AWM: Eating Frequency and Patterns (2013)

Citation:

Whybrow S, Mayer C, Kirk TR, Mazlan N, Stubbs RJ. Effect of two weeks' mandatory snack consumption on energy intake and energy balance. Obesity (Silver Spring). 2007; 15(3): 673-685.

 

PubMed ID: 17372318
 
Study Design:
Randomized Crossover Trial
Class:
A - Click here for explanation of classification scheme.
Quality Rating:
Neutral NEUTRAL: See Quality Criteria Checklist below.
Research Purpose:

To compare the effects of mandatory consumption of commercial snack products (CSPs) on energy intakes and energy balance in free-living adults and to assess the interaction between habitual level of CSP consumption and the intervention.

Inclusion Criteria:
  • All subjects were healthy, 19 to 50 years of age, non-smokers, not taking any medication known to affect appetite, not undertaking any dietary or exercise treatments to lose weight and had a recent history of weight stability
  • Body mass index of subjects in the lean group ranged from 19 to 25kg/m2 and in the overweight group ranged from 26 to 35kg/m2.
Exclusion Criteria:
  • 15 subjects dropped out before completing
  • 13 almost certainly under-reported their food and energy intakes and were excluded from the analyses.
Description of Study Protocol:

Recruitment

Subjects were recruited from Aberdeen by advertisements. Recruitment was not specifically targeted toward snackers or non-snackers and the usual level of snack and CSP consumption was not assessed before subjects were accepted into the study. 

Design

  • Randomized crossover trial
  • Each intervention period comprised of seven days of equilibration (Days one to seven) where the only manipulations were the consumption of the prescribed snacks each day and measurement of body weight on Day one.
  • For the following seven-day period (Days eight to 14), in addition to consuming the snacks, dietary data were collected using the seven-day weighted record method
  • With the exception of the snacking intervention, subjects consumed their normal diets and were not given any funds to purchase foods, nor were they given any additional instructions on how or what to eat
  • The degree on energy intake compensation during the two periods when subjects were consuming the intervention snacks was calculated as the mean of the changes in recorded energy intake relative to the control period
  • An individual who compensated completely for the energy content of the intervention snacks would have reported the same energy intake during each of the three intervention periods and the mean change in energy intake would be 0mJ
  • To assess whether the intervention snacks were simply displacing similar foods that normally would have been eaten, habitual CSP consumption was estimated from the food record during the control period.

Blinding Used

To disguise the manufacturer and brand and to remove all nutritional information, each snack was repackaged in plain wrappers and identified only by a random three-digit number and day of the week on which it was to be eaten.

Intervention

  • Three types of intervention snacks were used: HDL-cholesterol, mixed and H-fat
  • Each intervention comprised of six CSPs, four sweet and two savory which were supplied on a three-day rotating menu
  • Products were a mixture of chocolate confectionery, cakes, sweet and savory biscuits and potato chips normally available in individual portion sizes
  • Each subject received one type of snack by macronutrient composition but three manipulations of snack energy, 0mJ per day (control), 1.5mJ per day (low-energy, low-E, two snacks per day) and 3.0mJ per day (high-energy, high-E, four snacks per day)
  • Subjects were instructed to eat one half of the snacks mid-morning and one half mid-afternoon, but not at specific times
  • Each subject completed three interventions, which were two weeks in duration with a wash-out period between them of nominally two weeks. 

Statistical Analysis

  • Changes in body weight and energy intake were analyzed by ANOVA with type and amount of intervention snack, sex and body weight that were significantly different from zero body weight change
  • Kruskal-Wallis tests were used to compare non-parametric data
  • Statistical significance was accepted at the 5% probability level. 
Data Collection Summary:

Timing of Measurements

  • Body composition, height and dietary restraint were assessed during the pre-study tests  
  • Before the first intervention period, subjects completed the Three Factor Eating Inventory to assess degree of dietary restraint
  • Resting metabolic rate (RMR) was measured before the first intervention period and after the last intervention period
  • Energy expenditure was estimated during the three seven-day measurement periods by continuous heart rate monitor.

