Letters to the Editor Letters to the Editor are welcomed and will be published, iffound suitable, as space permits. Submission of a Letter to the Editor constitutes permission for its publication in the Journal. Letters should not duplicate similar material being submitted or published elsewhere. Letters referring to a recent Journal article should be received within three months of the article's publication. The editors reserve the right to edit and abridge letters, to publish replies, and to solicit responses from authors and others. Letters should be submitted in duplicate, double-spaced (including references), and should not exceed 400 words.
The Female Condom Changes in behavior and barrier methods of contraception are at present the only ways of slowing the sexual transmission of the human immunodeficiency virus (HIV). The female condom represents a new and potentially important addition to the existing choices. The preliminary study by Bounds, et al, I of 24 married or cohabiting couples found 63 percent of the men and 70 percent of the women reported the effect of sexual pleasure to be no different or better than with the male condom. Because the female condom is made ofpolyurethane, a material more durable than rubber, and covers a larger surface area, it may provide better protection than a standard latex condom. In addition, early reports on the US-made WPC-33 female condom (similar to the one reported by Bounds, et al,) suggest that the risk of female exposure to seminal fluid is less when WPC-33 is used as compared to a male condom.2 We tested the acceptability of the US-made WPC-33 female condom among 20 high-risk female sex workers3 in Khon Kaen, Thailand, trained by nurses to use the female condom; we supplied 20 unlubricated devices for each woman, as well as a supply of lubricant and their regular supply of male condoms. The participants were instructed about the risk of AIDS (acquired immunodeficiency syndrome) and advised that they could use the female condom as an alternative method of protection to the male con498
dom. The decision of which device to use, if any, was left to the participant. Two weeks later, participants reported using the female condom alone in a total of 78 (32 percent) of 247 episodes of vaginal intercourse, eight episodes of which were in conjunction with the male condom. The male condom was used in 90 (35 percent) of the episodes and no condom in 87 (34 percent) episodes. Two-thirds of the volunteers reported no aversion to the female condom while one-third disliked it. Mechanically, the female condom performed well. No rips or tears were reported during intercourse, and no woman reported severe pain. The most common objection to the 17 cm device for these Thai women was that it was too big. Also, the need to lubricate the condoms made their use messy and inconvenient. Nineteen participants said the female condom was less convenient to use than the male condom, and six said it was less comfortable. Most of these problems can be overcome by shortening and pre-lubricating the condoms. One other objection was difficult inserting (15 percent), a problem which may have been due to lack of experience with inserting the condoms. While the participants' own general assessment of the condom was fairly positive, most discontinued using the device because of male partner objection. Ten respondents reported that all partners with whom they used the female condoms objected to their use; eight said reactions were mixed; and two said all partners with whom they tried them reacted positively. Eighteen of 20 participants said they would advise other sex workers to try these female condoms. We are now preparing to repeat this study at the same site, using 15 cm pre-lubricated female condoms. Revised instructions will be provided by the manufacturer, and each participant will be required to practice inserting at least two of the devices before initiating the study. REFERENCES 1. Bounds W, Guillebaud J, Stewart L, Steele S: A female condom (Femshield'): A study of its
user acceptability. Br J Fam Plann 1988; 14:8387. 2. Leeper MA, Conrardy M, Henderson J: Evaluation of the WPC-33 female condom: Abstract No. 6305 of the V International Congress on AIDS, Montreal, Canada, June 1989. 3. Sakondhavat C: Consumer preference study of a female condom in a sexually active population at risk of contracting AIDS: Final Report. Khon Kaen, Thailand, July 1989.
