Childhood sexual abuse as a predictor of Complex Posttraumatic Stress Disorder: A meta-analysis

Marcelo Nvo-Fernández^1*, Valentina Miño-Reyes¹, Gastón González-Cabeza², Sofia Gálvez-Cienfuegos¹, Martina Ignacia C.C²

¹Laboratory of Methodology, Behavioural Sciences and Neuroscience, Faculty of Psychology, Universidad de Talca, Talca, Chile

²Universidad Autónoma de Chile, Talca, Chile

---

Abstract

Sexual abuse, especially when it occurs during childhood, is an experience that causes deep and lasting harm. Currently, its study as a risk factor for the development of trauma-related pathologies is of great relevance. In 2018, Complex Post-Traumatic Stress Disorder (CPTSD) was officially recognized as a distinct syndrome in the 11th Revision of the International Classification of Diseases (ICD-11), with the aim of distinguishing it from neurotic disorders secondary to stressful situations, somatoform disorders, and those specifically associated with stress. The inclusion of CPTSD in the ICD-11 marked the culmination of two decades of research dedicated to understanding its symptoms, treatments, and risk factors. This article aims to conduct a meta-analysis that explores the relationship between sexual abuse and the development of CPTSD. Fifteen studies were selected for analysis, and the results revealed several key risk factors associated with the development of CPTSD, with the primary one being childhood sexual abuse (k = 15; OR = 3.007).

---

Introduction

Childhood sexual abuse (CSA) is a globally contentious social issue, with many facets that require in-depth investigation. Reported incidences of sexual abuse have risen over the past few decades¹. The prevalence of CSA is estimated to range from 8% to 31% for girls and from 3% to 17% for boys². Numerous studies have documented associations between CSA and various disorders, including post-traumatic stress disorder (PTSD), depression, self-harm, and obesity^3,4.

Post-traumatic stress disorder (PTSD) is a mental health condition that can develop in individuals who have experienced or witnessed traumatic events, such as natural disasters, severe accidents, terrorist acts, war/combat, rape, or other violent personal assaults. The latest edition of the Diagnostic and Statistical Manual of Mental Disorders, DSM-5, emphasizes the behavioral symptoms accompanying PTSD and proposes four distinct diagnostic clusters: re-experiencing, avoidance, negative cognitions and mood, and hyperarousal. A previous study analyzing data from 26 surveys conducted in the World Health Organization's World Mental Health Surveys estimated that 3.9% of the global population has suffered from PTSD⁵.

However, in 2018, the 11th edition of the International Classification of Diseases (ICD-11) introduced a new category of psychopathology called "Disorders Specifically Associated with Stress"⁶. This revision distinguishes between Neurotic Disorders secondary to stressful situations and somatoform disorders, and those specifically related to stress^6,7. The updated classification also delineates the primary disorders associated with traumatic events: PTSD and Complex Post-Traumatic Stress Disorder (CPTSD). Although both conditions are linked to traumatic experiences, significant differences exist in their symptomatology and underlying dimensions⁸.

Post-traumatic stress disorder (PTSD) is a mental health condition that can develop in individuals who have experienced or witnessed traumatic events, such as natural disasters, severe accidents, terrorist acts, war/combat, rape, or other violent personal assaults. The latest edition of the Diagnostic and Statistical Manual of Mental Disorders, DSM-5, emphasizes the behavioral symptoms accompanying PTSD and proposes four distinct diagnostic clusters: re-experiencing, avoidance, negative cognitions and mood, and hyperarousal. A previous study analyzing data from 26 surveys conducted in the World Health Organization's World Mental Health Surveys estimated that 3.9% of the global population has suffered from PTSD⁵.

However, in 2018, the 11th edition of the International Classification of Diseases (ICD-11) introduced a new category of psychopathology called "Disorders Specifically Associated with Stress"⁶. This revision distinguishes between Neurotic Disorders secondary to stressful situations and somatoform disorders, and those specifically related to stress^6,7. The updated classification also delineates the primary disorders associated with traumatic events: PTSD and Complex Post-Traumatic Stress Disorder (CPTSD). Although both conditions are linked to traumatic experiences, significant differences exist in their symptomatology and underlying dimensions⁸. CPTSD places a strong emphasis on emotional dysregulation and difficulties in forming interpersonal relationships, which serves as a key distinction from PTSD⁶. Furthermore, the literature suggests that CPTSD is more likely to be associated with the accumulation of multiple "potentially traumatic" events rather than the occurrence of a single event, further distinguishing it from PTSD^9,10. While PTSD encompasses dimensions such as re-experiencing the traumatic event, avoidance of trauma-related memories, and hypervigilance, CPTSD includes these symptoms along with three additional dimensions: severe difficulties with emotional and behavioral regulation, persistent negative self-perceptions (e.g., feelings of being diminished, defeated, or worthless), and chronic challenges in forming or maintaining interpersonal relationships⁶.

