Introduction
Contemporary sexual behaviors have evolved, leading to extended periods characterized by multiple serial or concurrent sexual relationships, particularly for women [Johnson, Wadsworth, Wellings, & Field, 1994]. Data from national surveys reflect an increase in the number of sexual partners over recent decades [Aral, 2006], highlighting a shift in sexual patterns. While sexual exploration is common in young adulthood [CDC, 2005], the long-term psychological implications of engaging with Multi Sex Partners remain unclear.
Cross-sectional studies have indicated a correlation between sexual risk behaviors, such as early sexual initiation and having multi sex partners, and mental health challenges [Bachanas et al., 2002; Baskin-Sommers & Sommers, 2006; Brown et al., 2006; Erbelding, Hummel, Hogan, & Zenlman, 2001; Hutton, Lyketsos, Zenilman, Thompson, & Erbelding, 2004; Lavan & Johnson, 2002; Mazzaferro et al., 2006; Mota, Cox, Katz, & Sareen, 2010]. Research has often explored how pre-existing mental health issues might lead to risky sexual behaviors [Boden, Fergusson, & Horwood, 2010; Bohon, Garber, & Horowitz, 2007; Brook, Balka, & Whiteman, 1999; Dishion, 2000; Duncan, Strycker, & Duncan, 1999], suggesting that conditions like substance use and antisocial tendencies can predict such behaviors. However, fewer studies have investigated whether engaging in sexual risk behaviors, particularly with multi sex partners, can contribute to later mental health problems. Some existing research suggests that early sexual activity and sexually transmitted infections (STIs) are linked to subsequent depression and substance-related issues [Cornelius, Clark, Reynolds, Kirisci, & Tarter, 2007; Hallfors, Waller, Bauer, Ford, & Halpern, 2005; McGue & Iacono, 2005; Shrier, Harris, & Beardslee, 2002].
Sexual risk behavior is often defined by actions that increase the risk of STIs. However, the psychological consequences of having multi sex partners, whether in serial or concurrent relationships, warrant further investigation. The potential psychological impact could stem from the absence of a deep relational aspect in such sexual encounters, leading to feelings of impersonality [Långström & Hanson, 2006] and negative emotional outcomes. Alternatively, the emotional distress from the dissolution of multiple short-term relationships could also play a role. Furthermore, gender differences in experiencing these relationships may exist, as women are often found to engage in sexual relationships with a stronger emotional component compared to men [Meston & Buss, 2007].
This study aims to explore whether having multi sex partners, as one facet of sexual risk behavior, is a predictor for later diagnoses of common mental disorders, specifically anxiety, depression, and substance dependence disorders. This investigation was conducted across three distinct periods of young adulthood. Previous research has highlighted cross-sectional links at age 21 between having multi sex partners combined with inconsistent condom use, early sexual activity, and STIs, and various mental disorders, including anxiety, depression, and substance dependence [Ramrakha, Caspi, Dickson, Moffitt, & Paul, 2000]. Earlier work has also explored whether pre-existing mental health problems are associated with subsequent risky sexual behaviors, revealing connections between childhood antisocial behavior, low anxiety, and later sexual risk-taking [Ramrakha et al., 2007].
Method
Participants
The data for this study were drawn from the Dunedin Multidisciplinary Health and Development Study, a long-term project tracking the health and behavior of individuals born in Dunedin, New Zealand, between April 1972 and March 1973. The cohort, initially comprising 1,037 participants representative of the general population, was established at age 3. The cohort was primarily of white ethnicity (91%) and encompassed a wide range of socioeconomic backgrounds typical of New Zealand’s South Island. Regular assessments were carried out every two years until age 15, and subsequently at ages 18 (1990–1991), 21 (1993–1994), 26 (1998–1999), and 32 years (2003–2005). Participation rates remained high across assessments: age 18 (97%), age 21 (97%), age 26 (96%), and age 32 (96%). Ethical approvals were secured, and participant confidentiality was maintained throughout the study.
