- Email: [email protected]
Is extrinsic sugar a vehicle for dietary fat?
Is extrinsic sugar
a
vehicle for
dietary fat?
Summary
Introduction
Although many guidelines healthy eating recommend of extrinsic restriction of the intake (refined) sugar, there are concerns that such restriction might result in an increase in the amount and the proportion of dietary fats with a consequent possible increase in the risk of cardiovascular disease. We used regression analysis to examine the determinants of fat intake in subjects from a population survey who had weighed their food for 4 days. In men (n=77) and women (n=83), fat eaten was positively related to the intake of extrinsic sugar. When intakes were expressed as percent of calories the relation became negative. A survey in a semi-random sample of 739 men aged 40-69 yr and 976 women aged 25-69 yr showed that, in both sexes, an increase in extrinsic sugar was associated with a linear increase in the intake of sweetened fat and hence of fat combined with carbohydrate. This was due mainly to a higher intake of cakes and biscuits. Foods containing sugar and fat provided an extra 12·0 g per day of fat in the men and 13·8 g per day in the women when the highest quartile of extrinsic sugar consumers were compared with the lowest to
quartile. We conclude that lowering the intake of extrinsic sugar is unlikely to be associated with higher fat intake. Instead extrinsic sugar may act as a vehicle for fat intake, encouraging consumption by making the fat more
palatable. Lancet 1995; 345: 1537-40
All recent guidelines to healthy diet have recommended the limitation of fat intake, as highlighted by the UK government in "The Health of the Nation".’ Many guidelines also recommend limiting the intake of added or refined sugars, which consist chiefly of sucrose. Such sugars were renamed "non-milk extrinsic sugars" by the Committee on Medical Aspects of Food (COMA).2 We call them extrinsic sugar for brevity. The two aims of limiting fat and extrinsic sugar have been deemed incompatible on the grounds that, within a population, there is a negative correlation between fat intake expressed as percentage of energy or as g per 1000 kcal and extrinsic sugar intake expressed in the same way.23 The authors of the COMA report2 worried that "reduction in sugar intake might result in an increase in the amount and the proportion of dietary fats with a consequent possible increase in the risk of cardiovascular disease". The fear is that when people heed advice to cut down on sweet foods they will switch to fatty foods in their search for satiety and will end up eating more saturated fat, which would raise their low-densitylipoprotein-cholesterol. These notions, which are clearly a disincentive to reducing extrinsic sugar intake, have been subtly exploited by the sugar industry in the guise of nutrition education.4 Extrinsic sugar is combined with fats in the manufacture and home preparation of many popular delicacies and snacks (cakes, biscuits, pastries, ice creams, chocolate, fudge, candy bars). The popularity of these products remains undiminished despite health warnings about fats; indeed, in children, their intake may even be
increasing. 5,6 We hypothesised that, because of the widespread use of sweet-tasting products containing fat, fat intake and extrinsic sugar intake would be positively correlated in the adult population. We also suspected that consumers of a lot of extrinsic sugar might reject fruit as a source of We determined how fat intakes relate to intake of extrinsic sugar and other factors. We have also examined the contribution of different carbohydratecontaining food groups to fat intake.
sweetness.
Subjects and
methods men aged between
University Department of Medicine, Bristol Royal Infirmary, Bristol BS2 8HW, UK (P M Emmett BSc, K W Heaton FRCP)
Correspondence to:
K W Heaton
We studied 739 40 and 69 years, and 976 women aged between 25 and 69 years, between October, 1987 and March, 1989. All had taken part in an epidemiological survey (focused on gallstones)’ and successfully completed a food-frequency questionnaire. These people represented 88% and 92% of the volunteers who attended the survey and 66% and 65% of those invited. The invitees were a stratified random sample of all the adults in the stated age groups who were registered with nineteen general practitioners in East Bristol. A subsample of 160 subjects (77 men and 83 women) weighed their food and drink for 2 weekdays and 2 weekend days. This
1537
*Significant difference between highest and lowest intakes, p<0001. tRoast potatoes, chips, and crisps. *Boiled, baked or mashed, potatoes. Table 1: Daily intake in men (median) of four major food-groups in which fat is combined with carbohydrate, and of fat contained within these foods
subsample consisted of all subjects with gallstones who consented to further studies (86% of all subjects with gallstones, n=79), together with randomly chosen gallstone-free controls matched for age and sex (n=81). They used an automatic electronic weighing and recording system which minimised effort and error.8 The records were coded by a dietitian and converted by computer into daily intake of dietary components with the values for food composition published in standard tables."’" We calculated the mean daily intake of nutrients by allotting fivesevenths to the mean weekday values and two-sevenths to the mean weekend day-values. We developed the food-frequency questionnaire from a more wide-ranging one devised in the Medical Research Council Epidemiology Unit in Cardiff." Our questionnaire focused on foods and drinks containing carbohydrate and non-starch polysaccharide (dietary fibre) and did not cover all fat-containing foods. It has been validated against 4-day weighed intake records in the subjects of this study.For extrinsic sugar, the Pearson correlation coefficient between the two methods was 0-68, with only 5% of subjects classified in the first and third tertiles.11 Although the questionnaire did not cover all sources of fat, it included foods and drinks in which fat and carbohydrate are combined in manufacture or home preparation-cakes and buns, biscuits, chocolates, pastries, puddings, ice creams, milk shakes, potato chips (French fries), crisps, and savoury pastries. All but the last three contain extrinsic sugar. Returned questionnaires were coded and checked by experienced dietitians who entered numbers of portions and food codes into a computer. We were then able to calculate weights of food from portion sizes" and convert the data into mean daily intake of dietary components, including non-starch polysaccharide," by using values in published tables.12 We used forward-stepwise regression analysis, analysing men and women separately. We treated total fat intake as the