In terms of prevalence, studies conducted in Denmark, United States of America, Israel, and Germany have explored then feud epidemiology of CPTSD. In 2008 and 2009, Denmark’s National Center for Social Research¹¹ evaluated 2,980 individuals and found that approximately 1% of the population could be diagnosed with CPTSD, with women being three times more likely to experience the condition¹². A 2017 representative sample from United States of America households indicated a CPTSD prevalence of approximately 3.3%, with women being twice as likely to be affected as men¹⁰. In Israel, PTSD (9.0%) was found to be more common than CPTSD (2.6%). Similarly, a 2018 study in Germany revealed that 1% of a sample of 2,524 individuals had CPTSD symptoms, with women (0.7%) more affected than men (0.3%), aligning with the Danish findings^13,14.

A recent meta-analysis has highlighted and ranked the most significant risk factors in the development of Complex Post-Traumatic Stress Disorder (CPTSD)¹⁵. Among these, childhood sexual abuse emerges as the most relevant factor, revealing a direct and alarming connection with the disorder's onset. Considering this reality, the present study focuses on an in-depth exploration of this factor, aiming to clarify its impact and provide a more detailed understanding of how childhood sexual abuse contributes to the progression of CPTSD.

Method

Search strategy reporting methods and procedures were in accordance with the PRISMA checklist (Preferred Reporting Items for Systematic Reviews and Meta-Analyses). A flowchart is provided in Figure 1 (PRISMA Checklist)¹⁶. English-language articles published in peer-reviewed and indexed journals up to August 2024 were included and evaluated for this study. The searches were conducted in three scientific databases: Web of Science, Scopus, and PubMed, using combinations of terms related to CPTSD (‘complex PTSD’, ‘posttraumatic stress disorder’, PTSD, CPTSD) and risk factors (‘sexual abuse’, ‘risk factors’, ‘sexual violence’).

Due to the quantitative focus of this study, we excluded documents such as book chapters, theoretical reviews, systematic reviews, editorial comments, letters or notes, case studies, and other articles that provided non-quantitative information on risk factors for CPTSD. This study was not pre-registered.

Inclusion and Exclusion Criteria

Studies included in the meta-analysis met the following criteria: a) examine childhood sexual abuse as a potential risk factor (predictor) for the development of CPTSD; b) reported at least one of the following data: 1) Odds Ratio (OR) and corresponding 95% confidence intervals (CI); 2) Frequency of occurrence of at least one risk factor in the population with CPTSD and in the population without CPTSD, information with which data conversion is performed for further analysis.

In addition, studies were excluded for: a) addressing physical trauma (i.e., especially musculoskeletal pathologies) and not CPTSD; b) absence of meta-analyzable data (OR and Frequency); c) the study contained insufficient data to calculate univariate effect sizes, and such data could not be obtained from the study author; or d) absence of risk factors for CPTSD; e) the article was a review or qualitative study that did not present new data or only presented qualitative analyses; g) the study used a single-case design.

Statistical Analysis

Our analysis was based on ORs as the main effect size measure. ORs values were obtained directly from each study when reported by the authors. Alternatively, ORs were manually calculated based on frequency of exposure reported by the authors. In addition, other effect size measures such as correlations (between risk factors and CPTSD) or mean differences (between non-exposed vs exposed in terms of CPTSD, or between no-CPTSD vs CPTSD in terms of exposure to the risk factor) were transformed into OR using rESCMA, an open-source web-based calculator and effect size converter¹⁷.