Dependent Variables

  • Self-reported food intakes between Days eight and 14
  • Investigator-recorded body weights from Days one, eight and 15 of each intervention period
  • Daily energy expenditure estimated by heart rate monitoring.

Independent Variables

  • Subject group:
    • Lean males
    • Lean females
    • Overweight males
    • Overweight females
  • Type of CSP:
    • High carbohydrate
    • High fat
    • Mixed composition
  • Manipulation of Energy
    • Control: 0mJ
    • Low energy: 1.5mJ
    • High energy: 3.0mJ.

Control Variables

  • Age
  • Height
  • Weight
  • BMI
  • Dietary restraint. 
Description of Actual Data Sample:
  • Initial N: Recruitment continued until 72 subjects (18 lean males, 18 lean females, 18 overweight males, 18 overweight females) had completed the study
  • Attrition (final N): As above; 15 subjects dropped out before completing (21%) and 13 under-reported their food and energy intakes and were excluded from analyses.
  • Age:
  Men Women
  Lean Overweight Lean Overweight
Age 33.6±8.15 36.9±6.9 31.9±7.25 37.8±8.27

Anthropometrics:

  Men  Women 
  Lean Overweight Lean Overweight
Height (m) 1.8±0.07 1.8±0.06 1.65±0.07 1.65±0.06
Weight (kg) 69.4±6.92 95±12.21 58.8±7.13 78.4±8.11
BMI 21.5±1.52 29.4±2.79 21.7±1.45 28.8±2.15
  • Location: Aberdeen, Scotland.

 

Summary of Results:

Key Findings

  • Subject characteristics:
    • Males differed in age (35, 26 and 40 years for the H-CHO, mixed and H-fat interventions, respectively, P<0.01)
    • Lean females and overweight males scored differently for dietary restraint (three, 7.5 and seven, P<0.05 for lean females and four, nine and three, P<0.05 for overweight males for the H-CHO, mixed and H-fat interventions, respectively)
    • There were no other significant within-group differences for dietary restraint or for age, BMI or percentage body fat
  • Habitual snack consumption, energy intake and body weight status:
    •  Average CSP consumption reported during the control week was 19% (SD=10%) of total daily energy intake with a range from 0% to 47%
    • There was a significant correlation between mean daily total energy intake and absolute energy intake from CSPs (R2=0.204; P<0.001) but not with the percentage contribution of CSPs to total energy intake (R2=0.005; P<0.547)
    • Overweight subjects did not seem to consume more CSPs than did lean subjects, as there was no significant correlation between absolute energy intake from CSPs and BMI (R2=.003; P<0.664) or percentage body fat (R2=-0.002; P<0.706)
    • Thus, subjects with higher energy requirements though being active or having higher BMIs and therefore higher energy intakes, consumed more CSPs but not proportionally more so
    • A similar correlation was seen between consumption of all snacks and total energy intakes (R2=0.466; P<0.001)
    • There were very weak positive correlations between consumption of all snacks and BMI (R2=0.044; P<0.076)
    • There was a positive correlation between energy intake from all snacks and CSPs (R2=0.427; P<0.001)
  • Compliance with the intervention:
    • From the uneaten snacks and empty snack wrappers returned to the investigators, it appeared that compliance with the intervention was very good with at least 89% of the mandatory snacks being consumed on each treatment. Overall, 96% of the mandatory snacks were eaten.
    • Male subjects consumed significantly more of the prescribed snacks than did the females (97.8% vs. 93.7%, respectively, P=0.32) but there were no significant differences in the amount of intervention snacks eaten between lean and overweight subjects (P=0.998) or across types of snack intervention (P=0.712)
    • A smaller proportion of the mandatory snacks were consumed during the high-E treatment than during the low-E treatment (93.8% vs. 97.7%, respectively, P=0.023)
  • Effects of the interventions on energy intake:
    • As expected, male subjects reported significantly higher mean daily energy intakes at each level of intervention snack energy than female subjects (averaged over the three intervention periods of 12.7 and 9.3mJ per day, respectively, P=0.001, equal to 1.74 basal metabolic rate and 1.59 BMR
    • Also, overweight subjects reported higher energy intakes than lean subjects (averaged over the three interventions of 11.5 and 10.5mJ per day, respectively, P=0.017, equal to 1.64 BMR and 1.7 BMR). Thus, reported energy intakes were consistent with expected values.
    • Energy intakes increased as the intervention snacks were titrated into the diet. However, the increase was less than the energy content of the intervention snacks, indicating that subjects were partially compensating for energy by 54% and 64% (P=0.348) on the low-E and high-E treatments, respectively.
    • The mean compensation for the H-CHO, mixed and H-fat treatments were 84%, 67% and 31%, respectively (P=0.67)
    • There was no interaction between the amount and type of intervention snacks consumed [(F)4.138=1.52; standard error of difference = 0.76; P=0.201]. 
    • There was, however, evidence of a trend of increasing energy intake with increasing fat content of the interventions; the mean changes in energy intakes of the low-E and high-E treatments combined (compared with the control periods) were 0.41, 0.7, and 1.4mJ per day on the H-CHO, mixed and H-fat treatments, respectively (P=0.78)
    • The four groups of subjects responded in a similar way to the interventions; there were no significant interactions among the amount and type of intervention snacks and subject group [(F)4.12=1.24; P=0.298]
  • Energy expenditure and change in energy balance:
    • Compared with the control period and when considered across all subject groups and intervention snack types together, mean daily energy expenditure was significantly greater during the two intervention periods
    • There were no significant intervention snack type or amount and subject group interactions [F(4.117(3))=0.96; standard error of difference =1.4; P=0.435]
    • Similar differences were seen when energy expenditure was expressed as a multiple of RMR
    • There was no significant effect of the interventions on energy intake  - energy expenditure [(F)2.139(3))=0.16; P=0.96]
  • Effects of the intervention on body weight:
    • There was no significant effect on run order on change in body weight
    • There was a significant effect of study period on change in body weight in that, on average, subjects gained 0.14kg over the equilibration periods and lost 0.22kg over the measurement periods (P<0.001)
  • Weight change over the equilibration periods:
    • Although there was no significant effect of the level of intervention snack on change in body weight over the seven-day equilibration periods [F(2,136(2)0=1.29; P=0.273], there was evidence that body weight changes were more positive with increasing consumption of intervention snacks