Significant difference between highest and lowest intakes, *p<0001. tp<0’0001 tRoast potatoes, chips, and cnsps. §Boiled, baked or mashed, potatoes. Table 2: Daily intake (median) in women of four major foodgroups in which fat is combined with carbohydrate, and of fat contained within these foods
1538
Daily extrinsic sugar
intake in
lOg bands (x 10)
Figure 1: Intake of fat in men in all prepared carbohydrate-containing foods (fat with carbohydrates) and foods containing extrinsic sugar (sweetened fat) and intake of non-starch polysaccharide (dietary fibre) dependent variable and protein, starch, extrinsic sugar, intrinsic (intracellular) sugar, and non-starch polysaccharide as the predictor variables. We controlled for age, body mass index (kg/m2) and gallbladder status. We repeated the regression with fat energy expressed as a percentage of the total energy as the dependent variable. The other energy-providing nutrients were also expressed as a percentage of the total energy and used as predictor variables. In other studies of dietary intake we noted that some people’s energy intakes are lower than their predicted energy expenditure, even though they are neither slimming nor ill." This implies under-recording of food intake. If such under-recording had occurred in our subjects the findings of the study might be biased. We therefore took action to exclude such a bias. We calculated each subjects basal energy expenditure from their weight and age" and added 30% to allow for gentle physical activity. The resulting value was deemed to be the minimum plausible energy expenditure of that subject. Subjects in whom this value was more than their recorded energy intake were labelled as possible under-recorders. We repeated stepwise regression analysis with the label of possible under-recording as a dummy variable.
Results
Weighed intake data Mean fat intakes were 98 g per day, ranging from 35 to 182 in the men and 71 g per day (29-154) in the women. These values are similar to those obtained by similar methodology in a random sample of the British
population-the Adults’ Survey. 17 When daily intake of fat was the dependent variable, in men the predictors entered the regression analysis in the following order: protein (g), extrinsic sugar (g) and bodymass index (kg/m2). All gave positive associations (coefficients). 0-95, 0-18, and 1-68, respectively. The total variance predicted was 40-0%. Other variables were not significantly associated. In women starch was entered at the first step, then protein followed by extrinsic sugar. All gave positive associations (0-26, 0-74, and 0-23, respectively). Non-starch polysaccharide (g), when entered in the final step, gave a negative coefficient of 0-82. The total variance predicted was 58-0%. When we allowed for possible under-recording in men the association of fat intake with extrinsic intake became non-significant. Under-recording sugar itself was responsible for a substantial reduction in fat intake (33-4 g per day in men and 19-4 g per day in women). Under-recorders were heavier than the others and
women
The relation between extrinsic sugar intake and intake of fat combined with carbohydrate was linear at all ages. The same was true when the analysis was limited to foods containing extrinsic sugar (sweetened fat). The results for all age groups were combined and are shown in figures 1 and 2. Sweetened fat was largely responsible for the rise in consumption of carbohydrate-associated fat. Non-starch polysaccharide, used to represent the intake of healthy carbohydrate foods, did not increase as the intake of extrinsic sugar increased, nor did the intake of vitamin C
(data not shown). Discussion Daily extrinsic sugar intake in lOg bands (x 10)
Figure 2: Intake of fat in women in all prepared carbohydrate-containing foods (fat with carbohydrates) and foods containing extrinsic sugar (sweetened fat) and intake of non-starch polysaccharide (dietary fibre)
(in men, median body-mass index 27-7 25-8, p<0-001; in women, 26-2 vs 23-3, p<0-001). With energy from fat as the dependent variable, in men the predictors entered the regression analysis in the following order: percentage of energy from extrinsic sugar, percentage of energy from starch, percentage of energy from protein (all negative associations); age and body-mass index (both positive); and percentage of energy from extrinsic sugar (negative). Total variance was 58-6%. In women, energy from intrinsic sugars, starch, extrinsic sugars, and protein were all negatively associated and entered the equation in that order. No other associations were found and the total variance predicted was 54-7%. When we allowed for a possibility of underrecording, there was little difference to the prediction of percentage energy from fat. data Extrinsic sugar intakes varied between 5 and 380 g per day in men (median 81) and between 1 and 295g per day in women (median 52). The sources of extrinsic sugar were: table sugar and preserves (median 36% in men and 15% in women), non-alcoholic beverages (9% and 15%),
Questionnaire
confectionery (6% and 9%), cakes, pastries, and biscuits (10% and 13%), and puddings including ice cream (7% and 10%). The last three sources provided fat as well. Intakes of fat combined with carbohydrate of all kinds varied from 3 to 117 g per day in men (median 35) and from 2 to 126 g per day in women (median 28). Half of this fat was taken in sweetened form (median 46% in men and 49% in women). Tables 1 and 2 show the contribution of different foodgroups to fat intake in men and women when divided into groups according to their extrinsic sugar consumption. As sugar intake rose, so did the intake of all four major foodgroups in which carbohydrate is combined with fat. Foods containing sugar and fat provided an extra 12-0 g per day of fat in men and an extra 13-8 g per day in women in the highest versus the lowest quartile of extrinsic sugar consumption. By contrast, the extrinsic sugar intake rose, the intake of potatoes without fat and of vegetables remained stable whereas fruit intake fell. The last three food-groups were included as markers of healthy eating.
