people have been displaced from their homes, humanitarian relief from across the border has at least temporarily ensured better conditions.7 The current situation in Iraq is one of stalemate. Two of the principal aims of the UN's sanctions were to press the Iraqi government to yield up all its weapons of mass destruction, including its nuclear capability, and to stop persecuting the Kurds and other groups within the civilian population. These objectives are still far from being achieved, and the repressive regime of Saddam Husseim remains. Meanwhile the Iraqi government continues to refuse to release oil to enable the purchase of food and medical supplies for its own people and uses the increased infant mortality in propaganda to put pressure on the UN to lift sanctions.9 "' Although the extraordinary exertions of the UN relief agencies and donor countries have so far staved off disaster, their scale is inevitably limited and they cannot discharge the duties of the Iraqi government much longer. In any case, there remains the glaring paradox that at a time when the world faces major disasters in impecunious places such as the Horn of Africa the international emergency aid organisations are being asked to pour money into a country that has valuable oil reserves. Iraq's serious health problems require a political solution: the restoration of government with the interests of all sectors of its community at heart. In the interim the condition of children and other vulnerable people remains precarious and demands further measures. The options are few. Other countries should follow the United Kingdom's example and release frozen Iraqi assets for the purchase of items
of humanitarian relief approved by the UN Sanctions Committee. Relief organisations need further financial support to continue to operate throughout 1992. So far only $20m has been pledged of the $145m needed to finance the UN humanitarian agencies' plan of action for the first half of this year, and there is an urgent need to make up the shortfall." For its part the UN should continue its negotiations in Vienna with the Iraqi government to find a mutually acceptable way to implement resolutions 706 and 712. This would enable the provision of materials necessary to ensure the permanent restoration of safe drinking water and the distribution of supplies of medicines and food to those in need. E D ACHESON
Visiting professor in international health, London School of Hygiene and Tropical Medicine, London WC1E 7HT 1 Harvard Stud!' Team. The effect of the Gulf crisis on the children of Iraq. N Engl J Mled 1991 ;325:977-91. 2 Fuld JO, RuLssell RM. Nutrition mission to Iraq. Final report to Unicef b! Tufts University. Boston, Massachusetts: Human Nutrition Research Center, Tufts University, 1991. 3 Aga Khan S. Report to the secretart general on humanitarian needs in Iraq. New York: United Nations, 15 July 1991. 4 International Study Team. Schools of Law and Public Health Harvard University. Health and welfare in Iraq after the Gulf crisis. Child mortality and nutrition survey. A public health study. London: Royal College of Physicians, 22 Oct 1991. 5 Numenthaler M. Report on mission to Iraq on behalf of the International Committee of the Red Cross. Geneva: ICRC, 11 Nox 1991. 6 Iraq immunisation, diarrhoeal disease, maternal and childhood mortality survey. New York: Unicef, 1990. 7 OXFAM. Humanitarian crisis in Iraq. Lpdate No 3. Oxford: OXFAM, 12 Dec 1991. 8 Williamson D. Report on the water supplsy and sanitation situation in South Iraq. London: Save the Children Fund, 1991. 9 Sanctions leave 80000 children dead. Financial Times 1991 Dec 18:4 (cols 1-2). 10 Iraqi children die in battle ouer sanctions. Daily Telegraph 1991 Nov 22:12 (cols 1-7). 11 OXFAM. Humanitarian crisis in Iraq. Update No 4. Oxford: OXFAM, 3 Feb 1991.
