The reliability and validity of a short food frequency questionnaire among 9-11-year olds: a multinational study on three middle-income and high-income countries

OBJECTIVE

The main aim of this study was to assess the reliability and validity of a food frequency questionnaire with 23 food groups (I-FFQ) among a sample of 9-11-year-old children from three different countries that differ on economical development and income distribution, and to assess differences between country sites. Furthermore, we assessed factors associated with I-FFQ's performance.

METHODS

This was an ancillary study of the International Study of Childhood Obesity, Lifestyle and the Environment. Reliability (n=321) and validity (n=282) components of this study had the same participants. Participation rates were 95% and 70%, respectively. Participants completed two I-FFQs with a mean interval of 4.9 weeks to assess reliability. A 3-day pre-coded food diary (PFD) was used as the reference method in the validity analyses. Wilcoxon signed-rank tests, intraclass correlation coefficients and cross-classifications were used to assess the reliability of I-FFQ. Spearman correlation coefficients, percentage difference and cross-classifications were used to assess the validity of I-FFQ. A logistic regression model was used to assess the relation of selected variables with the estimate of validity. Analyses based on information in the PFDs were performed to assess how participants interpreted food groups.

RESULTS

Reliability correlation coefficients ranged from 0.37 to 0.78 and gross misclassification for all food groups was <5%. Validity correlation coefficients were below 0.5 for 22/23 food groups, and they differed among country sites. For validity, gross misclassification was <5% for 22/23 food groups. Over- or underestimation did not appear for 19/23 food groups. Logistic regression showed that country of participation and parental education were associated (P⩽0.05) with the validity of I-FFQ. Analyses of children's interpretation of food groups suggested that the meaning of most food groups was understood by the children.

CONCLUSION

I-FFQ is a moderately reliable method and its validity ranged from low to moderate, depending on food group and country site.

Associations of maximal strength and muscular endurance with cardiovascular risk factors

The aim was to study the associations of maximal strength and muscular endurance with single and clustered cardiovascular risk factors. Muscular endurance, maximal strength, cardiorespiratory fitness and waist circumference were measured in 686 young men (25±5 years). Cardiovascular risk factors (plasma glucose, serum high- and low-density lipoprotein cholesterol, triglycerides, blood pressure) were determined. The risk factors were transformed to z-scores and the mean of values formed clustered cardiovascular risk factor. Muscular endurance was inversely associated with triglycerides, s-LDL-cholesterol, glucose and blood pressure (β=-0.09 to - 0.23, p<0.05), and positively with s-HDL cholesterol (β=0.17, p<0.001) independent of cardiorespiratory fitness. Muscular endurance was negatively associated with the clustered cardiovascular risk factor independent of cardiorespiratory fitness (β=-0.26, p<0.05), whereas maximal strength was not associated with any of the cardiovascular risk factors or the clustered cardiovascular risk factor independent of cardiorespiratory fitness. Furthermore, cardiorespiratory fitness was inversely associated with triglycerides, s-LDL-cholesterol and the clustered cardiovascular risk factor (β=-0.14 to - 0.24, p<0.005), as well as positively with s-HDL cholesterol (β=0.11, p<0.05) independent of muscular fitness. This cross-sectional study demonstrated that in young men muscular endurance and cardiorespiratory fitness were independently associated with the clustering of cardiovascular risk factors, whereas maximal strength was not.

Effect of supervised and home exercise training on bone mineral density among breast cancer patients. A 12-month randomised controlled trial

The ability of combined step aerobic- and circuit-training to prevent bone loss after breast cancer treatments was related to skeletal site and patients' menopausal status. Among premenopausal breast cancer survivors, a 12-month exercise intervention completely prevented bone loss at the femoral neck, whereas no exercise effect was seen at lumbar spine or at neither site in postmenopausal women.

INTRODUCTION

The primary objective of this randomised clinical trial was to determine the preventive effect of supervised weight-bearing jumping exercises and circuit training on bone loss among breast cancer patients.

METHODS

Of 573 breast cancer survivors aged 35-68 years randomly allocated into exercise or control group after adjuvant treatments, 498 (87%) were included in the final analysis. The 12-month exercise intervention comprised weekly supervised step aerobic- and circuit-exercises and similar home training. Bone mineral density (BMD) at lumbar spine and femoral neck were measured by dual-energy X-ray absorptiometry. Physical performance was assessed by 2-km walking and figure-8 running tests, and the amount of physical activity was estimated in metabolic equivalent-hours/week.

