Rapid Responses to:

PAPERS: Douglas Lamont, Louise Parker, Martin White, Nigel Unwin, Stuart M A Bennett, Melanie Cohen, David Richardson, Heather O Dickinson, Ashley Adamson, K G M M Alberti, and Alan W Craft Risk of cardiovascular disease measured by carotid intima-media thickness at age 49-51: lifecourse study BMJ 2000; 320: 273-278 [Abstract] [Full text]

Rapid Responses: Submit a response to this article

Rapid Responses published:

Confusion requiring clarification

Martin R Miller   (4 February 2000)

Re: Confusion requiring clarification

Douglas Lamont   (7 February 2000)

Confusion requiring clarification 4 February 2000
Martin R Miller Send response to journal: Re: Confusion requiring clarification	The reader's understanding of this paper is not enhanced by '...socioeconomic position at birth was negatively associated with...' when a higher social class is numerically lower (1) than a lower social class (5). Nothing in the paper indicates conclusively in which way the authors have used social class in their analysis. Table 2 shows a conventional scale 1, 2, 3N, 3M, 4, 5. Their statement therefore suggests that women born into social class 1 have more carotid disease. I doubt if this is what the authors' meant. Can you clarify? Dr. Martin R Miller Senior Lecturer in Respiratory Medicine University Hospital Trust, Selly Oak Hospital, Birmingham B29 6JD Lamont D, et al.. Risk of cardiovascular disease... BMJ 2000; 320: 273-8.
Re: Confusion requiring clarification 7 February 2000
Douglas Lamont, Senior research associate in epidemiology Department of Child Health, University of Newcastle Send response to journal: Re: Re: Confusion requiring clarification	The descriptions attached to each social class in Table 2 of our paper are labels, not values, and were not intended to be interpreted as the numbers used in analysis. Each of the four social class variables used was scaled for analysis such that the highest numeric value was attached to the highest social class, ie social class 1=6, 2=5, 3N=4, 3M=3, 4=2 and 5=1 where social class 3 was split, and 1=5,...,5=1 otherwise. The negative regression coefficients obtained between social class at birth and the outcome measure, combined with the positive associations observed between poor housing conditions at birth and carotid intima-media thickness, then become interpretable as lower socieconomic position conferring higher risk, and vice versa. We could have said "Lower socioeconomic position at birth and lower birthweight were associated with higher values of carotid intima-media thickness....", but felt that it was sufficiently clear that a negative association between socioeconomic position and the outcome measure implied low socioeconomic position, high intima-media thickness and vice versa.