- Email: [email protected]
Psychiatric consultation and quality of decision making in euthanasia
RESEARCH LETTERS
The findings with respect to offspring birthweight and CVD mortality in this study are similar to those from previous studies.2,3 Associations between birth dimensions and mortality from smoking-related cancers have also been reported. We did not have data on smoking in this study, however since maternal smoking is related to lower offspring birthweight, shorter birth length, and lower ponderal index, it is likely that smoking contributes to the findings with respect to CVD mortality. Confounding by smoking should produce considerably greater effects for smoking-related cancer mortality than for CVD mortality, since the association of smoking with smoking-related cancer mortality is considerably stronger than with CVD mortality. Relative risks for smokers compared with those who have never smoked are 4·2 (95% CI 3·1–5·6) for smoking-related cancer mortality, and 1·7 (1·5–1·9) for CVD mortality among women in the Renfrew and Paisley study2 (Carole Hart, personal communication). In the present study, associations between birth dimension and CVD mortality, and between birth dimension and smoking-related cancer mortality, are of the same magnitude. Since confounding by smoking should produce much stronger associations with smoking-related cancer mortality than CVD mortality, the implication is that offspring birth dimensions have an influence on maternal CVD risk over and above that caused by smoking. This is in agreement with a previous study in which data on maternal smoking were available and in which adjustment for this variable had little effect on the association between birthweight and CVD mortality in the offspring.2 Previous studies of birth outcomes and long-term maternal mortality have not assessed the effects of premature delivery. We have shown a strong association of premature delivery with CVD mortality, and no association with mortality from smoking-related cancers, which suggests that confounding by smoking is not responsible for the former association. For breast cancer, case-series data have suggested that women who give birth to large babies are at increased risk.5 Maternal insulin-like growth factor-1 concentrations are related to both offspring ponderal index and to breast-cancer risk. Accident and violence mortality was included because unmeasured socioeconomic factors would be expected to produce associations between birth outcomes and such mortality. No such associations were seen. In conclusion, maternal factors which affect intrauterine growth and premature delivery seem to have important influences on long-term mortality risk in women. We thank Carole Hart for providing data on mortality risk among women in the Renfrew and Paisley study 1 2
3
4
5
Barker DJP. Mothers, babies and health in later life. London: Churchill-Livingstone, 1998. Davey Smith G, Hart C, Ferrell C, et al. Birthweight of offspring and mortality in the Renfrew and Paisley study: prospective observational study. BMJ 1997; 315: 1189–93. Davey Smith G, Harding S, Rosato M. Relationship between infants’ birth weight and mothers’ mortality: prospective observational study. BMJ 2000; 320: 839–40. Hemminki E, Gissler M, Toukomaa H. Exposure to female hormone drugs during pregnancy: effect on malformations and cancer. Br J Cancer 1999; 80: 1092–97. Berstein LM. Newborn macrosomy and cancer. Adv Cancer Res 1988; 50: 231–78.
Department of Social Medicine, University of Bristol, Canynge Hall, Bristol BS8 2PR, UK (Prof G D Smith DSc, E Whitley PhD); National Research and Development Centre for Welfare and Health, Health Services Research Unit, Helsinki, Finland (M Gissler DPhil, E Hemminki MD) Correspondence to: Prof George Davey Smith
THE LANCET • Vol 356 • December 16, 2000
Psychiatric consultation and quality of decision making in euthanasia Marjolein Bannink, Arthur R Van Gool, Agnes van der Heide, Paul J van der Maas The role of undiagnosed psychiatric disorders in patients requesting euthanasia is an important ethical and professional issue. From June, 1997, until June, 1999, we included a psychiatric consultation as part of our standard procedure for handling euthanasia requests. Of 22 cancer patients who requested euthanasia at short notice: ten had their longstanding and well-considered wish for euthanasia granted; six were denied euthanasia because they did not have such a wish; five were denied because of psychiatric problems; and one was granted the wish despite minor psychiatric symptoms. Our findings suggest that a psychiatrist should be consulted if the treatment team has doubts about a patient’s state of mind.
