down by just 1 point (on a scale of 10) in our study, the prevalence of severe or moderate impairment would have dropped to 67% in the hospital ward and to 44% in the nursing home. Conversely, if the cutoff point had been moved up by a point the prevalence in the hospital ward would have risen to 92% (close to that found by Teitelbaum and associates) and in the nursing home to 61%. With the projected doubling of our elderly population by the year 20214 most of those who are cognitively impaired will be resident in the community and will have to be cared for there. Canada has one of the highest rates of elderly people living in institutions,5 but even here almost half of people with moderate or severe cognitive impairment live at home, as do the vast majority of those with mild impairment.6,7 In our study, of people with severe or moderate cognitive impairment only 31% lived in institutions and 69% lived at home. Although the latter figure may be higher than in urban areas because of more extended kinship networks and a relative lack of available institutions it is clear that with the onset of "the silent epidemic" of dementia in Canada8 most of the caring will have to be done by families and family doctors. We cannot afford not to evaluate our community services for such people.9
4.
5. 6.
7. 8. 9.
prevalence rates. Psychol Med 1985; 15: about the fact that Canadian com771-788 munity services for demented paPopulation Projections for Canada and tients will be increasingly in dethe Provinces (cat no 91-514), Statistics mand and increasingly strained. Canada, Ottawa, 1974 Fact Book on Aging, Dept of National Therefore, the time for assessment Health and Welfare, Ottawa, 1983: 68- of our provision of such services 69 O'Connor DW, Pollitt PA, Hyde JB et is now upon us. al: The prevalence of dementia as measured by the Cambridge Mental Disor- Louise Teitelbaum, MD, FRCPC ders of the Elderly. Acta Psychiatr Hotel Dieu Hospital M. Lynne Ginsburg, MB, ChB, FRCPC Scand 1989; 79: 190-198 Preston GAN: Dementia in elderly Robert W. Hopkins, PhD adults: prevalence and institutionaliza- Kingston Psychiatric Hospital Kingston, Ont. tion. J Gerontol 1986; 41: 261-287 Beck JC: Dementia in the elderly: the silent epidemic. Ann Intern Med 1982; Reference 97: 231-241 Worrall G, Chambers LC: Can we af- 1. Folstein MF, Folstein SE, McHugh PR: ford not to evaluate services for persons Mini-Mental State: a practical method with dementia? Can Fam Physician for grading the cognitive state of pa1989; 35: 573-580 tients for the clinician. J Psychiatr Res 1975; 12: 189-198
[The authors respond.:
First, a score of less than 24 on the Mini-Mental State Examination indicates some degree of cognitive impairment' rather than moderate or severe cognitive impairment, as suggested by Dr. Worrall. Second, although the criteria for admission to the psychogeriatric wards in a provincial psychiatric hospital were not specifically indicated in our article (patients must be elderly and suffering from an organic mental disorder) the patients in them would likely be substantially different from the "residents (mostly psychogeriatric patients) of a longstay hospital ward." Third, since cutoff scores are (or should be) experimentally deGraham Worrall, MSc, CCFP rived points in a statistical distriDistrict medical officer bution their artificial movement Glovertown Medical Clinic Glovertown, Nfld. is not appropriate. If, however, minor score differences have the effect of greatly changing the References number of cases falling above or 1. Robertson D, Rockwood H, Stolee P: A below the cutoff score one should short mental status questionnaire. Can examine the validity of the instruJAging 1982; 1: 16-20 2. Jeans ER, Helmes E, Merskey H et al: ment. In our study such changes Some calculations on the prevalence of as plus or minus one point would dementia in Canada. Can J Psychiatry have had only minimal effects on 1987; 32: 81-88 the distributions (about 1% or 3. Kay DWK, Henderson AS, Scott R et
HIV infection among Quebec women giving birth to live infants I n the final paragraph of her reply (Can Med Assoc J 1991;
144: 955) to my letter (ibid: 954-955) Dr. Catherine A. Hankins states: "The only cases of HIV [human immunodeficiency virus] transmission through breast-feeding that have been documented worldwide have occurred in infants whose mothers were infected not during pregnancy or at birth, but, rather, post partum, in most cases through contaminated blood transfusions."