Procedure
Participants, irrespective of their location (within New Zealand or overseas), returned to the research unit for comprehensive data collection at each assessment point. Mental health data were gathered through confidential, standardized interviews conducted by trained interviewers unaware of the participants’ prior mental health history. Data on sexual behavior, including information about multi sex partners, were also collected via private, computer-based questionnaires, adapted from the 1990 British National Survey of Sexual Attitudes and Lifestyles [Johnson, Wadsworth, Wellings, & Field, 1994]. Interviewers were present to provide instructions and assistance but were positioned to ensure participant privacy during questionnaire completion.
Measures
Number of Sexual Partners
Participants were asked to report the number of opposite-sex partners they had engaged in penetrative sex with over different periods: the past 3 years (at age 21), 5 years (at age 26), and 6 years (at age 32). They also reported the number of partners in the last 12 months at each assessment. These data were used to calculate the number of sexual partners for three specific age periods: 18–20 years, 21–25 years, and 26–31 years. By subtracting the data from the last 12 months from the longer periods (3, 5, and 6 years), the researchers ensured that the predictor variables did not overlap with the outcome variable—mental disorder—which was assessed for the preceding 12 months at each age. The number of sexual partners per year was then categorized into three groups to facilitate comparison across assessment periods: 1.0 or fewer (reference group), 1.1–2.5, and more than 2.5 multi sex partners.
Mental Health
Mental health was assessed using the Diagnostic Interview Schedule [Robins, Helzer, Croughan, & Ratcliff, 1981], a standardized tool designed to diagnose disorders based on the criteria of the Diagnostic and Statistical Manual of Mental Disorders (DSM). The DSM versions used were DSM-III-R [APA, 1987] at ages 18 and 21, and DSM-IV [APA, 1994] at ages 26 and 32. Diagnosis required not only meeting symptom criteria but also an impairment rating of >2 on a scale from 1 (some impairment) to 5 (severe impairment). Diagnoses were made for each disorder independently of others. Mental disorders assessed over the past year included anxiety, depression, and substance (cannabis and/or alcohol) dependence disorders, which are the focus of this study.
Statistical Analysis
Logistic regression models were employed to analyze the relationship between the reported number of sexual partners and the subsequent onset of mental disorders (anxiety, depression, substance dependence). Analyses were conducted separately for each age period due to changes in sexual behavior over time [Humblet, Paul, & Dickson, 2003]. For each mental disorder, the interaction between gender and the number of multi sex partners was initially assessed. If no significant interaction was found, interaction terms were removed, and models were fitted adjusting for gender alone, and then for both gender and any mental disorder present at the preceding assessment (e.g., prior disorder at age 18 for outcomes at age 21). Adjustment for any prior mental disorder was used because comorbidity is common in mental disorders [Angold, Costello, & Erkanli, 1999; Kim-Cohen et al., 2003]. Predictive modeling using the continuous variable of reported partner numbers was also conducted for substance abuse, limited to those reporting 50 or fewer partners, to identify problematic interaction points for men and women.
Analyses also considered new cases of disorders at each assessment, defined as individuals diagnosed at ages 21, 26, or 32 who did not have the same disorder at the previous assessment.
Potential confounding by socioeconomic status in childhood and adulthood (age 32) was evaluated, but as it did not significantly alter results, it was not included in the final analyses.
Post-hoc analyses were conducted for significant findings to determine if early sexual initiation or STIs explained the observed relationships. As these adjustments did not substantially change the odds ratios or significance, only adjustments for prior disorder are presented in the tables.
Results
Table 1 presents the counts and percentages of men and women without mental disorders and with anxiety, depression, and substance dependence disorders, including new cases, at each assessment age.
Table 1.
