The relationship between anthropometric indices and occurrence of BCC has been little studied. To our knowledge, only two studies have investigated BCC in relation to anthropometric measures or indices [1
]. The study by Milan et al
. (2003) did not measure important potential confounders including phenotypic characteristics or sun exposure history, and although information on these potential confounders was collected in the study by Sahl et al
. (1995), they did not consider them in their analyses which was limited to comparing the mean body weight of a small number of cases and controls (49 and 46 respectively). Although the current study was also not designed specifically to examine the association between anthropometric measures and BCC, information on a large number of other risk factors, including phenotypic characteristics and sun exposure history, was collected and these risk factors were assessed as possible confounders in the relationship between BCC and anthropometric measures.
In the present large prospective study, there were no significant associations between any of the anthropometric measures or indices and BCC, though a suggestive positive association between short term weight gain and development of BCC in women was observed. This implies that a stable body weight may be more favourable with respect to the development of BCC then short-term changes in body weight in this study population. It is unclear why weight gain might be associated with the development of BCC. Weight gain is a consequence of higher energy intake than energy expenditure over time, and availability or restriction of energy can modulate the cellular replication process. There is evidence from studies of different animal models that suggests caloric restriction inhibits cell proliferation [27
] and carcinogenesis [28
]. One study in human subjects also supported this hypothesis [10
]. These data have been further supported by recent epidemiological studies that demonstrate the importance of energy balance and obesity in human cancer [3
The association between weight gain and BCC, if confirmed by other studies, may be mediated via alterations in the metabolism of endogenous hormones (sex steroids, insulin and insulin-like growth factors) (that lead to a change in the normal balance between cell proliferation, differentiation and apoptosis [11
]), or through reduction in immunity [14
], although this hypothesis is based on results of animal studies.
For weight and BMI at baseline, no association was observed. Some research has suggested that adult weight gain may be a better variable to assess adiposity and its metabolic consequences than body weight itself [33
], since weight gain largely reflects an increase in body fat, whereas weight reflects both lean and fat body mass. Height seemed to be associated inconsistently with an increased risk of developing BCC, but the point estimates were not significant, and there were no significant tests for trend. Tallness has been associated with increased risk of a number of other cancers, including cancers of the breast, colon, and prostate [3
] and the same biological mechanisms as described above for weight gain have been proposed to explain these associations. It may also be possible that there is some behavioural correlate of body height related to sun exposure, such as participation in outdoor sporting activities [34
], that increases the likelihood of developing BCC.
Strengths of this study include its large population sample with a high participation rate as well as its prospective design, and the ability to assess a large number of other risk factors as possible confounders of the association between anthropometric measures and BCC. Standardised methods for measuring weight and height, and comprehensive data on BCC incidence from 1992 -1996 were also features of this investigation. Our large prospective study did not have the limitations of the previous two studies. The case-control study by Sahl et al. (1995) was based on a very small number of clinic-based cases of BCC (less than 50), and thus it was limited in power, susceptible to selection bias due to non-representative cases and controls and to faulty recall of past weight and height measurements. Indeed both previous studies relied upon self-reported weight and height measurements. In the second study by Milan et al. (2003), subjects were drawn from a twin cohort and a single self-reported weight was the basis of their assessment of BMI in relation to incident BCC over the next 23 years. That is, no account of weight change or assessment of latent effect of BMI on disease was possible. Our study reported here was a prospective, population-based study of anthropometric measures and BCC and thus susceptible to neither recall nor selection biases. It is the first study to have used objectively measured height and weight, together with measured waist and hip circumference in relation to BCC, and the only study to have examined weight change. Also the potential influences of most known risk factors for BCC were thoroughly assessed.
Limitations of the study include reliance on self-report of BCC prior to 1986, and availability of only a single height measurement in 1992. In older adults (age 60+), height can decrease with age due to spinal deformity and thinning of the intervertebral discs [35
]. This decrease in height has been reported to range from 0.5–1.5 cm/decade [35
], to 2.5–5 cm/decade [36
]. We could therefore expect that there would have been some change in height for the 26.9% of study participants aged over 60 years (in 1992), however we would expect this change in height to be modest over the 4.5 year follow-up period.