So some say you should have plenty of fat [good sources] while you're low-carbing it so as not to let the body use too much protein for energy while others say your fat intake should be relatively low. What gives? Who is right? The high fat argument makes sense [and it seems more fun] but I'm thinking that on your high carb days you must conversely keep the fat intake low.

You're very late to the fat roundtable. Start digging through the more recent (07-08) threads. Like now.

Some interesting info. Basically what it appears to be saying is that those with higher RMR and greater SNS activity do better with higher fat intake.

Lean male high- and low-fat phenotypes¾different routes for achieving energy balance

J Cooling and J E Blundell

BioPyschology Group, School of Psychology, University of Leeds, Leeds, UK


Correspondence to: J Cooling, BioPyschology Group, School of Psychology, University of Leeds, Leeds LS2 9JT, UK.JohnC@psychology.leeds.ac.uk


Abstract

OBJECTIVE: This study investigated how energy expenditure may contribute to energy balance in lean male individuals consuming a diet either high or low in fat.

METHODS: Fifteen high-fat phenotypes (HF) and 15 low-fat phenotypes (LF) participated in the study. Energy intake and macronutrient intake variables were assessed using a food frequency questionnaire and 3 day food diaries. Total energy expenditure (TEE) was estimated from 24 h heart rate monitoring and factorial methods. Habitual physical activity was measured using the Baecke questionnaire.

RESULTS: There were no anthropometric differences between HF and LF. HF had a significantly higher heart rate over 24 h; this was particularly evident during the sleeping phase. There were no differences in TEE between HF and LF, but HF were more sedentary than LF.

CONCLUSIONS: In these young male subjects a high fat intake was associated with increased sedentariness; however, higher heart rates (basal and post-meal) could indicate that energy balance was achieved by relatively high basal metabolism and an increase in dietary-induced themogenesis (physiological route). In contrast LF could maintain energy balance through relatively high levels of physical activity (behavioural route).
International Journal of Obesity (2000) 24, 1561-1566


Keywords

diet; fat intake; energy expenditure; energy balance



Introduction

There is currently a debate concerning the role played by dietary fat in the development of a positive energy balance and obesity. On the one hand there is considerable evidence that high-fat foods either have a disproportionately weak control over appetite1,2,3 or actually stimulate appetite and lead to a positive energy balance. In contrast, it has been stated that 'Diets high in fat do not appear to be the primary cause of the high prevalence of excess body fat in society, and reductions in fat will not be a solution'.4 However, given the general problem of under-reporting in large-scale dietary surveys,5 the high level of under-reporting by the obese6 and the likely selective under-reporting of fat,7,8 there must be a suspicion that epidemiological data are unsafe. Although the argument is clearly not resolved, the weight of evidence suggests that dietary fat is one significant 'risk factor' for body weight gain.9,10

However, it is clear that the relationship between dietary fat and weight gain or obesity is not a biological inevitability,11 and that there exist a number of routes (combinations of nutritional and metabolic variables) which could give rise to body weight gain. It is possible that some individuals consume a high-fat diet for many years (perhaps throughout their lives) and could be protected behaviourally or physiologically from weight gain. Clearly not all high-fat consumers are obese or even overweight, although they may be in the process of slowly gaining weight.

During a series of investigations of the impact of habitual high- and low-fat consumption, a group of high-fat eaters were identified who were lean despite consuming a large amount of energy and fat. These groups of high- and low-fat consumers exhibited differences in a number of nutritional and physiological variables, which prompted the idea that this clustering of variables defines a specific phenotype. Our initial investigations examined two groups of young adult males who consumed different diets associated with different amounts of fat and carbohydrate (by definition), but additionally had differing meal patterns, food choices, control of appetite and expression of hunger.12 However, these phenotypes were indistinguishable from each other in physical appearance, having similar body weights, body mass indexes (BMIs) and percentage body fat. Clearly the groups were not achieving a state of energy balance in the same way. Subsequent studies found some evidence to confirm this: high-fat phenotypes (HF) had a higher basal metabolism and an increased ability to oxidize fat than low-fat phenotypes (LF)13 along with higher levels of fasting plasma leptin.14 Taken together, these features constitute a physiological mechanism offering at least partial protection against the weight-inducing potential of a high-fat diet. The contrast between the HF and LF is reminiscent of the distinction between energy-sparing and energy-profligate individuals.15 This can be coupled with the observation that there is considerable inter-subject variability in levels of energy expenditure at every level of body weight¾up to 25% of the variance may not be accounted for by age, gender and body composition.16 Some physiological mechanism for this unexplained variability may be partly responsible for the similarity of BMI, body weight and body composition of these young adult male high-fat consumers, who consume more fat and energy than their lean counterparts. One possibility for this mechanism might be explained by uncoupling proteins which are upregulated on a high-fat diet and in turn increase energy expenditure.17

The present study has been carried out to investigate whether any obvious differences in energy expenditure through the amount or energy cost of physical activity, or some other mechanism which could help to offset the high-energy intake of HF in order to keep them in or near energy balance. The study therefore estimated total energy expenditure (TEE) and physical activity patterns in high- and low-fat phenotypes by a number of different methods. The results provide some indication of the differing mechanisms which may operate to achieve energy balance in relation to habitual diet.