RESULTS

In premenopausal women, bone loss at the femoral neck was prevented by exercise, the mean BMD changes being -0.2% among the trainees vs. -1.4% among the controls (p = 0.01). Lumbar bone loss could not be prevented (-1.9% vs. -2.2%). In postmenopausal women, no significant exercise-effect on BMD was found either at the lumbar spine (-1.6% vs. -2.1%) or femoral neck (-1.1% vs. -1.1%).

CONCLUSIONS

This 12-month aerobic jumping and circuit training intervention completely prevented femoral neck bone loss in premenopausal breast cancer patients, whereas no effect on BMD was seen in postmenopausal women.

Physical activity, fitness and fatness: relations to mortality, morbidity and disease risk factors. A systematic review

The purpose of this systematic review was to study the relative health risks of poor cardio-respiratory fitness (or physical inactivity) in normal-weight people vs. obesity in individuals with good cardio-respiratory fitness (or high physical activity). The core inclusion criteria were: publication year 1990 or later; adult participants; design prospective follow-up, case-control or cross-sectional; data on cardio-respiratory fitness and/or physical activity; data on BMI (body mass index), waist circumference or body composition; outcome data on all-cause mortality, cardiovascular disease mortality, cardiovascular disease incidence, type 2 diabetes or cardiovascular and type 2 diabetes risk factors. Thirty-six publications filled the criteria for inclusion. The data indicate that the risk for all-cause and cardiovascular mortality was lower in individuals with high BMI and good aerobic fitness, compared with individuals with normal BMI and poor fitness. In contrast, having high BMI even with high physical activity was a greater risk for the incidence of type 2 diabetes and the prevalence of cardiovascular and diabetes risk factors, compared with normal BMI with low physical activity. The conclusions of the present review may not be applicable to individuals with BMI > 35.

Body composition in psychotic disorders: a general population survey

BACKGROUND

The literature suggests an association between obesity and schizophrenia but fat mass and fat-free mass, which have been shown to be more predictive of all-cause mortality than only waist circumference and obesity [body mass index (BMI) 30 kg/m2], have not been reported in psychotic disorders. We examined the detailed body composition of people with different psychotic disorders in a large population-based sample.

METHOD

We used a nationally representative sample of 8082 adult Finns aged 30 years with measured anthropometrics (height, weight, waist circumference, fat percentage, fat-free mass and segmental muscle mass). Psychiatric diagnoses were based on a consensus procedure utilizing the Structured Clinical Interview for DSM-IV (SCID)-interview, case-notes and comprehensive register data.

RESULTS

Schizophrenia (including schizo-affective disorder) was associated with obesity [odds ratio (OR) 2.3, 95% confidence interval (CI) 1.5-3.6], abdominal obesity (waist circumference 88 cm for women, 102 cm for men) (OR 2.2, 95% CI 1.3-3.6) and with higher fat percentage (mean difference 3.8%, 95% CI 2.0-5.7%), adjusted for age and gender, than in the remaining sample. The associations between schizophrenia and low fat-free mass and decreased muscle mass on trunk and upper limbs became statistically significant after adjusting for BMI. After further adjusting for current antipsychotic medication, education, diet and smoking, schizophrenia remained associated with obesity (OR 1.9, 95% CI 1.1-3.6) and abdominal obesity (OR 3.8, 95% CI 1.5-9.4). Participants with affective psychoses did not differ from the general population.

CONCLUSIONS

Individuals with schizophrenia have metabolically unfavorable body composition, comprising abdominal obesity, high fat percentage and low muscle mass. This leads to increased risk of metabolic and cardiovascular diseases.

Maintenance of exercise-induced benefits in physical functioning and bone among elderly women

SUMMARY

This study showed that about a half of the exercise-induced gain in dynamic balance and bone strength was maintained one year after cessation of the supervised high-intensity training of home-dwelling elderly women. However, to maintain exercise-induced gains in lower limb muscle force and physical functioning, continued training seems necessary.

INTRODUCTION

Maintenance of exercise-induced benefits in physical functioning and bone structure was assessed one year after cessation of 12-month randomized controlled exercise intervention.

METHODS

Originally 149 healthy women 70-78 years of age participated in the 12-month exercise RCT and 120 (81%) of them completed the follow-up study. Self-rated physical functioning, dynamic balance, leg extensor force, and bone structure were assessed.