The standard practice for handling euthanasia requests at the University Hospital Rotterdam-Daniel is based on the Royal Dutch Medical Association’s procedural guidelines, and establishes whether or not demands originate from a longstanding and well-considered wish to die.1 From June 1, 1997, until June 1, 1999, psychiatric consultation was added to this standard procedure, to determine whether appeals for euthanasia were influenced by psychiatric problems.2 During these 2 years, we (MB, AVG) were consulted by twelve female and ten male cancer patients who requested euthanasia at short notice. The mean age of these patients was 60 years (range 37–77). Following consultation, ten individuals were thought to have a longlasting and well-considered wish for euthanasia that did not originate from undue pressure from others or intolerable pain. Additionally, none of the patients had current psychiatric disorders, although three had a history of psychiatric problems. The patients’ requests were therefore granted, though three died from natural causes before the date set for euthanasia. The remainder died as a result of euthanasia either at home (four) or in hospital (three). In six of the 22 patients, all women with no psychiatric disorders, consultation with a psychiatrist established that the request for euthanasia had not been fully considered. One patient’s appeal seemed to be based on inaccurate medical information received before she was transferred to our hospital and, after adequate diagnostic testing and discussion of results, she withdrew her request. Four other patients did not sustain their requests for euthanasia after examination by the psychiatrist; they seemed to regard euthanasia, or having the option of euthanasia, as a way to control their feelings of despair, pain, and helplessness. According to the psychiatrist and the primary physician, their appeals were mainly aimed at explicitly discussing their situation with the attending physician, to reduce anxiety and insecurity. The sixth patient agreed to wait 4 days to be able to further discuss her request for euthanasia with the primary oncologist most acquainted with her case, who was abroad. She died before he returned. The final six patients, all men, showed psychiatric symptoms. Of these, five had cognitive or depressive disorders, which decreased their competence in decision making. Since these diagnoses concurred with doubts about the durability of the request, euthanasia was denied. Four of these patients died from natural causes within a week of psychiatric consultation, and one died from natural causes in a nursing home 4 months later. In two of these cases the primary physician had underestimated the influence of a psychiatric disorder on patient competence and decision-making abilities, and had considered complying with the request. In both cases the
2067
For personal use only. Not to be reproduced without permission of The Lancet.
RESEARCH LETTERS
psychiatrist’s involvement changed the policy and euthanasia was denied. One patient had depressive symptoms and slight cognitive impairment, but not to a degree which implicated impaired decision-making capacities.4 Because his request was in line with earlier expressed explicit opinions on his illness and his impending death, and his wish for autonomy, we agreed to comply with his wishes. This patient died as a result of euthanasia. Of 22 patients who requested euthanasia therefore, eight died according to their wishes and two were denied euthanasia as a direct result of psychiatric consultation. By not complying with the wish of these terminally ill patients we neither lengthened not shortened their lives, but did alter the way in which they died. In our opinion, this result does not automatically imply that standard psychiatric consultation should be mandatory.2,3 The benefits of such consultation should be balanced against the disadvantages of pushing the psychiatrist to the fore as the final gatekeeper.4,5 However, in general, staff appreciate the expertise, help, and support given by a consultant psychiatrist, who might add to the quality of the decisionmaking process in specific cases. 1
2
3 4 5
van der Maas PJ, van der Wal G, Haverkate I, et al. Euthanasia, physician-assisted suicide, and other medical practices involving the end of life in The Netherlands, 1990-1995. N Engl J Med 1996; 335: 1699–705. Baile WF, DiMaggio JR, Schapira DV, Janofsky JS. The request for assistance in dying: the need for psychiatric consultation. Cancer 1993; 72: 2786–91. Muskin PR. The request to die. JAMA. 1998; 279: 323–28. Zaubler TS. The unexamined death is not worth dying. Psychosomatics 2000; 41: 193–94. Sullivan MD, Ganzini L, Youngner SJ. Should psychiatrists serve as gatekeepers for physician-assisted suicide? Hastings Center Rep 1998; 28: 24–31.
Department of Psychosocial Oncology, University Hospital Rotterdam-Daniel, PO Box 5201, 3008 AE Rotterdam, Netherlands (M Bannink MD, A R Van Gool MD); and Department of Public Health, Erasmus University, Rotterdam (A van der Heide MD, Prof P J van der Maas MD) Correspondence to: Dr M Bannink (e-mail: [email protected])
Increased risk of stroke in patients with the A12308G polymorphism in mitochondria
Clinical features
Cases (%)
Cases with A12308G
Strokes Deafness Diabetes mellitus Myopathy Encephalopathy Retinitis pigmentosa Ophthalmoplegia Neuropathy MERRF
21 (44) 19 (40) 13 (27) 12 (25) 6 (13) 4 (8) 3 (6) 2 (4) 1 (2)
7 3 2 1 1 0 1 0 1
MERRF=myoclonic epilepsy and ragged red fibres. The cases who also had A12308G are indicated.
