Longitudinal associations of social cognition and substance use in childhood and early adolescence: findings from the Avon Longitudinal Study of Parents and Children

Longitudinal associations of social cognition and substance use in childhood and early... Eur Child Adolesc Psychiatry (2018) 27:739–752 https://doi.org/10.1007/s00787-017-1068-x ORIGINAL CONTRIBUTION Longitudinal associations of social cognition and substance use in childhood and early adolescence: findings from the Avon Longitudinal Study of Parents and Children 1,2,4 3 1,2 Meg E. Fluharty  · Jon Heron  · Marcus R. Munafò   Received: 1 April 2017 / Accepted: 12 October 2017 / Published online: 20 October 2017 © The Author(s) 2017. This article is an open access publication Abstract Substance use is associated with impaired social 1.92, 95% CI 1.43–2.58; cannabis: OR 1.54, 95% CI 1.16– cognition. Experimental studies have shown that acute 2.05). Overall, the relationship between social cognition and intoxication of alcohol, tobacco, and cannabis decreases substance use was different in each temporal direction. Poor the performance in non-verbal, social communication and non-verbal communication in childhood appeared protective theory of mind tasks. However, in epidemiological studies against later substance use, while adolescent substance use the temporal direction of this association has gone relatively was associated with decreased social cognitive performance. unstudied. We investigated both directions of association within an adolescent birth cohort: the association of social Keywords Social cognition · Substance use · cognition with subsequent substance use, and the association Adolescence · Epidemiology · ALSPAC of early substance use with subsequent social cognition. We used data from the Avon Longitudinal Study of Parents and Children, a UK birth cohort. Logistic regression indicated Introduction that poor childhood non-verbal communication was associ- ated with decreased odds of adolescent alcohol (OR 0.70, Alcohol, tobacco, and cannabis are the most commonly used 95% 0.54–0.91), tobacco (OR 0.62, 95% CI 0.47–0.83), and substances worldwide [1–3]. In 2016, the Global Drugs Sur- cannabis use (OR 0.62, 95% CI 0.46–0.83). Early adolescent vey found that 93% of respondents reported drinking, 60% substance use was associated with increased odds of poor smoking tobacco, and 63% using cannabis within the past social communication (alcohol: OR 1.46, 95% CI 0.99–2.14; 12 months [4]. Several studies have suggested that acute tobacco: OR 1.95, 95% CI 1.33–2.86) and poor social reci- administration of these substances, and/or prolonged use procity (alcohol: OR 1.57, 95% CI 1.18–2.09; tobacco: OR and abuse of these substances, is associated with deficits in social cognition (i.e., psychological processes involved in social interaction, comprising self-knowledge, perception of Electronic supplementary material The online version of this others, and motivational understanding). These deficits may article (doi:10.1007/s00787-017-1068-x) contains supplementary include social (i.e., pragmatic) or non-verbal (i.e., emotion material, which is available to authorized users. processing) communication, and/or Theory of Mind (ToM) * Meg E. Fluharty (i.e., the ability to attribute complex mental states to others) meg.fluharty@bristol.ac.uk processes, such as social reciprocity. 