RESULTS

During the intervention, exercise increased dynamic balance by 7% in the combination resistance and balance-jumping training group (COMB). At the follow-up, a 4% (95% CI: 1-8%) gain compared with the controls was still seen, while the exercise-induced isometric leg extension force and self-rated physical functioning benefits had disappeared. During the intervention, at least twice a week trained COMB subjects obtained a significant 2% benefit in tibial shaft bone strength index compared to the controls. A half of this benefit seemed to be maintained at the follow-up.

CONCLUSIONS

Exercise-induced benefits in dynamic balance and rigidity in the tibial shaft may partly be maintained one year after cessation of a supervised 12-month multi-component training in initially healthy elderly women. However, to maintain the achieved gains in muscle force and physical functioning, continued training seems necessary.

Influence of weight reduction on muscle performance and bone mass, structure and metabolism in obese premenopausal women

Muscle performance, body composition and bone mass at the lumbar spine and proximal femur with DXA, structural traits at the tibia and radius with pQCT, and biomarkers of bone metabolism were assessed at baseline and after a three-month weight reduction in obese premenopausal women. Associations between changes in weight loss and bone traits were analyzed by linear regression analysis. The mean (SD) weight loss was 4.3 (4.5) kg ranging from 14.8 kg loss to 2.1 kg gain. Muscle performance was well maintained, while no signs of bone loss or structural deterioration were observed. Changes in bone resorption were significantly associated with weight change (for CTX, r=-0.34; p=0.043, and for TRACP5b, r=-0.35; p=0.032). There were borderline (p<0.1) negative correlations between changes in biomarkers and bone traits. Reduced fat mass was associated with slight mean increase in cortical density of the radial shaft. Also total body BMC increased slightly. Changes in both fat and lean mass were associated with a change in BMC. Our findings suggest that mild-to-moderate weight reduction modulated bone turnover slightly, but they do not support the common notion that such a weight reduction would compromise bone rigidity, possibly partly due to well maintained muscle performance.

Dietary Intake, Serum Hormones, Muscle Mass and Strength During Strength Training in 49 - 73-Year-Old Men

Effects of dietary intake on serum hormones, muscle cross-sectional area (CSA) and strength during strength training were studied in two groups of men: 1) strength training + nutritional counseling (n = 22, 59.1 +/- 6.1 yrs), and 2) strength training (n = 23, 58.5 +/- 7.1 yrs). Both groups performed strength training twice a week for 21 weeks. Counseling increased carbohydrate (p < 0.01) and fiber intake (p < 0.001) and polyunsaturated/saturated fat-ratio (p < 0.05) and decreased fat intake (p < 0.01). Muscle strength and CSA increased by 16 - 20 % and by 5.4 - 5.9 % in both groups (p < 0.001). Changes in protein content of diet correlated with the changes in the acute postexercise concentrations of total (r = 0.64, p < 0.01) and free testosterone (r = 0.54, p < 0.05) after training in the counseling group. Moreover, changes in the free testosterone responses to heavy-resistance exercise correlated with the increases in the muscle CSA (r = 0.52, p < 0.05) in the counseling group. Serum basal testosterone/sex hormone-binding globulin-ratios correlated with the body mass normalized energy (kJ/kg: r = 0.54, p < 0.001), protein (g/kg: r = 0.42, p < 0.01) and fat (g/kg: r = 0.51, p < 0.01) intake in all participants during the training. The data indicate that protein and fat intake may influence serum testosterone concentrations and that the changes in exercise-induced testosterone responses may contribute to muscle mass development during strength training.

Effects of Strength Training and Reduced Training on Functional Performance and Metabolic Health Indicators in Middle-Aged Men

