Genetics of Mental Illness Research Paper

1. Twentieth Century Developments

Mendel was the first to postulate specific mechanisms of genetic transmission in plants. He was lucky in that he was working on traits which were due to a monogenic variation (one mutation, one changing trait). The relevance of this work was not immediately recognized. The redetection of the Mendelian law and the invention of the concept of genes in 1900 induced rapidly growing interest in the genetic hypothesis and stimulated its application to humans. Remarkably, the original concept of genes coding for physical and possibly also for psychological properties was hypothetical and was developed without any knowledge of the physical existence or structure of genes (which became apparent only several decades later through the work of Watson and Crick). The early concept of genes already included the possibility that a gene could have multiple variants (mutations) with differential functional consequences (polymorphism). Thus, the distinction between genetic and nongenetic influences became feasible. The familial transmission of patterns and disturbances of human behavior was also explored in a genetic perspective. The advantage of this development was the development of systematic family and twin studies and their application to mental diseases, mainly in the ‘Munich School’ headed by E. Rudin at the beginning of the twentieth century. Familial aggregation and genetic determination of most mental disorders were concluded.

These results were misinterpreted in the predominant one disease one gene perspective (monogenetic concept) which was valid mainly for rare genetic diseases but not so for common mental diseases. The eugenic movement in many countries was influenced by this misinterpretation, and stimulated programs to prohibit people affected from reproducing. Nazi Germany in particular established forced sterilization programs among patients with schizophrenia and affective disorders. This development is particularly surprising as it was noticed that many inherited behavioral traits did not stick to a Mendelian mode of familial transmission, and as the application of Mendelian genetic transmission to all inherited traits was called into question by the biometric school of geneticists, particularly in the UK. A polygenetic etiology to explain familial aggregation without Mendelian patterns of transmission was early proposed in a paper by the British statistician Sir Ronald Fisher in 1918. It is difficult to understand why leading psychiatric geneticists (e.g., E. Rudin in Germany) appeared to be ignorant of these insights and continued to claim the monogenic Mendelian origin of schizophrenia and manic-depressive disease. In a polygenetic perspective the rationale of the eugenic movement becomes fully invalid as those programs cannot change the frequency of most targeted mental disorders which are not transmitted in a Mendelian manner.

Despite excellent empirical work (e.g., Kallmann) the hypothesis of the genetic origin of mental disorders did not receive a lot of attention between the 1940s and 1970s. Several reasons can be cited for the lack of interest: first, the genetic hypothesis was seriously discredited by the morally disreputable eugenic movement focusing on these diseases; second, psychoanalytic thinking predominated in these decades in many countries (especially in the USA until about 1970); and third, the research tools for studying genetically influenced disorders without a Mendelian mode of transmission were very limited before the moleculargenetics era; finally, tools for understanding complex interactions of genetic and environmental influences were underdeveloped in these times. Given this lack of genetic research potential the environmental causes of mental diseases were overstressed.

Motivated by the emerging biological psychiatry during the 1970s again family-genetic approaches in searching for causes of mental diseases received growing attention. Advanced methods of family, twin, and adoption studies using the recent developments of epidemiological and biostatistical techniques were performed exploring the etiology of, and the relationship between, mental disorders. Particularly the landmark Danish adoption studies by Kety and Goodwin changed the predominating assumption of the psychosocial origin of psychoses and alcoholism and motivated a nature–nurture discussion which dominated the field for a decade. A refined arsenal of geneticepidemiological methods was used to demonstrate the interactions of environmental and genetic forces as the etiological basis of common mental diseases. Gene– environmental interactions turned out to be the rule but not an exception for the common mental diseases. The results strongly argued against the oversimplification of the nature vs. nurture debate.

Based on the detection of the biochemical structure of chromosomes as DNA strings, and the process of expression of genes in the functionally relevant proteins, the polymorphic nature of genes and the functional consequences became apparent in the middle of the twentieth century. The genetic variability might occur as different alleles of a gene (allelic variants) which might also result in different gene products ( proteins), or it might occur as different variants in promoter regions of genes which influence the degree of gene expression resulting in a variation of the quantity of the gene product. Thus, genetic causes of diseases could be identified on a DNA level as well as on a protein level. Genetics thus became a basic tool for unraveling the pathophysiology of diseases and for the understanding and development of effective treatments.