Chuanchom Sakondhavat, MD Department of Ob-Gyn, Khon Kaen University Hospital, Khon Kaen 40002, Thailand. Editor's Note: See also related commentary by Z. Stein, p 460 this issue. © 1990 American Journal of Public Health
Personality Traits and Addictive Disease It is not at all surprising that Yates, et al,I found that cocaine patients exhibited more eccentric, flamboyant, and anxious personality traits on the Per-
sonality Diagnostic Questionnaire (PDQ) than did either alcoholic patients or community controls. However, a slight reworking of their data reveals that alcoholic subjects averaged 1.87 (88/47) traits per subject for all PDQ clusters, almost as many as the mean of 2.17 (128/59) per cocaine abuser (many if not most of whom would almost certainly have met the diagnostic criteria for cocaine dependence had DSMIII-R been used rather than DSM-III). Community controls averaged only .62 (43/69) traits per subject. Nor is it surprising that cocaine subjects scored significantly higher on the narcissism trait. Alcohol abuse or dependent and cocaine abuse or dependent subjects have all been neurophysiologically "hotwired" by their drug of choice; they are all "emotionally augmented,"2.3 as the PDQ predictably revealed. But to conclude that narcissism or any other "personality factors appear to be worthy of further study as risk factors for the initiation and maintenance of cocaine or other drug abuse" or, worse, that "identification of highrisk personalities, and better understanding of the natural history of personality development, might allow for intensive preventive measures among AJPH April 1990, Vol. 80, No.4
LETTERS TO THE EDITOR
adolescents,"' based on data gathered from practicing or recently practicing alcoholics/addicts, is undoubtedly just plain "retrospective error"-confusing dependent and independent variables, outcome and cause. Yates and his colleagues seem to be making a valiant effort to resurrect a premorbid addictive personality in the etiology of addictive disease, which simply needs to be laid to rest, as Vaillant has done for the non-existent premorbid "alcoholic personality." "... most future alcoholics do not appear different from future asymptomatic drinkers in terms of premorbid psychological stability. However, not until several prospective studies were available, could such a hypothesis be seriously entertained. It was difficult to conceive that the 'alcoholic personality' might be secondary to the disorder, alcoholism. "4 In fact, cocaine itself probably can and does produce the emotionally augmented high level of narcissism seen in this sample. Only prospective studies can provide an answer for cocaine and possible antecedant personality risk factors. Such studies have clearly demonstrated that premorbid personality factors do not exist in the etiology of alcoholism. REFERENCES 1. Yates W, Fulton A, Gabel J, Brass C: Personality risk factors for cocaine abuse. Am J Public Health 1989; 79:891-892. 2. Milam J: The Emergent Comprehensive Concept of Alcoholism. Kirkland, WA: Alcoholism Center Associates, 1978; 26-28. 3. Rogers R, McMillin C: Freeing Someone You Love from Alcohol and other Drugs. Los Angeles, The Body Press, 1989; 150-169. 4. Vaillant G: The Natural History of Alcoholism. Cambridge, MA: Harvard University Press, 1983; 311-312.
Theodore Ernst, MDiv, DSW Clinical Director, Bethany Center, Honesdale, PA 18431
measure-this possibility is noted in the discussion section of our manuscript. Our objective was not to "resurrect a premorbid addictive personality in the etiology of addictive disease," as suggested by Dr. Ernst. Previous investigators have explored unitary causal hypotheses for alcoholism including personality characteristics. However, a consensus now exists that the pathways to alcohol and drug abuse are multifactorial. Our purpose was to explore discrete personality disorders as risk factors for some alcohol and cocaine abusers. Dr. Ernst has concluded that if a majority of alcoholics do not have premorbid personality disorders, premorbid personality disorders are not important factors in the development of alcohol or drug abuse in all individuals. I believe this is faulty reasoning. A similar false medical analogy would be: If most cancer patients do not have a history of exposure to asbestos, asbestos exposure is not important in the etiology of cancer. Finally, Dr. Ernst's complete dismissal of personality factors in the etiology of alcohol and drug abuse ignores important research findings. Vaillant's important work on the issues of personality and alcoholism includes evidence in the Core City sample of a role for antisocial personality disorder (sociopathy) in alcoholism, concluding: "many sociopaths later abuse alcohol as part of their antisocial behavior; but most alcoholics are not premorbidly sociopathic".' Additionally, a recent prospective study by Cloninger, et al,2 has given support to the predictive value of childhood personality in the risk assessment for adult alcohol abuse. In their study, high novelty seeking childhood personality traits predicted high risk for adult alcohol abuse. Interestingly, both antisocial personality disorder and narcissistic personality disorder are felt to
have high novelty seeking components.3 Dr. Cloninger's personality construct is linked to neurobiologic and genetic findings-this may explain the link between personality, behavior and risk for development of alcohol abuse for some alcoholics. These studies support our conclusion of the importance offurther study of personality measures in children and adolescents as one possible step in designing prevention strategies for alcohol and drug abuse. REFERENCES 1. Vaillant GE. The Natural History of Alcoholism. Cambridge MA: Harvard University Press, 1983; 82. 2. Cloninger CR, Sigvardsson S, Bohman M: Childhood personality predicts alcohol abuse in young adults. Alcoholism: Clin Exp Res 1988; 12:494-505. 3. Cloninger CR: A systematic method for clinical description and classification of personality variants: A proposal. Arch Gen Psychiatry 1987;
44:573-588. 4. Cloninger CR: Neurogenetic adaptive mechanisms in alcoholism. Science 1987; 236:410-416.