Our findings need to be confirmed in a randomly selected population as our subjects were not randomly selected and so cannot be assumed to be representative. However, our subjects had a wide distribution of fat intakes, similar to that of the general population.17 We found a consistent association of increased fat intake with increasing protein intake. This finding was predictable because of the fatty nature of many proteinrich foods. In the group as a whole, an increased fat intake was also associated with an increased extrinsic sugar intake. This was also predictable, in our view, because of the sugar content of many fat-rich foods. The latter association became non-significant when adjustment was made for possible under-recording of food intake. This may be because of the usual nature of under-recording, which can affect sugar and fat more than other elements of the diet.18 Such under-recording is prevalent in overweight people’S’8,’9 and the possible under-recorders in our study were much more likely to be overweight. People who are overweight generally know the foods they "should" avoid, especially cakes, biscuits, and chocolate, and the underrecording selectively involves these foods. Thus had no under-recording occurred in our study, the positive association between sugar and fat intakes may have remained intact. The questionnaire data (obtained on a near-random sample of the study population) showed that a significant element of fat intake was inseparable from extrinsic sugar intake, because of the widespread consumption of cakes, biscuits, pudding, and confectionery. Such fat and sugar combinations provided an extra 12-14 g fat in consumers of large amounts of extrinsic sugar. This fat is mostly saturated and usually contains trans fatty acids because hydrogenated oils are used in many manufactured products.20,21 Our data also showed that higher intakes of extrinsic sugar are associated with lower intakes of fruit (tables 1 and 2) and with static vegetable and non-starch polysaccharide intakes (figures 1 and 2). Together these findings suggest an overall deterioration in the quality of the diet in heavy users of extrinsic sugar. In any event, our subjects who recorded eating cakes, biscuits, and confectionery were taking in substantial amounts of fat that had been made palatable by being sweetened with sugar. We suggest that this fat-or most of it-would not otherwise have been eaten. Our multiple regression analysis confirmed that when intakes are expressed as percent of energy, then fat and extrinsic sugar are inversely related. However, the meaning of such a relation is questionable. At a given intake of energy, an increase in the percentage of one energy-giving nutrient must inevitably be at the expense of another such nutrient. The inclusion of under1539
recording in the analysis did not affect these relations. This implies that sugar and fat are co-variable, which supports our thesis. Our findings showed that it is unlikely that when people are advised to reduce their extrinsic sugar intakes they will increase the amount of fat eaten. Indeed the contrary may be true: extrinsic sugars may act as a vehicle for fat intake. study was supported by Grants from the South West Regional Health Authority and the Kellog company of Great Britain. We thank Mrs C L Symes, Mrs H A Cripps, and Mrs F E M Braddon for help. This
9
10
11
12
13
References
14
1 Anon. The health of the nation: strategy for health in London: HM Stationery Office, 1992.
England.
