How to hold values in variables while populating a column?

It's much easier: just use Fill Down and Replace Values on the Transform tab as demonstrated in this 30 sec video.

Well, it seems I was a bit fast with my answer...

In this case, the trick is to merge the table with itself so you will have Items and previous items on the same row.

The rest is pretty straightforward as you can see in this video.

During video recording, the following code was generated (in Power BI Desktop, with data from Excel):

let
    Source = Excel.Workbook(File.Contents("C:\Users\Marcel\Documents\Forum bijdragen\Power BI Community\How to hold variables etcetera.xlsx"), null, true),
    Tabel1_Table = Source{[Item="Tabel1",Kind="Table"]}[Data],
    #"Changed Type" = Table.TransformColumnTypes(Tabel1_Table,{{"Date", type date}, {"ItemID", Int64.Type}, {"Value", Int64.Type}}),
    #"Added Index" = Table.AddIndexColumn(#"Changed Type", "Index", 0, 1),
    #"Added Index1" = Table.AddIndexColumn(#"Added Index", "Index.1", 1, 1),
    #"Merged Queries" = Table.NestedJoin(#"Added Index1",{"Index"},#"Added Index1",{"Index.1"},"Prev",JoinKind.LeftOuter),
    #"Expanded Prev" = Table.ExpandTableColumn(#"Merged Queries", "Prev", {"ItemID"}, {"Prev.ItemID"}),
    #"Sorted Rows" = Table.Sort(#"Expanded Prev",{{"Index", Order.Ascending}}),
    #"Added Custom" = Table.AddColumn(#"Sorted Rows", "Custom", each if ([Prev.ItemID] = null or [Prev.ItemID] <> [ItemID]) then (if [Value] = null then 0 else [Value]) else [Value]),
    #"Filled Down" = Table.FillDown(#"Added Custom",{"Custom"}),
    #"Removed Columns" = Table.RemoveColumns(#"Filled Down",{"Value", "Index", "Index.1", "Prev.ItemID"}),
    #"Renamed Columns" = Table.RenameColumns(#"Removed Columns",{{"Custom", "Value"}}),
    #"Changed Type1" = Table.TransformColumnTypes(#"Renamed Columns",{{"Value", Int64.Type}})
in
    #"Changed Type1"