The rapidly developing field of molecular genetics shifted the focus from the nature–nurture debate to the search for the genetic origin of diseases in the variation on the DNA level (from about 1990). Stimulated by the progress in unraveling the causes of monogenic diseases, extensive efforts are now undertaken to identify genetic variants of etiological relevance. The resulting neglect of environmental influences in psychiatric research is compensated by the hope that once impacting genes are identified specific environmental contributions to the etiology will more easily be detected.

In the meantime the genetic causes of most monogenic diseases have been detected. As expected from this success story, the first successes in detection of gene mutations for mental diseases came for rare genetic diseases characterized by mental retardation and early onset dementia. Causal mutations of specific genes have now been identified for rare variants of mental retardation such as fragile X syndrome and Rett’s syndrome (Amir et al. 1999). Causal genes for rare variants of dementia of Alzheimer type characterized by early onset were also identified. Genes accounting for more common diseases have less deleterious effects. Thus, in spite of intensive research work, only very few genes influencing the common mental diseases have been identified. The major difficulty is the polygenetic origin of these disorders. The progress of the Human Genome Project will accelerate progress in finding disease genes in the future.

Among the common mental diseases schizophrenia, affective disorders, and addiction, and also anxiety disorders received most intensive study. The genetics of rare mental diseases, particularly early-onset Alzheimer’s disease and specific subtypes of mental retardation, have been extensively and very successfully studied; given the limitations of space these rare disorders are not included in this overview.

2. Specific Disorders

2.1 Schizophrenia

Among mental disorders schizophrenia is most intensively studied in a genetic perspective. Schizophrenia is a disabling lifelong disorder with the first signs in early childhood, with an often insidious onset in early childhood (lifetime prevalence 1 percent). The symptoms are heterogeneous and vary across the lifetime. An extensive body of evidence proposes that the symptoms of schizophrenia emerge from a maladaptive brain development. The familial-genetic basis had already been established at the beginning of the twentieth century. Recurrent risks among siblings is about 5–7 percent, resulting in a relative risk of about

10 (lifetime risk in the group of interest divided by lifetime risk in the general population). The risk among parents of schizophrenics is reduced (3–5 percent) because of a reduced fertility of schizophrenics (i.e., affected subjects have fewer children than random subjects in the general population). Currently, having a first-degree relative with schizophrenia shows the highest predictive power among all known risk factors for schizophrenia.

2.1.1 Genetic s. nongenetic sources.

Twin studies explore the genetic impact by systematically varying the genetic similarity between twins (i.e., by comparing mono and dizygotic twins). A higher concordance rate for schizophrenia was consistently found among monozygotic compared to dizygotic twins (50 percent to 10 percent). On the one hand, this difference proves the operation of genes. On the other hand, the concordance rate of monozygotic twins is far from 100 percent, arguing for the impact of nongenetic environmental forces. The application of variance-analytic methods to twin data combined with prevalence rates in the population make it possible to distinguish three sources of etiological variance under the assumption of specific modes of familial transmission (genetic, nongenetic familial environment, and nongenetic individual-specific environment). Model-dependent variance analyses propose about 50 percent of the etiological variance to be due to genetic factors, whereas the remainder is mainly allocated to individual-specific environment. Adoption studies systematically vary environment by teasing apart the biological background (i.e., being a biological child of affected parents) and foster environment (i.e., being adopted away into another familial environment). Following this strategy a strong genetic influence interacting with environmental forces was concluded. The familial environment cannot be excluded on the basis of adoption studies. Up to now there have been different conclusions from twin compared to adoption studies with regard to the relevance of familial environment for the emergence of schizophrenia; whereas this problem is currently unresolved, the relevance of strong genetic influences is unquestionable (Kety et al. 1994, Tienari et al. 1994).