1 Studies indicate that alcohol, tobacco, and cannabis may MRC Integrative Epidemiology Unit (IEU) at the University of Bristol, Bristol, UK disrupt non-verbal communication: acute intoxication from 2 alcohol is associated with decreased reactivity to threat cues School of Experimental Psychology, UK Centre for Tobacco and Alcohol Studies, University of Bristol, Bristol, UK [5], while alcohol-dependent individuals display reduced 3 accuracy in judging sadness and disgust, and require greater School of Community and Social Medicine, University of Bristol, Bristol, UK emotional intensity to detect fear and anger [6]. These 4 impairments persist when alcohol-dependent individuals School of Experimental Psychology, University of Bristol, 12a Priory Road, Bristol BS8 1TU, UK are detoxified [ 7], and can be sustained up to 2 months into Vol.:(0123456789) 1 3 740 Eur Child Adolesc Psychiatry (2018) 27:739–752 sobriety [8]. In daily cigarette smokers, deficits become This study, conducted using data from the Avon Lon- apparent when individuals are tobacco-deprived. Acute gitudinal Study of Parents and Children (ALSPC), inves- withdrawal in smokers is associated with diminished pro- tigated the temporal associations between poor social cessing of happy faces relative to neutral faces [9], and dis- cognitive function (non-verbal communication, social rupted attentional bias to facial stimuli [10]. Additionally, communication, and social reciprocity) and substance use chronic cannabis use is associated with a reduced ability behaviours (current, frequent, and age of onset). We exam- to identify emotions, particularly negative emotions [11]. ined the association of poor childhood social cognition However, the acute effects of different cannabinoids are with subsequent adolescent substance use, and the asso- distinct, with ∆-9-tetrahydrocannabinol (THC) impairing ciation of early substance use behaviour with subsequent affect recognition, but cannabidiol (CBD) improving affect social cognition. We hypothesised that there would be recognition [12]. associations between poor social cognition and substance Experimental studies have also shown that acute intoxi- use in both temporal directions. cation with alcohol results in ToM deficits [13]. Alcohol- dependent individual display ToM deficits, as they have dif- ficulty identifying their own mental states and that of social partners [14, 15]. While chronic cannabis users display no Methods change in ToM task performance compared to healthy con- trols, when compared at the neuroanatomical level they show Participants differential network activation. Heavy cannabis users display less activation in the left parahippocampal gyrus, right pre- The Avon Longitudinal Study of Parents and Children cuneus and cuneus, but greater activation in the left cuneus (ALSPAC) is a prospective, population-based birth cohort and right anterior cingulate gyrus, suggesting changes at study that recruited 14,541 pregnant women resident in the physiological level [16]. This indicates an aberrant or Avon, UK, with expected delivery dates from April 1st greater activity of ToM network, and similar changes have 1991 to December 31st 1992 (http://www.alspac.bris. been observed in at-risk psychosis populations [16, 17]. ac.uk). Information has been collected on the participants Long-term cannabinoid exposure may result in changes and and their offspring from over 60 questionnaires and 9 functionality of the endocannabinoid system, and subsequent clinic assessments [25]. The study website contains details desensitisation of CB receptors may explain the compensa- of all the data that is available through a fully searchable tory elevated CB receptors elsewhere in the striatum [18] data dictionary (http://www.bris.ac.uk/alspac/research- noted in heavy cannabis users compared to controls [19]. ers/data-access/data-dictionar y/). The study included However, it remains unclear whether it is substance use 13,617 mother–offspring pairs from singleton live births itself causing these deficits, or whether these deficits lead who survived to at least 1  year; only these are consid- to substance use (for example, to enhance certain aspects ered here. Ethics approval for the study was obtained from of social cognition). One argument for the latter is that chil- the ALSPAC Ethics and Law Committee and the Local dren that have received social-cognitive interventions within Research Ethics Committee. schools and the home have lower rates of substance abuse in