Panel on Dietary Sugars. Dietary sugars and human disease. London: HM Stationery Office, Reports on Health and Social Subjects, no 37, 1989. 3 Gibney MJ. Dietary guidelines: a critical appraisal. J Hum Nutr Diet 1980; 3: 245-54. 4 Anon. The fats of life. Good Dietary Practice News 1992; 3: 1-3. 5 Rugg-Gunn AJ, Adamson AJ, Appleton DR, Butler TJ, Hackett AF. Sugars consumption by 379 11-12-year-old English children in 1990 compared with results in 1980. J Hum Nutr Diet 1993; 6: 419-31. 6 Adamson AJ, Rugg-Gunn AJ, Appleton DR, Butler TJ, Hackett AF. Dietary sources of energy, protein, unavailable carbohydrate and fat in 11-12-year-old English children in 1990 compared with results in 1980. JHum Nutr Diet 1992; 5: 371-85. 7 Heaton KW, Braddon FEM, Mountford RA, Hughes AO, Emmett PM. Symptomatic and silent gallstones in the community. Gut 1991; 32: 316-20. 8 Barker ME, McKenna PG, Reid NG, et al. A comparison of the Petra food recording system with a conventional weighed inventory technique. J Hum Nutr Diet 1988; 1: 179-86. 2
15
16
17
18
19 20 21
Paul AA, Southgate DAT. McCance and Widdowson’s the composition of foods, 4th ed. London: HM Stationery Office, 1978. Wiles SJ, Nettleton PA, Black AE, Paul AA. The nutrient intake of some cooked dishes eaten in Britain: a supplementary food composition table. J Hum Nutr 1980; 34: 189-223. Yarnell JWG, Fehily AM, Millbank JE, Sweetnam PM, Walker CL. A short dietary questionnaire for use in an epidemiological survey: comparison with weighed dietary records. J Hum Nutr Diet 1983; 37A: 103-12. Emmett P, Symes C, Braddon F, Heaton K. Validation of a new questionnaire for assessing habitual intake of starch, non-starch polysaccharides, sugars and alcohol. J Hum Nutr Diet 1992; 5: 245-54. Emmett PM, Symes CL, Heaton KW. Dietary intakes and sources of non-starch polysaccharide in English men and women. Eur J Clin Nutr 1993; 47: 20-30. Black AE, Goldberg SA, Jebb SA, et al. Critical evaluation of energy intake data using fundamental principles of energy physiology 2: evaluating the results of published surveys. Eur J Clin Nutr 1991; 45: 583-99. Pryer JA. Vrijheid M, Nichols R, Elliot P. Who are the ’low energy reporters’ in the Dietary and Nutritional Survey of British Adults? Proc Nutr Soc 1994; 53: 235A. Panel on Dietary Reference Values. Dietary reference values for food energy and nutrients for the United Kingdom. London: HM Stationery Office, Reports on Health and Social Subjects, no 41, 1991. Gregory J, Foster K, Tyler H, Wiseman M. The dietary and nutritional survey of British adults. London: HM Stationery Office, 1990. Bingham S, Welch A, Cassidy A, Runswick S. The use of 24 h urine nitrogen to detect bias in the reported habitual food intake of individuals assessed from weighed dietary records. Proc Nutr Soc 1991; 50: 32A. Prentice AM, Black AE, Coward WA, et al. High levels of energy expenditure in obese women. BMJ 1986; 292: 983-87. Mann GV. Metabolic consequences of dietary trans fatty acids. Lancet 1994; 343: 1268-71. Willett WC, Ascherio A. Trans fattty acids: are the effects only marginal? Am J Public Health 1994; 84: 722-24.
the first 20 years in the UK. Poor prognostic factors include presence and severity of hypertension, high concentration of protein in urine, and extent of tissue damage at biopsy.’Increased production of angiotensin II over
in enzyme gene and
Polymorphisms
angiotensin-convertingprogression of IgA
Lancet 1995; 345: 1540-42
of glomerulonephritis by mechanisms and growth-promoting haemodynamic effects on glomerular cells.2 Polymorphic regions have been identified in the genes that encode components of the renin-angiotensin system. These polymorphisms may act as markers of activity of this system.3-’ An insertiondeletion polymorphism in intron 16 of the angiotensinconverting enzyme (ACE) gene is associated with variation in circulating levels of ACE.3,4 Individuals homozygous for the deletion (D) allele have the highest plasma enzyme concentrations.4 Although ACE is not thought to be rate-limiting in the production of angiotensin II, associations have been found between the D and allele various conditions,3 including cardiomyopathy,’ myocardial infarction,6 and nephropathy in patients with insulin-dependent diabetes mellitus.’7 These observations are consistent with increased pressor and growth-promoting actions of angiotensin II, although the mechanisms remain to be established. Because angiotensin II causes tubular and mesangial cell proliferation and matrix productionand is also a major determinant of arterial and glomerular hypertension, we have investigated the effects of polymorphisms in the ACE gene on the development and progression of IgA disease.
IgA nephropathy is the commonest glomerulonephritis, leading to end-stage renal failure in about 20% of patients
ACE genotype was established in 100 consecutive patients with biopsy-proven IgA nephropathy who were attending the nephrology, dialysis, and transplantation clinics of the Western
nephropathy
We have investigated the influence of the functional insertion (I) and deletion (D) polymorphism in intron 16 of the gene for angiotensin-converting enzyme (ACE) in a retrospective study of 100 patients with IgA nephropathy. There was no difference in genotype frequency compared with normal subjects. However, patients homozygous for the D allele tended to present at an earlier age (medians: DD, 33; ID, 34; II, 42 years) and to require renal
replacement therapy at a younger age (medians 37, 42, and 48 years, respectively). The rate of progression was significantly worse in patients homozygous for the D allele. The DD genotype is associated with increased severity of disease in patients with IgA nephropathy.