A random-effects model was used, which assumes that the actual potential effect varies among the included studies. Heterogeneity among effects was assessed using Cochran's Q test and the heterogeneity index (I²), where a significant p-value in the Q test would indicate heterogeneity among effects, while the I² index indicates the percentage of existing heterogeneity¹⁸. Each analysis was conducted following the current recommendations for OR-based meta-analysis¹⁹, transforming the raw ORs values into its natural logarithmic expression (logOR) for a more robust approach towards the precision of the pooled estimates, and then back-transforming them into ORs for interpretation. The results would be free of publication bias. Moreover, we examined the funnel plots and Egger's Z regression to check for evidence of potential publication bias and asymmetry on the effects' distribution. All analyses were conducted in RStudio version 2023.06.0 with R version 4.3.1, using the metafor package^20,21.

Results

This literature search yields a preliminary database of 2.337 published articles, which are subsequently reviewed for inclusion in the meta-analysis using various inclusion and exclusion criteria, ending with a total of 15 studies for analysis. Figure 1 presents the flow diagram for the study selection process.

Figure 1: Flow Diagram

Table 1: Characteristics of studies included with risk factors for CPTSD

ID	Authors	Sample type	Country	Total sample	% female	Sample CPTSD	Sample Exposed	Risk factor`s	OR	VarLog OR
01	(Ben-Ezra et al., 2018)13	General population	Israel	1003	51.7%	26	128	Childhood sexual abuse	1.89	0.0135
02	(Cloitre et al., 2019)22	General population	United States	1839	52%	70	-	Childhood sexual abuse	3.88	0.0081
03	(Gallagher, 2022)23	General population	United States	553	100%	-	166	Childhood sexual abuse	4.43	0.0278
04	(Facer-Irwin et al., 2002)24	Prisoners	United Kingdom	221	0%	37	32	Childhood sexual abuse	2.5	0.0636
05	(Ferrajao et al., 2024)25	General population	Uganda	401	50.6%	145	79	Childhood sexual abuse	1.93	0.0339
06	(Frost et al., 2019)26	Refugee	United States	308	56%	15	30	Childhood sexual abuse	7.60	0.0562
07	(Frost, Hyland et al., 2019)10	General population	United Kingdom	1051	68.4%	204	49	Childhood sexual abuse	1.84	0.0128
08	(Hyland et al., 2017)27	General population	Denmark	2591	54.6%	26	66	Childhood sexual abuse	9.43	0.0070
09	(Kairyte et al., 2022)28	General population	Lithuania	158	85.4%	59	48	Childhood sexual abuse	3.05	0.0917
10	(Karatzias et al., 2017)29	Clinical population	Scotland	193	65.1%	146	93	Childhood sexual abuse	3.40	0.0763
11	(Karatzias et al., 2019)30	General population	Israel	521	78%	49	49	Childhood sexual abuse	2.65	0.0271
12	(Kvedaraite et al., 2002)31	General population	Lithuania	885	63.4%	16	21	Childhood sexual abuse	2.35	0.0193
13	(Maercker et al., 2018)14	General population	Germany	530	55%	13	27	Childhood sexual abuse	1.72	0.0254
14	(Spitzer et al., 2006)32	Clinical population	Germany	82	7.5%	14	31	Childhood sexual abuse	1.15	0.1637
15	(Truskauskaite et al., 2023)33	Clinical population	Lithuania	1626	68.2%	-	-	Childhood sexual abuse	5.60	0.0096

Childhood sexual abuse was identified as a significant risk factor for the development of CPTSD in the meta-analysis (n = 10,330; k = 15; OR = 3.07; CI = 0.8262–1.415; 95%; PI = 0.007–2.23; I² = 93.71; Q = 2,867.664; p = .000; Figures 2 and 3). However, the confidence interval (CI) is relatively wide (0.8262–1.415), which suggests some degree of uncertainty in this estimate. This interval indicates that, while childhood sexual abuse increases the risk of developing CPTSD, the precision of the estimate varies considerably. To clarify, this wide range implies that in some studies the effect could be stronger or weaker.

Figure 2: Forest plots for childhood sexual abuse as a risk factor for CPTSD based on back-transformed effect sizes

Figure 3: Funnel plot for childhood sexual abuse as a risk factor for CPTSD

Discussion

The primary finding of this study has expanded and updated the understanding of childhood sexual abuse (CSA) as a risk factor for the development of Complex Post-Traumatic Stress Disorder (CPTSD). Individuals who have experienced CSA demonstrate a prevalence of CPTSD that is nearly three times higher compared to those who have not undergone such experiences. This aligns with the findings of major studies on CPTSD^9,27,34,35.

