The thrust of the idea is that various nutritional recommendations have been shown to work in some of these groups but not in others. Lumping all groups together may mask an effect which occurs in one group but not the other whereas restricting attention to those groups where the nutritional recommendations work would be expected to better detect the effect.
Along the indolent/aggressive axis we have dairy and calcium consumption. High dairy consumption or high calcium consumption increases the risk of aggressive disease but does not pose a risk for organ confined disease. If one were to mix together indolent and aggressive patients one might miss the effect since it would not be present in the indolent ones. By restricting attention to the aggressive ones we would more likely detect the effect.
Along the young/old axies we have a variety of supporting items:
- high body mass index reduces risk by 50% at age 21 but increases risk slightly at age 60
- hormones is a driving factor behind PCa in youger men while inflammation is a driving factor in older men (note: this has been disputed -- William Nelson suggests inflammation determines whether you get PC or not whereas hormones determine how aggressive it will be)
- as men age they tend to get heavier which in turn results in greater inflammation
- exercise reduces risk of prostate cancer among older men but not younger men consistent with exercise reducing inflammation
- lycopene has a relative risk of 67% for older men but only 89% for younger men. Again this is consistent with its ability to reduce inflammation.
Dividing men into a younger group and an older group would ideally be done based on a hormonal event paralleling menopause in women.
According to in cites Walter Willett was the second most cited researcher in clinical medicine in the 1995-2005 period with 516 papers and 29,311 citations although citations to him tend to be in the area of Diabetes, not Prostate Cancer.
News item on Willett