1540
may
cause
progression
a
vehicle for
dietary fat?
Summary
Introduction
Although many guidelines healthy eating recommend of extrinsic restriction of the intake (refined) sugar, there are concerns that such restriction might result in an increase in the amount and the proportion of dietary fats with a consequent possible increase in the risk of cardiovascular disease. We used regression analysis to examine the determinants of fat intake in subjects from a population survey who had weighed their food for 4 days. In men (n=77) and women (n=83), fat eaten was positively related to the intake of extrinsic sugar. When intakes were expressed as percent of calories the relation became negative. A survey in a semi-random sample of 739 men aged 40-69 yr and 976 women aged 25-69 yr showed that, in both sexes, an increase in extrinsic sugar was associated with a linear increase in the intake of sweetened fat and hence of fat combined with carbohydrate. This was due mainly to a higher intake of cakes and biscuits. Foods containing sugar and fat provided an extra 12·0 g per day of fat in the men and 13·8 g per day in the women when the highest quartile of extrinsic sugar consumers were compared with the lowest to
quartile. We conclude that lowering the intake of extrinsic sugar is unlikely to be associated with higher fat intake. Instead extrinsic sugar may act as a vehicle for fat intake, encouraging consumption by making the fat more
palatable. Lancet 1995; 345: 1537-40
All recent guidelines to healthy diet have recommended the limitation of fat intake, as highlighted by the UK government in "The Health of the Nation".’ Many guidelines also recommend limiting the intake of added or refined sugars, which consist chiefly of sucrose. Such sugars were renamed "non-milk extrinsic sugars" by the Committee on Medical Aspects of Food (COMA).2 We call them extrinsic sugar for brevity. The two aims of limiting fat and extrinsic sugar have been deemed incompatible on the grounds that, within a population, there is a negative correlation between fat intake expressed as percentage of energy or as g per 1000 kcal and extrinsic sugar intake expressed in the same way.23 The authors of the COMA report2 worried that "reduction in sugar intake might result in an increase in the amount and the proportion of dietary fats with a consequent possible increase in the risk of cardiovascular disease". The fear is that when people heed advice to cut down on sweet foods they will switch to fatty foods in their search for satiety and will end up eating more saturated fat, which would raise their low-densitylipoprotein-cholesterol. These notions, which are clearly a disincentive to reducing extrinsic sugar intake, have been subtly exploited by the sugar industry in the guise of nutrition education.4 Extrinsic sugar is combined with fats in the manufacture and home preparation of many popular delicacies and snacks (cakes, biscuits, pastries, ice creams, chocolate, fudge, candy bars). The popularity of these products remains undiminished despite health warnings about fats; indeed, in children, their intake may even be
increasing. 5,6 We hypothesised that, because of the widespread use of sweet-tasting products containing fat, fat intake and extrinsic sugar intake would be positively correlated in the adult population. We also suspected that consumers of a lot of extrinsic sugar might reject fruit as a source of We determined how fat intakes relate to intake of extrinsic sugar and other factors. We have also examined the contribution of different carbohydratecontaining food groups to fat intake.
sweetness.
Subjects and
methods men aged between
University Department of Medicine, Bristol Royal Infirmary, Bristol BS2 8HW, UK (P M Emmett BSc, K W Heaton FRCP)
Correspondence to:
K W Heaton
We studied 739 40 and 69 years, and 976 women aged between 25 and 69 years, between October, 1987 and March, 1989. All had taken part in an epidemiological survey (focused on gallstones)’ and successfully completed a food-frequency questionnaire. These people represented 88% and 92% of the volunteers who attended the survey and 66% and 65% of those invited. The invitees were a stratified random sample of all the adults in the stated age groups who were registered with nineteen general practitioners in East Bristol. A subsample of 160 subjects (77 men and 83 women) weighed their food and drink for 2 weekdays and 2 weekend days. This
1537
*Significant difference between highest and lowest intakes, p<0001. tRoast potatoes, chips, and crisps. *Boiled, baked or mashed, potatoes. Table 1: Daily intake in men (median) of four major food-groups in which fat is combined with carbohydrate, and of fat contained within these foods
subsample consisted of all subjects with gallstones who consented to further studies (86% of all subjects with gallstones, n=79), together with randomly chosen gallstone-free controls matched for age and sex (n=81). They used an automatic electronic weighing and recording system which minimised effort and error.8 The records were coded by a dietitian and converted by computer into daily intake of dietary components with the values for food composition published in standard tables."’" We calculated the mean daily intake of nutrients by allotting fivesevenths to the mean weekday values and two-sevenths to the mean weekend day-values. We developed the food-frequency questionnaire from a more wide-ranging one devised in the Medical Research Council Epidemiology Unit in Cardiff." Our questionnaire focused on foods and drinks containing carbohydrate and non-starch polysaccharide (dietary fibre) and did not cover all fat-containing foods. It has been validated against 4-day weighed intake records in the subjects of this study.For extrinsic sugar, the Pearson correlation coefficient between the two methods was 0-68, with only 5% of subjects classified in the first and third tertiles.11 Although the questionnaire did not cover all sources of fat, it included foods and drinks in which fat and carbohydrate are combined in manufacture or home preparation-cakes and buns, biscuits, chocolates, pastries, puddings, ice creams, milk shakes, potato chips (French fries), crisps, and savoury pastries. All but the last three contain extrinsic sugar. Returned questionnaires were coded and checked by experienced dietitians who entered numbers of portions and food codes into a computer. We were then able to calculate weights of food from portion sizes" and convert the data into mean daily intake of dietary components, including non-starch polysaccharide," by using values in published tables.12 We used forward-stepwise regression analysis, analysing men and women separately. We treated total fat intake as the