Methods

Subjects

Fifteen habitual high-fat consumers (HF) and 15 habitual low-fat consumers (LF) were recruited from the staff-student population of Leeds University. HF and LF were defined as obtaining >43% or <33% of energy from fat, respectively, on the basis of a food frequency questionnaire.18 This criterion was coupled with the requirement that all HF must consume a greater absolute weight of fat than any LF. Therefore, a dual criterion was imposed with HF required to show a high absolute intake of fat (g) and a high energy percentage. All volunteers were in the age range 18-25 and had a BMI<26 kg/m2. None of the subjects was a regular smoker. Each subject was required to read and sign a participant information/ethics form as required by the University Ethics Committee, following ethical approval of this study.

Design

The purpose of this study was to assess physical activity patterns and total daily energy expenditure of HF and LF using 24 h continuous heart rate monitoring and activity questionnaires. All volunteers participated in 24 h continuous heart rate monitoring on three non-consecutive days. Energy expenditure was later calculated from heart rate readings following a calibration of energy expenditure (measured by indirect calorimetry) and heart rate performed in the laboratory.

Eight HF and eight LF additionally completed food diaries during the time of heart rate recording, while the remaining seven HF and seven LF completed activity questionnaires and an activity diary during the time of heart rate recording.

Procedure

Twenty-four-hour heart rate recording. Twenty-four-hour heart rate was recorded in all subjects on three non-consecutive days, typically over the space of 10 days. One test day was required to be on a weekend as both dietary and physical activity patterns may differ between weekend and midweek days.

Subjects were instructed on the operation of the heart rate monitor (Polar Sports Tester PE4000, Finland) and were required to fit and start the monitor immediately after rising on each study day. The monitor was set to record heart rate every 60 s. Subjects were reminded that the monitor should be worn for at least 24 h (including during bathing and showering) and that they were free to behave freely on the study day, with no restriction on exercise or sleep.

During the periods of heart rate recording, eight HF and eight LF were additionally required to complete a household measures food diary which was analysed using Comp-eat software, while seven HF and seven LF completed a daily activity diary which was converted to TEE (factorial method)19 and the Baecke questionnaire of habitual physical activity.20 The Baecke questionnaire consists of several questions scored on a 1-5 Likert scale regarding the components of work, leisure and sports; the maximum Baecke score is 15.

Conversion of 24 h heart rate to TEE. On a day when 24 h heart rate was not measured, subjects were asked to attend the Human Appetite Research Unit (HARU), University of Leeds, at approximately 12:00 midday having fasted for 2 h. Heart rate was recorded every 5 s during the session (Polar Sports Tester PE4000, Finland). Indirect calorimetry and heart rate were recorded simultaneously during seven stages of increasing physical demand: lying down, sitting, standing, walking at 2.8 miles/h, walking at 3.6 miles/h; jogging at 5 miles/h, and running at 6 miles/h. On the first stage subjects were asked to lay down for 5 min, during the last minute of which oxygen (O2) consumption and carbon dioxide (CO2) production was measured using an indirect calorimetry system (SensorMedics Vmax 29). Subjects then performed the next level of physical demand (ie sitting down for 5 min, and so on) and the procedure was repeated. Exercise was performed on a treadmill (Powerjog).

A flex heart rate (FLEXHR) was calculated for each subject: this was an average of the highest average heart rate observed during the last minute of resting variables (laying, sitting, standing), and the lowest average heart rate during the last minute of exercise variables +10.21

FLEXHR=((highest heart rate at rest+lowest heart rate during exercise)/2)+10

An average value of resting energy expenditure (calculated from indirect calorimetry) from the three levels of resting was calculated for each subject (REST). A regression of energy expenditure (calculated from indirect calorimetry) and heart rate during the four levels of exercise was calculated for each subject.