Table 1: Frequency of clinical features in 48 patients with A3243G
patients with the A3243G mutation have strokes. Other A3243G phenotypes include myopathy, chronic progressive external ophthalmoplegia (CPEO), diabetes mellitus, and deafness. The mechanisms underlying this diversity remain unclear and the factors which lead to strokes have not been identified. In 1998, El Meziane and colleagues1 identified a G12300A mutation in the mitochondrial transfer RNA for the leucine tRNALeu(CUN) gene in lung cancer cybrid cells containing the A3243G mutation. The G12300A mutation was shown to ameliorate the impaired respiratory-chain function caused by the A3243G mutation, suggesting the possibility that tRNALeu(CUN) polymorphisms may be relevant to the phenotypic diversity seen in human beings.1,2 We investigated the possible role of mitochondrial tRNALeu(CUN) gene polymorphisms in influencing 3243associated phenotypes in 48 unrelated people with A3243G (table 1). Automated sequencing of tRNALeu(CUN) failed to identify the G12300A mutation in these people. Because El Meziane and colleagues reported that as little as 11% of G12300A could correct the biochemical abnormalities in their cell system, we also did fluorescent-based PCR and restriction fragment length polymorphism analysis which allows detection of very low proportions of this mutation (down to at least 5%). Even with this assay, none of the 48 cases had detectable G12300A, indicating that this mutation may not be relevant to disease associated with A3243G. We did, however, identify one homoplasmic change, A12308G, in the tRNALeu(CUN) gene compared with the Cambridge mitochondrial sequence in nine patients (18·8%).3 We analysed the link between clinical features and this change. Patients with the A12308G transition had
T Pulkes, M G Sweeney, M G Hanna Factors which increase the risk of stroke in patients with the A3243G (mitochondrial encephalomyopathy, lactic acidosis, and stroke [MELAS]) mutation in human mitochondrial DNA are unclear. Previous work on lung-cancer cells with an A3243G mutation showed that a mutation in the mitochondrial transfer gene for leucine tRNALeu(CUN) was able to ameliorate the A3243G-induced biochemical phenotype. We analysed the tRNALeu(CUN) gene in 48 unrelated A3243G cases. We showed that a polymorphism, A12308G, in tRNALeu(CUN) increases the risk of developing stroke in patients with the A3243G mutation (relative risk=2·17). This may have implications for genetic counselling.
Mitochondrial encephalomyopathy, lactic acidosis, and stroke (MELAS) is a life-threatening mitochondrial encephalomyopathy characterised by strokes, and is commonly diagnosed at a young age. The commonest mitochondrial DNA point mutation associated with MELAS is the A3243G mutation. However, only 50% of
2068
Patient Phenotype number
Mitochondrial transfer DNA haplogroup
1
16298 72 195 200
2 3 4 5 6 7 8 9
Encephalopathy, deafness, diabetes mellitus MELAS, deafness MELAS Myopathy, ophthalmoplegia MELAS MELAS MELAS, deafness, diabetes mellitus MELAS MELAS, MERRF
57 61 146 183 16129 16192 16256 16270 16399 73 16192 16256 16270 16362 16399 16428 16129 16224 16298 16318 16519 73 195 16129 16140 16224 16298 16362 16519 73 152 16069 16126 73 185 195 228 295 16126 16294 16296 16304 16519 73 16093 16224 73 152
Haplotypes are given as the Cambridge mitochondrial sequence.3 The base changes were C→T at 16069, 16192, 16256, 16270, 16294, 16296, 61; T→C at 16093, 16126, 16140, 16224, 16298, 16304, 16311, 16362, 16519, 57, 72, 146, 152, 195; G→A at 16129, 16428, 185, 228; A→G at 16318, 16399, 72, 183, 200. MERRF=myoclonic epilepsy and ragged red fibres.
Table 2: Mitochondrial transfer DNA sequence haplogroups in patients with A3243G and A12308G
THE LANCET • Vol 356 • December 16, 2000
For personal use only. Not to be reproduced without permission of The Lancet.
