Abstract
Introduction: The Diogenes study investigated the effect of protein and glycaemic index (GI) on weight maintenance after ‡8% weight loss at eight centres across Europe. This study explores how dietary patterns (DP) 4 weeks into the intervention predict weight change at 6 months.
Method: DP were analysed for participants who completed 6 months intervention and 3-day food diaries at week 4 (n = 422, 66% women; BMI before weight loss = 34 ± 5 kg/m2 ). Reduced rank regression was applied to reported intake of 44 food groups with protein, GI, fat and fibre as response variables. Associations between DP and weight change were analysed using multiple regression models, adjusting for age, sex, family structure, initial weight loss, diet arm, total energy intake and study centre.
Results: DP1 was positively loaded (Pos) for fruit, vegetables, high- fibre bread and cereals and negatively loaded (Neg) for high fat cheese and yoghurt. Higher score predicted greater weight regain when adjusting for confounders (P = 0.004), but not when adding study centre to the model (P = 0.130). Adjusted for all confounders DP2 (Pos: meat, fish; Neg: sugar and honey, fats) was not associated with weight change. Higher DP3-scores (Pos: red meat, bread, potatoes; Neg: fruit, pulses, yoghurt, pasta) predicted more weight regain (P = 0.004), whereas higher DP4-scores (Pos: high-fibre bread, nuts, cheese; Neg: low fiber bread, low fat milk) predicted less weight regain (P = 0.036).
Conclusion: DP analyses identified combinations of foods that were related to weight regain and which may inform dietary guidelines on weight maintenance.
Conflict of interest: None disclosed.
Funding: Research relating to this research was funded by the EU contract nr: Food -2005-CT-513946.
Method: DP were analysed for participants who completed 6 months intervention and 3-day food diaries at week 4 (n = 422, 66% women; BMI before weight loss = 34 ± 5 kg/m2 ). Reduced rank regression was applied to reported intake of 44 food groups with protein, GI, fat and fibre as response variables. Associations between DP and weight change were analysed using multiple regression models, adjusting for age, sex, family structure, initial weight loss, diet arm, total energy intake and study centre.
Results: DP1 was positively loaded (Pos) for fruit, vegetables, high- fibre bread and cereals and negatively loaded (Neg) for high fat cheese and yoghurt. Higher score predicted greater weight regain when adjusting for confounders (P = 0.004), but not when adding study centre to the model (P = 0.130). Adjusted for all confounders DP2 (Pos: meat, fish; Neg: sugar and honey, fats) was not associated with weight change. Higher DP3-scores (Pos: red meat, bread, potatoes; Neg: fruit, pulses, yoghurt, pasta) predicted more weight regain (P = 0.004), whereas higher DP4-scores (Pos: high-fibre bread, nuts, cheese; Neg: low fiber bread, low fat milk) predicted less weight regain (P = 0.036).
Conclusion: DP analyses identified combinations of foods that were related to weight regain and which may inform dietary guidelines on weight maintenance.
Conflict of interest: None disclosed.
Funding: Research relating to this research was funded by the EU contract nr: Food -2005-CT-513946.
| Original language | English |
|---|---|
| Article number | T4:PO. 14 |
| Pages (from-to) | 303 |
| Number of pages | 1 |
| Journal | Obesity Reviews |
| Volume | 11 |
| Issue number | S1 |
| DOIs | |
| Publication status | Published - Jul 2010 |