A Mendelian pattern of familial transmission cannot be observed and the precise nature of the transmission mechanism remains obscure. Thus, schizophrenia is a complex disease like insulin-dependent diabetes, coronary heart disease or autoimmune disorders. Schizophrenia shares a series of features with other complex diseases:

(a) Environmental factors are operating in concert with genetic factors (evidenced by less than 100 percent concordance among monozygotic twins).

(b) Penetrance of the phenotype is incomplete: the offspring of an unaffected twin of a schizophrenic case (discordant monozygotic twins) have the same risk (10 percent) of transmitting the disorder to their children as monozygotic twins with both being affected.

(c) The boundaries of the transmitted familial phenotype of schizophrenia are not distinct. Also related syndromes (other psychoses) and isolated symptoms are aggregating. Among the relatives without any psychiatric disorder during lifetime neurobiological characteristics of schizophrenia occur more often than expected by chance (e.g., attention deficit, deviant patterns of evoked potentials, memory problems, slow pursuit eye movement disturbances) (Tsuang 2000).

2.1.2 Molecular approaches.

The strong support for a genetic influence on schizophrenia encouraged the search for predisposing genetic variants. Despite several promising leads there is not yet a definite association of a genetic variant with the illness. Although many genes coding for proteins which are considered as relevant for the pathophysiology or as treatment targets are polymorphic, an association of these variants with the disease was not consistently found. However, meta-analyses demonstrated very mild effects of genetic variants of the genes for serotonin receptor 2a and for dopamine receptor D3 (relative risks about 1.2).

The genetic linkage strategy working on families with multiple affected individuals was enormously successful in monogenic diseases and was applied also to complex diseases. The strategy makes use of the transmission of genetic information via

(a) chromosomes where genes are placed in a fixed order, and

(b) crossing over between chromosomal pairs of maternal and paternal origin.

The cosegregation of the disease and the variation at a genetic locus is explored by this strategy. Cosegregating markers point to the location of the impacting gene in close spatial neighborhood. A major advantage of this strategy is that linkage can be explored genome-wide by applying a limited number of equally spaced markers (at least 400).

In the first stage the application of this strategy identified several candidate regions on the genome hosting susceptibility genes which have yet to be identified. The first candidate region which was confirmed later on was reported by Straub et al. (1995). In the meantime several genome-wide linkage scans were completed, and multiple candidate regions were reported. Some of the candidate regions were confirmed in different linkage studies. Currently candidate regions on 1q, 5q, 6p, 6q, 8p, 10p, 13q, 18p and 22q are most well confirmed across the various linkage studies. Taken together, evidence for a major gene locus was not found in the vast majority of the studies. The identified candidate regions for schizophrenia and other complex diseases are broad, including hundreds of genes. It will take a long time to identify the susceptibility genes in a systematic manner. The progress in mapping genes and in characterizing their function (e.g., due to the Human Genome Project which is being conducted in North America and Europe) will accelerate this search for genes.

It can be concluded from the multiplicity of candidate regions that there is no single causal or major gene that explains most of the genetic variance but that multiple susceptibility genes influence the risk of schizophrenia. The demonstrated polygenetic basis of schizophrenia explains:

(a) the complex pattern of familial aggregation not fitting to a Mendelian mode of transmission, and

(b) that the prevalence of schizophrenia remains stable over time although this genetically influenced disease with an onset in adolescence and early childhood is associated with a significant reduction of fertility (Gershon 2000, Maier et al. 2000).

2.2 Affective Disorders

Affective disorders run in families and are genetically influenced. The familial pattern of aggregation of specific affective syndromes proposed a genetic-epidemiological split between bipolar disorder (manic episodes in combination with depressive episodes) and unipolar disorder (recurrent depressive episodes only). More than three decades ago Angst and Winokur observed that bipolar disorders were more common than in the general population, 1 percent among parents, children, and siblings of patients with bipolar disorder, 7 percent with the same disorder, but not so among relatives with unipolar depression (1–2 percent); in contrast, unipolar depression aggregates in families with both syndromes 20 percent compared to about 10 percent in the general population) (Winokur et al. 1995).

