Don’t dismiss the data: a fatter Britain really is consuming fewer calories

Three years ago, a study found that people in Britain drink considerably more than they say they do. It became an international news story even though it is common knowledge among researchers that drinkers greatly under-report their intake.

This week a report found that people in Britain eat considerably more than they say they do. This, too, has received an enormous amount of news coverage despite the fact that anybody who has spent more than five minutes with the data knows that under-reporting is endemic in this area too.

In the case of alcohol, we don’t need surveys to tell us how much we’re drinking because we’ve got the tax receipts. All we have to do is estimate how much has been spilled or abandoned and we can work out average intake with relative ease.

Food is more difficult. We have surveys showing what people say they eat and we have surveys showing what people say they bought, but people cannot be trusted to recall what they have eaten and a lot of food is thrown away. If we took the Living Costs and Food Survey at face value, the average Briton eats only 2,192 calories a day. This is implausible when 25 per cent of us are obese. (The government recommends men and women eat 2,500 and 2,000 calories a day respectively).

All sources are agreed on one point, however: we are eating fewer calories as a fat nation than we did as a slim nation. World War Two rations were designed to give civilians 3,000 calories a day. In the late 1940s, scientists found that people lost weight if they got less than 2,900 calories a day. A survival diet 70 years ago would be an obesogenic diet today.

It is safe to assume that calorie consumption rose when rationing ended, although statistics are thin on the ground until the Living Costs and Food Survey started up in 1974. In its first year, this survey reported that 2,534 calories per day were consumed in the home (it did not look at eating out). By 2012, with eating out now included, this had fallen to 2,192 calories.

The exact numbers found in these surveys are misleading due to under-reporting, but the trend is clear. The Institute for Fiscal Studies has written about these data before, as have I for the Institute of Economic Affairs. Not everybody is convinced, however. The Behavioural Insights Team — commonly known as the Nudge Unit and now privatised — yesterday released a report claiming that ‘calorie consumption has not significantly decreased over time’.

Their argument is that under-reporting is endemic and has got worse in recent decades. They suggest several reasons why people today might be more inclined to forget or lie about what they have eaten, such as the tendency of obese people to under-report more than slim people and the general awareness of obesity as a health issue. These are worthwhile considerations and I agree that they have played a part, but researchers are well aware of them (they are mentioned in both the IFS and IEA reports). In my view, it is very unlikely that they can explain all of the self-reported decline in calorie consumption, let alone that they have masked a rise in calorie consumption. The self-reported decline is just too steep.

With the exception of the mavericks at Public Health Collaboration, nobody seriously believes people are consuming 400 fewer calories than they did in 1974. The question is whether we are eating more calories than they did back then. The Nudge report accepts that we are not. Instead it says that, after correcting for under-reporting, ‘the decline since 1974 is a lot smaller than previously stated — around 200kcal per day compared to around 400kcal’.

This is a significant admission. Given that obesity rates were very low in the 1970s, it leads to the obvious conclusion that a decline in physical activity, not an increase in food intake, has driven the rise in obesity. But this is the opposite of the Nudge Unit’s thesis and so they spend several pages trying to debunk the notion that people are less physically active than they used to be.

This is the weakest part of the report because they resort to a tactic used by our old friend Aseem Malhotra and use leisure time exercise as a proxy for physical activity. Surveys show an increase in the number of people who follow the Chief Medical Officer’s recommendation of exercising for 30 minutes five times a week. The data-set only goes back to 1997 and the Nudgers acknowledge that it contains the same kind of self-reported figures that they have just criticised, but that is only part of the problem. There is no contradiction between a third of the population taking part in regular leisure time exercise and a quarter of the population being obese. They are not the same people.

Moreover, people only exercise in their spare time when they are not getting enough physical activity in their daily lives. People did not need to go to the gym 70 years ago because they had naturally active lifestyles. Sure enough, there are figures tucked away in the appendix of the Behavioural Insights Team’s report showing that men expend 100 fewer calories at work today than they did in the early 1980s. Combine that with the rise in car ownership, the decline of walking and the rise of labour-saving devices and it is easy to see why physical activity has declined by 24 per cent since the 1960s, as Public Health England says it has.

Despite torturing the data, the Nudge Unit team fails to extract a confession. After throwing everything they can at the statistics, the best they can manage is the claim that calorie consumption has not fallen as much in the last 40 years as a naive interpretation of food surveys might suggest. But we are already knew that, just as we know that people drink a lot more than they say they do. The fact remains that calorie consumption has fallen over the long term and so has physical activity. If, as the Nudge Unit admits, we are eating fewer calories than we did in the 1970s, a decline in physical activity is the only possible culprit.

The authors of the report seem to believe that there are competing views on how to tackle obesity with one side claiming that it is futile to reduce calorie consumption because it has failed to reduce obesity in the recent past. I’m not sure that is true. It is certainly not my position. Reducing calorie consumption is as valid an approach as increasing physical activity. Indeed, reducing calorie consumption might be a more realistic option for many people. Whether the government has the ability or mandate to reduce calorie consumption is another issue, but regardless of where you stand on that question we should not rewrite history. The notion that obesity started rising after 1980 because the nation was consuming more and more calories remains a myth.


  • annewareham

    Calories do not all have the same effects. Sugar v fat consumption?

