A double-blind, placebo-controlled, parallel design was adopted. The experimental design consisted of an intervention period lasting 87d. During days 1–87, all subjects’ background caffeine intake was standardised at 300mg/d (the equivalent of about three 150 ml cups of coffee) in order to maintain their habit- ual caffeine intake (van Dam & Feskens, 2002; Hulshof et al. 2003; van Boxtel et al. 2003; Kovacs et al. 2004). The subjects consumed a cup of coffee before breakfast (08.00 h), in the morn- ing (10.00h) and in the afternoon (14.00h). The subjects were asked to abstain from other caffeine-rich products such as tea, cola-type soft drinks and energy drinks.
During days 1–3, subjects consumed a standardised energy- balance diet at 100% of EE estimated by means of the ‘body weight’ computer simulation programme, which includes body weight, an average physical activity level of 1·6 and the estimated basal energy expenditure (Westerterp et al. 1995a).