Numbers and percentages of men and women with a mental disorder, and a new disorder, at each assessment age
| Age | No disorder | Men | No disorder | Women |
|---|---|---|---|---|
| N | (%) | Anxiety | Depression | Substance dependence |
| N | (%) | N | (%) | N |
| 21 | 299 | (61.3) | 65 | (13.3) |
| 26 | 262 | (52.7) | 101 | (20.3) |
| 32 | 309 | (57.8) | 92 | (18.8) |
| New case of disordera | ||||
| 21 | 33 | (7.2) | 34 | (7.4) |
| 26 | 63 | (13.1) | 42 | (8.8) |
| 32 | 45 | (9.3) | 40 | (8.3) |
aNew cases are defined as those who did not have the disorder at the previous assessment phase
Table 2 presents the associations between the annual number of sexual partners and subsequent anxiety and depression. No significant link was found between engaging with multi sex partners and an increased risk of anxiety or depression at ages 21, 26, or 32, after adjusting for prior disorder. Gender interaction was not significant, thus combined results are shown.
Table 2.
The relationship between the reported number of sex partners and later anxiety and depression at three age periods
| Partners per year | N (%) | Anxiety | Depression |
|---|---|---|---|
| Odds ratio (95 % CI) | |||
| N (%) | AORa | AORb | N (%) |
| 18–20 years | At age 21 years | ||
| 0–1 | 512 (55.7) | 107 (21.0) | 1.0 (reference) |
| 1.1–2.5 | 293 (31.8) | 53 (18.1) | 0.9 (0.6, 1.3) |
| > 2.5 | 115 (12.5) | 27 (23.5) | 1.4 (0.8, 2.3) |
| 21–25 years | At age 26 years | ||
| 0–1 | 450 (47.7) | 111 (24.8) | 1.0 (reference) |
| 1.1–2.5 | 412 (43.6) | 94 (22.9) | 1.0 (0.7, 1.3) |
| >2.5 | 82 (8.7) | 18 (22.0) | 1.0 (0.6, 1.8) |
| 26–31 years | At age 32 years | ||
| 0–1 | 513 (56.1) | 109 (21.1) | 1.0 (reference) |
| 1.1–2.5 | 350 (38.0) | 73 (20.9) | 1.1 (0.8, 1.5) |
| >2.5 | 54 (5.9) | 17 (31.5) | 2.3 (1.2, 4.3)* |
*p
aAOR = Adjusted, for sex, odds ratio
bAOR = Adjusted, for sex and prior disorder, odds ratio
Table 3 displays the association between the number of partners and later substance dependence disorders, separated by gender due to significant gender interaction effects. For women, a statistically significant positive correlation was observed between the number of multi sex partners and substance dependence disorder across all age periods, with odds ratios increasing with partner numbers. For men, this trend was significant at ages 21 and 32, but not at 26. Women reporting more than 2.5 partners annually showed considerably higher odds of substance dependence disorder compared to those with 0–1 partners, with adjusted odds ratios (95% CIs) of 9.6 (4.4–20.9), 7.3 (2.5–21.3), and 17.5 (3.5–88.1) at ages 21, 26, and 32, respectively. The effect was most pronounced at age 32, though based on smaller sample sizes. Predictive models further indicated that while having multi sex partners was linked to substance dependence for both genders, men with few or no partners were more likely to have a disorder, whereas women with more than approximately 10 partners in the same period had a much higher likelihood of disorder than men (see Fig. 1).
Table 3.