Daily energy expenditure was calculated for each subject from 24 h heart rate by substituting any heart rate<FLEXHR with an average resting value of energy expenditure (REST) (ie beats per minute (bpm) converted to kcal/min). The regression was applied to any heart rate>FLEXHR and the calculated value of energy expenditure (kcal/min) substituted for the value of heart rate.

If the whole day period contained more than 10% missing data then that day was excluded from the main analysis. Any missing heart rate values (eg due to momentary loss of heart rate telemetry signal or interference from electrical signals) were replaced by mean values taken from 5 min either side of the missing value. Heart rate values which were spuriously high (above 200 bpm) or low (below 30 bpm) were removed and were replaced by mean values taken from 5 min either side of the missing value. This formulation resulted in 1440 values (ie minutes in a day) of energy expenditure, the sum of which was a calculated value of 24 h energy expenditure. Sedentary activity was defined as the percentage of TEE accounted for by REST values.

Data analysis

Data are presented as means±s.e.m. Heart rates were averaged over 30 min periods (ie 48 periods per day) and analysed using ANOVA with time as the within-subjects factor and high- or low-fat consumer group as the between-subjects factor. Individual differences in heart rate at a given time period, dietary variables, activity variables and daily energy expenditure were analysed using independent t-tests. All analysis was performed by SPSS for Windows program 6.1 (SPSS Inc., USA).


Results

Dietary patterns

Details of the 30 recruited subjects, 15 high-fat and 15 low-fat consumers, can be seen in Table 1. The macronutrient and energy intake variables were derived from analysis of the FFQ. Significant differences were found between the groups on fat intake, percentage fat intake and percentage CHO intake: the high- and low-fat consumers met the criteria for inclusion within the groups. This pattern of consumption was confirmed from the analysis of the 3 day food diaries completed in eight HF and eight LF¾there were no significant differences between dietary variables measured by FFQ and food diary.

Twenty-four-hour heart rate

Following the exclusion of missing heart rate data >10%, mean heart rate profiles were calculated for each group (n=33 and 35 days for HF and LF, respectively) and can be seen in Figure 1. ANOVA revealed a main effect of group for the whole day period (F[1,66]=6.77, P=0.01). The mean heart rate over 24 h was significantly higher in HF than in LF: 74.7±1.3 and 69.6±1.5 bpm in HF and LF, respectively (mean±s.e., t=2.6, P=0.01). During the night/sleeping period the profiles clearly separate, with HF having a consistently higher heart rate than LF. Independent t-tests revealed that HF had a significantly higher heart rate at time points (as indicated in Figure 1) 23, 28-39, 42 and 44.

Twenty-four-hour energy expenditure

The total calculated daily energy expenditure, heart rate method (HR: n=15 HF, n=15 LF) and factorial method (F: n=7 HF, n=7 LF), can be seen in Table 2. LF had a higher estimated TEE than HF for both methods of assessment, but these differences were not significant. However, HF were more sedentary; using the heart rate method, a significantly greater percentage of TEE was taken up by sedentary activities (ie) resting: heart rate<flex; t=2.77, P<0.01). Similarly, while using the factorial method a significantly greater percentage of TEE was also expended in sedentary activities (ie activities reported as lying down and sitting down, t=3.13, P<0.01).

TEE values were significantly higher than estimates of energy intake; however, it must be stressed that the FFQ used in this study was not designed to measure energy intake and dietary assessment methods may be unsuitable to assess energy intake of diets constituting the extreme ends of dietary fat intake.22 The FFQ is validated only for percentage macronutrient intake.

Baecke activity questionnaire

There were no significant differences between HF and LF for the total Baecke score or the three components: work, leisure and sports. However LF scores were consistently higher than HF (Table 3).


Discussion

The results from this study have indicated how energy balance might be achieved by different combinations of dietary variables, physical activity and physiological factors. The behaviour of the high-fat phenotype contains variables with the potential to promote a positive energy balance and weight gain¾the consumption of a high-fat diet and related changes in appetite control12 along with a tendency for increased sedentariness. Taken together, these two risk factors (high-fat diet and sedentariness) suggest that weight gain is likely to occur. However, the high-fat phenotype was indistinguishable from the low-fat phenotype on all anthropometrical measures. Indeed, within this cohort of subjects the high-fat phenotype had a slightly lower body mass index than the low-fat phenotype. Therefore, despite certain lifestyle risk factors these young adult male HF appear to be in energy balance and protected from weight gain. How is this occurring?