Changes in muscular fitness and metabolic health indicators were examined in 22 men (57.9 +/- 6.6 years, BMI 24.5 +/- 2.6 kg/m (2)) and 21 control men (58.2 +/- 6.1 years, BMI 25.4 +/- 2.8 kg/m (2)) during two consecutive 21-week periods: 1) whole body progressive strength training (ST: twice a week), and 2) continued reduced training (CRT: 3 ST sessions/2 weeks, n = 17 + 17). After the 21-week ST period, maximal strength of leg extensors increased in the ST group by 19.6 +/- 7.6 % vs. 2.8 +/- 4.4 % (p < 0.001) and also 10-m walking time and 10-step stair-climbing time shortened by - 17.2 +/- 7.6 % vs. 4.1 +/- 3.9 % (p < 0.01) and by - 8.2 +/- 6.8 % vs. - 3.0 +/- 6.8 % (p < 0.05) compared to the controls. Systolic blood pressure (BP) decreased in the ST group by - 4.4 +/- 6.7 % vs. 1.3 +/- 9.5 % (p < 0.05) compared to the controls after the ST period. Muscle strength as well as walking and stair-climbing times remained the same during the CRT. However, the changes in diastolic BP (- 8.9 +/- 8.7 % vs. - 1.0 +/- 6.6 %, p < 0.05) and fasting blood glucose concentrations (0.1 +/- 0.4 mmol/L vs. 0.3 +/- 0.4 mmol/L, p < 0.05) differed between the ST and control groups after the whole 42-week study period. Strength training has positive health effects in aging men by increasing maximal strength and functional performance and by decreasing resting blood pressure and high-intensity reduced strength training can maintain these health benefits.

Sleep-related disturbances and physical inactivity are independently associated with obesity in adults

OBJECTIVE

To study relationships between obesity, physical inactivity and sleep-related disturbances (obstructive sleep apnea (OSA), sleep duration, sleep disturbances concomitant with daytime tiredness) in adults (> or =30 years).

DESIGN

Cross-sectional study with a random population sample.

PARTICIPANTS

A total of 3377 men (mean age 52.3, s.d. 14.8, years) and 4264 women (56.4, s.d. 17.2, years).

MAIN OUTCOME MEASURES

Dependent variables, measured: Waist circumference (WC) and body mass index (BMI). Independent variables, from a detailed interview/questionnaire: probable OSA, other sleep-related disturbances, sleep duration, type and frequency of leisure physical activity. Age, mental health, smoking and education were included in analyses as potential confounders.

RESULTS

In men, OSA and physical inactivity increased likelihood for abdominal obesity (WC > or =102 cm). Physical inactivity also increased, but long (> or =9 h/day) sleep decreased likelihood for abdominal overweight (WC: 94-101 cm) in men. In women, abdominal obesity (WC > or =88 cm) was associated positively with OSA, moderate sleep-related disturbances, and physical inactivity. Education modulated the influence of age on abdominal obesity in both genders. Using BMI as the dependent variable did not change the general information obtained by the model. In addition, abdominal obesity was found to be an independent risk factor also in multivariable models predicting categories of a combined sleep duration and sleep disturbances.

CONCLUSIONS

Sleep duration and sleep-related disturbances are associated with obesity, even after controlling for OSA and physical inactivity. The results support the hypothesis of vicious circle between sleep and obesity.

Physical fitness in adolescents with normal weight and overweight

We studied the associations of overweight (OW, BMI > or =85th percentile) and physical activity (PA) with physical fitness in adolescents. The nationally representative sample was 1120 boys and 1146 girls, aged 15-16 years. Height and weight were self-reported. The level of PA was based on self-reported frequency and duration of sweating during organized and non-organized activity. Fitness was measured by sit-ups, sit-and-reach, five-jump, back-and-forth jumping, ball skills, coordination and endurance shuttle run tests. The fitness index was calculated as the mean of z-scores for individual tests. The prevalence of OW was 17.3% in boys and 11.8% in girls. The main effect of PA (in analysis of variance) on all fitness tests was significant (P< or =0.005). The main effect of OW was significant (P<0.002) for all tests, except for sit-and-reach. According to linear regression models, the association between PA and fitness was stronger than that between OW and fitness. Sit-ups, endurance shuttle-run and fitness index showed the strongest association with PA (standardized beta coefficients 0.31-0.49). OW was not associated with sit-and-reach test (coefficient 0.04) and only weakly with the ball skills test (coefficients -0.10 to -0.12). In conclusion, OW had the most negative association with cardiorespiratory and muscle endurance, and explosive power tests.

Stakeholder appraisal of policy options for responding to obesity in Finland

The aim of the Finnish Policy Options for Responding to the Growing Challenge of Obesity Research Project study was to use a multi-criteria mapping method to assess stakeholder opinions on policies that might have relevance when planning policies targeting obesity prevention. When evaluating policies, criteria relating to 'Positive societal benefits' and 'Additional health benefits' were often considered important, along with the more obvious 'Efficacy in addressing obesity'. Other criteria (like economic factors) were of less importance. All policy options targeting at 'Educational and Research Initiatives' were highly ranked in the analysis by all stakeholders. Policies aimed at improving 'Food-Related Informational Initiatives' were also ranked well on average, although more variation between stakeholders existed. Policies targeting increased physical activity and easier access/provision to healthy foods received also relatively good rankings but with wide variation and doubts in the appraisals. Policies encouraging new technological innovations and institutional reforms were mostly ranked poorly and were not seen as solutions to obesity problem. Irrespective of the stakeholder group, it seems that while traditional educational policies were most highly ranked, several other policies are also viable options in the policy portfolio of obesity prevention.

