Most chocolate bars, crisps, cakes, pastries and other ‘unhealthy’ snacks have a high content of saturated fat, salt and refined sugars. These dietary factors can then elevate total cholesterol, blood pressure and body weight, the major risk factors for CHD, stroke and diabetes. CVD and diabetes cause over 170 000 deaths in the UK every year(1).
People are therefore putting themselves at risk by snacking on unhealthy foods that are high in salt, saturated fats and sugar. In 2004, a large study found that nine out of every ten people snacked between meals, with 43 % snacking on chocolate, 41 % on crisps, 34 % favouring cakes and pastries, and 21 % consuming pies and sausage rolls(2). In 2005, there were approximately 25 billion snacking occasions in the UK. This represented over 410 snacks per person per year. Furthermore, approximately 6·2 billion snacking occasions each year featured a chocolate bar, sugar confectionary product, bag of crisps or another type of salted snack(3). This commercial market totals over £820 million per annum, and the associated advertising is both extensive and intensive(3).
According to the National Dietary and Nutrition Survey(4), most adults in the UK have a saturated fat intake well above the recommended 10 % of energy intake(4). If current unhealthy snacking (on chocolate and crisps and other unhealthy snacks) could be partially replaced by dried fruit, fresh fruit, natural nuts and seeds, it might substantially decrease total fat intake, perhaps by as much as 10 g/d (equivalent to one small packet of crisps). This could potentially reduce total energy from saturated fats to 10 % or less. Furthermore, if sources of nuts and seeds were chosen wisely (walnuts, brazil nuts, pine nuts, pumpkin seeds or flaxseed), this would also result in a valuable increase in the omega 3 and 6 fat consumption(4). The omega fats have several well-documented health effects, most importantly lowering blood cholesterol levels(Reference Hu, Manson and Willett5).
However, research into the potential public health effects of changing snack behaviour is scarce. A UK study of habitual snackers found that while total daily energy intake did not vary between high- and low-fat snack consumption, percentage of total daily energy intake from fat increased significantly with high-fat snacks from 37 % to 41 %; conversely, low-fat snacks reduced daily fat intake to 33·5 %(Reference Lawton, Delargy, Smith, Hamilton and Blundell6). A US study examined the impact of replacing unhealthy snacks with healthy snacks in vending machines in schools. Initial findings suggest positive reactions to the changes; however, data relating to changes in diet quality are yet to be published(Reference Davee, Blum, Devore, Beaudoin, Kaley, Leiter and Wigand7).
We therefore examined the potential public health impact on CHD and stroke mortality of replacing one ‘unhealthy’ snack with one ‘healthy’ snack per person, per day, across the UK population.
Methods
In the present study, ‘unhealthy’ snacks were defined as chocolate bars, crisps, pies and sausage rolls, cakes and pastries; ‘healthy’ snacks were defined as dried fruit, unsalted nuts, seeds and fresh fruit such as apples, oranges and bananas. Nutritional information (saturated fat, salt, sugar and energy) for different healthy and unhealthy snack products was obtained from McCance & Widdowson’s Food Composition Tables(8) and from the manufacturer’s information (Appendix A). The nutritional values were then transformed from nutritional value per 100 g into average nutritional values per portion (Appendix B).
The expected changes in dietary intake were then calculated, based on one average ‘healthy’ snack replacing one ‘unhealthy snack’, per person, per day, across the whole UK population. The subsequent mean change in total blood cholesterol levels in the general population was estimated using the Keys equation(Reference Keys, Anderson and Grande9). Keys’ predictive equation enables quantification of the effects of fatty acids and dietary cholesterol on plasma cholesterol concentrations, showing that total cholesterol and LDL-cholesterol are increased by saturated fat and decreased by polyunsaturated fat. The Keys equation is: change in serum cholesterol concentration (mmol/l) = 0·031 × (2D SAT − D PUFA) + 1·50
, where D SAT is the change in the percentage of dietary energy derived from saturates, D PUFA is the change in the percentage of dietary energy derived from polyunsaturates and D CHOL is the change in the dietary cholesterol intake.
The effect of changing cholesterol and salt levels on CHD deaths was then calculated using the equations provided by the Law and He meta-analyses, respectively(Reference Law, Wald and Thompson10, Reference He and MacGregor11). The quoted reduction (95 % confidence interval) in CHD mortality per 0·6 mmol/l reduction in total cholesterol was 25 % (15 %, 35 %) for randomised trials and 27 % (23 %, 32 %) for long-term cohort studies(Reference Law, Wald and Thompson10). For stroke deaths, we used a conservative estimate of an overall reduction of 13 % (6 %, 19 %) per 1·0 mmol/l reduction in total cholesterol(Reference Law, Wald and Rudnicka12). The risk ratio for fatal and non-fatal events was very similar(13). A 3 g reduction in daily salt intake would reduce coronary deaths by approximately 11 % and stroke deaths by 13 %(Reference He and MacGregor11).
The estimated reductions in cardiovascular deaths were then tested in a sensitivity analysis(Reference Briggs, Sculpher and Buxton14). Values for deaths were rounded to the nearest zero or five.
Results
Portions of ‘unhealthy’ snacks typically contained 1·0 to 9·1 g of saturated fat (average 5·1 g) and 0·1 to 1·4 g of salt (average 0·56 g). Portions of ‘healthy’ snacks typically contained 0 to 2·5 g of saturated fat (average 0·69 g) and 0 to 0·15 g of salt (average 0·05 g).
