Many of us may want to forget how much pavlova we ate on Christmas Day. Now researchers have pinpointed the foods most likely to slip our minds.
Snack-foods such as biscuits, muesli bars and chips top the list, according to a University of Auckland study that placed cameras on 40 people to monitor what they ate and asked them to log their food intake.
It is well established that dietary surveys under-report the quantities people consume because participants want to be seen to be eating healthily.
But the researchers say, in the European Journal of Clinical Nutrition, that up until now there has been little objective evidence on which foods and meals people are likely to under-report.
Images from the cameras, set to shoot at least two pictures a minute, reminded the participants of 265 food or drink items they had consumed but omitted to report in their diaries.
These items added 8 per cent to the self-reported energy intake of the men and 6 per cent for the women.
The greatest number of initially unreported foods or drinks consumed were snack-foods (64 items), followed by 50 instances of eating condiments such as jam or Marmite, most often at breakfast.
When measured by time of day, 75 of the events - again the greatest number - involved "afternoon snacks", which included biscuits and other snack-foods, as well as drinks and fruit.
"Our results suggest that the most commonly under-reported foods that have substantial impact on energy intake are snack foods, condiments and beverages.
"And the meals at which foods are most likely to be under-reported are afternoon snacks and dinner," researchers Luke Gemming and Cliona Ni Mhurchu say in the journal.
Professor Ni Mhurchu said people may be more likely to forget afternoon snacks because they are often consumed with less formality than a sit-down meal.
She said snack-foods tended to be less healthy than staple foods so, either consciously or unconsciously, people may not report them as often as other foods.
The study's findings could improve clinical practice and population health measurement. Knowing the kinds of foods and meals patients misreport could refine the advice dietitians give.
And in national surveys, "we know that under-reporting is a huge problem and that what we are getting a lot of the time is really poor data".
"In a previous study we showed that, despite people's weight increasing over 10 years, they are reporting eating less.
"If we can't accurately estimate what people are consuming, we can't put effective policies and interventions into place that will be able to tackle the problems that we have."