One possibility could be that the physiological system responds to the consumption of a high-fat, energy-dense, obesigenic diet and sedentariness by increasing basal metabolism to match energy expenditure with energy intake; evidence of this has been found in other studies. A high basal metabolic rate, low respiratory quotient (high fat oxidation) and high plasma leptin have been reported previously in the high-fat phenotype.13,14 The phenomenon of a high basal metabolism in the high-fat phenotype is also apparent in the current study from the profiles of 24 h heart rate recording, particularly during the night/ sleeping phase of the day. Overall HF had a significantly higher mean heart rate than LF, despite being less physically active. During the sleeping hours the heart rate in HF was approximately 15% higher than the values for LF subjects. This finding is consistent with the concept of a high basal metabolism in high-fat eaters. The markedly different heart rate during sleeping is one of the clearest distinctive features of HF and LF. This difference could be due to endogenous physiological activation or to increased restlessness during sleep.

It must be emphasized however, that this study was cross-sectional; it is quite possible that a physiological increase in metabolism is insufficient to maintain energy balance on high-fat diets, therefore the high-fat phenotypes may be slowly gaining weight and would become obese over a number of years.23 This is supported by the observation that obesity clearly increases with advancing age,24 which is partly due to declining resting metabolic rate (RMR) and decreasing rate of fat oxidation with increasing age.25 However, in contrast to the idea that a habitual high-fat diet leads to physiological adjustments, it should be considered that the reverse of this 'adaptive response' hypothesis could also be true. This would mean that a naturally occurring (genetic) high level of basal metabolism 'causes' an individual to eat a more energy dense diet and to become less active. This theory is not favoured as even short-term physiological adaptations to a high-fat diet have been reported in studies in animals26 and in humans.27

Paradoxically, low-fat phenotypes who are consuming a prudent diet could also be at risk of weight gain. Genetic studies have shown that both a relatively low basal metabolic rate (BMR) and a high RQ (typical of low-fat phenotypes) are recognized as metabolic risk factors for weight gain,28,29 although it remains to be determined whether the observation seen in LF is a dietary or genetic effect. In any case LF will be prone to fat deposition (and weight gain) in response to any periodic high-fat eating episodes as it takes 3-5 days of a constant high-fat diet for fat oxidation to match fat intake.30 However, in addition to the consumption of a low-fat diet, low-fat phenotypes report greater amounts of physical activity. Although the doubly labelled water method is the gold standard of TEE measurement in free-living conditions,31 both the estimation of TEE from heart rate and factorial methods have been validated and shown to be reasonably accurate, particularly for group comparisons.16,32,33 No significant difference in TEE was observed between the groups for any method of measurement, although there is a trend towards a higher TEE value in the low-fat phenotype. However, even identical TEE values would indicate more physical activity in low-fat phenotypes due to their lower BMR, which would result in a higher physical activity level (PAL) (TEE/BMR ratio, an indicator of physical activity) than in high-fat phenotypes. More significant was the calculated proportion of TEE accounted for by sedentary activity, which was consistently higher in high-fat phenotypes. The additional activity in low-fat phenotypes is not apparent from initial examination of heart rate profiles because of their low resting heart rate value. However, low-fat phenotypes raise average daytime heart rate to the level of high fat phenotype values, despite having a 15% lower heart rate during rest. Consistently higher scores in the Baecke activity questionnaire confirm that low-fat phenotypes are more physically active than high-fat phenotypes. This may indicate that LF are more physically fit than HF, which would offer an explanation for lower resting heart rate values in LF; however, preliminary studies in our laboratory to examine VO2max (direct measurement and estimation) in HF and LF have found no group differences.

One other interesting feature observed from examination of Figure 1 is the occasional peaks in heart rate observed in high-fat phenotypes during the daytime, whereas in low-fat phenotypes the heart rate profile remains fairly constant. These peaks may represent increased dietary induced thermogenesis (DIT) or increased meal-related activity since they occur at approximately 09:00, 13:00 and 18:00, which are standard meal times in the UK. Such a difference in DIT between the groups could be another example of high-fat phenotypes being more wasteful with energy than low-fat phenotypes. This hypothesis is consistent with the concept of 'energy sparing' and 'energy profligate' individuals as described by Goldberg15 in marginally and very well nourished women, respectively.

In summary it would appear that high- and low-fat phenotypes attain energy balance by differing combinations of behavioural and physiological variables. The behavioural profile of the high-fat phenotype is a risk factor for a positive energy balance but balance appears to be achieved by physiological increases in basal energy expenditure. In contrast the physiological profile of the low-fat phenotype is itself a risk factor for weight gain, while the behaviour (diet and activity) are protective. It remains to be determined if these mechanisms are endogenous or adaptations to a habitual diet. It also needs to be determined whether these mechanisms occur in women and whether they endure with age.