A multi-component exercise regimen to prevent functional decline and bone fragility in home-dwelling elderly women: randomized, controlled trial

SUMMARY

This study showed that combination of strength, balance, agility and jumping training prevented functional decline and bone fragility in home-dwelling elderly women. The finding supports the idea that it is possible to maintain good physical functioning by multi-component exercise program and thus postpone the age-related functional problems.

INTRODUCTION

This 1-year randomized, controlled exercise intervention trial assessed the effects of two different training programs and their combination on physical functioning and bone in home-dwelling elderly women.

METHODS

One hundred and forty-nine healthy women aged 70-78 years were randomly assigned into: group 1-resistance training (RES), group 2-balance-jumping training (BAL), group 3-combination of resistance and balance-jumping training (COMB), and group 4-controls (CON). Self-rated physical functioning, leg extensor force, dynamic balance, and bone mass and structure were measured.

RESULTS

Self-rated physical functioning improved in the COMB group, but was reduced in the CON group; the mean inter-group difference was 10% (95% CI: 0-22%). Mean increase in the leg extensor force was higher in the RES (14%; 4-25%) and COMB (13%; 3-25%) compared with the CON groups. Dynamic balance improved in the BAL (6%; 1-11%) and in the COMB (8%; 3-12%) groups. There were no inter-group differences in BMC at the proximal femur. In those COMB women who trained at least twice a week, the tibial shaft structure weakened 2% (0-4%) less than those in the CON group.

CONCLUSIONS

Strength, balance, agility, and jumping training (especially in combination) prevented functional decline in home-dwelling elderly women. In addition, positive effects seen in the structure of the loaded tibia indicated that exercise may also play a role in preventing bone fragility.

Preventing excessive weight gain during pregnancy - a controlled trial in primary health care

OBJECTIVE

To investigate whether individual counselling on diet and physical activity during pregnancy can have positive effects on diet and leisure time physical activity (LTPA) and prevent excessive gestational weight gain.

DESIGN

A controlled trial.

SETTING

Six maternity clinics in primary health care in Finland. The clinics were selected into three intervention and three control clinics.

SUBJECTS

Of the 132 pregnant primiparas, recruited by 15 public health nurses (PHN), 105 completed the study.

INTERVENTIONS

The intervention included individual counselling on diet and LTPA during five routine visits to a PHN until 37 weeks' gestation; the controls received the standard maternity care.

RESULTS

The counselling did not affect the proportion of primiparas exceeding the weight gain recommendations or total LTPA when adjusted for confounders. The adjusted proportion of high-fibre bread of the total weekly amount of bread decreased more in the control group than in the intervention group (difference 11.8%-units, 95% confidence interval (CI) 0.6-23.1, P=0.04). The adjusted intake of vegetables, fruit and berries increased by 0.8 portions/day (95% CI 0.3-1.4, P=0.004) and dietary fibre by 3.6 g/day (95% CI 1.0-6.1, P=0.007) more in the intervention group than in the control group. There were no high birth weight babies (>or=4000 g) in the intervention group, but eight (15%) of them in the control group (P=0.006).

CONCLUSIONS

The counselling helped pregnant women to maintain the proportion of high-fibre bread and to increase vegetable, fruit and fibre intakes, but was unable to prevent excessive gestational weight gain.

Self-guided brisk walking training with or without poles: a randomized-controlled trial in middle-aged women

Walking with poles (Nordic walking, NW) has become popular. We compared training responses of brisk walking (W) or NW on cardiorespiratory and neuromuscular fitness. We randomized 121 non-obese sedentary women (aged 50-60) to an NW or W group (NWG, WG), to train 40 min four times weekly for 13 weeks. Intensity was based on subjective perception of exertion. Cardiorespiratory performance was assessed in four levels corresponding to 50%, 65%, 80% and 100% of peak VO(2). Fifty-four NWG and 53 WG subjects completed the study. The mean intensity was about 50% of heart rate (HR) reserve. The baseline peak VO(2) was 25.8 (SD 3.9) mL/min/kg. Both groups improved peak VO(2) similarly (NWG 2.5 mL/min/kg, 95% confidence interval (CI) 1.9-3.3; WG 2.6, CI 1.9-3.3). In the submaximal stages while walking with or without poles, HR and lactate decreased after training in both groups, but the changes were not statistically significantly different between the groups. Of the neuromuscular tests after training, the only significant difference between the groups was in the leg strength in the one-leg squat, favoring WG. In conclusion, both training modes improved similarly health-enhancing physical fitness, and they were feasible and safe.