When should asymptomatic patients with HIV infection be treated with zidovudine? Rate offall of CD4 count may be a poor guide Infection with HIV is followed by an asymptomatic period of variable length. The best laboratory marker currently available to predict the development of AIDS during this period is the CD4 count' despite variation both diurnally and with age and with time.2 It is unlikely that the circulating CD4 count alone will be totally accurate in this prediction as CD4 cells make up only a small fraction of the total lymphocyte pool and important functional changes in T cells may occur independently of the total count.3 The CD4 count is also being used to indicate when to start antiretroviral treatment in asymptomatic patients. Treatment with zidovudine in asymptomatic people4 and those with early symptoms' delays the development of AIDS in patients with CD4 counts below 0 5 x 1IO/1. As with other large scale trials these studies have been extensively criticised6 -perhaps most cogently because the delay in progression to AIDS may not be translated into longer life. The concern that early treatment might encourage the rapid development of viral resistance to zidovudine seems to be less well founded.7 If these studies had continued for longer-with the CD4 count in treated patients remaining stable for prolonged periods - they might have provided more convincing evidence of the benefit of early treatment. A further indication that early use of zidovudine may delay the development of AIDS was published recently in this journal.' Fewer patients in the Royal Free Hospital's haemophilic cohort developed AIDS in 1989 and 1990 than 456
was predicted from a previous projection,9 possibly because of the increased use of zidovudine. Results from the much larger multicentre AIDS cohort suggest that the benefit of zidovudine is confined to patients with CD4 counts below 035x 109/1.10 The present licensed indication for the use of zidovudine in the United Kingdom includes "asymptomatic patients who have a rapidly falling OK T4 [CD4] count." At seroconversion the CD4 count falls rapidly." In some studies the rate of decline is linear and predicts the rate at which AIDS develops," but in others AIDS is heralded by a period of relative stability followed by a rapid terminal decline in the CD4 count.'3 In the largest study of serial CD4 counts in HIV positive patients computer modelling showed that the rate of fall was linear only over short periods of follow up. ' Whether the rate at which the count falls predicts the risk of AIDS developing was addressed in a study of American HIV positive haemophilic patients published recently in this journal.'5 The rate of fall in the CD4 count, measured on three occasions over nine months, was faster in those who developed AIDS, although this rate of fall was not predictable from the first two counts, casting doubt on the value of one of the licensed indications for the use of zidovudine in the United Kingdom. No consensus currently exists in the United Kingdom about when to start zidovudine in asymptomatic patients with HIV infection. If the drug is effective for only a short time BMJ VOLUME 304
22 FEBRUARY 1992
then treatment should begin when the risk of AIDS is highthat is, when the CD4 count falls below 0 35 x 109/1.10 If the benefits of zidovudine operate for longer in the asymptomatic phase then treatment of patients with counts above this should also be considered. The Anglo-French (Concorde) study, which is due to continue for at least another six months, may help to resolve the dilemma.3 If survival is similar in these two groups the relative merits of extending life expectancy during the asymptomatic phase compared with extending life expectancy when symptoms have supervened are likely to be a matter of continuing debate. This controversy will be fuelled by the recent publication from the Veterans Affairs showing no apparent survival benefit with early, as opposed to delayed, treatment. 6 B G GAZZARD
Consultant physician, Westminster Hospital, London SW1P 2AP 1 Burcham J, Marmor M, Dubin N, Tindall B, Cooper DA, Berrv G, et al. CD4% is the best predictor of development of AIDS in a cohort of HIV-infected homosexual men. AIDS
1991;5:365-72. 2 Malone JL, Simms TE, Gray GC, Wagner KF, Burge JR, Burke DS. Sources of variability in repeated T-helper lymphocyte counts from human immunodeficiency virus type 1-infected patients: total lymphocyte count fluctuations and diurnal cycle are important. J Acquir Immune Defic Syndr 1990;3:144-51.