Hankins appears to be unaware of work by Blanche and associates and the HIV Infection in Newborns French Collaborative Study Group.' In a prospective study of infants born to women seropositive for HIV type 1, 32 of 117 infants (27%) were seropositive after 18 months; "5 of the 6 infants who were breast-fed became seropositive, as compared with 25 of 99 who were not (p < 0.01)." 2%). It would seem that this findal: Dementia and depression among the Finally, we wholeheartedly ing is also at variance with Hankelderly living in the Hobart community: the effect of diagnostic criteria on the agree with Worrall's comments ins' next statement: "It has not
198
CAN MED ASSOC J 1991; 145 (3)
LE
lI' AO(JT 1991
been demonstrated that breastfeeding poses an incremental risk to infants who have received placentally acquired antibodies during gestation." With respect, I find myself in complete disagreement with her final statement. I feel it is high time that our future generations be protected by appropriate public health measures, which include HIV testing in pregnancy. James E. Parker, MB, FRCPC 303-2151 McCallum Rd. Abbotsford, BC
Reference 1. Blanche S, Rouzioux C, Moscato MLG et al and the HIV Infection in Newborns French Collaborative Study Group: A prospective study of infants born to women seropositive for human immunodeficiency virus type 1. N Engl J Med 1989; 320: 1643-1648
[Dr. Hankins responds.]
HIV from mother to infant; however, they call for further studies. Although the authors do not comment on the strength of the association that they observed an unrecognized or unidentified risk factor that influences the maternal-fetal transmission rate and that is more likely to be present among breast-feeding mothers may be responsible for the increased rate observed. No data are presented regarding the characteristics of the mothers who breast-fed compared with those who did not. I discussed the results of the study with Dr. Blanche on several occasions before and after publication, most recently last December, when I was invited to visit him and Dr. Claude Griscelli at the Hopital Necker in Paris. As a general rule it is wise to avoid drawing conclusions and developing policy from one study, particularly when the number of subjects is small. To suggest that HIV testing in pregnancy as a public health measure is justified by the results of the collaborative study is to take a giant step. In Canada HIV testing is available free of charge for any woman, pregnant or not, who is capable of providing informed consent for the procedure after adequate pretest counselling. Clinicians should be encouraged to discuss the risk of HIV and other sexually transmitted diseases with sexually active women as a matter of course. In Canada routine screening of pregnant women without appropriate counselling and informed consent is just not on.
Unfortunately, Dr. Parker does not appear to have read the discussion accompanying the results that he quotes from the study by Blanche and associates and the HIV Infection in Newborns French Collaborative Study Group.' The authors are quite clear that the role of breast-feeding in HIV transmission remains controversial. They point to the fact that breast-feeding is a major mode of transmission for another retrovirus, human T-cell lymphotropic virus type 1, and they indicate - as did I - that breastfeeding appears to be the route of transmission to infants whose mothers acquire HIV through blood transfusion after delivery. Catherine A. Hankins, MD, FRCPC "At the same time, other prospec- Public health epidemiologist tive or retrospective surveys re- Department of Community Health General Hospital port no difference in the outcome Montreal Montreal, Que. in infants with respect to the mode of feeding." They comment that their own data, involving a Reference small number of infants (6 S, Rouzioux C, Moscato MLG breast-fed v. 99 bottle-fed), tend 1. etBlanche al and the HIV Infection in Newto suggest that breast-feeding may borns French Collaborative Study Group: A prospective study of infants have a role in the transmission of AUGUST 1, 1991
bom to women seropositive for human immunodeficiency virus type 1. N Engi JMed 1989; 320: 1643-1648
Advertising meal replacement products A
recent issue of CMAJ featured a four-page advertisement for Slim-Fast diet products (1991; 144: 633-636). One of the references for the information provided is a personal communication from the Eating Disorders Clinic of Toronto General Hospital. We wish to clear up any possible sense of endorsement from our program for this or any other diet product. Indeed, our whole orientation is toward discouraging dieting behaviour. The advertisement begins with a recognition of the hazards of repetitive weight loss and gain, although it fails to mention that dieting is the commonest precipitant of anorexia nervosa and bulimia nervosa. It soon switches, however, to a discussion of obesity and its health risks without any consideration of what degree of obesity poses medical problems. The generalization regarding obesity is highlighted by the accompanying photograph of a young woman exercising on a stationary bicycle. Next to the picture it is stated that weight management "is a strategy for enhancing the quality of life, and preventing the most common cause of death among Canadians." We see many patients who began with the shape of the model in the advertisement; the quality of their lives has been seriously eroded by dieting. Estimates of rates of death from eating disorders across recent long-term outcome studies are up to 18%.' What is strikingly absent from this advertisement is any evidence to support a beneficial role for the diet product. However, its appearance in a medical journal among pharmaceutical adCAN MED ASSOC J 1991; 145 (3)
199
4.
