This sounds like a familiar question posted recently?
What is the clustering index or the index used for reorgs. Maybe that has changed since when you had good response time? If somedate is the clustering index, it is more likely to pick that one for access. Make sure the table is reorged, then do runstats again.
You could also try different query optimization levels (higher or lower).
Lastly, I would check for APAR's in FP4 or APAR's yet to be fixed.
http://www-3.ibm.com/cgi-bin/db2www/...lib.d2w/report