Significant difference between highest and lowest intakes, *p<0001. tp<0’0001 tRoast potatoes, chips, and cnsps. §Boiled, baked or mashed, potatoes. Table 2: Daily intake (median) in women of four major foodgroups in which fat is combined with carbohydrate, and of fat contained within these foods
1538
Daily extrinsic sugar
intake in
lOg bands (x 10)
Figure 1: Intake of fat in men in all prepared carbohydrate-containing foods (fat with carbohydrates) and foods containing extrinsic sugar (sweetened fat) and intake of non-starch polysaccharide (dietary fibre) dependent variable and protein, starch, extrinsic sugar, intrinsic (intracellular) sugar, and non-starch polysaccharide as the predictor variables. We controlled for age, body mass index (kg/m2) and gallbladder status. We repeated the regression with fat energy expressed as a percentage of the total energy as the dependent variable. The other energy-providing nutrients were also expressed as a percentage of the total energy and used as predictor variables. In other studies of dietary intake we noted that some people’s energy intakes are lower than their predicted energy expenditure, even though they are neither slimming nor ill." This implies under-recording of food intake. If such under-recording had occurred in our subjects the findings of the study might be biased. We therefore took action to exclude such a bias. We calculated each subjects basal energy expenditure from their weight and age" and added 30% to allow for gentle physical activity. The resulting value was deemed to be the minimum plausible energy expenditure of that subject. Subjects in whom this value was more than their recorded energy intake were labelled as possible under-recorders. We repeated stepwise regression analysis with the label of possible under-recording as a dummy variable.
Results
Weighed intake data Mean fat intakes were 98 g per day, ranging from 35 to 182 in the men and 71 g per day (29-154) in the women. These values are similar to those obtained by similar methodology in a random sample of the British
population-the Adults’ Survey. 17 When daily intake of fat was the dependent variable, in men the predictors entered the regression analysis in the following order: protein (g), extrinsic sugar (g) and bodymass index (kg/m2). All gave positive associations (coefficients). 0-95, 0-18, and 1-68, respectively. The total variance predicted was 40-0%. Other variables were not significantly associated. In women starch was entered at the first step, then protein followed by extrinsic sugar. All gave positive associations (0-26, 0-74, and 0-23, respectively). Non-starch polysaccharide (g), when entered in the final step, gave a negative coefficient of 0-82. The total variance predicted was 58-0%. When we allowed for possible under-recording in men the association of fat intake with extrinsic intake became non-significant. Under-recording sugar itself was responsible for a substantial reduction in fat intake (33-4 g per day in men and 19-4 g per day in women). Under-recorders were heavier than the others and
women
The relation between extrinsic sugar intake and intake of fat combined with carbohydrate was linear at all ages. The same was true when the analysis was limited to foods containing extrinsic sugar (sweetened fat). The results for all age groups were combined and are shown in figures 1 and 2. Sweetened fat was largely responsible for the rise in consumption of carbohydrate-associated fat. Non-starch polysaccharide, used to represent the intake of healthy carbohydrate foods, did not increase as the intake of extrinsic sugar increased, nor did the intake of vitamin C
(data not shown). Discussion Daily extrinsic sugar intake in lOg bands (x 10)
Figure 2: Intake of fat in women in all prepared carbohydrate-containing foods (fat with carbohydrates) and foods containing extrinsic sugar (sweetened fat) and intake of non-starch polysaccharide (dietary fibre)
(in men, median body-mass index 27-7 25-8, p<0-001; in women, 26-2 vs 23-3, p<0-001). With energy from fat as the dependent variable, in men the predictors entered the regression analysis in the following order: percentage of energy from extrinsic sugar, percentage of energy from starch, percentage of energy from protein (all negative associations); age and body-mass index (both positive); and percentage of energy from extrinsic sugar (negative). Total variance was 58-6%. In women, energy from intrinsic sugars, starch, extrinsic sugars, and protein were all negatively associated and entered the equation in that order. No other associations were found and the total variance predicted was 54-7%. When we allowed for a possibility of underrecording, there was little difference to the prediction of percentage energy from fat. data Extrinsic sugar intakes varied between 5 and 380 g per day in men (median 81) and between 1 and 295g per day in women (median 52). The sources of extrinsic sugar were: table sugar and preserves (median 36% in men and 15% in women), non-alcoholic beverages (9% and 15%),
Questionnaire
confectionery (6% and 9%), cakes, pastries, and biscuits (10% and 13%), and puddings including ice cream (7% and 10%). The last three sources provided fat as well. Intakes of fat combined with carbohydrate of all kinds varied from 3 to 117 g per day in men (median 35) and from 2 to 126 g per day in women (median 28). Half of this fat was taken in sweetened form (median 46% in men and 49% in women). Tables 1 and 2 show the contribution of different foodgroups to fat intake in men and women when divided into groups according to their extrinsic sugar consumption. As sugar intake rose, so did the intake of all four major foodgroups in which carbohydrate is combined with fat. Foods containing sugar and fat provided an extra 12-0 g per day of fat in men and an extra 13-8 g per day in women in the highest versus the lowest quartile of extrinsic sugar consumption. By contrast, the extrinsic sugar intake rose, the intake of potatoes without fat and of vegetables remained stable whereas fruit intake fell. The last three food-groups were included as markers of healthy eating.