Twin and adoption studies strongly suggest a genetic influence which is stronger for the bipolar than for the unipolar variant. Mean monozygotic concordance rates are 30–70 percent for unipolar depression and 50–80 percent for bipolar disorder. Dizygotic concordance rates are 10–30 percent for unipolar depression and 10–20 percent for bipolar disorder. The calculated heritability rates are 30–50 percent for unipolar depression and 50–80 percent for bipolar disorder, leaving space for environmental risk factors (Sullivan et al. 2000).

Given the relatively high lifetime prevalence of unipolar depression in the general population, very informative longitudinal studies exploring the relationship between different risk factors from various domains are feasible. A population-based prospective twin study (Kendler et al. 1990) proposed that genetic and environmental risk factors (such as early parental loss, perceived parental warmth in childhood and critical live events and social support later on) interact. The risk factors occurring already in childhood also influence personality features (e.g., neuroticism) and coping strategies which also operate as risk factors for depression and mediate the early environmental and genetic sources for the final disease status.

Mendelian patterns of transmission were only observed in selected extended families; most families with more cases show a more complex pattern. Also the recently feasible systematic search for genes using genome-wide linkage approaches primarily focused on bipolar disorder. In 1994, the first candidate region for bipolar disorder obtained by molecular-genetic tools was found on chromosome 18p (Berrettini et al. 1994); this result was confirmed by other studies. Other confirmed candidate regions are on 4p, 13q, 18q, and 21q.

A lot of genetic association studies comparing allele frequencies between cases and controls were performed mainly for unipolar depression focusing on polymorphic genes which are believed to be involved in the pathophysiology or pharmacology. A promoter variant of the 5-HT transporter received particular attention. Classical linkage studies using monogenic phenotypes as markers—e.g., red–green color blindness, which is localized on the X chromosome—in bipolar disorder produced ambiguous results. As with schizophrenia, multiple regions on the genome were proposed by linkage studies to host genes contributing to the risk of bipolar disorder. Some of these regions were confirmed by independent groups. Specific susceptibility genes have not yet been identified (Craddock and Jones 1999).

2.3 Anxiety Disorders

Some behavioral disorders can phenomenologically be considered as extremes of behavioral variants with a broad variation in the general population. For example, anxiety is a complex behavioral reaction physiologically revealed in dangerous situations; another example is eating disorders (anorexia, bulimia) which might be considered as variants of dieting. The physiological reactions are expressed in an interindividually, quantitatively, and qualitatively variable manner under the same situational context. The interindividual variation of these behavioral traits is partly under genetic control, as evidenced by twin studies. The degree of genetic impact is variable across traits with strong effects on anxiety proneness and smaller effects on dieting. Thus, it did not come as a surprise that the phenomenologically related disorders demonstrate familial similarity and genetic influences. However, the magnitude of genetic influence may vary along the behavioral continuum, and a qualitatively additional effect might operate on the extremes (i.e., on the disorders). An additional genetic effect was indeed observed for anorexia whereas a qualitatively additional effect on anxiety disorders seems to be less likely.

Anxiety disorders display a phenomenologically heterogeneous and variable symptomatology overlapping with nearly all other psychological disorders. Subtyping of anxiety disorders is widely accepted on the basis of distinct phenomenological features. The various clinical variants (generalized anxiety disorders, panic disorders, phobias) reveal specific familial aggregation in family studies, although substantial intrafamilial cosegregation between various specific anxiety disorders and also with depression (especially generalized anxiety disorder but panic disorder substantially less so) and addictive behavior ( panic disorder, phobias) was observed. The absolute prevalence rates for specific disorders vary considerably between studies due to methods of case identification and sampling. The reported relative risks vary between 2 and 10. Twin studies including multiple anxiety  affective disorders demonstrated that:

(a) Generalized anxiety disorder, panic disorder and specific phobic disorders are under genetic influence (with heritability rates between 30 percent and 45 percent); obsessive-compulsive disorder seems to have the lowest level of genetic influence.

