In industrialised countries, most people regulate their energy expenditure poorly. Individual energy expenditure may differ up to 20-fold between resting conditions and high physical activity, but such differences have until now been weakly correlated to energy intake at subsequent meals [
The body’s own physiological signaling system is hunger. Blood glucose concentration (BG) is a reliable index of energy availability to body cells [
Subjects can be trained to predict when BG is low by attending to their subjective experience of hunger [
In this study, we tested the hypothesis that the IHMP is associated with improvements in metabolic biomarkers, in particular insulin sensitivity.
The Paediatric Gastroenterology Unit of Florence University recruited 143 subjects to this study from 1996 to 2000. This unit diagnoses and treats celiac disease in children and adults. Aged 18 to 60 years, subjects suffered from symptoms of functional bowel disorders such as dyspepsia, abdominal pain, and diarrhoea (Figure
The trained group continued their regular work or recreational activities under tutorial assistance for seven weeks and maintained the IHMP for a further three months independently (Figure
Subjects were trained in the IHMP, first by identifying IH, which was guided by consistency in subjective sensations and the association of these sensations with BG measurement. During training, subjects measured capillary blood by portable glucometer (Glucocard Memory; Menarini Diagnostics; Florence, Italy) in the 15 min before a meal. Accuracy of measurements by the glucometer was validated against periodic measurements by hospital autoanalyzer. Seven-day home diaries reported BG measurements and presence or absence of IH before the three main meal times. Also recorded in the diary were energy and vegetable intake, hours in bed, and hours spent during physical and outdoor activities (weekly mean and SD). Subjects were advised that BG measurements after taking small quantities of food (even a few grams), after changes in ambient temperature, after physical activity such as walking or cycling, and when under psychological stress would be misleading since we had previously found that BG and IH do not correlate well under these conditions [
Subjects reported IH as gastric pangs, sensations of emptiness and hollowness, and mental or physical weakness [
120 subjects who completed the study were assessed for blood parameters at baseline (before training), after the first 7 weeks of training, and at the end of the investigation after a further three months (total duration of the investigation: 5 months). During the glucose tolerance test, after a 12-hour overnight fast, all subjects were given a 75 g-oral glucose load. Venous blood samples were taken immediately before glucose was administered, and 30, 60, 90, 120, and 180 min thereafter to determine plasma glucose and serum insulin. Serum insulin was measured with the IMx insulin assay (Abbott Lab. Diagn. Div. USA) [
The primary endpoint was the change in insulin sensitivity [
Analyses were also performed on beta cell function [
Previous work in similar patients found that the insulin sensitivity index in the intervention group was greater by 3 than that in the control group, with a standard deviation (SD) of 3.0 [
A list was divided into blocks of 1 to 4 places, and the blocks were randomly assigned using Armitage even and odds random numbers on a 3 : 1 ratio to either training or control groups. A dietician kept the list and subsequently assigned each recruited subject to the first empty list place. Control or training destination was revealed after the first visit (Figure
Values are expressed as means
A training effect and correlations between the two body size parameters (weight and BMI), the two energy-balance parameters (arm and skinfold thickness), the four metabolic indexes (mean BG and HbA1c values, and BG and insulin AUCs), and three intake factors (energy, fruit, and vegetable) were longitudinally investigated (i.e., on post minus predifferences) by simple, linear correlation and regression analyses in all of the 120 subjects completing the study (Figure
Figure
In this study the protocol was to follow the IHMP. We do not have data on the extent to which IH was present pre-meal for each meal, that is, we do not know how closely each subject adhered to the IHMP. Achieving the IHMP appeared to be difficult for 12 subjects who had high pretraining mean BG concentrations (e.g., around 100 mg/dL) or participated in heavy manual labour, especially in cold conditions. Although some subjects may not have been faithful to the IHMP for all meals, we have included all those who completed the study in the final analysis, since it was our intention to treat them [
Twenty-three subjects (18 trained and 5 control) did not complete the study (dropouts). All were contacted by telephone. Their given reasons were that they “required no further training” or had “busy schedules.” To ascertain whether these biases could have affected the generalisability of the study’s conclusions, we performed a sensitivity analysis using baseline and 7-week data from all 23 dropouts. The 18 trained dropouts significantly decreased mean BG (from
Since no significant gender difference in baseline mean BG concentrations was observed in the control group (females:
Baseline values of mean age, school education years, body weight, BMI, arm and leg skinfold thickness, and blood values did not significantly differ between control and trained groups (Tables
Group composition and effects of training on anthropometry.