Successful weight maintenance preserves lower levels of oxidized LDL achieved by weight reduction in obese men

OBJECTIVE

We studied the association between weight maintenance, oxidized low-density lipoprotein (ox-LDL) and other lipoproteins in obese men.

METHODS

A 2-month weight reduction phase (WRP) with a very-low-energy diet was followed by a 6-month weight maintenance period and an unsupervised 2-year follow-up. Ninety men entered and 68 (76%) completed the study. Subjects were analyzed as one group and after division into two subgroups: 20 most successful men in maintaining the lost weight (subgroup 1) and the remaining (n=48) men (subgroup 2). Ox-LDL was measured by quantifying the amount of conjugated dienes in LDL particles.

RESULTS

The mean (+/-s.d.) weight reduction at the end of the WRP (n=68) was 14% (confidence interval (CI) 12.9-14.7%, 14.5+/-4.2 kg, P<0.001). Ox-LDL decreased by 22% (CI 16.9-28.1, 12.3+/-15.4 micromol/l, P<0.001). At the end of the 2-year follow-up, the regain in weight from the end of the WRP was 11% (CI 9.0-12.4, 9.6+/-6.2 kg, P<0.001). The regain in ox-LDL was 30% (CI 18.7-41.2, 8.2+/-15.4 micromol/l, P<0.001). In subgroup 1 vs 2, the respective regains were 3% (CI 0.9-4.2, 2.2+/-3.0 kg, P=0.006) vs 14% (CI 12.7-15.6, 12.9+/-4.0 kg, P<0.001) regarding weight and 9% (2.0+/-6.9 micromol/l, P=NS) vs 39% (CI 23.7-53.9, 11.2+/-17.2 micromol/l, P</=0.001) in ox-LDL.

CONCLUSION

The favorable changes seen in ox-LDL particles and serum lipids during weight reduction could be maintained by keeping the weight reduced, which may indicate decreased risk of atherosclerosis. But weight regain causes a resurge of ox-LDL.

High-fibre, low-fat diet predicts long-term weight loss and decreased type 2 diabetes risk: the Finnish Diabetes Prevention Study

AIMS/HYPOTHESIS

The aim of this study was to investigate the association of dietary macronutrient composition and energy density with the change in body weight and waist circumference and diabetes incidence in the Finnish Diabetes Prevention Study.

SUBJECTS AND METHODS

Overweight, middle-aged men (n=172) and women (n=350) with impaired glucose tolerance were randomised to receive either 'standard care' (control) or intensive dietary and exercise counselling. Baseline and annual examinations included assessment of dietary intake with 3-day food records and diabetes status by repeated 75-g OGTTs. For these analyses the treatment groups were combined and only subjects with follow-up data (n=500) were included.

RESULTS

Individuals with low fat (<median) and high fibre (>median) intakes lost more weight compared with those consuming a high-fat (>median), low-fibre (<median) diet (3.1 vs 0.7 kg after 3 years). In separate models, hazard ratios for diabetes incidence during a mean follow-up of 4.1 years were (highest compared with lowest quartile) 0.38 (95% CI 0.19-0.77) for fibre intake, 2.14 (95% CI 1.16-3.92) for fat intake, and 1.73 (95% CI 0.89-3.38) for saturated-fat intake, after adjustment for sex, intervention assignment, weight and weight change, physical activity, baseline 2-h plasma glucose and intake of the nutrient being investigated. Compared with the low-fat/high-fibre category, hazard ratios were 1.98 (95% CI 0.98-4.02), 2.68 (95% CI 1.40-5.10), and 1.89 (95% CI 1.09-3.30) for low-fat/low-fibre, high-fat/high-fibre, and high-fat/low-fibre, respectively.

CONCLUSIONS/INTERPRETATION

Dietary fat and fibre intake are significant predictors of sustained weight reduction and progression to type 2 diabetes in high-risk subjects, even after adjustment for other risk factors.