Performance Tuning Using Index. How To Force The Query Engine To Use It?

Hello all.
I have the following table

Create Table Item(
I_AssetCode NVarChar(40) Primary Key NOT NULL,
I_Name NVarChar(160),
I_BC nvarchar(20),
I_Company nvarchar(20)
);

Create Index ind_Item_Name on Item(I_Name);
Create Index ind_Item_BC on Item(I_BC);
Create Index ind_Item_Company on Item(I_Company);

It is populated with 50 000 records.
Searching on indexed columns is fast, but I've run into the following problem:
I need to get all distinct companies in the table.
I've tried with these two queries, but they both are very slow!

1. "select I_Company from item group by I_Company " - This one takes 19 seconds

2. "select distinct(I_Company) from item" -This one takes 29 secons

When I ran them through the SQL Management Studio and checked the performance plan, I saw that the second one doesn't use index at all ! So I focused on the first...
The first one used index (it took it 15% of the time), but then it ran the "stream aggregate" which took 85% of the time !
Actully 15% of 19 seconds - about 2 seconds is pretty much enough for me. But it looks that aggregate function is run for nothing!
So is it possible to force the query engine of the SSCE not to run it, since there is actually no aggregate functions in my select clause?
According to SQL CE Books online:
Group By


"Specifies the groups (equivalence classes) that output rows are to be placed in. If aggregate functions are included in the SELECT clause <select list>, the GROUP BY clause calculates a summary value for each group."
It seems the aggregate is run every time, not only when there is an aggregate function.

Is this a bug?

Thanks in advance,
TipoMan