Richard Archer wrote:
>
> At 14:01 +1000 3/5/1999, Aaron Wieland wrote:
>
> >the amount of vitamin C in meat is *assumed* to be zero.  Period.
>
> And the amount of carbohydrate is assumed to be whatever remains once
> fat, protein, water and ash have been extracted.

And protein is often estimated from nitrogen content, etc.  Yes, there
are many assumptions hidden from the casual browser.  Of course, I
understand why these shortcuts are used; an exhaustive biochemical
analysis of every food could be prohibitively expensive and
time-consuming.

> Another problem: who is to say the product the USDA tested bears any
> resemblance to the product available in your area? For example, the
> amount of minerals may vary widely depending on the soil depletion.

Yes; I believe the USDA points this out.

> The USDA database has its shortcomings, but it is still the most
> comprehensive nutritional resource I have come across.
>
> But as with anything, you need to be aware of the shortcomings,
> and keep this in mind while examining the data.

Agreed.

> >reading about the USDA and Canadian Nutrient File databases;
>
> Is there a separate Canadian database?
> If so, is it available on the Web?

There is indeed a separate Canadian database (though many of its entries
have been imported from the USDA's system).  It is not yet available on
the WWW, but there are tentative plans to Web-enable it.  For now,
you'll have to shell out some money ($100-150 Canadian, I think; I
forget the exact price) for some ASCII or DBF files.

Cheers,
-- Aaron Wieland