William R. Yates, MD Assistant professor of Psychiatry, University of Iowa College of Medicine, and Psychiatric Hospital, 500 Newton Road, Iowa City, IA 52242. e 1990 American Journal of Public Health
AIDS as a Cause of Death in Children, Adolescents, and Young Adults We wish to add to the article by Fingerhut and Kleinman on mortality among children and youth by providing data on deaths due to the acquired immunodeficiency syndrome (AIDS).' Using national AIDS case and death reports (through July 1989) and annual intercensus population estimates, we calculated yearly death rates (deaths/100,000) for 1980 through 1987 for persons under 1 year of age, ages 1-4, 5-14, and 15-24 years of age (Table
© 1990 American Journal of Public Health TABLE 1-AIDS Mortality Rates and Numbers of Deaths by Year for Pediatric and Young Adult Populations, United States, 1980-87
Response from Dr. Yates Dr. Ernst's letter addresses important issues in the study of personality disorders as risk factors for alcohol and non-alcohol drug abuse. I agree with many of Dr. Ernst's comments includ-
ing the need for prospective design studies. Although I cannot share Dr. Ernst's certainty, I acknowledge our finding of increased narcissism in cocaine abusers could result from an effect of the drug on the personality AJPH April 1990, Vol. 80, No.4
Age Groups Deaths/1 00,000 (Number) 5-14
Year
The Female Condom Changes in behavior and barrier methods of contraception are at present the only ways of slowing the sexual transmission of the human immunodeficiency virus (HIV). The female condom represents a new and potentially important addition to the existing choices. The preliminary study by Bounds, et al, I of 24 married or cohabiting couples found 63 percent of the men and 70 percent of the women reported the effect of sexual pleasure to be no different or better than with the male condom. Because the female condom is made ofpolyurethane, a material more durable than rubber, and covers a larger surface area, it may provide better protection than a standard latex condom. In addition, early reports on the US-made WPC-33 female condom (similar to the one reported by Bounds, et al,) suggest that the risk of female exposure to seminal fluid is less when WPC-33 is used as compared to a male condom.2 We tested the acceptability of the US-made WPC-33 female condom among 20 high-risk female sex workers3 in Khon Kaen, Thailand, trained by nurses to use the female condom; we supplied 20 unlubricated devices for each woman, as well as a supply of lubricant and their regular supply of male condoms. The participants were instructed about the risk of AIDS (acquired immunodeficiency syndrome) and advised that they could use the female condom as an alternative method of protection to the male con498
dom. The decision of which device to use, if any, was left to the participant. Two weeks later, participants reported using the female condom alone in a total of 78 (32 percent) of 247 episodes of vaginal intercourse, eight episodes of which were in conjunction with the male condom. The male condom was used in 90 (35 percent) of the episodes and no condom in 87 (34 percent) episodes. Two-thirds of the volunteers reported no aversion to the female condom while one-third disliked it. Mechanically, the female condom performed well. No rips or tears were reported during intercourse, and no woman reported severe pain. The most common objection to the 17 cm device for these Thai women was that it was too big. Also, the need to lubricate the condoms made their use messy and inconvenient. Nineteen participants said the female condom was less convenient to use than the male condom, and six said it was less comfortable. Most of these problems can be overcome by shortening and pre-lubricating the condoms. One other objection was difficult inserting (15 percent), a problem which may have been due to lack of experience with inserting the condoms. While the participants' own general assessment of the condom was fairly positive, most discontinued using the device because of male partner objection. Ten respondents reported that all partners with whom they used the female condoms objected to their use; eight said reactions were mixed; and two said all partners with whom they tried them reacted positively. Eighteen of 20 participants said they would advise other sex workers to try these female condoms. We are now preparing to repeat this study at the same site, using 15 cm pre-lubricated female condoms. Revised instructions will be provided by the manufacturer, and each participant will be required to practice inserting at least two of the devices before initiating the study. REFERENCES 1. Bounds W, Guillebaud J, Stewart L, Steele S: A female condom (Femshield'): A study of its