3 Clerici M, Stocks NI, Zajac RA, Boswell RN, Lucev DR, Via CS, et al. Detection of three distinct patterns of T-helper cell dysfunction in asymptomatic, human immunodeficiency virusseropositive patients. J Clin Investigation 1989;84:1892-9. 4 Volberding PA, Lagakos SW, Koch MA, Pettinelli C, Myers MW, Booth DK, et al. Zidovudine in asymptomatic human immunodeficiency infection-a controlled trial in persons with fewer than 500 CD4-positive cells per cubic millimeter. N EnglJ Med 1990;322:941-9. 5 Fischl MA, Richman DD, Hanson M, Collier AC, Cavey IT, Pava MF, et al. The safety and efficacy of zidovudine (AZT) in the treatment of patients with mildly symptomatic human immuno-deficiency virus type I (HIV infection: a double-blind placebo controlled trial). Ann Intern Med 1990;112:727-37. 6 Zidovudine for symptomless HIV infection [editorial]. Lancet 1990;335:821-2. 7 Richman DD, Grimes JM, Lagakos SW. Effect of stage of disease and drug dose on zidovudine susceptibilities of isolates of human immunodeficiency virus. J Acquir Immune Defic Svndr 1990;3:743-6. 8 Lee CA, Philips AN, Elford J, Janossv G, Griffiths P, Kurnoff P. Progression of HIV disease in a haemophilic cohort followed for 11 years and the effect of treatment. BMJ 1991;303:1093-6. 9 Philips A, Lee CA, Elford J, Janossy G, Bofill M, Timms A, et al. Prediction of progression of AIDS by analysis of CD4 lymphocyte counts in a haemophiliac cohort. AIDS 1989;3:373-41. 10 Graham MMH, Zeger SL, Park LP, Phair JP, Detels R, Vermund SH, et al. Effect of zidovudine in Pneumocvstis carinii pneumonia prophylaxis on progression of HIV infection to AIDS. Lancet
1991;338:265-9. 11 Lang W, Perkins H, Anderson RE, Royce R, Jewell N, Winkelstein W Jr. Patterns of T-lymphocyte changes with human immunodeficiency virus infection: from seroconversion to the development of AIDS. J Acquir Immune Defic Svndr 1989;2:63-9. 12 Phillips AN, Lee CA, Elford J, Janossy G, Timms A, Bofill M, Kernoff PBA. Serial CD4 lymphocyte counts and development of AIDS. Lancet 1991;337:389-92. 13 Kaplan JE, Spira TJ, Fishbein DB, Bozeman LH, Pinksy PF, Shonberger LB. A 6 year follow-up of HIV infected homosexual men with lymphadenopathy. JAMA 1988;260:2694-7. 14 Taylor JMG, Tan SJ, Detels R, Giorgi JV. Applications of a computer simulation model of the
natural history of CD4 T-cell number in HIV-infected individuals. AIDS 1991;5:159-67. 15 Aledort LM, Hilgartner MW, Pike MC, Gjerset GF, Koerper MA, Lian EYC, et al. Variability in serial CD4 counts and relation to progression of HIV-I infection to AIDS in haemophilic patients. BMJ 1992;304:212-6. 16 Hamilton JD, Hartigan PM, Simberkoff MS, Day PL, Diamond GR, Dickinson GM. A controlled trial of early versus late treatment with zidovudine in symptomatic human immunodeficiency virus infection. N Engl3' Med 1992;326:437-43.
Marital breakdown and health More than a broken heart Should the government adopt marital breakdown as one of the key areas in its strategy for improving the health of the nation?' One plus One, an organisation dedicated to "marriage and partnership research," argues that the question deserves careful consideration by doctors and the government, traditionally reluctant to accord social factors an important role in health. Its publication, Marital Breakdown and the Health of the Nation,2 marshals detailed epidemiological evidence for a link between breakdown in relationships and poor physical and mental health and consequent increased mortality. That divorcees of all ages and sexes are at greater risk of premature death than married people has been shown for every country with accurate health statistics for all ages and both sexes.3 For men between the ages of 35 and 45 the risk is doubled.4 Statistics from general practice also show a consistent overall increase in morbidity among divorcees compared with married people.5 This morbidity may result from stress and loss, increased susceptibility to disease, smoking and drinking, and psychological symptoms. Psychiatric consequences of marital breakdown include mainly affective and anxiety disorders, parasuicide, and misuse of alcohol. When the commonest cases of death are investigated divorcees, especially men, are shown to have higher mortality from cardiovascular and cerebrovascular disease,6 cancer,7 suicide,8 and accidental death.9 The impact on the health of children of divorced parents is especially severe, with a higher risk of ill health from the time of parental separation until adult life; children under 5 when their parents divorce are especially vulnerable. Children of divorced parents are much more susceptible to subsequent psychiatric illness. Those whose parents have divorced are more likely to become divorced themselves. Although an association between marital breakdown and BMJ
VOLUME 304
22 FEBRUARY 1992
