Inter-arm blood pressure differences compared with ambulatory monitoring: a manifestation of the 'white-coat' effect?