The impact of CSA, particularly during childhood, surpasses that of physical and emotional abuse when considered separately¹⁵. This heightened effect is likely because CSA often encompasses multiple forms of abuse, including physical and emotional manipulation. This observation underscores the intense trauma inherent in sexual abuse, which typically involves more than just a singular abusive event, such as physical coercion or emotional manipulation³⁶. Therefore, sexual abuse does not occur in isolation but is frequently accompanied by other adverse experiences. In fact, children who have experienced sexual abuse are more likely to have suffered other forms of abuse compared to those who have not been sexually abused³⁷.

Moreover, the literature on trauma and psychopathology indicates that early interpersonal abuse, particularly of a sexual nature, contributes to a wide array of psychological difficulties. CSA often leads to emotional dysregulation, increasing the likelihood of dissociative episodes, hallucinations, and eating disorders³⁸, and significantly raising the risk of suicide attempts in adulthood³⁹. It is also clearly related (CSA) to the presence of emotional neglect and physical abuse during childhood. Therefore, CSA constitutes a critical risk factor for mental health and the development of psychopathology, including CPTSD^40,41.

Another risk factor with a similarly profound impact is physical abuse. The literature consistently suggests that childhood abuse experiences are especially relevant in relation to CPTSD, and our results corroborate this, indicating that both childhood and adulthood physical abuse are linked to an increased probability of developing CPTSD¹⁵. Notably, childhood physical abuse is as influential as CSA in this regard.

Furthermore, previous studies, including systematic reviews and meta-analyses, have identified CSA as a significant risk factor for the development of Post-Traumatic Stress Disorder (PTSD) in the general population (OR = 2.34; 95% CI, 1.59–3.43)⁴². However, our findings suggest that CSA may be more closely associated with the development of Complex Post-Traumatic Stress Disorder (CPTSD) as a potentially more likely outcome of this adverse childhood event. Nevertheless, further research is needed to determine specific factors, such as patient demographics and the duration, type, and severity of the abuse, that may predispose individuals to develop one condition over the other.

A limitation in the literature regarding the study of risk factors for CPTSD is that most of the studies are conducted in developed countries, likely due to the greater feasibility of conducting research in these settings. However, it is essential to study affected populations in diverse contexts and cultures to analyze how factors, such as abuse, manifest in different environments. This is illustrated by a study from Frost et al., (2019), where a population exposed to multiple adverse factors (e.g., war, abuse, terrorism) reports a significantly higher likelihood of trauma occurrence compared to other populations” (Frost et al., 2019).

It is important to mention that our results present a high level of statistical heterogeneity (I² = 93.71%), mainly attributable to two factors. First, the samples from each study have not been exposed exclusively to sexual abuse as the sole factor; in many cases, we find multi-traumatized samples and others with fewer risk factors. This generates clinical heterogeneity, which is common in studies of this nature. Second, statistical heterogeneity is observed due to the differences in sample sizes across the analyzed studies”.