(b) The genetic contribution to each anxiety syndrome is neither highly specific nor highly unspecific (partly syndrome-specific and partly shared by other anxiety syndromes and by unipolar depression).

(c) Different anxiety disorders are genetically heterogeneous with at least two genetically distinct groups: panic disorder, phobias and bulimia defining a group of disorders with broad overlap of influencing genetic factors, and generalized anxiety and depression defining a separate genetically overlapping group (Kendler et al. 1995).

The phenotype transmitted in families is not only restricted to specific clinical syndromes. Increased anxiety proneness, behavioral disinhibition, increased sensitivity to hyperventilation or elevated autonomic reaction were also observed more commonly than expected by chance among healthy relatives (Merikangas et al. 1999).

The search for specific genes influencing anxiety disorders has been unsuccessful up to now. Some genome-wide linkage studies were performed for panic disorder without providing conclusive results on the localization of susceptibility genes. In any case, a major gene effect was not found. Thus, it is very likely that multiple genes influence the risk for panic disorder, each with an effect too small to be detected by linkage analysis. On the other side, associations with variants of candidate genes which are known to be involved in the pathophysiology of anxiety could also not be detected up to now (Van den Heuvel et al. 2000).

2.4 Alcoholism

Alcoholism is a brain disease. In order to emerge, drinking alcohol is a prerequisite. Drinking is common in the general population and does not necessarily induce alcoholism as a disease. Only in a subgroup does drinking proceed to alcoholism. Twin studies in the general population have demonstrated that drinking of alcohol itself is genetically determined in a complex manner. Three behavioral dimensions are widely independently influenced by different factors:

(a) time pattern of abstinence,

(b) frequency of drinking,

(c) quantity of drinking.

Particularly the frequency and the quantity of drinking seem to be under genetic control in the general population but counter to expectancy both traits are not influenced by the same genes.

Alcoholism is a complex behavioral condition characterized by compulsive drinking; crucial signs are abuse (consumption in spite of anticipated adversity) and or dependence addiction (e.g., loss of control on drinking, tolerance to alcohol, unsuccessful attempts to quit drinking, repeated withdrawal syndromes, inability to abstain and continuous consumption even in the morning). Both conditions require the availability and the consumption of alcohol. Given the variability of these conditions across countries and milieu conditions, the prevalence rates vary due to sociocultural sources. Consistently, alcoholism is less common among females. Although these nonbiological factors explain a limited degree of familial aggregation of alcoholism which is reported in a series of family studies (relative risks varied between 2 and 10 with a very broad range of absolute lifetime prevalences), genetic factors are an even stronger contributor (at least among males). Five twin studies report heritability rates between 30 and 60 percent with mainly stronger effects in males.

Some adoption studies also point in the same direction. Adoption studies have also revealed the genetic heterogeneity of alcoholism (Cloninger 1987, Cadoret et al. 1995). It was proposed that inability to abstain (combined with early onset of the disease) on the one hand, and lack of control on the other hand were genetically independent. It was also suggested that a common subtype of alcoholism characterized by a combination with antisocial personality disorder is genetically distinct from alcoholism beyond a familial loading with antisocial personality disorders.

These observations motivated subtype classification of alcoholism with an early onset variant, inability to abstain, and antisocial and criminal behavior as a subtype with the strongest genetic basis which is qualitatively and quantitatively different from the genetic forces determining drinking in the general population; the other subtypes are also genetically influenced but less strongly so (Heath et al. 1991, 1997, Cadoret et al. 1995).

The metabolism of alcohol is under the control of enzymes which present as various isoforms each with various genetic variants with differential activity (aldehyde dehydrogenase—ALDH, alcohol dehydrogenase—ADH). Carriers of two copies of the less active variant of one isoform of ALDH react to alcohol in an aversive manner (flushes in the face, nausea) creating a barrier for excessive or long-term use of alcohol with a reduced risk for alcoholism as a consequence. Similarly, carriers of genetic ADH variants with reduced activity are associated with lower risks for alcoholism. These influential allelic variants for alcoholism are protective against alcoholism. Thus, apart from late-onset Alzheimer’s disease, alcoholism is the only common mental disorder with well confirmed susceptibility genes. However, the frequency of allelic variants associated with reduced metabolic activity vary across populations with relatively high frequencies in Asian populations and with neglectible prevalences in Caucasian (European) populations.