    • Change in body weight were +0.02kg (not significantly different from zero, T-test), +0.15kg (P=0.078) and +0.25kg (P=0.12) for the control, low-E and high-E treatments, respectively
    • There was no significant interaction between level of intervention snack energy, sex or weight category over the three equilibration periods [F(2,118)=2.64; P=0.075]
    • There were no significant differences between intervention snack types for change in body weights [F(4,136(2)=1.82; P=0.101]
  • Weight change over the measurement period
    • There was no significant effect of the level of intervention snack on the change in body weight (P=0.293) over three measurement periods
    • However, as with the changes in body weight over the equilibration periods, there was evidence that body weight changes were more positive with increasing level of intervention snack energy over the three measurement periods; changes in body weight over the control, low-E and high-E treatments, respectively, were -0.26kg (significantly different from zero, T-test, P=0.002), -0.24kg (P=0.003) and -0.14kg (not significant)
    • Again, there was no effect of intervention snack type on body weight change during the three measurement periods [F(4,138)=1.03; P=0.664]
  • Effects of baseline energy intake and baseline CSP consumption on energy intake compensation
    • Individuals with higher recorded energy intakes did not compensate any more completely than those with lower intakes
    • Individuals with higher habitual energy intakes from CSPs compensated better than those with lower intakes
    • The regression line crosses the horizontal axis at 3.8mJ, suggesting that an individual who habitually consumed 3.8mJ per day of CSPs would be predicted to compensate completely for the mean of 2.25mJ per day of intervention snacks. 
Author Conclusion:
  • Habitual snack consumption and body weight status:
    • The present study investigated the effects of incorporating CSPs into the diet on energy intake and energy balance. CSP consumption as estimated in this study would have included some foods that would have been consumed as part of main meals. Using CSP consumption to classify subjects is the same as using snack patterns. It would be possible to have a high-E intake from snacks but to consume no CSPs at all. However, there was a significant correlation between energy intake from all snacks and energy intake form CSPs.
    • In this sample of Scottish adults, there was no evidence that habitual CSP consumption was associated with obesity, BMI or percentage of body fat. The very weak relationship between energy intake from all snacks and BMI is consistent with earlier cross-sectional studies that tend not to support a positive relationship between feeding frequency and obesity. When examining feeding frequency and energy balance, reporting is often flawed and overweight individuals reduce their snacking in the belief that it will help reduce energy intakes and lead to weight loss. In this study, subjects who were suspected of grossly misreporting their food intakes were excluded from analysis. This does not mean than the remaining data are completely free from effects of low-E reporting or changes in feeding behavior during the recording periods as it is becoming increasing clear that all subjects misreport to some degree when self-reporting their food intakes.
    • Energy intake from all snacks and from CSPs correlated positively with total energy intakes suggesting that individuals with higher energy requirements consume snacks to help achieve energy balance. Individuals with extremely high-E expenditures seem to eat more frequently than those who are less active. Four groups were recruited for this study. As groups, their responses to the snacking interventions were similar; both lean and overweight subjects compensated to a similar degree. The main factor associated with energy intake compensation was habitual CSP consumption (although this association was not strong). The greater compensatory trends seem to be reflective of the particular aspects of habitual feeding behavior patterns which become established over time.
  • Energy intake:
    • The snacking intervention produced an increase in mean daily average intake, but this was less than the energy content of the snacks. Subjects compensated for 54% of the low-E and 64% of the high-E intervention. A tendency for a similar degree of compensation has been reported elsewhere. Energy intake compensation tended to be greater for snacks consumed before the evening meal compared with those consumed before lunch. Compensation for snacks may be through delaying the onset of the subsequent meal, rather than in meal size reduction and the timing of snacks relative to fixed  meal times such as those imposed by working patterns may influence the effects of snacking on energy balance. The degree of energy intake compensation may have been different if the snack times had been prescribed or if subjects were completely free to determine when the intervention snacks were consumed.