    • lolexplosm

      Calories do have pretty similar effects and the clinical evidence repeatedly agrees with this. Sure there are things like fat increasing satiety to an extent and it’s never a good idea to have too much sugar but in the grand scheme of things too many calories, regardless of macronutrient composition, will cause weight gain. Dietary fat is stored very efficiently as fat, excess carbs are stored as fat. An extra calorie of fat will cause as much weight gain as an extra calorie of carbs/sugar.

      The only people that suggest otherwise are ones that have the latest diet plans and books to sell.

      • fundamentallyflawed

        If you read body building sites you will find that this is both true and untrue… calorie is a unit of energy so 1 cal of anything is theoretically the same (the tonne of feathers argument)
        However to micro manage weight they are very big on managing Insulin sensitivity and sugar plays a large part of this…

        • lolexplosm

          Bodybuilding would be moving into a more extreme form of the argument, they tend to be on much stricter diets and calorie intakes depending on their goals. This is about the general population and your average Joe public so I don’t think bodybuilding is a fair comparison.

          If you are referring to sugar increasing insulin and if this increased insulin leads to resistance, we should consider that protein would also play a large part and bodybuilders eat a lot of protein. Bodybuilders have different approaches too, some say if it fits your macros, or simply calories in calories out, lo carb, paleo etc so who knows who is right in this regard. Carb cycling is one method for helping lose some body fat when already at low-ish levels but that includes all carbs and not restricted to limiting sugar.

  • Stanlycam
    • lolexplosm

      Well it was a small study and a poor one due to its multiple confounding factors such as differences in calorific deficit and weight loss. Then food intake and ratios were uncontrolled, the high carb diet ended up reduced carb intake by about 20%. The data is probably fine but the arguments and conclusions being suggested can’t really be made.

      As an aside, it was partially funded by the Atkins Foundation of which some of the authors are members. These same authors also have several Atkins/low carb diet books. I’m sure that’s irrelevant though.

      • Stanlycam

        Dietary Carbohydrate Restriction Enhances Weight Loss and Reduces Adiposity Out of Proportion to the Caloric Deficit
        Despite similar reductions in calories, weight loss in the CRD group was, on average, twofold greater than in the low-fat control (10.1 kg vs 5.2 kg). This apparent decrease in caloric efficiency with carbohydrate restriction has been observed many times (reviews: [38, 39]), although the current report is one of the more dramatic demonstrations. There was substantial individual variation, but 9 of 20 subjects in the CRD group lost 10% of their starting weight, more than all of the subjects in the LFD group. Indeed, none of the subjects following the LFD lost as much weight as the average weight loss for the experimental group. The number of subjects who lost >5% of body weight was 19 of 20 subjects for the CRD compared to 12 of 20 for the LFD. Despite greater absolute fat intake and similar total caloric intake, whole body fat mass decreased significantly more in subjects following the CRD (5.7 kg) than in subjects following the LFD (3.7 kg) (Table 2). Fat mass in the abdominal region, associated with many features of the insulin resistance syndrome, was similarly decreased significantly more in subjects consuming the CRD than subjects following the LFD (−828 g vs −506 g).
        >>>>>

        • lolexplosm

          Copy and paste the article as much as you like, the criticisms and limitations are still valid.

          They weren’t “similar reductions in calories” at all, the CRD group on “average” decreased calorie intake 20% more than LFD. I say “average” because of the huge ranges involved, the lowest intake was 1700 kcal and the highest was 3000 kcal in the CRD group alone.

          It’s really quite easy to find statistical significances with such a large variation and number of confounding factors.

  • GreenWyvern

    One word: Antibiotics

    Farmers feed antibiotics to animals because it makes them put on weight.

    It’s standard practice, and it’s been used for decades. It works for cattle, sheep, pigs, chickens, and almost any animal.

    If people eat those animals filled with antibiotics… and if people are always being prescribed antibiotics for every minor little illness… they will gain weight too. It’s just plain common sense.

    That’s not to say that we should eliminate antibiotics – nobody wants to die young from some easily curable infection – but we should be very careful to use them only when really necessary. And think about getting them out of the food chain.

    • Patrick Lawless

      …and while the mechanism is poorly understood it is possibly related to less diversity of bacteria in the gut, antibiotics being catch all. Add into the mix fibre light diets, as per the mainstream western diet, which deprive bacteria of a beneficial food source and one can easily speculate we might find there is a third element to the calories/exercise equation which will help fully answer the serious national health crisis knocking on our door.

      Even better the research behind the newly discovered role bacteria plays in our well being is rigorously scientific and (currently) largely devoid of cranks proclaiming miracle cures for the obseity epidemic.

      http://www.scientificamerican.com/article/how-gut-bacteria-help-make-us-fat-and-thin/

      • lolexplosm

        I’m not so sure, the whole alternative scene has proclaimed they’ve been harping on about this for years and allopathic medicine has just caught up. There is a lot of woo involved with microbiomes and healing/treating/prompting it seems to be used as a panacea.

        It certainly is an interesting area of research but more needs to be done to confirm these preliminary findings in my opinion. It could end up changing a lot of things in medicine.

    • lolexplosm

      But food from animals have a withdrawal period so that there is hardly any trace of drug metabolites nevermind the original compound. What reason is there to believe that:

      Animal antibiotics are being passed to humans in strong enough doses to cause an effect?

      That animal antibiotics can have a significant effect on human gut flora?

      That because animals gain weight via use of antibiotics, assuming it is the actual drug and not just a lack of illness, humans will gain weight too?

      Of course resistance, overuse etc are valid concerns otherwise.