June 2003, Vol 93, No. 6 | American Journal of Public Health 857-858
© 2003 American Public Health Association
PAUL RESPONDS
Jay P. Paul, PhD
Correspondence: Requests for reprints should be sent to Jay P. Paul, PhD, University of California, San Francisco, Center for AIDS Prevention Studies, 74 New Montgomery St, San Francisco, CA 94105 (e-mail: jpaul{at}psg.ucsf.edu).
Pierce’s letter suggests a misreading or misunderstanding of our article and our findings. The concerns about the Table 2 title might have been eased if we had included a table note clarifying that "never attempters" were the 0 value and suicide attempters (after the age of 17 years) were the 1 value for the dependent variable. The analysis on which Table 2 was based excluded a considerable proportion of our respondents who reported suicide attempts. Several of our independent variables were framed as occurring before the age of 17 years. To avoid temporal confounding with suicide attempts, we excluded respondents who reported a first suicide attempt before the age of 17 (which we noted on pages 1340 and 1343). Pierce’s unease about how we report the study findings and the contention that we have made "erroneous" claims appears to be based on confusion of the discussion of Table 3 results (midparagraph, page 1342) with the odds ratios and confidence intervals reported in Table 2. As the 2 tables examine different questions and refer to different subsets of the entire sample, their findings would not be expected to be entirely consistent. The results of Table 2 are discussed on page 1343 (first paragraph); we do not make the claims suggested by Pierce there.
It is somewhat unclear in what way 2 of our questions concerning "suicide ideation in relation to age" were not adequately addressed. No in-depth analysis of suicidal ideation was done; we presented only univariate correlates of ever making a suicide plan. In terms of suicidal behavior (rather than ideation), the critique of the analyses seems to reflect how one chooses to examine such questions. Contrary to the contention that "no investigation of suicide . . . attempts that occurred after the age of 25" was made, in Table 3 we compared all those who made first (or only) suicide attempts before the age of 25 with those who made such attempts after the age of 24. Our focus in these analyses was on age at first attempt, rather than age at later attempts for multiple attempters. Pierce seems to be suggesting an analysis based on age of most recent attempt. Although this is an interesting suggestion, it has its own attendant problems, as age of latest attempt is in part determined by the age of the respondent. As our data set only included age up to the fourth attempt, some recent attempts would be missing. Nevertheless, approximately half of those who made multiple attempts made all identified attempts (up to 4) prior to the age of 25. It is therefore likely that the findings from the analysis suggested by Pierce would not deviate markedly from those shown in Table 3.
Copyright © 2003 by the American Public Health Association