The association between reported number of sex partners and later substance dependence disorder for women and men at three age periods
| Partner per year | N (%) | Substance dependence disorder |
|---|---|---|
| N (%) | Unadjusted | AORa |
| Women | ||
| 18–20 years | At 21 years | |
| 0–1 | 281 (61.8) | 21 (7.5 %) |
| 1.1–2.5 | 131 (28.8) | 22 (16.8) |
| > 2.5 | 43 (9.5) | 20 (46.5) |
| 21–25 years | At 26 years | |
| 0–1 | 264 (56.2) | 20 (7.6) |
| 1.1–2.5 | 185 (39.4) | 28 (15.1) |
| > 2.5 | 21 (4.5) | 7 (33.3) |
| 26–31 years | At 32 years | |
| 0–1 | 306 (65.9) | 12 (3.9) |
| 1.1–2.5 | 150 (32.3) | 17 (11.3) |
| > 2.5 | 8 (1.7) | 4 (50.0) |
| Men | ||
| 18–20 years | At 21 years | |
| 0–1 | 231 (49.7) | 42 (18.4) |
| 1.1–2.5 | 162 (34.8) | 60 (37.0) |
| > 2.5 | 72 (15.5) | 38 (52.8) |
| 21–25 years | At 26 years | |
| 0–1 | 186 (39.2) | 35 (18.8) |
| 1.1–2.5 | 227 (47.9) | 71 (31.4) |
| > 2.5 | 61 (12.9) | 16 (26.2) |
| 26–31 years | At 32 years | |
| 0–1 | 210 (46.1) | 22 (10.5) |
| 1.1–2.5 | 200 (43.9) | 35 (17.6) |
| > 2.5 | 46 (10.1) | 14 (30.4) |
*p
aAOR = Adjusted, for prior disorder, odds ratio
Fig. 1.
Predicted probability of substance dependence disorder from a logistic model including reported number of partners, sex, and their interaction. Reported number of partners was used as a continuous variable and restricted to those with partners ≤50. The p value shown is for the interaction term
The effects of alcohol and cannabis dependence disorders were examined separately and showed similar patterns for both men and women combined (Table 4). Separate results by gender were not feasible due to small cell sizes.
Table 4.
The association between reported number of sexual partners and alcohol and cannabis dependence disorders at three age periods
| Partners/year | N (%) | Alcohol dependence | Cannabis dependence |
|---|---|---|---|
| Odds ratio (95 % CI) | |||
| AORa | AORb | N (%) | AORa |
| 18–20 years | At 21 years | ||
| 0–1 | 26 (5.1) | 1.0 (reference) | 26 (5.1) |
| 1.1–2.5 | 39 (13.3) | 2.7 (1.6, 4.5)* | 2.4 (1.4, 4.2)* |
| > 2.5 | 24 (20.9) | 4.4 (2.4, 8.0)* | 3.4 (1.8, 6.4)* |
| 21–25 years | At 26 years | ||
| 0–1 | 31 (6.9) | 1.0 (reference) | 31 (6.9) |
| 1.1–2.5 | 72 (17.5) | 2.6 (1.6, 4.0)* | 2.6 (1.6, 4.1)* |
| > 2.5 | 21 (25.6) | 3.5 (1.9, 6.7)* | 3.3 (1.7, 6.5)* |
| 26–32 years | At 32 years | ||
| 0–1 | 27 (5.3) | 1.0 (reference) | 11 (2.1) |
| 1.1–2.5 | 31 (8.9) | 1.6 (0.9, 2.8) | 1.3 (0.7, 2.2) |
| > 2.5 | 12 (22.2) | 4.2 (1.9, 9.1)* | 2.8 (1.2, 6.1)* |
*p
aAOR = Adjusted, for sex, odds ratio
bAOR = Adjusted, for sex and prior disorder, odds ratio
Repeating the analyses for new cases of disorder, defined as those who developed a disorder at the current assessment but did not have it at the previous assessment, yielded similar results. Table 5 indicates no significant associations between the number of multi sex partners and new cases of anxiety or depression, except for a link between depression and the 1.1–2.5 partners per year category at age 26. However, this finding was not supported by significance for those with >2.5 partners per year, lacking a clear dose-response relationship. Furthermore, continuous variable analysis did not confirm this association.
Table 5.