subsequent ill health seems incontrovertible, the claim for a causal relation is less secure. Disturbed relationships resulting in separation or divorce; increased smoking and drinking; and physical illness associated with increased mortality may all result from affective or neurotic disorders. 10-12 Suffering from a potentially lethal illness may be the immediate precipitant of divorce in some cases. Sorting out what is cause and what is effect is therefore likely to be extremely complicated. And further questions arise. Which has the greater adverse effect on health-the emotional effects of the breakdown in relationships or the socioeconomic consequences of the changed legal status? The more a form of behaviour deviates from current social norms the more likely are its perpetrators to differ from the rest of the population. Thus as marital breakdown, with separation and divorce, has become more common the differences in ill health experienced by divorced and married people compared with other people may have diminished. Some of the references quoted in Marital Breakdown and the Health of the Nation date from the 1960s and '70s, and replication of the studies now would provide valuable information on the importance of the breakdown of relationships to ill health. More recent studies, however, still show substantially better health status for those in continuing relationships. If most of the data on which these authors base their polemic are well known and well validated and point so clearly to the advantage of marital over divorced status why has this not been more generally acknowledged and included in health education? The answer, once again, is complexlying partly in doctors' insistence on accepting new evidence only when it is thoroughly proved and their reluctance to accept information that has practical bearings on individual behaviour. (Cigarette smoking and alcohol misuse spring immediately to mind.) Whereas the faintest suspicion of risk 457
of humanitarian relief approved by the UN Sanctions Committee. Relief organisations need further financial support to continue to operate throughout 1992. So far only $20m has been pledged of the $145m needed to finance the UN humanitarian agencies' plan of action for the first half of this year, and there is an urgent need to make up the shortfall." For its part the UN should continue its negotiations in Vienna with the Iraqi government to find a mutually acceptable way to implement resolutions 706 and 712. This would enable the provision of materials necessary to ensure the permanent restoration of safe drinking water and the distribution of supplies of medicines and food to those in need. E D ACHESON
Visiting professor in international health, London School of Hygiene and Tropical Medicine, London WC1E 7HT 1 Harvard Stud!' Team. The effect of the Gulf crisis on the children of Iraq. N Engl J Mled 1991 ;325:977-91. 2 Fuld JO, RuLssell RM. Nutrition mission to Iraq. Final report to Unicef b! Tufts University. Boston, Massachusetts: Human Nutrition Research Center, Tufts University, 1991. 3 Aga Khan S. Report to the secretart general on humanitarian needs in Iraq. New York: United Nations, 15 July 1991. 4 International Study Team. Schools of Law and Public Health Harvard University. Health and welfare in Iraq after the Gulf crisis. Child mortality and nutrition survey. A public health study. London: Royal College of Physicians, 22 Oct 1991. 5 Numenthaler M. Report on mission to Iraq on behalf of the International Committee of the Red Cross. Geneva: ICRC, 11 Nox 1991. 6 Iraq immunisation, diarrhoeal disease, maternal and childhood mortality survey. New York: Unicef, 1990. 7 OXFAM. Humanitarian crisis in Iraq. Lpdate No 3. Oxford: OXFAM, 12 Dec 1991. 8 Williamson D. Report on the water supplsy and sanitation situation in South Iraq. London: Save the Children Fund, 1991. 9 Sanctions leave 80000 children dead. Financial Times 1991 Dec 18:4 (cols 1-2). 10 Iraqi children die in battle ouer sanctions. Daily Telegraph 1991 Nov 22:12 (cols 1-7). 11 OXFAM. Humanitarian crisis in Iraq. Update No 4. Oxford: OXFAM, 3 Feb 1991.
When should asymptomatic patients with HIV infection be treated with zidovudine? Rate offall of CD4 count may be a poor guide Infection with HIV is followed by an asymptomatic period of variable length. The best laboratory marker currently available to predict the development of AIDS during this period is the CD4 count' despite variation both diurnally and with age and with time.2 It is unlikely that the circulating CD4 count alone will be totally accurate in this prediction as CD4 cells make up only a small fraction of the total lymphocyte pool and important functional changes in T cells may occur independently of the total count.3 The CD4 count is also being used to indicate when to start antiretroviral treatment in asymptomatic patients. Treatment with zidovudine in asymptomatic people4 and those with early symptoms' delays the development of AIDS in patients with CD4 counts below 0 5 x 1IO/1. As with other large scale trials these studies have been extensively criticised6 -perhaps most cogently because the delay in progression to AIDS may not be translated into longer life. The concern that early treatment might encourage the rapid development of viral resistance to zidovudine seems to be less well founded.7 If these studies had continued for longer-with the CD4 count in treated patients remaining stable for prolonged periods - they might have provided more convincing evidence of the benefit of early treatment. A further indication that early use of zidovudine may delay the development of AIDS was published recently in this journal.' Fewer patients in the Royal Free Hospital's haemophilic cohort developed AIDS in 1989 and 1990 than 456