The replacement of one ‘unhealthy’ snack a day with one ‘healthy’ snack would result in an average reduction of approximately 4·41 g saturated fat and 0·51 g salt intake per person per day. This would represent a decrease of approximately 11·7 % and 6·7 % in average daily intake, respectively (Table 1).
Change in dietary intake: replacing one ‘unhealthy’ snack with one ‘healthy’ snack
*Source: Office for National Statistics (2005) Expenditure and Food Survey 2003/04. London: The Stationery Office.
*Source of data for sugar, fibre & energy: FSA (2003) The National Diet and Nutrition Survey: Adults Aged 19–64 Years. vol. 2. London: The Stationery Office.
†Assuming 2 1 monounsaturated: polyunsaturated split.
Baseline situation
In 2004, 105 845 CHD deaths and 60 458 stroke deaths were reported in the UK.
Ideal scenario, fat intake: If one unhealthy snack was replaced by one healthy snack, the 4·41 g reduction in saturated fat consumption would lead to a reduction in blood cholesterol levels of approximately 0·054 mmol/l. This 0·054 mmol/l decrease would result in approximately 2400 fewer deaths from CHD per year (minimum estimate 1435, maximum estimate 3355) and 425 fewer stroke deaths (minimum estimate 195, maximum estimate 625) (Table 2).
Reduction in serum cholesterol concentration, and hence in CHD and stroke mortality, with an 11·7 % decrease in saturated fat intake
Ideal scenario, salt intake: If one unhealthy snack was replaced by one healthy snack, the 510 mg decrease in salt intake would result in approximately 1790 fewer coronary deaths (minimum estimate 1155, maximum estimate 2505) and 1330 fewer deaths from strokes (minimum estimate 860, maximum estimate 25501) (Table 3).
Reduction in CHD and stroke with a 0·51 g decrease in dietary salt intake
In total, approximately 6000 cardiovascular deaths a year could be prevented from this minor change to diet (minimum estimate 3650, maximum estimate 7785).
Discussion
Simply replacing one unhealthy snack with one healthy snack per day might prevent approximately 6000 cardiovascular deaths every year. The corresponding reductions in cardiovascular morbidity, obesity and diabetes would also be valuable, particularly given the recent alarming trends(15). These are not trivial benefits. In the USA, the Baltimore Longitudinal Study of Aging recently reported that CHD mortality was approximately 75 % lower among men with both low saturated fat and high fruit and vegetable intake (compared with those with low fruit and vegetable consumption and high saturated fat intake); furthermore, this combination was considerably more protective than either behaviour alone(Reference Tucker, Hallfrisch, Qiao, Muller, Andres and Fleg16).
Cholesterol remains the major reversible risk factor for CVD(1, Reference Hu, Manson and Willett5, Reference Law, Wald and Thompson10, Reference Law, Wald and Rudnicka12). In Finland, a logical combination of national policies and local interventions reduced the population average total cholesterol by over 1 mmol/l, cardiovascular deaths subsequently fell by over 70 %(Reference Laatikainen, Critchley, Vartiainen, Salomaa, Ketonen and Capewell17).
Furthermore, He and McGregor(Reference He and MacGregor11) recently estimated major reductions in CHD and stroke mortality in the UK population simply by decreasing dietary salt. A conservative estimate of salt intake reduction by 3 g/d (from 12 to 9 g/d) could potentially prevent approximately 7800 stroke deaths and 11 500 CHD deaths per year. If salt intake was reduced from the current average intake of 12 to 3 g/d (as in the successful DASH Trial level), this might prevent approximately 20 500 stroke deaths and 31 400 CHD deaths every year(Reference He and MacGregor11).
Yet, persuading large numbers of UK citizens to switch from unhealthy to healthy snacks may not be realistic using health education alone. Such information is easily buried by the immense volume of food industry advertising with an annual budget now exceeding £700 million(18). However, simple legislation may be much more effective; for instance, banning TV junk food adverts prior to 21.00 hours(19).
This first, simple analysis of a complex dietary phenomenon has obvious limitations. Data quality was imperfect. Furthermore, the methodology would clearly benefit from future refinement; for instance, modelling of separate age, sex and socio-economic categories. Consideration might also be given to quantifying the beneficial effect of these same modest dietary changes on obesity, diabetes and relevant cancers. However, having acknowledged these limitations, we would emphasise that these calculations used conservative estimates, because individuals might well change by more than one snack per day. Also, the true population benefit could be even greater, not least because of the additional unquantified harm from trans fats. In conclusion, even small changes to diet (such as choosing healthy snacks) could lead to potentially large reductions in cardiovascular deaths.
Acknowledgements
Sources of funding: Modi Mwatsama received £2000 from Whitworths to support analysis.
Conflicts of interest: The authors have no commercial, personal, political or academic conflicts of interest. Modi Mwatsama received £2000 from Whitworths to support this analysis. All other authors have no financial conflicts of interest.
Author contributions: F.L.-W. participated in the analysis of the data, the literature review and writing and editing of the paper. M.M. participated in obtaining, collating and transforming the nutritional information, analysing the data and editing the paper. R.I. participated in checking the nutritional information, and writing and editing of the paper. S.C. participated in model development and analysis and writing and editing of the paper.
Appendix A
Nutritional value (average value per 100 g) of typical ‘unhealthy’ and ‘healthy’ snacks
Appendix B
Nutritional value (average values per portion) of typical ‘unhealthy’ and ‘healthy’ snacks