    • When snackers were asked to consume at least 25% of their daily energy intake from four types of CSPs (low-fat or H-fat and sweet or non-sweet) for three-week periods, there was almost complete compensation with no significant change in energy intake or body weight when compared with a control period. However, the intervention would have been providing approximately the customary level of snack energy intake and the compensation is likely to have been because the intervention snacks were replacing existing snacks in the diets of regular snackers.  A similar effect was seen in the present study but to varying degrees because of the differences in usual level of CSP consumption.
    • In the present study, the intervention was acute relative to habitual patterns of feeding behavior and customary levels of CSP consumption. While the short-term intervention elevated energy intake and resulted in increased body weight, data from the control period measurements suggest that this effect would level out in the longer term; there were no associations between CSP consumption and body weight status. Once snacking has been incorporated into the eating pattern, it was likely to be less of a risk for over-consumption.  There was clear evidence in this study that it was the frequent consumer of CSPs who compensated most when CSPs were included in the diet as mandatory supplements. This was because the intervention snacks were partly replacing snack foods that the subjects would normally have consumed.
  • Macronutrient composition and energy density:
    • There was a trend for increasing degree of compensation with decreasing energy density of the interventions, which was greater than might have been anticipated given the quite small differences in energy density of the three types of snacks, although this effect was not statistically significant. 
    • Reporting included the normal variation food intake across two or more time-points. The snacks also differed in macronutrient composition and CHO is more satiating than fat, at least in the short to medium term. This combined with the effects of differences in energy density, probably contributed to the apparent differences in compensation.
  • Energy balance and energy expenditure: The apparently higher energy expenditure when subjects were consuming the intervention snacks compared with the control period was interesting to the authors. They note the following points:
    • There are considerable limits in using heart rate to estimate energy expenditure and although some of these limitations were minimized by the with-in subject comparisons and the counterbalanced order that subjects received the treatments, it is possible that the apparent differences in energy expenditure were attributable to the method. Data showed that subjects wore the heart rate monitors for the same amount of time during each of the three intervention periods (P=0.537).
    • Dietary-induced thermogenesis was likely to have been lower during the control periods than the low-E and high-E periods because of the differences in energy intakes. Assuming that the dietary-induced thermogenesis was 10% to 15% of food energy, the difference between the high-E and control periods would have been approximately 0.12mJ per day.
    • It is possible that subjects were altering their volitional activity during the study. In the present study, subjects were, on average, in negative energy balance during the control and low-E treatment measurement periods and in approximate energy balance during the high-E treatment measurement period. Lean men reduce their volitional activity in response to very low-E diets and starvation. Conversely, it has been suggested that some individuals increase their spontaneous activity when overfed. It is also possible that some subjects deliberately increased volitional activity in a conscious attempt to limit weight gain during the study.
    • Therefore, the differences in energy expenditure were likely to have been the cumulative effects of several mechanisms all of which tended to increase energy expenditure during  the low-E and high-E intervention periods. While this effect is significant, any change in energy expenditure that can be attributed to the increased snack consumption was insufficient to counter the additional energy content of the intervention snacks.
  • Energy balance and body weight:
    • Body weight changes were more negative over the measurement periods than over the equilibrium periods. This made it clear that subjects behaved differently over the two periods. Body weight was measured at the beginning and end of the equilibration period and this was used as an indication of energy balance over the equilibration period, when food intake was not recorded, to assess whether subjects compensated for the additional energy of the intervention. An estimate of the expected weight gain, had no compensation occurred, was calculated from the energy content of the intervention snacks consumed over the equilibration period and comparing this to the actual weight change. If the intervention snacks had been simply added to the diet, with no compensation for their energy content, the estimated weight gain would have been 0.59kg. It was clear that the average increase in body weight was only about a one half of that expected had no energy compensation occurred. Subjects were partially compensating for the energy content of the snacks. Body weight change over the equilibration period of the low-E intervention was not significantly different. Over the equilibration period and on average, subjects partially compensated for the energy content of the intervention snacks, but this was insufficient to prevent a significant increase in body weight on the high-E intervention and a non-significant trend for weight gain on the low-E intervention. 