The association between the reported number of sex partners and new cases of anxiety and depression at three age periods
| Partners/year | Adjusted odds ratioa (95 % CI) |
|---|---|
| N (%) | New anxietyb |
| 18–20 years | At 21 years |
| 0–1 | 47 (9.8) |
| 1.1–2.5 | 22 (7.9) |
| > 2.5 | 10 (9.3) |
| 21–25 years | At 26 years |
| 0–1 | 55 (12.6) |
| 1.1–2.5 | 58 (14.4) |
| > 2.5 | 11 (13.9) |
| 26–31 years | At 32 years |
| 0–1 | 54 (10.5) |
| 1.1–2.5 | 35 (10.1) |
| > 2.5 | 8 (14.8) |
*p
aAdjusted for sex
bNew cases are defined as those who did not have the disorder at the previous assessment phase
Table 6 presents the relationship between the number of multi sex partners and new cases of later substance dependence disorder. While gender interaction was not significant at ages 21 or 26, it was significant at age 32 (p = .01). Consistent with earlier findings (Table 3), results are shown separately by gender. These results showed similar, though less pronounced trends: women were more likely than men to develop new substance dependence disorders with increasing numbers of partners.
Table 6.
The association between the reported number of sex partners and later new cases of substance dependence disorder at three age periods
| Partners/year | N (%) | Men | Women |
|---|---|---|---|
| New case of substance disordera | |||
| Odds ratio (95 % CI) | N (%) | Odds ratio (95 % CI) | |
| 18–20 years | At 21 years | ||
| 0–1 | 24 (11.1) | 1.0 (reference) | 13 (4.9) |
| 1.1–2.5 | 35 (22.9) | 2.4 (1.4, 4.2)* | 17 (13.5) |
| > 2.5 | 16 (24.6) | 2.6 (1.3, 5.3)* | 9 (20.9) |
| 21–25 years | At 26 years | ||
| 0–1 | 15 (8.4) | 1.0 (reference) | 8 (3.1) |
| 1.1–2.5 | 26 (11.8) | 1.5 (0.7, 2.9) | 16 (8.8) |
| > 2.5 | 7 (11.9) | 1.5 (0.6, 3.8) | 2 (11.1) |
| 26–31 years | At 32 years | ||
| 0–1 | 11 (5.3) | 1.0 (reference) | 8 (2.6) |
| 1.1–2.5 | 15 (7.6) | 1.5 (0.7, 3.3) | 7 (4.7) |
| > 2.5 | 3 (6.5) | 1.2 (0.3, 4.6) | 3 (35.5) |
*p
aNew cases are defined as those who did not have a substance dependence disorder at the previous assessment phase
Discussion
The findings of this longitudinal study reveal a significant association between an increasing number of multi sex partners and a heightened risk of later substance dependence disorders, particularly among women, even after considering prior mental health status. Conversely, no consistent links were found with subsequent anxiety or depression across the examined age periods. These results were robust, confirmed when analyzing new cases of disorders.
The robust association between having multi sex partners and the subsequent development of substance dependence disorders during young adulthood is noteworthy. Existing research has often pointed in the opposite direction, showing that frequent and heavy alcohol consumption increases the likelihood of risk-taking behaviors, including having multi sex partners [Cavazos-Rehg et al., 2007; Dogan, Stockdale, Widaman, & Conger, 2010; Ferguson & Lynsky, 1996; Lavan & Johnson, 2002; Valois, Oeltmann, Waller, & Hussey, 1999; Zenilman, Hook, Shepherd, Rompalo, & Celentano, 1994]. This study, focusing on substance use at a disorder level, demonstrates that the association persists beyond prior substance use and general mental health problems. Furthermore, a dose-response relationship appears to exist, where the risk of substance dependence disorder escalates with the number of multi sex partners. This trend was consistently observed for women across all age groups. These associations remained significant even after accounting for socioeconomic status, age at first intercourse, or a history of STIs.