5. 6.
7. 8. 9.
prevalence rates. Psychol Med 1985; 15: about the fact that Canadian com771-788 munity services for demented paPopulation Projections for Canada and tients will be increasingly in dethe Provinces (cat no 91-514), Statistics mand and increasingly strained. Canada, Ottawa, 1974 Fact Book on Aging, Dept of National Therefore, the time for assessment Health and Welfare, Ottawa, 1983: 68- of our provision of such services 69 O'Connor DW, Pollitt PA, Hyde JB et is now upon us. al: The prevalence of dementia as measured by the Cambridge Mental Disor- Louise Teitelbaum, MD, FRCPC ders of the Elderly. Acta Psychiatr Hotel Dieu Hospital M. Lynne Ginsburg, MB, ChB, FRCPC Scand 1989; 79: 190-198 Preston GAN: Dementia in elderly Robert W. Hopkins, PhD adults: prevalence and institutionaliza- Kingston Psychiatric Hospital Kingston, Ont. tion. J Gerontol 1986; 41: 261-287 Beck JC: Dementia in the elderly: the silent epidemic. Ann Intern Med 1982; Reference 97: 231-241 Worrall G, Chambers LC: Can we af- 1. Folstein MF, Folstein SE, McHugh PR: ford not to evaluate services for persons Mini-Mental State: a practical method with dementia? Can Fam Physician for grading the cognitive state of pa1989; 35: 573-580 tients for the clinician. J Psychiatr Res 1975; 12: 189-198
[The authors respond.:
First, a score of less than 24 on the Mini-Mental State Examination indicates some degree of cognitive impairment' rather than moderate or severe cognitive impairment, as suggested by Dr. Worrall. Second, although the criteria for admission to the psychogeriatric wards in a provincial psychiatric hospital were not specifically indicated in our article (patients must be elderly and suffering from an organic mental disorder) the patients in them would likely be substantially different from the "residents (mostly psychogeriatric patients) of a longstay hospital ward." Third, since cutoff scores are (or should be) experimentally deGraham Worrall, MSc, CCFP rived points in a statistical distriDistrict medical officer bution their artificial movement Glovertown Medical Clinic Glovertown, Nfld. is not appropriate. If, however, minor score differences have the effect of greatly changing the References number of cases falling above or 1. Robertson D, Rockwood H, Stolee P: A below the cutoff score one should short mental status questionnaire. Can examine the validity of the instruJAging 1982; 1: 16-20 2. Jeans ER, Helmes E, Merskey H et al: ment. In our study such changes Some calculations on the prevalence of as plus or minus one point would dementia in Canada. Can J Psychiatry have had only minimal effects on 1987; 32: 81-88 the distributions (about 1% or 3. Kay DWK, Henderson AS, Scott R et
HIV infection among Quebec women giving birth to live infants I n the final paragraph of her reply (Can Med Assoc J 1991;
144: 955) to my letter (ibid: 954-955) Dr. Catherine A. Hankins states: "The only cases of HIV [human immunodeficiency virus] transmission through breast-feeding that have been documented worldwide have occurred in infants whose mothers were infected not during pregnancy or at birth, but, rather, post partum, in most cases through contaminated blood transfusions."
