental animals. However, using standard P values to assess the significance of results at the end of either stage will lead to an increase in risk of false positive conclusions. This paper provides a possible protocol for two-stage experiments and a method for adjusting P values. It is shown that, for experiments with reasonable power (>80%), the expected reduction in animal numbers will be at least 20%.

Automatic recognition of maize cell types using context information.

When classifying objects in images of biological specimens, it is usual for there to be some dependence among neighbouring objects. This can in theory be used to augment the information available for classifying each object. However, much of the methodology developed for this type of contextual classification assumes a fixed number of neighbours, such as is found on a regular grid. In this paper, we show how Markov random fields can be used in the case where the number of neighbours varies, and we illustrate this with an application in the classification of cells types in microscope images of plant stems.

Rate and extent of compensatory changes in energy intake and expenditure in response to altered exercise and diet composition in humans.

We assessed the effect of no exercise (Nex; control) and high exercise level (Hex; approximately 4 MJ/day) and two dietary manipulations [a high-fat diet (HF; 50% of energy, 700 kJ/100 g) and low-fat diet (LF; 20% of energy, 300 kJ/100 g)] on compensatory changes in energy intake (EI) and energy expenditure (EE) over 7-day periods. Eight lean men were each studied four times in a 2 x 2 randomized design. EI was directly quantified by weight of food consumed. EE was assessed by heart rate (HR) monitoring. Body weight was measured daily. Mean daily EE was 17.6 and 11.5 MJ/day (P < 0.001) on the pooled Hex and Nex treatments, respectively. EI was higher on HF diets (13.4 MJ/day pooled) compared with the LF diets (9.0 MJ/day). Regression analysis showed that these energy imbalances induced significant compensatory changes in EB over time of approximately 0.3-0.4 MJ/day (P < 0.05). These were due to changes in both EI and EE in the opposite direction to the perturbation in energy balance. These changes were significant, small but persistent, amounting to approximately 0.2 and approximately 0.35 MJ/day for EI and EE, respectively.

Slow fiber cluster pattern in pig longissimus thoracis muscle: implications for myogenesis.

Recent evidence implicates fiber type proportions as playing a role in meat eating quality, and in pigs it has been suggested that the slow oxidative fibers contribute to both juiciness and tenderness. The fiber distribution in pigs is different from that found in most other species, in which the various types of skeletal muscle fiber are distributed in a "checkerboard" pattern, because in pigs the slow oxidative fibers have a clustered distribution. The initial processes leading to fiber clustering are likely to occur during myogenesis, but the precise mechanistic aetiology of this patterning and whether the slow oxidative fiber clusters occur in a random or ordered fashion is unknown. In the present study longissimus thoracis muscle from Large White crossbred pigs was sampled at the 10th rib, 48 h postmortem. Transverse cryo-sections were cut and histochemically stained to allow the identification of the main muscle fiber types: slow oxidative, fast glycolytic, and fast oxidative glycolytic. Images of the sections were captured and analyzed using point processes and Voronoi Tesselations to examine the randomness and spatial distribution of the clusters of slow oxidative fibers found in pig longissimus thoracis muscle. The results showed that an assumption of complete spatial randomness can be rejected and that a mathematical model incorporating a minimum distance of 1.7 to 2.0 microm between cluster centers produced fiber patterns similar to those observed in the original transverse sections of the muscle. In addition, if it assumed that the central fiber in each cluster is derived from primary myoblast progenitors, these results suggest that there may be some degree of repulsion between the primary fibers during the initial stages of cluster formation. The mechanistic basis of such repulsion is not clear, but it is speculated that secreted factors, such as sonic hedgehog or myostatin may play a role.

Predicting basal metabolic rate in the obese is difficult.

OBJECTIVE: To reassess the validity of the Schofield equations for predicting basal metabolic rate (BMR) in the obese. DATA: The data collection compiled from many studies by Schofield in 1985. METHODS: Nonparametric curve fitting and comparison of groups. RESULTS: BMR increases more slowly at heavier weights in men (above about 75 kg) and women (above about 65 kg), and to ignore this is to predict overestimates of BMR. Data obtained from studies of different populations show substantial differences, which are influential when data are combined. CONCLUSIONS: The Schofield equations are unsuitable for obese populations. Given that current Western populations exhibit prevalences of obesity many times greater than those in the Schofield database, this indicates the need for further study of suitable predictors for these individuals.

The effect of graded levels of exercise on energy intake and balance in free-living men, consuming their normal diet.