References

1. Radford L, Corral S, Bradley C, Fisher HL. The prevalence and impact of child maltreatment and other types of victimization in the UK: Findings from a population survey of caregivers, children, young people, and young adults. Child Abuse Negl. 2013;37(10):801-13. doi: 10.1016/j.chiabu.2013.02.004.
2. Barth J, Bermetz L, Heim E, Trelle S, Tonia T. The current prevalence of child sexual abuse worldwide: A systematic review and meta-analysis. Birkhauser Verlag AG; 2013. doi: 10.1007/s00038-012-0426-1.
3. Sarwer DB, Gustafson TB, Sarwer DB. Childhood sexual abuse and obesity. 2004.
4. Ramirez JC, Milan S. Childhood Sexual Abuse Moderates the Relationship Between Obesity and Mental Health in Low-Income Women. Child Maltreat. 2016;21(1):85-9. doi: 10.1177/1077559515611246.
5. Koenen KC, et al. Posttraumatic stress disorder in the World Mental Health Surveys. Cambridge University Press; 2017. doi: 10.1017/S0033291717000708.
6. World Health Organization. International Statistical Classification of Diseases and Related Health Problems. 2018.
7. World Health Organization. International Statistical Classification of Diseases and Related Health Problems. 2016.
8. Brewin CR. Complex post-traumatic stress disorder: a new diagnosis in ICD-11. BJPsych Adv. 2020;26(3):145-52. doi: 10.1192/bja.2019.48.
9. Jowett S, Karatzias T, Shevlin M, Hyland P. Psychological trauma at different developmental stages and ICD-11 CPTSD: The role of dissociation. J Trauma Dissociation. 2022;23(1):52-67. doi: 10.1080/15299732.2021.1934936.
10. Frost R, Hyland P, McCarthy A, Halpin R, Shevlin M, Murphy J. The complexity of trauma exposure and response; profiling PTSD and CPTSD among a refugee sample. 2019. Available from: psychotraumanetwork.com
11. Center for Security Studies. Department of Humanities. Sciences, Social and Political. 2017.
12. Hyland P, et al. Validation of post-traumatic stress disorder (PTSD) and complex PTSD using the International Trauma Questionnaire. Acta Psychiatr Scand. 2017;136(3):313-22. doi: 10.1111/acps.12771.
13. Ben-Ezra M, et al. Posttraumatic stress disorder (PTSD) and complex PTSD (CPTSD) as per ICD-11 proposals: A population study in Israel. Depress Anxiety. 2018;35(3):264-74. doi: 10.1002/da.22723.
14. Maercker A, Hecker T, Augsburger M, Kliem S. ICD-11 Prevalence Rates of Posttraumatic Stress Disorder and Complex Posttraumatic Stress Disorder in a German Nationwide Sample. J Nerv Ment Dis. 2018;206(4):270-6. doi: 10.1097/NMD.0000000000000790.
15. Leiva-Bianchi M, Nvo-Fernandez M, Villacura-Herrera C, Miño-Reyes V, Parra Varela N. What are the predictive variables that increase the risk of developing a complex trauma? A meta-analysis. 2023; doi: 10.1016/j.jad.2023.10.002.
16. Hutton B, et al. The PRISMA extension statement for reporting of systematic reviews incorporating network meta-analyses of health care interventions: Checklist and explanations. Ann Intern Med. 2015;162(11):777-84. doi: 10.7326/M14-2385.
17. Author(s) Unknown. A brief summary on effect size conversion for meta-analysis. 2020. Available from: https://doi.org/10.31219/osf.io/8np9d
18. Sánchez-Meca J, Botella J. Revisions sistemáticas y meta-análisis: Herramientas para la práctica profesional. Available from: http://www.cop.es/papeles
19. Higgins JP, White IR, Anzures-Cabrera J. Meta-analysis of skewed data: Combining results reported on log-transformed or raw scales. Stat Med. 2008;27(29):6072-92. doi: 10.1002/sim.3427.
20. Viechtbauer W. Conducting meta-analyses in R with the metafor package. J Stat Softw. 2010. doi: 10.18637/jss.v036.i03.
21. Viechtbauer W, Cheung MWL. Outlier and influence diagnostics for meta-analysis. Res Synth Methods. 2010;1(2):112-25. doi: 10.1002/jrsm.11.
22. Cloitre M, et al. ICD-11 Posttraumatic Stress Disorder and Complex Posttraumatic Stress Disorder in the United States: A Population-Based Study. J Trauma Stress. 2019;32(6):833-42. doi: 10.1002/jts.22454.
23. Gallagher Y. Relative Effects of Childhood Trauma, Intimate Partner Violence, and Other Traumatic Life Events on Complex Posttraumatic Stress Disorder Symptoms. Traumatology. 2022. doi: 10.1037/trm0000379.supp.