The analysis of the association between childhood social adolescence [20, 21]. It is also possible that the relationship cognition and subsequent substance use was restricted to between substance use and social cognition may be due to the offspring of parents who had completed the Social and shared risk factors (genetic or environmental). Communication Disorders Checklist (SCDC) (N = 3,007), The relationship between substance use and social cog- SCDC sub-scale (N = 3,058) at age 7, and/or offspring nition is therefore complex, as some deficits occur rapidly who had completed the Diagnostic Assessment of Non- with intoxication while others may arise only after longer Verbal Accuracy (DANVA) (N = 2,985) at age 8, and off- periods of use. Furthermore, despite evidence of associa- spring who had taken part in the substance use computer tions of poor social cognition with substance use, there has task at age 18 (N = 3,820). The analysis of the association been relatively little research into the temporal relationships between early adolescent substance and subsequent social between the two to date. As individuals are most likely to cognition was further restricted to the offspring who had experiment and initiate substance use during their adolescent taken part in the substance use computer task (N = 5,009) period [22–24], and several studies have suggested social at age 15, and offspring whose parents had completed the cognitive problems among hardened users, it is important Social and Communication Disorders Checklist (SCDC) to further understand whether substance use in early ado- (N = 5,506) at age 17. Flow diagrams (Figs. 1 and 2) dis- lescence is associated with later social cognitive deficits, play the final sample size for each temporal association or whether poor social cognition in childhood is associated analysis (see Supplementary Figure 1 for a longitudinal with later substance use. representation of assessments). 1 3 Eur Child Adolesc Psychiatry (2018) 27:739–752 741 Fig. 1 Flow diagram of final sample size in analysis of childhood social cognition (age 7/8) predicting adolescent substance use (age 18) Measures Social cognition Fig. 2 Flow diagram of final sample size in analysis of adolescent Non-verbal communication at age 8 was measured via com- substance (age 15) use predicting social cognition (age 18) puter session during a clinic visit using the faces subset of the DANVA [26]. This contains 24 photographs of children’s faces displaying an either high or low intensity version of current users of each respective substance. Additionally, age 18 measures of alcohol, tobacco and cannabis use were the following emotions: happy, sad, fear, or anger. Each photograph was displayed to the children for 2 s and they collected via a computer-based assessment during a clinic visit. Individuals were classified as users of each substance, responded as to what emotion they perceived. Scoring ≥ 7 total errors on the DANVA was coded as poor performance and a user of all three substances if appropriate. Individuals scoring ≥ 8 on the Alcohol Use Disorders Identic fi ation Test [26]. Social communication was measured by maternal com- pletion of SCDC at offspring age 7 and 17 via questionnaire, (AUDIT), smoking cigarettes in the past 30 days, or using cannabis in the past 12 months were classified as users of scoring ≥ 8 out of a possible of 24 was coded as poor perfor- mance [27]. Social reciprocity at age 7 and 17 was derived each respective substance. Due to widespread acceptance of alcohol use in the UK, the alcohol use variable was restricted from five questions on the SCDC that were specifically designed to measure social reciprocity [28, 29]. Responses to hazardous use on the AUDIT rather than an ever/never response, as never drinkers may differ in regards to other of yes to ≥ three questions were coded as poor performance. societal factors comparable to social drinkers (e.g., abstain- ers