    • Body weight changes over the equilibration and measurement periods were markedly different. Weight change was generally more negative over the measurement periods than the equilibration periods. A possible explanation is that the intervention snacks resulted in an initial positive energy balance, which subjects adjusted to over the seven days of the equilibration period and then corrected for over the subsequent seven days of the measurement period. However, this was highly unlikely given the similar pattern of weight change over the control period. It was more likely that subjects changed their feeding behavior when they were recording their food intakes. Body weight change over the measurement periods were consistent with subjects changing their diets while recording their food intake, under-eating rather then grossly misreporting their actual food intake. 
    • Measuring changes in body weight over seven days when subjects were not self-reporting their food intakes and clearly changing their feeding behavior provided better insights into the effects of the interventions. Considering weight change only over the measurement periods or over the whole 14 days could have led to the conclusion that snacking protects against weight loss.
  • In conclusion, the authors note:
    • This study demonstrated that consumption of 1.5mJ and 3.0mJ of mandatory commercially available snack foods per day elevated energy intake over seven-day periods but that there was partial energy intake compensation
    • Changes in body weight tended to be more positive with increasing amount of intervention snack consumption
    • Energy intakes were elevated when subjects were consuming each of the three types of snacks, although there was some evidence that the H-CHO snacks produced smaller increases then the mixed composition snacks, which in turn, produced smaller increases than the H-fat snacks
    • The interventions were less effective in elevating energy intakes in subjects who were frequent consumers of commercial snack products compared with less-frequent consumers
    • This, together with the lack of correlation between usual level of CSP consumption, or energy intake from all snacks, and measures of obesity, suggest that snacking can become established as part of an individual's normal feeding behavior and present less of a risk for over consumption than if these foods were only infrequently consumed.
Funding Source:
Government: Scottish Executive Environment and Rural Affairs Department
Industry:
Buscuit, Cake, Chocolate and Confectionery Association
Commodity Group:
Reviewer Comments:
  • The authors note the following limitations:
    • The use of energy intake-to-RMR ratios to identify mis-reporters of food and energy intake is arbitrary and is biased toward failing to detect mis-reports. As it appears based on other studies that all subjects (overweight and lean, men and women) mis-report their food intake, this is likely to alter the apparent energy intakes.
    • There are limitations in estimating energy expenditure from heart rate monitoring, especially in sedentary populations. The effects of the amounts of intervention snacks (control, low-E and high-E) on energy intake and balance were, however, made using within-subject comparisons, which reduced the effect of bias.
    • The precision of energy intake and energy expenditure assessments was a best around ±1.0mJ and estimating energy balance from change in body weight over seven-day periods can even less precise
    • The use of snack foods of similar macronutrients composition during each treatment was artificial; subjects would probably have selected from a variety of foods having a much wider range of energy densities and macronutrient composition
    • Furthermore, as the study was designed to evaluate the effects that they had of elevating energy balance cannot be generalized as evidence that consumption of commercial snack products inevitably results in positive energy balances
    • The evidence of changes in feeding behavior that were sufficient to affect body weight between the equilibration and measurement periods is credible, but this could have been strengthened by randomizing the order of the measurement and equilibration periods or by including an additional post-measurement intervention period
  • Study groups differed in age and dietary restraint at baseline. The age range of subjects was 19 to 50, which may not have been a representative sample. There may have been bias as the study was supported in part by the Biscuit, Cake, Chocolate and Confectionery Association.
Quality Criteria Checklist: Primary Research
Relevance Questions
  1. Would implementing the studied intervention or procedure (if found successful) result in improved outcomes for the patients/clients/population group? (Not Applicable for some epidemiological studies) Yes
  2. Did the authors study an outcome (dependent variable) or topic that the patients/clients/population group would care about? Yes
  3. Is the focus of the intervention or procedure (independent variable) or topic of study a common issue of concern to dieteticspractice? Yes
  4. Is the intervention or procedure feasible? (NA for some epidemiological studies) Yes
 