The findings with respect to offspring birthweight and CVD mortality in this study are similar to those from previous studies.2,3 Associations between birth dimensions and mortality from smoking-related cancers have also been reported. We did not have data on smoking in this study, however since maternal smoking is related to lower offspring birthweight, shorter birth length, and lower ponderal index, it is likely that smoking contributes to the findings with respect to CVD mortality. Confounding by smoking should produce considerably greater effects for smoking-related cancer mortality than for CVD mortality, since the association of smoking with smoking-related cancer mortality is considerably stronger than with CVD mortality. Relative risks for smokers compared with those who have never smoked are 4·2 (95% CI 3·1–5·6) for smoking-related cancer mortality, and 1·7 (1·5–1·9) for CVD mortality among women in the Renfrew and Paisley study2 (Carole Hart, personal communication). In the present study, associations between birth dimension and CVD mortality, and between birth dimension and smoking-related cancer mortality, are of the same magnitude. Since confounding by smoking should produce much stronger associations with smoking-related cancer mortality than CVD mortality, the implication is that offspring birth dimensions have an influence on maternal CVD risk over and above that caused by smoking. This is in agreement with a previous study in which data on maternal smoking were available and in which adjustment for this variable had little effect on the association between birthweight and CVD mortality in the offspring.2 Previous studies of birth outcomes and long-term maternal mortality have not assessed the effects of premature delivery. We have shown a strong association of premature delivery with CVD mortality, and no association with mortality from smoking-related cancers, which suggests that confounding by smoking is not responsible for the former association. For breast cancer, case-series data have suggested that women who give birth to large babies are at increased risk.5 Maternal insulin-like growth factor-1 concentrations are related to both offspring ponderal index and to breast-cancer risk. Accident and violence mortality was included because unmeasured socioeconomic factors would be expected to produce associations between birth outcomes and such mortality. No such associations were seen. In conclusion, maternal factors which affect intrauterine growth and premature delivery seem to have important influences on long-term mortality risk in women. We thank Carole Hart for providing data on mortality risk among women in the Renfrew and Paisley study 1 2
3
4
5
Barker DJP. Mothers, babies and health in later life. London: Churchill-Livingstone, 1998. Davey Smith G, Hart C, Ferrell C, et al. Birthweight of offspring and mortality in the Renfrew and Paisley study: prospective observational study. BMJ 1997; 315: 1189–93. Davey Smith G, Harding S, Rosato M. Relationship between infants’ birth weight and mothers’ mortality: prospective observational study. BMJ 2000; 320: 839–40. Hemminki E, Gissler M, Toukomaa H. Exposure to female hormone drugs during pregnancy: effect on malformations and cancer. Br J Cancer 1999; 80: 1092–97. Berstein LM. Newborn macrosomy and cancer. Adv Cancer Res 1988; 50: 231–78.
Department of Social Medicine, University of Bristol, Canynge Hall, Bristol BS8 2PR, UK (Prof G D Smith DSc, E Whitley PhD); National Research and Development Centre for Welfare and Health, Health Services Research Unit, Helsinki, Finland (M Gissler DPhil, E Hemminki MD) Correspondence to: Prof George Davey Smith
THE LANCET • Vol 356 • December 16, 2000
Psychiatric consultation and quality of decision making in euthanasia Marjolein Bannink, Arthur R Van Gool, Agnes van der Heide, Paul J van der Maas The role of undiagnosed psychiatric disorders in patients requesting euthanasia is an important ethical and professional issue. From June, 1997, until June, 1999, we included a psychiatric consultation as part of our standard procedure for handling euthanasia requests. Of 22 cancer patients who requested euthanasia at short notice: ten had their longstanding and well-considered wish for euthanasia granted; six were denied euthanasia because they did not have such a wish; five were denied because of psychiatric problems; and one was granted the wish despite minor psychiatric symptoms. Our findings suggest that a psychiatrist should be consulted if the treatment team has doubts about a patient’s state of mind.