for religious reasons [30, 31], or individuals with high Substance use anxiety [32]). First, individuals using all three substances were additionally classified as multi-substance users, while Current use of alcohol, tobacco, and cannabis at age 15 was collected via computer session during a clinic visit. individuals using one to two substances were classified as non multi-substance users. Second, frequency of use was Individuals were classified as either current or non-users of each substance. Individuals reporting  ≥  20 drinks in categorised as either non-weekly or weekly use. Finally, age of onset was a categorical measure based on self-reported the past 6 months, smoking cigarettes in the past 30 days, or using cannabis in the past 12 months were classified as first use of each respective substance. 1 3 742 Eur Child Adolesc Psychiatry (2018) 27:739–752 Confounders Secondary analysis Based on the literature, risk factors for poor social cog- Additionally, a secondary analysis was conducted after nition and substance use were considered as potential initial investigation of the DANVA exposure results. This confounders. These included: (1) pre-birth/demographic followed the same statistical procedure as above but inves- confounders (sex [33, 34], parity [35, 36] and socioeco- tigated response accuracy to individual emotions (happy, nomic measures [37–42] including maternal social class, sad, fear anger) and level of affect intensity (low to high) of maternal education status, maternal home ownership sta- emotions as opposed to task accuracy as a whole. tus, and maternal age) as measured by baseline maternal questionnaire; (2) maternal substance use [43, 44] con- founders (maternal cannabis use at offspring age 9, mater - Results nal binge drinking and smoking at offspring age 12) col- lected via maternal questionnaire at offspring ages 9 and/ Characteristics of participants or 12; (3) childhood confounders (IQ [34, 45] measured by the Wechsler Intelligence Scale for Children-III [46], Data were available on N = 3058 participants for the analy- victimisation [47–49] measured by a modified version sis of childhood social cognition with subsequent substance of the Bullying and Friendship Interview Schedule [50], use, and N = 3613 for the analysis of early adolescent sub- borderline personality [51, 52] measured via interview, stance use with subsequent social cognition. Characteris- and peer problems [53] measured via interview, and The tics of these participants are shown in Table 1. Confounder Strengths and Difficulties Questionnaire [54]) all collected characteristics and associations with each outcome are pre- via clinic assessment at age 8 or maternal questionnaire. sented in Supplementary Table S1. The results presented Additionally, for the analysis of early substance use and below are from the fully adjusted models. Unadjusted and subsequent social cognition, confounders included (4) pre- partially adjusted models are presented in Supplementary vious incidence of poor social cognition (age 7 SCDC and Tables (S2–S4). In general, sex-stratified analyses did not SCDC sub-scale scores, as described above). indicate any clear differences in the strength of association observed for males and females separately. The results are therefore presented unstratified, except where indicated, with Statistical Analysis sex-stratified analyses presented in Supplementary Tables S5–S8. First, we examined the association of social cognition at age 7/8 (exposure) with subsequent substance use behaviour at age 18 (outcome). Next, we examined the Association of childhood social cognition (age 7/8) association of early substance use behaviours at age 15 with adolescent substance use (age 18) (exposure) with subsequent social cognition at age 17 (out- come). We assessed both temporal relationships before and Non‑verbal communication after adjustment for covariates using logistic regression. We examined