Abstract
Background/Objectives:
Diet cost could influence dietary patterns, with potential health consequences. Assigning a monetary cost to diet is challenging, and there are contrasting methods in the literature. This study compares two methods—a food cost database linked to 4-day diet diaries and an individual cost calculated from household till receipts.
Subjects/Methods:
The Diet and Nutrition Tool for Evaluation (DANTE) had supermarket prices (cost per 100 g) added to its food composition table. Agreement between diet costs calculated using DANTE from food diaries and expenditure recorded using food purchase till receipts for 325 individuals was assessed using correlation and Bland Altman (BA) plots.
Results:
The mean difference between the methods’ estimates was £0.10. The BA showed 95% limits of agreement of £2.88 and -£3.08. Excluding the highest 5% of diet cost values from each collection method reduced the mean difference to £0.02, with limits of agreement ranging from £2.31 to -£2.35. Agreement between the methods was stronger for males and for adults.
Conclusions:
Diet cost estimates using a food price database with 4-day food diaries are comparable to recorded expenditure from household till receipts at the population or group level. At the individual level, however, estimates differed by as much as £3.00 per day. The methods agreed less when estimating diet costs of children, females or those with more expensive diets.
Diet cost could influence dietary patterns, with potential health consequences. Assigning a monetary cost to diet is challenging, and there are contrasting methods in the literature. This study compares two methods—a food cost database linked to 4-day diet diaries and an individual cost calculated from household till receipts.
Subjects/Methods:
The Diet and Nutrition Tool for Evaluation (DANTE) had supermarket prices (cost per 100 g) added to its food composition table. Agreement between diet costs calculated using DANTE from food diaries and expenditure recorded using food purchase till receipts for 325 individuals was assessed using correlation and Bland Altman (BA) plots.
Results:
The mean difference between the methods’ estimates was £0.10. The BA showed 95% limits of agreement of £2.88 and -£3.08. Excluding the highest 5% of diet cost values from each collection method reduced the mean difference to £0.02, with limits of agreement ranging from £2.31 to -£2.35. Agreement between the methods was stronger for males and for adults.
Conclusions:
Diet cost estimates using a food price database with 4-day food diaries are comparable to recorded expenditure from household till receipts at the population or group level. At the individual level, however, estimates differed by as much as £3.00 per day. The methods agreed less when estimating diet costs of children, females or those with more expensive diets.
| Original language | English |
|---|---|
| Pages (from-to) | 1072-1076 |
| Journal | European Journal of Clinical Nutrition |
| Volume | 67 |
| DOIs | |
| Publication status | Published - 11 Sept 2013 |
Bibliographical note
AcknowledgementsThe authors would like to acknowledge Kevin Tarbutt (funded via a Rank Prize Fund vacation studentship) and Edmund Parks, who created and updated the DANTE cost database; and Claire Oyston and Joan Ransley, who created the SNIP till receipt dataset. Thanks also go to an earlier peer reviewer for their comments. Authors MAM and KAT are funded via ESRC/MRC PhD Studentships.