The standard practice for handling euthanasia requests at the University Hospital Rotterdam-Daniel is based on the Royal Dutch Medical Association’s procedural guidelines, and establishes whether or not demands originate from a longstanding and well-considered wish to die.1 From June 1, 1997, until June 1, 1999, psychiatric consultation was added to this standard procedure, to determine whether appeals for euthanasia were influenced by psychiatric problems.2 During these 2 years, we (MB, AVG) were consulted by twelve female and ten male cancer patients who requested euthanasia at short notice. The mean age of these patients was 60 years (range 37–77). Following consultation, ten individuals were thought to have a longlasting and well-considered wish for euthanasia that did not originate from undue pressure from others or intolerable pain. Additionally, none of the patients had current psychiatric disorders, although three had a history of psychiatric problems. The patients’ requests were therefore granted, though three died from natural causes before the date set for euthanasia. The remainder died as a result of euthanasia either at home (four) or in hospital (three). In six of the 22 patients, all women with no psychiatric disorders, consultation with a psychiatrist established that the request for euthanasia had not been fully considered. One patient’s appeal seemed to be based on inaccurate medical information received before she was transferred to our hospital and, after adequate diagnostic testing and discussion of results, she withdrew her request. Four other patients did not sustain their requests for euthanasia after examination by the psychiatrist; they seemed to regard euthanasia, or having the option of euthanasia, as a way to control their feelings of despair, pain, and helplessness. According to the psychiatrist and the primary physician, their appeals were mainly aimed at explicitly discussing their situation with the attending physician, to reduce anxiety and insecurity. The sixth patient agreed to wait 4 days to be able to further discuss her request for euthanasia with the primary oncologist most acquainted with her case, who was abroad. She died before he returned. The final six patients, all men, showed psychiatric symptoms. Of these, five had cognitive or depressive disorders, which decreased their competence in decision making. Since these diagnoses concurred with doubts about the durability of the request, euthanasia was denied. Four of these patients died from natural causes within a week of psychiatric consultation, and one died from natural causes in a nursing home 4 months later. In two of these cases the primary physician had underestimated the influence of a psychiatric disorder on patient competence and decision-making abilities, and had considered complying with the request. In both cases the
2067
For personal use only. Not to be reproduced without permission of The Lancet.
RESEARCH LETTERS
psychiatrist’s involvement changed the policy and euthanasia was denied. One patient had depressive symptoms and slight cognitive impairment, but not to a degree which implicated impaired decision-making capacities.4 Because his request was in line with earlier expressed explicit opinions on his illness and his impending death, and his wish for autonomy, we agreed to comply with his wishes. This patient died as a result of euthanasia. Of 22 patients who requested euthanasia therefore, eight died according to their wishes and two were denied euthanasia as a direct result of psychiatric consultation. By not complying with the wish of these terminally ill patients we neither lengthened not shortened their lives, but did alter the way in which they died. In our opinion, this result does not automatically imply that standard psychiatric consultation should be mandatory.2,3 The benefits of such consultation should be balanced against the disadvantages of pushing the psychiatrist to the fore as the final gatekeeper.4,5 However, in general, staff appreciate the expertise, help, and support given by a consultant psychiatrist, who might add to the quality of the decisionmaking process in specific cases. 1
2
3 4 5
van der Maas PJ, van der Wal G, Haverkate I, et al. Euthanasia, physician-assisted suicide, and other medical practices involving the end of life in The Netherlands, 1990-1995. N Engl J Med 1996; 335: 1699–705. Baile WF, DiMaggio JR, Schapira DV, Janofsky JS. The request for assistance in dying: the need for psychiatric consultation. Cancer 1993; 72: 2786–91. Muskin PR. The request to die. JAMA. 1998; 279: 323–28. Zaubler TS. The unexamined death is not worth dying. Psychosomatics 2000; 41: 193–94. Sullivan MD, Ganzini L, Youngner SJ. Should psychiatrists serve as gatekeepers for physician-assisted suicide? Hastings Center Rep 1998; 28: 24–31.
Department of Psychosocial Oncology, University Hospital Rotterdam-Daniel, PO Box 5201, 3008 AE Rotterdam, Netherlands (M Bannink MD, A R Van Gool MD); and Department of Public Health, Erasmus University, Rotterdam (A van der Heide MD, Prof P J van der Maas MD) Correspondence to: Dr M Bannink (e-mail: [email protected])
Increased risk of stroke in patients with the A12308G polymorphism in mitochondria
Clinical features
Cases (%)
Cases with A12308G
Strokes Deafness Diabetes mellitus Myopathy Encephalopathy Retinitis pigmentosa Ophthalmoplegia Neuropathy MERRF
21 (44) 19 (40) 13 (27) 12 (25) 6 (13) 4 (8) 3 (6) 2 (4) 1 (2)
7 3 2 1 1 0 1 0 1
MERRF=myoclonic epilepsy and ragged red fibres. The cases who also had A12308G are indicated.