Prior studies have explored the prospective link between early sexual initiation and later substance dependence. McGue and Iacono [2005] reported that early adolescent sexual intercourse was associated with subsequent substance use disorder. Cornelius et al. [2007] also found early sexual initiation in men to be predictive of later substance disorder, for both alcohol and cannabis. This study uniquely contributes to the field by examining the effect of multi sex partners on mental health in young adulthood within a general population sample.
The underlying mechanisms for this relationship are likely multifaceted. Several potential explanations warrant consideration. Firstly, sexual risk-taking and substance use could be components of a broader pattern of risk-taking behaviors common in adolescence and young adulthood [Arnett, 1992; Boyer et al., 2000; Caspi et al., 1997; Desiderato & Crawford, 1995; Donovan & Jessor, 1985; Taylor, Fulop, & Green, 1999]. Individuals with impulsive traits might be predisposed to both activities, increasing their vulnerability to substance dependence. Secondly, substance use can create opportunities for sexual behavior due to its disinhibitory effects and impaired risk perception [Crowe & George, 1989; Fromme, D’Amico, & Katz, 1999]. The interplay between substance use and sexual behavior, particularly regarding condom use, is complex [Weinhardt and Carey, 2000]. Thirdly, shared social contexts may be influential. Young individuals often encounter potential sexual partners in environments where alcohol is prevalent, potentially fostering both sexual behavior and multi sex partners.
A fourth intriguing possibility is that the act of having multi sex partners itself contributes to the risk of substance disorder. This could be linked to the impersonal nature of such relationships or the emotional strain from multiple relationship failures, potentially leading to anxiety about forming new relationships. Substance use may serve as a form of self-medication to manage this interpersonal anxiety [Khantzian, 1997; Stoner, George, Peters, & Norris, 2006]. Feelings of loneliness and hopelessness have been linked to substance use [Page, Allen, Moore, & Hewitt, 1993], and using alcohol to cope with negative emotions is known to contribute to alcohol-related problems [Cooper, Shapiro, & Powers, 1998; Taylor et al., 1999].
Contrary to earlier cross-sectional findings, no clear association was found between having multi sex partners and subsequent anxiety or depressive disorders. This null finding aligns with a review indicating that negative emotional states, including depression and anxiety, are not consistently related to sexual risk-taking [Crepaz & Marks, 2000]. However, other studies have reported links between STI diagnoses and later depression [Shrier et al. 2002], and between having multi sex partners in adolescence and increased risk of depression, especially in women [Hallfors et al., 2005]. It is possible that any transient depression or anxiety resulting from multi sex partners may resolve within the 12-month assessment period, indicating no long-term impact on these conditions in this study. Alternatively, the causal direction may primarily run from anxiety and depressive disorders to sexual behavior, which warrants further investigation.
Sex Differences
While having multi sex partners increased the odds of substance dependence disorders for both sexes, the probability of disorder differed by gender based on partner numbers. Men were more likely to experience substance dependence with fewer partners (up to 10), whereas women with more than 10 partners showed a substantially higher probability of disorder compared to men. This pattern persisted across a decade, from early 20s to early 30s. Although both substance dependence and having multi sex partners were more common among men than women, the relationship between having multi sex partners and substance dependence was stronger for women at each age.
Research indicates increasing convergence in sexual behavior between young men and women [Grunseit, Richters, Crawford, Song, & Kippax 2005], as well as in substance use patterns [Sarigiani, Ryan, & Petersen, 1999]. These findings suggest that while women may engage in behaviors similar to men, they might experience greater internal conflict between their actions and personal or societal gender expectations. Gender differences in sexual attitudes include varying views on casual sex: men generally exhibit more acceptance of casual sex and hold more permissive sexual attitudes, while women are more likely to acknowledge societal double standards [Crawford & Popp, 2003; Oliver & Hyde, 1993; Sprecher & Hatfield, 1996]. Motivations for sex may also differ, with women more often citing love, commitment, and emotional reasons, while men, although sharing these motivations, may be more inclined to engage in sex for purely physical reasons [Carroll, Volk, & Hyde 1985; Meston & Buss, 2007]. Some women might use alcohol to lower inhibitions and facilitate sexual encounters [Taylor et al., 1999]. Tolman [2002] suggests young women may struggle to recognize and assert their sexual desires, acting primarily on partner desires and suppressing their own due to fears of pregnancy and disease, leading to confusion and anxiety. Studies have shown that some young women engage in unwanted sex due to fear of partner anger [Blythe, Fortenberry, Temkit, Tu, and Orr 2006]. Therefore, frequent casual sex for women could lead to complex emotions like shame, fear, and dissatisfaction, with substance use potentially serving to alleviate negative feelings and enable continued engagement in such encounters.