Other risk factors for alcoholism and other substance abuse are personality features such as antisocial behaviour (disorder) or novelty seeking. These personality patterns are partly under genetic control and may influence the use, abuse and addiction of alcoholism on a genetic basis (Cloninger 1987, Cadoret et al. 1995). The genetic impact of personality on alcoholism, however, is via a gene–environmental interaction as the availability of alcohol or other substances is a prerequisite.

Drug addiction is currently the only common psychiatric disorder with available valid animal disease models. Genetic manipulations of addictive behavior became feasible using transgenic techniques. Thus, major progress in unraveling the genetic basis of drug addiction in mice has been made and will elucidate the molecular genetics of alcoholism in the near future (Nestler 2000).

3. Conclusion

The genetics of mental disorders has been the topic of a long and controversial debate. Only various rare variants (prevalences substantially lower than 1 percent) of mental retardation and early onset Alzheimer’s disease have been demonstrated to be classical genetic diseases with mutations at one gene locus causing the clinical syndrome (monogenic diseases). Most mental diseases are highly prevalent in the general population, such as schizophrenia, affective disorders, anxiety disorders or addiction (at least 1 percent lifetime prevalence), also called common diseases. Common mental disorders share a series of features with common diseases (e.g., hypertension, cardiac heart disease, diabetes mellitus):

(a) All common mental diseases are familial (i.e., relatives of patients are more likely to be affected with the same disease than random subjects in the general population).

(b) Twin studies were performed for all common mental diseases pointing to genetic causes.

(c) There is evidence that influences on the manifestation of the disease at least partly derives from the genetically influenced underlying behavioral traits (e.g., personality dimensions as neuroticism) ranging between mental health and illness (Bouchard 1994).

(d) The concordance rate among monozygotic twins is far from 100 percent; thus, the etiology is multifactorial with environmental as well as genetic factors contributing.

(e) The phenotype as transmitted in families is variable between relatives; unlike monogenic diseases a clear distinction between affected and healthy status is impossible.

(f) The familial aggregation of one disease often goes together with the co-aggregation of another disease (e.g., generalized anxiety disorder and depression); common genetic factors are mainly responsible for this coaggregation in families and partly explain the excess comorbidity among patients.

(g) The genetics of each common mental disease is complex and associated with genetic heterogeneity; the genetics does not follow a clear Mendelian mode of transmission, but genome-wide linkage studies argue for the contribution of multiple genes to each disorder. Consequently, the contribution of a single genetic variation is not causal but only probabilistic (susceptibility genes influencing the risk of the disease).

(h) Multiple susceptibility genes for a disease may either emerge from multiple monogenic, clinically unrecognized subtypes or from the simultaneous and cumulative contribution of several genes. Whereas the first possibility cannot be excluded definitively, the second possibility is substantially more plausible given the lack of Mendelian transmission. Whereas the contributing gene variants are likely to be rare under the first condition, the contrary should be the case under the second condition. Each of these common gene variants influencing the risk is likely to be of ancient origin dating back about 100,000 years. The functional consequence of each variant is presumably modest, facilitating overcoming the process of selection. In contrast, rare genetic variants causing monogenic diseases are more recent in origin.

(i) Strong efforts using very similar techniques are currently being undertaken for each of these diseases to identify the contributing genes. Recently, major progress has come from linkage studies providing knowledge of the localization of susceptibility genes. Currently only very few susceptibility genes are identified. Due to the progress of the Human Genome Project and the development of high throughput techniques the detection of susceptibility genes for most common mental disorders can be expected in the near future.