Our findings need to be confirmed in a randomly selected population as our subjects were not randomly selected and so cannot be assumed to be representative. However, our subjects had a wide distribution of fat intakes, similar to that of the general population.17 We found a consistent association of increased fat intake with increasing protein intake. This finding was predictable because of the fatty nature of many proteinrich foods. In the group as a whole, an increased fat intake was also associated with an increased extrinsic sugar intake. This was also predictable, in our view, because of the sugar content of many fat-rich foods. The latter association became non-significant when adjustment was made for possible under-recording of food intake. This may be because of the usual nature of under-recording, which can affect sugar and fat more than other elements of the diet.18 Such under-recording is prevalent in overweight people’S’8,’9 and the possible under-recorders in our study were much more likely to be overweight. People who are overweight generally know the foods they "should" avoid, especially cakes, biscuits, and chocolate, and the underrecording selectively involves these foods. Thus had no under-recording occurred in our study, the positive association between sugar and fat intakes may have remained intact. The questionnaire data (obtained on a near-random sample of the study population) showed that a significant element of fat intake was inseparable from extrinsic sugar intake, because of the widespread consumption of cakes, biscuits, pudding, and confectionery. Such fat and sugar combinations provided an extra 12-14 g fat in consumers of large amounts of extrinsic sugar. This fat is mostly saturated and usually contains trans fatty acids because hydrogenated oils are used in many manufactured products.20,21 Our data also showed that higher intakes of extrinsic sugar are associated with lower intakes of fruit (tables 1 and 2) and with static vegetable and non-starch polysaccharide intakes (figures 1 and 2). Together these findings suggest an overall deterioration in the quality of the diet in heavy users of extrinsic sugar. In any event, our subjects who recorded eating cakes, biscuits, and confectionery were taking in substantial amounts of fat that had been made palatable by being sweetened with sugar. We suggest that this fat-or most of it-would not otherwise have been eaten. Our multiple regression analysis confirmed that when intakes are expressed as percent of energy, then fat and extrinsic sugar are inversely related. However, the meaning of such a relation is questionable. At a given intake of energy, an increase in the percentage of one energy-giving nutrient must inevitably be at the expense of another such nutrient. The inclusion of under1539
recording in the analysis did not affect these relations. This implies that sugar and fat are co-variable, which supports our thesis. Our findings showed that it is unlikely that when people are advised to reduce their extrinsic sugar intakes they will increase the amount of fat eaten. Indeed the contrary may be true: extrinsic sugars may act as a vehicle for fat intake. study was supported by Grants from the South West Regional Health Authority and the Kellog company of Great Britain. We thank Mrs C L Symes, Mrs H A Cripps, and Mrs F E M Braddon for help. This
9
10
11
12
13
References
14
1 Anon. The health of the nation: strategy for health in London: HM Stationery Office, 1992.
England.