This study has limitations. The data originate from a single country, although prevalence rates for mental disorders and sexual behaviors are broadly consistent with other Western nations [Humblet et al., 2003; Kim-Cohen et al., 2003; Moffitt et al., 2007]. Replication in diverse countries and cultures is needed. Sexual risk measures were self-reported, and contextual information was limited. While adjustments were made for prior substance disorder, prior heavy drinking short of disorder was not accounted for. The analysis focused on heterosexual partners, potentially misclassifying individuals with multi sex partners of the same sex as having 0–1 partners, although this was rare in the sample. Dependence on drugs other than alcohol and cannabis was excluded at age 21 due to data limitations, but analyses at ages 26 and 32 including all substance dependence yielded similar results. Key strengths include the prospective longitudinal design, establishing temporality, consistent measures across age periods, and consideration of prior mental disorder status. The high cohort retention minimizes potential bias.
In conclusion, this study indicates that while having multi sex partners is not associated with anxiety and depression, it is strongly linked to substance dependence across young adulthood, especially for women. This highlights the possibility that interpersonal anxieties from multiple sexual relationships may contribute to substance abuse problems, particularly in women. Societal shifts in sexual norms and alcohol consumption necessitate further research to understand these relationships in different cultural contexts and elucidate the underlying mechanisms.
Acknowledgments
The Dunedin Multidisciplinary Health and Development Research Unit is supported by the Health Research Council of New Zealand. This research also received support from the United Kingdom Medical Research Council (Grant G0100527) and from the National Institute of Mental Health (Grants MH45070 and MH49414). Terrie E. Moffitt and Avshalom Caspi are Royal Society Wolfson Research Merit Award holders. The authors are grateful to Richie Poulton, Director of the Research Unit, for valuable comments on earlier drafts of this paper. We thank Antony Ambler for assistance with the analyses. The authors are indebted to Phil Silva, the founder of the Dunedin Study, and to the Study members and their families for their long-term involvement.
Contributor Information
Sandhya Ramrakha, Email: [email protected], Dunedin Multidisciplinary Health and Development Research Unit, Department of Preventive and Social Medicine, Dunedin School of Medicine, University of Otago, Dunedin, New Zealand.
Charlotte Paul, Department of Preventive and Social Medicine, Dunedin School of Medicine, University of Otago, Dunedin, New Zealand.
Melanie L. Bell, Department of Preventive and Social Medicine, Dunedin School of Medicine, University of Otago, Dunedin, New Zealand
Nigel Dickson, Department of Preventive and Social Medicine, Dunedin School of Medicine, University of Otago, Dunedin, New Zealand.
Terrie E. Moffitt, Departments of Psychology and Neuroscience and Psychiatry and Behavioral Sciences and Institute for Genome Sciences and Policy, Duke University, Durham, NC, USA. Social, Genetic, and Developmental Psychiatry Research Centre, Institute of Psychiatry, Kings College, London, UK
Avshalom Caspi, Departments of Psychology and Neuroscience and Psychiatry and Behavioral Sciences and Institute for Genome Sciences and Policy, Duke University, Durham, NC, USA. Social, Genetic, and Developmental Psychiatry Research Centre, Institute of Psychiatry, Kings College, London, UK.
