Age effects and health appraisal: a meta-analysis

The purpose was to clarify the effects of methodological variables in the research on age differences in self-rated health and specifically the effects of: (a) the item type used to solicit the appraisal, (b) the approach to sampling, and (c) the criteria for assembling the age groups for study.

Studies differing on these dimensions were compared using meta-analytic techniques. Seventeen usable samples were grouped according to the age comparisons--seven compared "young" and "old" samples while ten compared "young-old" and "old-old" groups.

The results failed to support the frequently reported health optimism of old-old samples when compared with young-old. Further, the findings suggest that the item type used in eliciting the rating of health may be instrumental in the respondent's appraisal process. For instance, use of the age-comparative item tends to favor "old" groups when compared to young, whereas the global health item has an effect in the other direction. Effect size was also related to a study's sampling procedure and to the method used in assembling groups (i.e., age ranges used to represent young and old).

Practical implications and areas of needed research are discussed.