Bibliography:

1. Amir R E, Van den Veyver I B, Wan M, Tran C Q, Francke U, Zoghbi H Y 1999 Rett syndrome is caused by mutations in Xlinked MECP2, encoding methyl-CpG-binding protein 2. Nature Genetics 23: 185–8
2. Berrettini W H, Ferraro T N, Goldin L R, Weeks D E, DeteraWadleigh S, Nurnberger J I, Gershon E S 1994 Chromosome 18 DNA markers and manic-depressive illness: Evidence for a susceptibility gene. Proceedings of the National Academy of Science USA 91: 5918–21
3. Bouchard T 1994 Genes, environment, and personality. Science 61: 1700–1
4. Cadoret R J, Yates W R, Troughton E, Woodworth G, Stewart M A 1995 Adoption study demonstrating two genetic pathways to drug abuse. Archi es of General Psychiatry 52: 42–52
5. Cloninger C R 1987 Neurogenetic adaptive mechanism in alcoholism. Science 236: 412–20
6. Craddock N, Jones I 1999 Genetics of bipolar disorder. Journal of Medical Genetics 36: 585–94
7. Gershon E S 2000 Bipolar illness and schizophrenia as oligogenic diseases: Implications for the future. Biological Psychiatry 47: 240–4
8. Heath A C, Bucholz K K, Madden P A F, Dinwiddie S H, Slutske W S, Bierut L J, Statham D J, Dunne M P, Whitfield J B, Martin N G 1997 Genetic and environmental contributions to alcohol dependence risk in a national twin sample: Consistency of findings in women and men. Psychological Medicine 27: 1381–96
9. Heath A C, Meyer J, Jardine R, Martin N G 1991 The inheritance of alcohol consumption patterns in a general population twin sample: II. Determinants of consumption frequency and quantity consumed. Journal of Studies on Alcohol 52: 425–33
10. Kendler K S, Kessler R C, Neale M C, Heath A C, Eaves L J 1990 The prediction of major depression in women: Toward an integrated etiologic model. American Journal of Psychiatry 150: 1139–48
11. Kendler K S, Walters E E, Neale M C, Kessler R C, Heath A C, Eaves L J 1995 The structure of the genetic and environmental risk factors for six major psychiatric disorders in women. Phobia, generalized anxiety disorder, panic disorder, bulimia, major depression, and alcoholism. Archi es of General Psychiatry 52: 374–83
12. Kety S S, Wender P H, Jacobsen B, Ingraham L J, Jansson L, Faber B, Kinney D K 1994 Mental illness in the biological and adoptive relatives of schizophrenic adoptees. Archi es of General Psychiatry 51: 442–55
13. Maier W, Schwab S, Rietschel M 2000 The genetics of schizophrenia. Current Opinion in Psychiatry 13: 3–9
14. Merikangas K R, Avenevoli S, Dierker L, Grillon C 1999 Vulnerability factors among children at risk for anxiety disorders. Biological Psychiatry 46: 1523–35
15. Nestler E J 2000 Genes and addiction. Nature Genetics 26: 277–81
16. Straub R E, MacLean C J, O’Neill F O, Burke J, Murphy B, Duke F, Shinkwin R, Webb B T, Zhang J, Walsh D et al. 1995 A potential vulnerability locus for schizophrenia on chromosome 6p24-22: Evidence for genetic heterogeneity. Nature Genetics 11: 287–93
17. Sullivan P, Neale M C, Kendler K S 2000 Genetic epidemiology of major depression: Review and meta-analysis. American Journal of Psychiatry 157: 1552–62
18. Tienari P, Wynne L C, Moring J, Lahti I, Naarala M, Sorr A, Wahlberg K-E, Saarento O, Seitamaa M, Kaleva M, Laksy K 1994 The Finnish adoptive family study of schizophrenia. British Journal of Psychiatry, Supplement 23: 20–6
19. Tsuang M 2000 Schizophrenia: genes and environment. Biological Psychiatry 47: 210–20
20. Van den Heuvel O A, Van de Wetering J M, Veltman D J, Pauls D L 2000 Genetic studies on panic disorder: A review. Journal of Clinical Psychiatry 61: 756–66
21. Winokur G, Coryell W, Keller M, Endicott J, Leon A 1995 A family study of manic-depressive (bipolar I ) disease. Archi es of General Psychiatry 52: 367–73