OBJECTIVE: To assess the effect of graded increases in exercised-induced energy expenditure (EE) on appetite, energy intake (EI), total daily EE and body weight in men living in their normal environment and consuming their usual diets. DESIGN: Within-subject, repeated measures design. Six men (mean (s.d.) age 31.0 (5.0) y; weight 75.1 (15.96) kg; height 1.79 (0.10) m; body mass index (BMI) 23.3(2.4) kg/m(2)), were each studied three times during a 9 day protocol, corresponding to prescriptions of no exercise, (control) (Nex; 0 MJ/day), medium exercise level (Mex; approximately 1.6 MJ/day) and high exercise level (Hex; approximately 3.2 MJ/day). On days 1-2 subjects were given a medium fat (MF) maintenance diet (1.6 x resting metabolic rate (RMR)). MEASUREMENTS: On days 3-9 subjects self-recorded dietary intake using a food diary and self-weighed intake. EE was assessed by continual heart rate monitoring, using the modified FLEX method. Subjects' HR (heart rate) was individually calibrated against submaximal VO(2) during incremental exercise tests at the beginning and end of each 9 day study period. Respiratory exchange was measured by indirect calorimetry. Subjects completed hourly hunger ratings during waking hours to record subjective sensations of hunger and appetite. Body weight was measured daily. RESULTS: EE amounted to 11.7, 12.9 and 16.8 MJ/day (F(2,10)=48.26; P<0.001 (s.e.d=0.55)) on the Nex, Mex and Hex treatments, respectively. The corresponding values for EI were 11.6, 11.8 and 11.8 MJ/day (F(2,10)=0.10; P=0.910 (s.e.d.=0.10)), respectively. There were no treatment effects on hunger, appetite or body weight, but there was evidence of weight loss on the Hex treatment. CONCLUSION: Increasing EE did not lead to compensation of EI over 7 days. However, total daily EE tended to decrease over time on the two exercise treatments. Lean men appear able to tolerate a considerable negative energy balance, induced by exercise, over 7 days without invoking compensatory increases in EI.

Statistical analysis of nutritional studies.

Statistical analysis of data collected in experimental or observational studies is an important part of nutritional research. If it is not done appropriately, there is a risk that information in the data is lost, or that conclusions are misleading. The present article does not attempt a full review of the subject of statistics, but attempts to list aspects where care is needed.

Effect of diets rich in oleic acid, stearic acid and linoleic acid on postprandial haemostatic factors in young healthy men.

The aim of the present study was to investigate the effects of stearic acid-, oleic acid- and linoleic acid-rich meals on postprandial haemostasis in young healthy volunteers whose background diets had been controlled for 14 d in a residential study. Six healthy male volunteers were assigned randomly to consume diets rich in stearic acid, oleic acid or linoleic acid for 14 d. On day 15, plasma lipids and haematological variables were measured in the fasted state, and 3 and 7 h (factor VII and prothrombin activation peptide fragments, 1 and 2 only) after consumption of a test meal. Test meals provided 40 % of the subjects' daily energy requirement, with 41 % of the energy provided as fat, 17 % energy as protein and 42 % energy as carbohydrate. The mean fat content of the meal was 45 (sd 5) g. Significant alterations from fasted values were observed for activated factor VII after 7 h), factor VII antigen after 7 h), prothrombin activation peptide fragments 1 and 2 after 7 h) and plasminogen activator inhibitor type 1 activity after 3 h) after consumption of each of the three meals. No significant differences were observed in haemostatic values (factor VII coagulant activity, factor VII antigen, tissue plasminogen activator activity prothrombin activation peptide fragment and plasminogen activator inhibitor type-1) with regard to diet except for activated factor VII at 3 h; values were higher after the oleic acid- and linoleic acid-rich meals than after the stearic acid-rich meal After consumption of each of the three meals, chylomicrons contained proportionately more palmitic acid than the lipids ingested. The present study shows that there are demonstrable changes in postprandial haemostasis when young healthy volunteers with controlled dietary backgrounds are challenged with a physiological fat load. These changes are independent of the fatty acid composition of the test meals.

Subclinical infection with the nematode Trichostrongylus colubriformis increases gastrointestinal tract leucine metabolism and reduces availability of leucine for other tissues.

Gastrointestinal (GI) tract leucine metabolism was measured in 6- to 9-mo-old lambs subjected to trickle infection with Trichostrongylus colubriformis larvae and in separate animals that were not infected. Animals prepared with a jejunal catheter and with indwelling catheters into the aorta and the portal- (PDV) and mesenteric- (MDV) drained viscera were infused simultaneously with [1-13C] and [5,5,5-2H3] leucine to determine GI tract sequestration of leucine from arterial and luminal amino acid pools by tracer and tracee arteriovenous concentration differences. Leucine oxidative losses and net fluxes were also determined across the GI tract. Infection had no detectable effect on whole-body leucine flux, but it increased total GI tract leucine sequestration by 24% (P<.05) and GI tract oxidative losses of leucine by 22 to 41% (P<.01). Net PDV fluxes of leucine were decreased by 20 to 32% during the infection. The infection did not alter either the proportion of precursor leucine used by GI tract metabolism that was derived from the arterial leucine pool (.84 to .88) or the proportional sequestration of digesta-derived leucine during "first pass" absorptive metabolism (.12 to .18). These findings help to elucidate the metabolic basis for the reduced growth rates and nitrogen retention observed when animals are subjected to subclinical nematode infection.