Control | Trained | |||
Baseline | After 5 mo. | Baseline | After 5 mo. | |
Number of subjects and Gender | 14 F + 17 M | 46 F + 43 M | ||
Schooling (years)1 | ||||
Age (years)1 | ||||
BMI | ||||
Weight (Kg) | ||||
Arm skinfold thickness (mm) | ||||
Leg skinfold thickness (mm) |
Values are expressed as means
Effects of training on metabolic and intake parameters.
Control | Trained | |||
Baseline | After 5 mo. | Baseline | After 5 mo. | |
Mean pre-meal BG (mg/dL) | ||||
BG diary SD (mg/dL)1 | ||||
Glycated Hb (%) | ||||
Insulin AUC2 (mU L-13 h-1) | ||||
Insulin peak (mU L-1) | ||||
Insulin sens. (index)3 | ||||
BG AUC (mg/dL) | ||||
BG peak (mg/dL) | ||||
Energy intake (Cal/d) | ||||
Meals per day4 | ||||
Vegetable intake (g/d) | ||||
Fruit intake (g/d) |
1Diary SD refers to BG SD of 21 measurements reported by each of 7 d diary.
2AUC: area under GTT curve.
3Whole body insulin sensitivity index [
4Meal was an event of higher energy intake than 20 kcal.
Values are expressed as mean
Significant decreases among trained subjects compared to controls were found in insulin sensitivity index, insulin and BG peaks, insulin at 60 minutes and 90 minutes during GTT, glycated haemoglobin, mean pre-meal BG, BG diary standard deviation (SD), energy intake, BMI, body weight, arm and leg skinfold thickness.
Index of beta cell function changed from
A significant decrease of preprandial BG mean values achieved during training was maintained three months after the training period ceased (baseline:
The absolute pre/post change (increase or decrease) in 31 control subjects was
Trained subjects reported few negative effects when adjusting their food intake and in accommodating irregular intermeal intervals in the first few days of trial and error. The reported adverse effects included a slightly depressed BG (below 60 mg/dL (3.3 mmol/l)) and weakness or abdominal pain.
The high number of dropouts is an important limitation of this study. However, from our sensitivity analysis, we conclude that the dropout subjects are unlikely to represent a significantly different population with respect to the endpoint measures of this study and that the absence of final data from these subjects is unlikely to have significantly affected the overall results.
Our findings are from subjects who attended a gastroenterology clinic over a 5-month period. Further investigation will be necessary to evaluate the effect of the IHMP in other populations and what “reminder” training might be necessary to ensure compliance with the IHMP over years.
A seven-week training program to establish the IHMP led to significant decreases in insulin sensitivity index, insulin and BG peaks, glycated haemoglobin, mean pre-meal BG and BG diary SD. Energy intake, BMI, and body weight also significantly decreased.
IH may represent an important afferent arm of a physiological regulation mechanism that provides meal-by-meal feedback on energy need thus optimizing energy intake. The observed improved insulin sensitivity may reflect lowered energy intake resulting from the IHMP.
Before training, mean pre-meal BG showed high intersubject variability, in agreement with other authors’ findings. This variability has engendered a perception that BG has no relevance to food intake regulation [
We suggest the IHMP offers a viable alternative to low fat and low carbohydrate diets [
The ramifications of improved insulin sensitivity extend well beyond glucose homoeostasis [
Our data suggest that (i) IH provides meal-by-meal feedback allowing the conscious formation of a new eating pattern (IHMP) and sustained self-regulation of energy intake, and (ii) over a five-month period the IHMP is associated with improvement in insulin sensitivity, LBG, HbA1c, and other cardiovascular risk factors.
These findings, together with those of an associated study on weight [
Initial hunger meal pattern
Area under curve
Body mass index
Blood glucose concentration
Oral glucose tolerance test
Low blood glucose
Mean pre-meal blood glucose standard deviation reported by seven day diary
C reactive protein.
The authors thank Laura Chiesi and Stefania Bini MD for dietary analyses and Stephen Buetow, Tim Kenealy, Chris Harshaw, Simon Thornton, Kent Berridge, James Gibbs, Charlotte Erlanson-Albertsson, and Michael Hermanussen for helpful insights on earlier drafts of this paper. This research was supported by the Italian Ministry of University, Research, Science and Technology grants for the years 1998–2002 and ONLUS Nutrizione e Prevenzione, Firenze for years 2003–2008. The authors declare that they have no competing interests.