Validity Questions
1. Was the research question clearly stated? Yes
  1.1. Was (were) the specific intervention(s) or procedure(s) [independent variable(s)] identified? Yes
  1.2. Was (were) the outcome(s) [dependent variable(s)] clearly indicated? Yes
  1.3. Were the target population and setting specified? Yes
2. Was the selection of study subjects/patients free from bias? ???
  2.1. Were inclusion/exclusion criteria specified (e.g., risk, point in disease progression, diagnostic or prognosis criteria), and with sufficient detail and without omitting criteria critical to the study? Yes
  2.2. Were criteria applied equally to all study groups? Yes
  2.3. Were health, demographics, and other characteristics of subjects described? Yes
  2.4. Were the subjects/patients a representative sample of the relevant population? ???
3. Were study groups comparable? No
  3.1. Was the method of assigning subjects/patients to groups described and unbiased? (Method of randomization identified if RCT) Yes
  3.2. Were distribution of disease status, prognostic factors, and other factors (e.g., demographics) similar across study groups at baseline? No
  3.3. Were concurrent controls or comparisons used? (Concurrent preferred over historical control or comparison groups.) Yes
  3.4. If cohort study or cross-sectional study, were groups comparable on important confounding factors and/or were preexisting differences accounted for by using appropriate adjustments in statistical analysis? N/A
  3.5. If case control study, were potential confounding factors comparable for cases and controls? (If case series or trial with subjects serving as own control, this criterion is not applicable.) N/A
  3.6. If diagnostic test, was there an independent blind comparison with an appropriate reference standard (e.g., "gold standard")? N/A
4. Was method of handling withdrawals described? Yes
  4.1. Were follow-up methods described and the same for all groups? Yes
  4.2. Was the number, characteristics of withdrawals (i.e., dropouts, lost to follow up, attrition rate) and/or response rate (cross-sectional studies) described for each group? (Follow up goal for a strong study is 80%.) Yes
  4.3. Were all enrolled subjects/patients (in the original sample) accounted for? Yes
  4.4. Were reasons for withdrawals similar across groups? Yes
  4.5. If diagnostic test, was decision to perform reference test not dependent on results of test under study? N/A
5. Was blinding used to prevent introduction of bias? Yes
  5.1. In intervention study, were subjects, clinicians/practitioners, and investigators blinded to treatment group, as appropriate? Yes
  5.2. Were data collectors blinded for outcomes assessment? (If outcome is measured using an objective test, such as a lab value, this criterion is assumed to be met.) No
  5.3. In cohort study or cross-sectional study, were measurements of outcomes and risk factors blinded? N/A
  5.4. In case control study, was case definition explicit and case ascertainment not influenced by exposure status? N/A
  5.5. In diagnostic study, were test results blinded to patient history and other test results? N/A
6. Were intervention/therapeutic regimens/exposure factor or procedure and any comparison(s) described in detail? Were interveningfactors described? Yes
  6.1. In RCT or other intervention trial, were protocols described for all regimens studied? Yes
  6.2. In observational study, were interventions, study settings, and clinicians/provider described? N/A
  6.3. Was the intensity and duration of the intervention or exposure factor sufficient to produce a meaningful effect? Yes
  6.4. Was the amount of exposure and, if relevant, subject/patient compliance measured? Yes
  6.5. Were co-interventions (e.g., ancillary treatments, other therapies) described? Yes
  6.6. Were extra or unplanned treatments described? N/A
  6.7. Was the information for 6.4, 6.5, and 6.6 assessed the same way for all groups? Yes
  6.8. In diagnostic study, were details of test administration and replication sufficient? N/A
7. Were outcomes clearly defined and the measurements valid and reliable? Yes
  7.1. Were primary and secondary endpoints described and relevant to the question? Yes
  7.2. Were nutrition measures appropriate to question and outcomes of concern? Yes
  7.3. Was the period of follow-up long enough for important outcome(s) to occur? Yes
  7.4. Were the observations and measurements based on standard, valid, and reliable data collection instruments/tests/procedures? Yes
  7.5. Was the measurement of effect at an appropriate level of precision? Yes
  7.6. Were other factors accounted for (measured) that could affect outcomes? Yes
  7.7. Were the measurements conducted consistently across groups? Yes
8. Was the statistical analysis appropriate for the study design and type of outcome indicators? Yes
  8.1. Were statistical analyses adequately described and the results reported appropriately? Yes
  8.2. Were correct statistical tests used and assumptions of test not violated? Yes
  8.3. Were statistics reported with levels of significance and/or confidence intervals? Yes
  8.4. Was "intent to treat" analysis of outcomes done (and as appropriate, was there an analysis of outcomes for those maximally exposed or a dose-response analysis)? N/A
  8.5. Were adequate adjustments made for effects of confounding factors that might have affected the outcomes (e.g., multivariate analyses)? Yes
  8.6. Was clinical significance as well as statistical significance reported? Yes
  8.7. If negative findings, was a power calculation reported to address type 2 error? N/A
9. Are conclusions supported by results with biases and limitations taken into consideration? Yes
  9.1. Is there a discussion of findings? Yes
  9.2. Are biases and study limitations identified and discussed? Yes
10. Is bias due to study's funding or sponsorship unlikely? No
  10.1. Were sources of funding and investigators' affiliations described? Yes
  10.2. Was the study free from apparent conflict of interest? No