was predicted from a previous projection,9 possibly because of the increased use of zidovudine. Results from the much larger multicentre AIDS cohort suggest that the benefit of zidovudine is confined to patients with CD4 counts below 035x 109/1.10 The present licensed indication for the use of zidovudine in the United Kingdom includes "asymptomatic patients who have a rapidly falling OK T4 [CD4] count." At seroconversion the CD4 count falls rapidly." In some studies the rate of decline is linear and predicts the rate at which AIDS develops," but in others AIDS is heralded by a period of relative stability followed by a rapid terminal decline in the CD4 count.'3 In the largest study of serial CD4 counts in HIV positive patients computer modelling showed that the rate of fall was linear only over short periods of follow up. ' Whether the rate at which the count falls predicts the risk of AIDS developing was addressed in a study of American HIV positive haemophilic patients published recently in this journal.'5 The rate of fall in the CD4 count, measured on three occasions over nine months, was faster in those who developed AIDS, although this rate of fall was not predictable from the first two counts, casting doubt on the value of one of the licensed indications for the use of zidovudine in the United Kingdom. No consensus currently exists in the United Kingdom about when to start zidovudine in asymptomatic patients with HIV infection. If the drug is effective for only a short time BMJ VOLUME 304
22 FEBRUARY 1992
then treatment should begin when the risk of AIDS is highthat is, when the CD4 count falls below 0 35 x 109/1.10 If the benefits of zidovudine operate for longer in the asymptomatic phase then treatment of patients with counts above this should also be considered. The Anglo-French (Concorde) study, which is due to continue for at least another six months, may help to resolve the dilemma.3 If survival is similar in these two groups the relative merits of extending life expectancy during the asymptomatic phase compared with extending life expectancy when symptoms have supervened are likely to be a matter of continuing debate. This controversy will be fuelled by the recent publication from the Veterans Affairs showing no apparent survival benefit with early, as opposed to delayed, treatment. 6 B G GAZZARD
Consultant physician, Westminster Hospital, London SW1P 2AP 1 Burcham J, Marmor M, Dubin N, Tindall B, Cooper DA, Berrv G, et al. CD4% is the best predictor of development of AIDS in a cohort of HIV-infected homosexual men. AIDS
1991;5:365-72. 2 Malone JL, Simms TE, Gray GC, Wagner KF, Burge JR, Burke DS. Sources of variability in repeated T-helper lymphocyte counts from human immunodeficiency virus type 1-infected patients: total lymphocyte count fluctuations and diurnal cycle are important. J Acquir Immune Defic Syndr 1990;3:144-51.
3 Clerici M, Stocks NI, Zajac RA, Boswell RN, Lucev DR, Via CS, et al. Detection of three distinct patterns of T-helper cell dysfunction in asymptomatic, human immunodeficiency virusseropositive patients. J Clin Investigation 1989;84:1892-9. 4 Volberding PA, Lagakos SW, Koch MA, Pettinelli C, Myers MW, Booth DK, et al. Zidovudine in asymptomatic human immunodeficiency infection-a controlled trial in persons with fewer than 500 CD4-positive cells per cubic millimeter. N EnglJ Med 1990;322:941-9. 5 Fischl MA, Richman DD, Hanson M, Collier AC, Cavey IT, Pava MF, et al. The safety and efficacy of zidovudine (AZT) in the treatment of patients with mildly symptomatic human immuno-deficiency virus type I (HIV infection: a double-blind placebo controlled trial). Ann Intern Med 1990;112:727-37. 6 Zidovudine for symptomless HIV infection [editorial]. Lancet 1990;335:821-2. 7 Richman DD, Grimes JM, Lagakos SW. Effect of stage of disease and drug dose on zidovudine susceptibilities of isolates of human immunodeficiency virus. J Acquir Immune Defic Svndr 1990;3:743-6. 8 Lee CA, Philips AN, Elford J, Janossv G, Griffiths P, Kurnoff P. Progression of HIV disease in a haemophilic cohort followed for 11 years and the effect of treatment. BMJ 1991;303:1093-6. 9 Philips A, Lee CA, Elford J, Janossy G, Bofill M, Timms A, et al. Prediction of progression of AIDS by analysis of CD4 lymphocyte counts in a haemophiliac cohort. AIDS 1989;3:373-41. 10 Graham MMH, Zeger SL, Park LP, Phair JP, Detels R, Vermund SH, et al. Effect of zidovudine in Pneumocvstis carinii pneumonia prophylaxis on progression of HIV infection to AIDS. Lancet