Hankins appears to be unaware of work by Blanche and associates and the HIV Infection in Newborns French Collaborative Study Group.' In a prospective study of infants born to women seropositive for HIV type 1, 32 of 117 infants (27%) were seropositive after 18 months; "5 of the 6 infants who were breast-fed became seropositive, as compared with 25 of 99 who were not (p < 0.01)." 2%). It would seem that this findal: Dementia and depression among the Finally, we wholeheartedly ing is also at variance with Hankelderly living in the Hobart community: the effect of diagnostic criteria on the agree with Worrall's comments ins' next statement: "It has not
198
CAN MED ASSOC J 1991; 145 (3)
LE
lI' AO(JT 1991
been demonstrated that breastfeeding poses an incremental risk to infants who have received placentally acquired antibodies during gestation." With respect, I find myself in complete disagreement with her final statement. I feel it is high time that our future generations be protected by appropriate public health measures, which include HIV testing in pregnancy. James E. Parker, MB, FRCPC 303-2151 McCallum Rd. Abbotsford, BC
Reference 1. Blanche S, Rouzioux C, Moscato MLG et al and the HIV Infection in Newborns French Collaborative Study Group: A prospective study of infants born to women seropositive for human immunodeficiency virus type 1. N Engl J Med 1989; 320: 1643-1648
[Dr. Hankins responds.]
HIV from mother to infant; however, they call for further studies. Although the authors do not comment on the strength of the association that they observed an unrecognized or unidentified risk factor that influences the maternal-fetal transmission rate and that is more likely to be present among breast-feeding mothers may be responsible for the increased rate observed. No data are presented regarding the characteristics of the mothers who breast-fed compared with those who did not. I discussed the results of the study with Dr. Blanche on several occasions before and after publication, most recently last December, when I was invited to visit him and Dr. Claude Griscelli at the Hopital Necker in Paris. As a general rule it is wise to avoid drawing conclusions and developing policy from one study, particularly when the number of subjects is small. To suggest that HIV testing in pregnancy as a public health measure is justified by the results of the collaborative study is to take a giant step. In Canada HIV testing is available free of charge for any woman, pregnant or not, who is capable of providing informed consent for the procedure after adequate pretest counselling. Clinicians should be encouraged to discuss the risk of HIV and other sexually transmitted diseases with sexually active women as a matter of course. In Canada routine screening of pregnant women without appropriate counselling and informed consent is just not on.
Unfortunately, Dr. Parker does not appear to have read the discussion accompanying the results that he quotes from the study by Blanche and associates and the HIV Infection in Newborns French Collaborative Study Group.' The authors are quite clear that the role of breast-feeding in HIV transmission remains controversial. They point to the fact that breast-feeding is a major mode of transmission for another retrovirus, human T-cell lymphotropic virus type 1, and they indicate - as did I - that breastfeeding appears to be the route of transmission to infants whose mothers acquire HIV through blood transfusion after delivery. Catherine A. Hankins, MD, FRCPC "At the same time, other prospec- Public health epidemiologist tive or retrospective surveys re- Department of Community Health General Hospital port no difference in the outcome Montreal Montreal, Que. in infants with respect to the mode of feeding." They comment that their own data, involving a Reference small number of infants (6 S, Rouzioux C, Moscato MLG breast-fed v. 99 bottle-fed), tend 1. etBlanche al and the HIV Infection in Newto suggest that breast-feeding may borns French Collaborative Study Group: A prospective study of infants have a role in the transmission of AUGUST 1, 1991
bom to women seropositive for human immunodeficiency virus type 1. N Engi JMed 1989; 320: 1643-1648
Advertising meal replacement products A
recent issue of CMAJ featured a four-page advertisement for Slim-Fast diet products (1991; 144: 633-636). One of the references for the information provided is a personal communication from the Eating Disorders Clinic of Toronto General Hospital. We wish to clear up any possible sense of endorsement from our program for this or any other diet product. Indeed, our whole orientation is toward discouraging dieting behaviour. The advertisement begins with a recognition of the hazards of repetitive weight loss and gain, although it fails to mention that dieting is the commonest precipitant of anorexia nervosa and bulimia nervosa. It soon switches, however, to a discussion of obesity and its health risks without any consideration of what degree of obesity poses medical problems. The generalization regarding obesity is highlighted by the accompanying photograph of a young woman exercising on a stationary bicycle. Next to the picture it is stated that weight management "is a strategy for enhancing the quality of life, and preventing the most common cause of death among Canadians." We see many patients who began with the shape of the model in the advertisement; the quality of their lives has been seriously eroded by dieting. Estimates of rates of death from eating disorders across recent long-term outcome studies are up to 18%.' What is strikingly absent from this advertisement is any evidence to support a beneficial role for the diet product. However, its appearance in a medical journal among pharmaceutical adCAN MED ASSOC J 1991; 145 (3)
199