the impact of confounding by comparing Poor non-verbal communication was associated with mod- unadjusted results with those adjusted for pre-birth/demo- erately decreased odds of alcohol (fully adjusted OR 0.70, graphics confounders (model 1), and then additionally and 95% CI 0.54–0.91, P = 0.007), tobacco (fully adjusted OR cumulatively maternal substance use (model 2), childhood 0.62, 95% CI 0.47–0.83, P = 0.001), and cannabis use (fully confounders (model 3), and (for the association of early adjusted OR 0.62, 95% CI 0.46–0.83, P = 0.001). These adolescent substance use with subsequent social cognition) results are shown in Table  2. No clear evidence of asso- history of social cognition at age 7/8 (model 4). Finally, ciation was observed for age of onset, or frequency of use we ran a second set of confounder-adjusted analyses only (non-weekly/weekly) at age 18 (see Supplementary Tables including the complete cases from model 3 (for the asso- S2–S3). ciation of childhood social cognition with subsequent sub- stance use) or 4 (for the association of early adolescent Social communication and social reciprocity substance use with subsequent social cognition). Both analyses were conducted unstratified and stratified by sex. There was no clear evidence of an association of either Each analysis was conducted in full (total sample) and poor social communication or social reciprocity with alco- complete cases (sample restricted to data available at both hol, tobacco, cannabis, or all substance use. These results time-points). Analyses were conducted in Stata version 13 are shown in Table 2. Additionally, no clear evidence of (Stata Corp LP, College Station TX USA). 1 3 Eur Child Adolesc Psychiatry (2018) 27:739–752 743 Table 1 Characteristics of participants N Normal Poor a,b,c Childhood social cognitive ability (age 7/8)  Social communication 7907 90% (7138) 10% (6814)  Social reciprocity 8058 84% (6757) 16% (1301)  Non-verbal communication 6814 78% (5290) 22% (1524) N Normal Poor a,b Adolescent social cognitive ability (age 18)  Social communication 5468 88% (4833) 12% (4300)  Social reciprocity 5571 77% (4300) 23% (1271) Current use N No Yes d,e,f Early adolescent substance use (age 15)  Cannabis 5048 81% (4064) 19% (984)  Tobacco 5107 83% (4214) 17% (893)  Alcohol 5051 81% (4077) 19% (974) Current use Frequency N No Yes N ≥ Weekly < Weekly d,f,g,h,i Late adolescent substance use (age 18)  Cannabis 3820 70% (2656) 30% (1164) 1187 85% (1014) 15% (173)  Tobacco 3820 71% (2702) 29% (1118) 1181 61% (716) 39% (465)  Alcohol 3820 57% (2196) 43% (1624) 3886 74% (2874) 25% (1012)  Multi-substance 3820 86% (3268) 14% (552) Age N Cannabis Tobacco Alcohol Age of first substance  Six 1443 0% (0) 0.10% (1) 0.20% (3)  Seven 1443 0% (0) 0.14% (2) 0.69% (10)  Eight 1443 0.10% (1) 0.30% (4) 0.90% (13)  Nine 1443 0.14% (2) 0.50% (7) 1% (21)  Ten 1443 0.14% (2) 2% (21) 6% (81)  Eleven 1443 1% (16) 5% (70) 7% (96)  Twelve 1443 4% (51) 11% (160) 17% (250)  Thirteen 1443 9% (133) 17% (246) 23% (335)  Fourteen 1443 17% (246) 21% (307) 25% (354)  Fifteen 1443 24% (345) 20% (293) 15% (212)  Sixteen 1443 31% (447) 17% (242) 4% (60)  Seventeen 1443 12% (447) 6% (81%) 0.50% (8)  Eighteen 1443 1% (19) 0.60% (8) 0% (0)  Nineteen 1443 0.14% (2) 0.10% (1) 0% (0) Poor social communication: total score of ≥ 16 on the SCDC Poor social reciprocity: scoring yes on ≥ 3 from 5 sub questions on social reciprocity on the SCDC Poor non-verbal communication: ≥ 7 total errors on the DANVA Current tobacco use: use of tobacco is past 30 days Current alcohol use: ≥ 20 drinks in past 6 months Current cannabis use: use of cannabis in past 12 months Current alcohol use: ≥ 8 AUDIT Multi-substance users were classified as being current users of all three substances Frequency of use: measure of less or more than weekly use Age of first use: categorical age of first use as measured by computerised interview 1 3 744 Eur Child Adolesc Psychiatry (2018) 27:739–752 association was observed for age of onset, or frequency of Discussion use (non-weekly/weekly) at age 18 (see Supplementary Tables S2–S3). Our results indicate that, in this cohort, poor non-verbal communication at age 8 is associated with decreased alcohol, tobacco, and cannabis use. Adjustment for pre- Secondary analyses birth/demographic, maternal, and childhood confounders strengthened the associations for tobacco and cannabis use, To further investigate the association of non-verbal com- but weakened the associations for alcohol. We analysed munication and current substance use, we investigated individual emotions within the DANVA to identify whether the DANVA by individual emotion and intensity. There sensitivity to specific emotions were driving this association. was no clear pattern of association across the individual No pattern of association was found for individual emotions, emotions (see Supplementary Table S4). However, indi- although poor identification of both low and high intensity of viduals displaying reduced ability to identify emotions in emotional expression was associated with alcohol, tobacco, general, as demonstrated by poor identification of both cannabis, and all substance use. Adjustment for confound- ‘low’ and ‘high’ intensity emotionally expressive faces, ers strengthened the associations for alcohol, tobacco, and had decreased odds of substance use onset, similar to the cannabis, but weakened the association for all substance use. results seen above. Poor identification of low and high Interestingly, poor non-verbal communication appeared to intensity faces was associated with decreased odds of be protective against later substance use; thus the deficits in alcohol, tobacco, and cannabis use, and this was robust to non-verbal communication previously reported in substance adjustment (see Table 3 for details). users are more likely to be the outcome of prolonged use [6–8, 10, 11], as opposed to reflecting self-medication of these deficits. In the opposite temporal direction, our results Association of early adolescent substance use (age 15) indicate that current alcohol, tobacco, and/or cannabis use with later social cognition (age 18) at age 15 is associated with poor social communication and social reciprocity at 17. In all cases, adjustment for pre-birth, Social communication maternal, childhood, or previous indication of poor social cognition (age 7) did not substantially alter these associa- Increased odds of poor social communication was associ- tions. As both analyses adjust for previous indication of ated with earlier adolescent alcohol (fully adjusted OR poor social cognition prior to the onset of any substance 1.46, 95% CI 0.99–2.14, P = 0.051), and tobacco use (fully use (age 7), this suggests that being a current user of alco- adjusted OR 1.95, 95% CI 1.33–2.86, P = 0.001). There hol, tobacco, and cannabis may have a substantial impact on was no clear evidence of an association of poor social social cognitive abilities. communication with earlier cannabis use. These results Generally, these analyses suggest that social cognitive are shown in Table 4. In stratified analyses, associations deficits may result from the initiation and/or regular use of were slightly stronger for males, with respect to tobacco these substances. While previous literature has suggested outcomes. these social cognitive deficits can arise during periods of acute intoxication [5, 12] or withdrawal [10], our results suggest these deficits remain present over longer periods of Social reciprocity time among users. Alcohol dependence has been associated with impaired semantic memory (i.e., deficits general knowl- Increased odds of poor social reciprocity was associated edge accumulated through personal experience). As seman- with earlier adolescent alcohol (fully adjusted OR 1.57, tic memory may be necessary for the maintenance of social 95% CI 1.18–2.09, P = 