Setting up - Balanced Panel Data with time gaps

Hello all,

I have a panel in the following general form:

unit year
1 1980
1 1985
1 1990
1 1995
1 2001
1 2006
1 2009

2 1980
2 1985
2 1990
2 1995
2 2001
2 2006
2 2009

3 1980
3 1985... etc

In summary, I have a balanced panel for the years available (all units have the same number of observations). When I apply xtset unit year, it says balanced panel... years with gaps.

The adequate approach for my case seems to be running a model with all variables in first differences. However, as you can see from the data, as I do not have all years, it would have to be a "first" difference, from t to t-k, a difference to the last available observation.

From my research in this forum, it seems that if the gaps were the same (for example, every 5 years), I could take a "long 5-difference" and this would be okay.
My question is: as I don't have all the gaps with same length, would it be possible to take differences between t and "the last period available"?

( 2009-2006; 2006-2001; 2001-1995; 1995-1990... etc?)

I have set up my years as being a sequence year = (1,2,3,4,5,6,7) and it all runs smoothly. I am just not sure if there is any implication of having periods of different length, or if this will be affecting my estimations.

Thank you!
Bertha