user acceptability. Br J Fam Plann 1988; 14:8387. 2. Leeper MA, Conrardy M, Henderson J: Evaluation of the WPC-33 female condom: Abstract No. 6305 of the V International Congress on AIDS, Montreal, Canada, June 1989. 3. Sakondhavat C: Consumer preference study of a female condom in a sexually active population at risk of contracting AIDS: Final Report. Khon Kaen, Thailand, July 1989.
Chuanchom Sakondhavat, MD Department of Ob-Gyn, Khon Kaen University Hospital, Khon Kaen 40002, Thailand. Editor's Note: See also related commentary by Z. Stein, p 460 this issue. © 1990 American Journal of Public Health
Personality Traits and Addictive Disease It is not at all surprising that Yates, et al,I found that cocaine patients exhibited more eccentric, flamboyant, and anxious personality traits on the Per-
sonality Diagnostic Questionnaire (PDQ) than did either alcoholic patients or community controls. However, a slight reworking of their data reveals that alcoholic subjects averaged 1.87 (88/47) traits per subject for all PDQ clusters, almost as many as the mean of 2.17 (128/59) per cocaine abuser (many if not most of whom would almost certainly have met the diagnostic criteria for cocaine dependence had DSMIII-R been used rather than DSM-III). Community controls averaged only .62 (43/69) traits per subject. Nor is it surprising that cocaine subjects scored significantly higher on the narcissism trait. Alcohol abuse or dependent and cocaine abuse or dependent subjects have all been neurophysiologically "hotwired" by their drug of choice; they are all "emotionally augmented,"2.3 as the PDQ predictably revealed. But to conclude that narcissism or any other "personality factors appear to be worthy of further study as risk factors for the initiation and maintenance of cocaine or other drug abuse" or, worse, that "identification of highrisk personalities, and better understanding of the natural history of personality development, might allow for intensive preventive measures among AJPH April 1990, Vol. 80, No.4
LETTERS TO THE EDITOR
adolescents,"' based on data gathered from practicing or recently practicing alcoholics/addicts, is undoubtedly just plain "retrospective error"-confusing dependent and independent variables, outcome and cause. Yates and his colleagues seem to be making a valiant effort to resurrect a premorbid addictive personality in the etiology of addictive disease, which simply needs to be laid to rest, as Vaillant has done for the non-existent premorbid "alcoholic personality." "... most future alcoholics do not appear different from future asymptomatic drinkers in terms of premorbid psychological stability. However, not until several prospective studies were available, could such a hypothesis be seriously entertained. It was difficult to conceive that the 'alcoholic personality' might be secondary to the disorder, alcoholism. "4 In fact, cocaine itself probably can and does produce the emotionally augmented high level of narcissism seen in this sample. Only prospective studies can provide an answer for cocaine and possible antecedant personality risk factors. Such studies have clearly demonstrated that premorbid personality factors do not exist in the etiology of alcoholism. REFERENCES 1. Yates W, Fulton A, Gabel J, Brass C: Personality risk factors for cocaine abuse. Am J Public Health 1989; 79:891-892. 2. Milam J: The Emergent Comprehensive Concept of Alcoholism. Kirkland, WA: Alcoholism Center Associates, 1978; 26-28. 3. Rogers R, McMillin C: Freeing Someone You Love from Alcohol and other Drugs. Los Angeles, The Body Press, 1989; 150-169. 4. Vaillant G: The Natural History of Alcoholism. Cambridge, MA: Harvard University Press, 1983; 311-312.