24. Facer-Irwin E, Karatzias T, Bird A, Blackwood N, MacManus D. PTSD and complex PTSD in sentenced male prisoners in the UK: prevalence, trauma antecedents, and psychiatric comorbidities. Psychol Med. 2022;52(13):2794-804. doi: 10.1017/S0033291720004936.
25. Ferrajão P, Frias F, Elklit A. Exploring Independent and Cumulative Effects of Adverse Childhood Experiences on PTSD and CPTSD a Study in Ugandan Adolescents. Children. 2024;11(5). doi: 10.3390/children11050517.
26. Frost R. The distribution of psychosis, ICD-11 PTSD and complex PTSD symptoms among a trauma-exposed UK general population sample. Psychosis. 2019.
27. Hyland P, et al. Variation in post-traumatic response: the role of trauma type in predicting ICD-11 PTSD and CPTSD symptoms. Soc Psychiatry Psychiatr Epidemiol. 2017;52(6):727-36. doi: 10.1007/s00127-017-1350-8.
28. Kairyte A, Kvedaraite M, Kazlauskas E, Gelezelyte O. Exploring the links between various traumatic experiences and ICD-11 PTSD and Complex PTSD: A cross-sectional study. Front Psychol. 2022;13. doi: 10.3389/fpsyg.2022.896981.
29. Karatzias T, et al. PTSD and complex PTSD: ICD-11 updates on concept and measurement in the UK, USA, Germany and Lithuania. Eur J Psychotraumatol. 2017;8. doi: 10.1080/20008198.2017.1418103.
30. Karatzias T, et al. Risk factors and comorbidity of ICD-11 PTSD and complex PTSD: Findings from a trauma-exposed population based sample of adults in the United Kingdom. Depress Anxiety. 2019;36(9):887-94. doi: 10.1002/da.22934.
31. Kvedaraite M, Gelezelyte O, Kairyte A, Roberts NP, Kazlauskas E. Trauma exposure and factors associated with ICD-11 PTSD and complex PTSD in the Lithuanian general population. Int J Soc Psychiatry. 2022;68(8):1727-36. doi: 10.1177/00207640211057720.
32. Spitzer C, Chevalier C, Gillner M, Freyberger HJ, Barnow S. Complex posttraumatic stress disorder and child maltreatment in forensic inpatients. J Forensic Psychiatry Psychol. 2006;17(2):204-16. doi: 10.1080/14789940500497743.
33. Truskauskaite I, et al. ICD-11 PTSD and Complex PTSD in Lithuanian University Students: Prevalence and Associations with Trauma Exposure. Psychol Trauma. 2023;15(5):772-80. doi: 10.1037/tra0001436.
34. Levin Y, et al. Fatalism and ICD-11 CPTSD and PTSD diagnoses: results from Nigeria, Kenya & Ghana. Eur J Psychotraumatol. 2021;12(1). doi: 10.1080/20008198.2021.1988452.
35. Løkkegaard SS, Elklit A, Vang ML. Examination of ICD-11 PTSD and CPTSD using the International Trauma Questionnaire–Child and Adolescent version (ITQ-CA) in a sample of Danish children and adolescents exposed to abuse. Eur J Psychotraumatol. 2023;14(1). doi: 10.1080/20008066.2023.2178761.
36. Vachon DD, Krueger RF, Rogosch FA, Cicchetti D. Assessment of the harmful psychiatric and behavioral effects of different forms of child maltreatment. JAMA Psychiatry. 2015;72(11):1135-42. doi: 10.1001/jamapsychiatry.2015.1792.
37. Vaillancourt-Morel MP, Godbout N, Bédard MG, Charest É, Briere J, Sabourin S. Emotional and Sexual Correlates of Child Sexual Abuse as a Function of Self-Definition Status. Child Maltreat. 2016;21(3):228-38. doi: 10.1177/1077559516656069.
38. Wohab MA, Akhter S. The effects of childhood sexual abuse on children’s psychology and employment. Procedia Soc Behav Sci. 2010;144-9. doi: 10.1016/j.sbspro.2010.07.063.
39. Maniglio R. The impact of child sexual abuse on health: A systematic review of reviews. 2009. doi: 10.1016/j.cpr.2009.08.003.
40. Jaroenkajornkij N, Lev-Wiesel R, Binson B. Use of Self-Figure Drawing as an Assessment Tool for Child Abuse: Differentiating between Sexual, Physical, and Emotional Abuse. Children. 2022;9(6). doi: 10.3390/children9060868.
41. Dong M, Anda RF, Dube SR, Giles WH, Felitti VJ. The relationship of exposure to childhood sexual abuse to other forms of abuse, neglect, and household dysfunction during childhood. Child Abuse Negl. 2003;27(6):625-39. doi: 10.1016/S0145-2134(03)00105-4.
42. Chen LP, et al. Sexual abuse and lifetime diagnosis of psychiatric disorders: Systematic review and meta-analysis. Mayo Clin Proc. 2010;85(7):618-29. doi: 10.4065/mcp.2009.0583.