Panel on Dietary Sugars. Dietary sugars and human disease. London: HM Stationery Office, Reports on Health and Social Subjects, no 37, 1989. 3 Gibney MJ. Dietary guidelines: a critical appraisal. J Hum Nutr Diet 1980; 3: 245-54. 4 Anon. The fats of life. Good Dietary Practice News 1992; 3: 1-3. 5 Rugg-Gunn AJ, Adamson AJ, Appleton DR, Butler TJ, Hackett AF. Sugars consumption by 379 11-12-year-old English children in 1990 compared with results in 1980. J Hum Nutr Diet 1993; 6: 419-31. 6 Adamson AJ, Rugg-Gunn AJ, Appleton DR, Butler TJ, Hackett AF. Dietary sources of energy, protein, unavailable carbohydrate and fat in 11-12-year-old English children in 1990 compared with results in 1980. JHum Nutr Diet 1992; 5: 371-85. 7 Heaton KW, Braddon FEM, Mountford RA, Hughes AO, Emmett PM. Symptomatic and silent gallstones in the community. Gut 1991; 32: 316-20. 8 Barker ME, McKenna PG, Reid NG, et al. A comparison of the Petra food recording system with a conventional weighed inventory technique. J Hum Nutr Diet 1988; 1: 179-86. 2
15
16
17
18
19 20 21
Paul AA, Southgate DAT. McCance and Widdowson’s the composition of foods, 4th ed. London: HM Stationery Office, 1978. Wiles SJ, Nettleton PA, Black AE, Paul AA. The nutrient intake of some cooked dishes eaten in Britain: a supplementary food composition table. J Hum Nutr 1980; 34: 189-223. Yarnell JWG, Fehily AM, Millbank JE, Sweetnam PM, Walker CL. A short dietary questionnaire for use in an epidemiological survey: comparison with weighed dietary records. J Hum Nutr Diet 1983; 37A: 103-12. Emmett P, Symes C, Braddon F, Heaton K. Validation of a new questionnaire for assessing habitual intake of starch, non-starch polysaccharides, sugars and alcohol. J Hum Nutr Diet 1992; 5: 245-54. Emmett PM, Symes CL, Heaton KW. Dietary intakes and sources of non-starch polysaccharide in English men and women. Eur J Clin Nutr 1993; 47: 20-30. Black AE, Goldberg SA, Jebb SA, et al. Critical evaluation of energy intake data using fundamental principles of energy physiology 2: evaluating the results of published surveys. Eur J Clin Nutr 1991; 45: 583-99. Pryer JA. Vrijheid M, Nichols R, Elliot P. Who are the ’low energy reporters’ in the Dietary and Nutritional Survey of British Adults? Proc Nutr Soc 1994; 53: 235A. Panel on Dietary Reference Values. Dietary reference values for food energy and nutrients for the United Kingdom. London: HM Stationery Office, Reports on Health and Social Subjects, no 41, 1991. Gregory J, Foster K, Tyler H, Wiseman M. The dietary and nutritional survey of British adults. London: HM Stationery Office, 1990. Bingham S, Welch A, Cassidy A, Runswick S. The use of 24 h urine nitrogen to detect bias in the reported habitual food intake of individuals assessed from weighed dietary records. Proc Nutr Soc 1991; 50: 32A. Prentice AM, Black AE, Coward WA, et al. High levels of energy expenditure in obese women. BMJ 1986; 292: 983-87. Mann GV. Metabolic consequences of dietary trans fatty acids. Lancet 1994; 343: 1268-71. Willett WC, Ascherio A. Trans fattty acids: are the effects only marginal? Am J Public Health 1994; 84: 722-24.
the first 20 years in the UK. Poor prognostic factors include presence and severity of hypertension, high concentration of protein in urine, and extent of tissue damage at biopsy.’Increased production of angiotensin II over
in enzyme gene and
Polymorphisms
angiotensin-convertingprogression of IgA
Lancet 1995; 345: 1540-42
of glomerulonephritis by mechanisms and growth-promoting haemodynamic effects on glomerular cells.2 Polymorphic regions have been identified in the genes that encode components of the renin-angiotensin system. These polymorphisms may act as markers of activity of this system.3-’ An insertiondeletion polymorphism in intron 16 of the angiotensinconverting enzyme (ACE) gene is associated with variation in circulating levels of ACE.3,4 Individuals homozygous for the deletion (D) allele have the highest plasma enzyme concentrations.4 Although ACE is not thought to be rate-limiting in the production of angiotensin II, associations have been found between the D and allele various conditions,3 including cardiomyopathy,’ myocardial infarction,6 and nephropathy in patients with insulin-dependent diabetes mellitus.’7 These observations are consistent with increased pressor and growth-promoting actions of angiotensin II, although the mechanisms remain to be established. Because angiotensin II causes tubular and mesangial cell proliferation and matrix productionand is also a major determinant of arterial and glomerular hypertension, we have investigated the effects of polymorphisms in the ACE gene on the development and progression of IgA disease.
IgA nephropathy is the commonest glomerulonephritis, leading to end-stage renal failure in about 20% of patients
ACE genotype was established in 100 consecutive patients with biopsy-proven IgA nephropathy who were attending the nephrology, dialysis, and transplantation clinics of the Western
nephropathy
We have investigated the influence of the functional insertion (I) and deletion (D) polymorphism in intron 16 of the gene for angiotensin-converting enzyme (ACE) in a retrospective study of 100 patients with IgA nephropathy. There was no difference in genotype frequency compared with normal subjects. However, patients homozygous for the D allele tended to present at an earlier age (medians: DD, 33; ID, 34; II, 42 years) and to require renal
replacement therapy at a younger age (medians 37, 42, and 48 years, respectively). The rate of progression was significantly worse in patients homozygous for the D allele. The DD genotype is associated with increased severity of disease in patients with IgA nephropathy.
1540
may
cause
progression