Theodore Ernst, MDiv, DSW Clinical Director, Bethany Center, Honesdale, PA 18431
measure-this possibility is noted in the discussion section of our manuscript. Our objective was not to "resurrect a premorbid addictive personality in the etiology of addictive disease," as suggested by Dr. Ernst. Previous investigators have explored unitary causal hypotheses for alcoholism including personality characteristics. However, a consensus now exists that the pathways to alcohol and drug abuse are multifactorial. Our purpose was to explore discrete personality disorders as risk factors for some alcohol and cocaine abusers. Dr. Ernst has concluded that if a majority of alcoholics do not have premorbid personality disorders, premorbid personality disorders are not important factors in the development of alcohol or drug abuse in all individuals. I believe this is faulty reasoning. A similar false medical analogy would be: If most cancer patients do not have a history of exposure to asbestos, asbestos exposure is not important in the etiology of cancer. Finally, Dr. Ernst's complete dismissal of personality factors in the etiology of alcohol and drug abuse ignores important research findings. Vaillant's important work on the issues of personality and alcoholism includes evidence in the Core City sample of a role for antisocial personality disorder (sociopathy) in alcoholism, concluding: "many sociopaths later abuse alcohol as part of their antisocial behavior; but most alcoholics are not premorbidly sociopathic".' Additionally, a recent prospective study by Cloninger, et al,2 has given support to the predictive value of childhood personality in the risk assessment for adult alcohol abuse. In their study, high novelty seeking childhood personality traits predicted high risk for adult alcohol abuse. Interestingly, both antisocial personality disorder and narcissistic personality disorder are felt to
have high novelty seeking components.3 Dr. Cloninger's personality construct is linked to neurobiologic and genetic findings-this may explain the link between personality, behavior and risk for development of alcohol abuse for some alcoholics. These studies support our conclusion of the importance offurther study of personality measures in children and adolescents as one possible step in designing prevention strategies for alcohol and drug abuse. REFERENCES 1. Vaillant GE. The Natural History of Alcoholism. Cambridge MA: Harvard University Press, 1983; 82. 2. Cloninger CR, Sigvardsson S, Bohman M: Childhood personality predicts alcohol abuse in young adults. Alcoholism: Clin Exp Res 1988; 12:494-505. 3. Cloninger CR: A systematic method for clinical description and classification of personality variants: A proposal. Arch Gen Psychiatry 1987;
44:573-588. 4. Cloninger CR: Neurogenetic adaptive mechanisms in alcoholism. Science 1987; 236:410-416.
William R. Yates, MD Assistant professor of Psychiatry, University of Iowa College of Medicine, and Psychiatric Hospital, 500 Newton Road, Iowa City, IA 52242. e 1990 American Journal of Public Health
AIDS as a Cause of Death in Children, Adolescents, and Young Adults We wish to add to the article by Fingerhut and Kleinman on mortality among children and youth by providing data on deaths due to the acquired immunodeficiency syndrome (AIDS).' Using national AIDS case and death reports (through July 1989) and annual intercensus population estimates, we calculated yearly death rates (deaths/100,000) for 1980 through 1987 for persons under 1 year of age, ages 1-4, 5-14, and 15-24 years of age (Table
© 1990 American Journal of Public Health TABLE 1-AIDS Mortality Rates and Numbers of Deaths by Year for Pediatric and Young Adult Populations, United States, 1980-87
Response from Dr. Yates Dr. Ernst's letter addresses important issues in the study of personality disorders as risk factors for alcohol and non-alcohol drug abuse. I agree with many of Dr. Ernst's comments includ-
ing the need for prospective design studies. Although I cannot share Dr. Ernst's certainty, I acknowledge our finding of increased narcissism in cocaine abusers could result from an effect of the drug on the personality AJPH April 1990, Vol. 80, No.4
Age Groups Deaths/1 00,000 (Number) 5-14
Year