1991;338:265-9. 11 Lang W, Perkins H, Anderson RE, Royce R, Jewell N, Winkelstein W Jr. Patterns of T-lymphocyte changes with human immunodeficiency virus infection: from seroconversion to the development of AIDS. J Acquir Immune Defic Svndr 1989;2:63-9. 12 Phillips AN, Lee CA, Elford J, Janossy G, Timms A, Bofill M, Kernoff PBA. Serial CD4 lymphocyte counts and development of AIDS. Lancet 1991;337:389-92. 13 Kaplan JE, Spira TJ, Fishbein DB, Bozeman LH, Pinksy PF, Shonberger LB. A 6 year follow-up of HIV infected homosexual men with lymphadenopathy. JAMA 1988;260:2694-7. 14 Taylor JMG, Tan SJ, Detels R, Giorgi JV. Applications of a computer simulation model of the
natural history of CD4 T-cell number in HIV-infected individuals. AIDS 1991;5:159-67. 15 Aledort LM, Hilgartner MW, Pike MC, Gjerset GF, Koerper MA, Lian EYC, et al. Variability in serial CD4 counts and relation to progression of HIV-I infection to AIDS in haemophilic patients. BMJ 1992;304:212-6. 16 Hamilton JD, Hartigan PM, Simberkoff MS, Day PL, Diamond GR, Dickinson GM. A controlled trial of early versus late treatment with zidovudine in symptomatic human immunodeficiency virus infection. N Engl3' Med 1992;326:437-43.
Marital breakdown and health More than a broken heart Should the government adopt marital breakdown as one of the key areas in its strategy for improving the health of the nation?' One plus One, an organisation dedicated to "marriage and partnership research," argues that the question deserves careful consideration by doctors and the government, traditionally reluctant to accord social factors an important role in health. Its publication, Marital Breakdown and the Health of the Nation,2 marshals detailed epidemiological evidence for a link between breakdown in relationships and poor physical and mental health and consequent increased mortality. That divorcees of all ages and sexes are at greater risk of premature death than married people has been shown for every country with accurate health statistics for all ages and both sexes.3 For men between the ages of 35 and 45 the risk is doubled.4 Statistics from general practice also show a consistent overall increase in morbidity among divorcees compared with married people.5 This morbidity may result from stress and loss, increased susceptibility to disease, smoking and drinking, and psychological symptoms. Psychiatric consequences of marital breakdown include mainly affective and anxiety disorders, parasuicide, and misuse of alcohol. When the commonest cases of death are investigated divorcees, especially men, are shown to have higher mortality from cardiovascular and cerebrovascular disease,6 cancer,7 suicide,8 and accidental death.9 The impact on the health of children of divorced parents is especially severe, with a higher risk of ill health from the time of parental separation until adult life; children under 5 when their parents divorce are especially vulnerable. Children of divorced parents are much more susceptible to subsequent psychiatric illness. Those whose parents have divorced are more likely to become divorced themselves. Although an association between marital breakdown and BMJ
VOLUME 304
22 FEBRUARY 1992
subsequent ill health seems incontrovertible, the claim for a causal relation is less secure. Disturbed relationships resulting in separation or divorce; increased smoking and drinking; and physical illness associated with increased mortality may all result from affective or neurotic disorders. 10-12 Suffering from a potentially lethal illness may be the immediate precipitant of divorce in some cases. Sorting out what is cause and what is effect is therefore likely to be extremely complicated. And further questions arise. Which has the greater adverse effect on health-the emotional effects of the breakdown in relationships or the socioeconomic consequences of the changed legal status? The more a form of behaviour deviates from current social norms the more likely are its perpetrators to differ from the rest of the population. Thus as marital breakdown, with separation and divorce, has become more common the differences in ill health experienced by divorced and married people compared with other people may have diminished. Some of the references quoted in Marital Breakdown and the Health of the Nation date from the 1960s and '70s, and replication of the studies now would provide valuable information on the importance of the breakdown of relationships to ill health. More recent studies, however, still show substantially better health status for those in continuing relationships. If most of the data on which these authors base their polemic are well known and well validated and point so clearly to the advantage of marital over divorced status why has this not been more generally acknowledged and included in health education? The answer, once again, is complexlying partly in doctors' insistence on accepting new evidence only when it is thoroughly proved and their reluctance to accept information that has practical bearings on individual behaviour. (Cigarette smoking and alcohol misuse spring immediately to mind.) Whereas the faintest suspicion of risk 457
