Using DATEDIFF in measure

// For this to work you have to have a correct model.
// If you want to see what a correct model means you
// can ask Google this phrase "star schema in power bi"
// and a Microsoft page should appear near the top of the
// page, if not at the top. Please do READ IT.
// You need at least 3 tables with the following
// characteristics:
// 1. Sales table that will be your fact table (hidden).
// 2. Stores table that will hold all your stores.
//    this is your dimension.
// 3. Dates table that will be the calendar in the
//    model.
// Of course, a correct model will also very likely 
// have a dimension called Customers. The Date table
// will join to the fact table on the sales date.
// I think in your case the column with the sales date
// is called just Sales[Date]. With all this in place,
// you can then write:

[Sales Filtered] =
// The assumption is that the period in the
// current context does not span more than
// 1 year. Otherwise the measure returns BLANK.
var __currentYear = SELECTEDVALUE( Dates[Year] )
// This variable will store StoreID's of stores
// as visible in the current context but such
// that have their Stores[OpenedDate] before
// the __currentYear.
var __storesThatOpenedBeforeTheYear =
    CALCULATETABLE(
        DISTINCT( Stores[StoreID] ),
        KEEPFILTERS(
            YEAR( Stores[OpenedDate] ) < __currentYear
        )
    )
var __output =
    CALCULATE(
        SUM( Sales[Sale] ),
        __storesThatOpenedBeforeTheYear
    )
return
    __output