Table 1: Frequency of clinical features in 48 patients with A3243G
patients with the A3243G mutation have strokes. Other A3243G phenotypes include myopathy, chronic progressive external ophthalmoplegia (CPEO), diabetes mellitus, and deafness. The mechanisms underlying this diversity remain unclear and the factors which lead to strokes have not been identified. In 1998, El Meziane and colleagues1 identified a G12300A mutation in the mitochondrial transfer RNA for the leucine tRNALeu(CUN) gene in lung cancer cybrid cells containing the A3243G mutation. The G12300A mutation was shown to ameliorate the impaired respiratory-chain function caused by the A3243G mutation, suggesting the possibility that tRNALeu(CUN) polymorphisms may be relevant to the phenotypic diversity seen in human beings.1,2 We investigated the possible role of mitochondrial tRNALeu(CUN) gene polymorphisms in influencing 3243associated phenotypes in 48 unrelated people with A3243G (table 1). Automated sequencing of tRNALeu(CUN) failed to identify the G12300A mutation in these people. Because El Meziane and colleagues reported that as little as 11% of G12300A could correct the biochemical abnormalities in their cell system, we also did fluorescent-based PCR and restriction fragment length polymorphism analysis which allows detection of very low proportions of this mutation (down to at least 5%). Even with this assay, none of the 48 cases had detectable G12300A, indicating that this mutation may not be relevant to disease associated with A3243G. We did, however, identify one homoplasmic change, A12308G, in the tRNALeu(CUN) gene compared with the Cambridge mitochondrial sequence in nine patients (18·8%).3 We analysed the link between clinical features and this change. Patients with the A12308G transition had
T Pulkes, M G Sweeney, M G Hanna Factors which increase the risk of stroke in patients with the A3243G (mitochondrial encephalomyopathy, lactic acidosis, and stroke [MELAS]) mutation in human mitochondrial DNA are unclear. Previous work on lung-cancer cells with an A3243G mutation showed that a mutation in the mitochondrial transfer gene for leucine tRNALeu(CUN) was able to ameliorate the A3243G-induced biochemical phenotype. We analysed the tRNALeu(CUN) gene in 48 unrelated A3243G cases. We showed that a polymorphism, A12308G, in tRNALeu(CUN) increases the risk of developing stroke in patients with the A3243G mutation (relative risk=2·17). This may have implications for genetic counselling.
Mitochondrial encephalomyopathy, lactic acidosis, and stroke (MELAS) is a life-threatening mitochondrial encephalomyopathy characterised by strokes, and is commonly diagnosed at a young age. The commonest mitochondrial DNA point mutation associated with MELAS is the A3243G mutation. However, only 50% of
2068
Patient Phenotype number
Mitochondrial transfer DNA haplogroup
1
16298 72 195 200
2 3 4 5 6 7 8 9
Encephalopathy, deafness, diabetes mellitus MELAS, deafness MELAS Myopathy, ophthalmoplegia MELAS MELAS MELAS, deafness, diabetes mellitus MELAS MELAS, MERRF
57 61 146 183 16129 16192 16256 16270 16399 73 16192 16256 16270 16362 16399 16428 16129 16224 16298 16318 16519 73 195 16129 16140 16224 16298 16362 16519 73 152 16069 16126 73 185 195 228 295 16126 16294 16296 16304 16519 73 16093 16224 73 152
Haplotypes are given as the Cambridge mitochondrial sequence.3 The base changes were C→T at 16069, 16192, 16256, 16270, 16294, 16296, 61; T→C at 16093, 16126, 16140, 16224, 16298, 16304, 16311, 16362, 16519, 57, 72, 146, 152, 195; G→A at 16129, 16428, 185, 228; A→G at 16318, 16399, 72, 183, 200. MERRF=myoclonic epilepsy and ragged red fibres.
Table 2: Mitochondrial transfer DNA sequence haplogroups in patients with A3243G and A12308G
THE LANCET • Vol 356 • December 16, 2000
For personal use only. Not to be reproduced without permission of The Lancet.