0.002), tobacco (fully adjusted OR networks [55], this may subsequently lead to more specific 1.92, 95% CI 1.43–2.58, P = < 0.001), and cannabis use social cognitive deficits [ 14]. Prolonged nicotine exposure (fully adjusted OR 1.54, 95% CI 1.16–2.05, P = 0.003). may dysregulate the hypothalamic–pituitary–adrenal system, These results are shown in Table 4. In stratified analyses, resulting in the hypersecretions of cortisol and alterations in associations were slightly stronger for males, with respect the activity of the associated monoamine neurotransmitter to tobacco outcomes (see Supplementary Tables S8). system, which contributes to stress-regulation [56]. This may result in individuals being more susceptible to environmen- tal stressors and associated difficulties with affect and emo- tional regulation [57, 58]. Finally, evidence from imaging studies indicate neuroanatomical changes in heavy cannabis uses associated with prolonged endocannabinoid exposure, 1 3 Eur Child Adolesc Psychiatry (2018) 27:739–752 745 1 3 Table 2 Associations of poor childhood social cognition (age 7/8) with adolescent current substance use (age 18) Unadjusted Adjustment 1 pre-birth/demographic Adjustment 2 maternal Adjustment 3 offspring N OR 95% CI p N OR 95% CI p N OR 95% CI p N OR 95% CI p Full case analysis  Poor non-verbal communication   Alcohol 2985 0.74 (0.62– 0.001 2395 0.74 (0.61– 0.004 1890 0.77 (0.62– 0.027 1567 0.70 (0.54– 0.007 0.89) 0.91) 0.97) 0.91)   Tobacco 2985 0.83 (0.68– 0.059 2395 0.77 (0.62– 0.024 1890 0.70 (0.54– 0.009 1567 0.62 (0.47– 0.001 1.01) 0.97) 0.92) 0.83)   Cannabis 2985 0.72 (0.59– 0.001 2395 0.75 (0.60– 0.009 1890 0.72 (0.56– 0.010 1567 0.62 (0.46– 0.001 0.88) 0.93) 0.92) 0.83)   Multi-sub- 2985 0.73 (0.56– 0.024 2395 0.75 (0.56– 0.060 1890 0.80 (0.57– 0.191 1567 0.67 (0.46– 0.047 stance 0.96) 1.01) 1.12) 1.00)  Poor social communication   Alcohol 3007 1.33 (1.01– 0.045 2543 1.27 (0.93– 0.126 2025 1.29 (0.91– 0.155 1609 1.42 (0.95– 0.089 1.77) 1.73) 1.84) 2.12)   Tobacco 3007 1.39 (1.03– 0.030 2543 1.32 (0.95– 0.100 2025 1.09 (0.74– 0.667 1609 1.09 (0.70– 0.708 1.87) 1.82) 1.61) 1.70)   Cannabis 3007 1.57 (1.17– 0.002 2543 1.38 (1.00– 0.047 2025 1.32 (0.91– 0.146 1609 1.56 (1.02– 0.039 2.09) 1.91) 1.91) 2.37)   Multi-sub- 3007 1.55 (1.08– 0.017 2543 1.27 (0.93– 0.126 2025 1.18 (0.73– 0.508 1609 1.30 (0.73– 0.394 stance 2.22) 1.73) 1.90) 2.20)  Poor social reciprocity   Alcohol 3058 1.11 (0.91– 0.305 2586 1.12 (0.90– 0.304 2061 1.06 (0.82– 0.640 1638 1.10 (0.82– 0.544 1.37) 1.41) 1.38) 1.47)   Tobacco 3058 1.23 (0.99– 0.064 2586 1.21 (0.95– 0.120 2061 1.17 (0.89– 0.263 1638 1.14 (0.83 0.409 1.53) 1.54) 1.56) to1.58)   Cannabis 3058 1.27 (1.02– 0.029 2586 1.28 (1.01– 0.043 2061 1.29 (0.98– 0.067 1638 1.29 (0.94– 0.109 1.58) 1.62) 1.69) 1.77)   Multi-sub- 3058 1.24 (0.94– 0.126 2586 1.19 (0.88– 0.256 2061 1.09 (0.76– 0.641 1638 1.02 (0.67– 0.923 stance 1.64) 1.62) 1.56) 1.55) Complete case analysis  Poor non-verbal communication   Alcohol 1567 0.68 (0.53– 0.003 1567 0.66 (0.52– 0.002 1567 0.58 (0.44– <0.001 1567 0.70 (0.54– 0.007 0.88) 0.86) 0.78) 0.91)   Tobacco 1567 0.66 (0.49– 0.004 1567 0.64 (0.48– 0.003 1567 0.65 (0.48– 0.004 1567 0.62 (0.47– 0.001 0.88) 0.86) 0.87) 0.83) 746 Eur Child Adolesc Psychiatry (2018) 27:739–752 1 3 Table 2 (continued) Unadjusted Adjustment 1 pre-birth/demographic Adjustment 2 maternal Adjustment 3 offspring N OR 95% CI p N OR 95% CI p N OR 95% CI p N OR 95% CI p   Cannabis 1567 0.63 (0.48– 0.001 1567 0.58 (0.44– <0.001 1567 0.58 (0.44– <0.001 1567 0.62 (0.46– 0.001 0.83) 0.78) 0.78) 0.83)   Multi-sub- 1567 0.68 (0.47– 0.050 1567 0.64 (0.44– 0.025 1567 0.66 (0.45– 0.035 1567 0.67 (0.46– 0.047 stance 1.00) 0.95) 0.97) 1.00)  Poor social communication   Alcohol 1609 1.28 (0.87– 0.212 1609 1.26 (0.85– 0.245 1609 1.27 (0.96– 0.228 1609 1.42 (0.95– 0.089 1.89) 1.86) 1.89) 2.12)   Tobacco 1609 1.10 (0.72– 0.648 1609 1.13 (0.73– 0.581 1609 1.14 (0.73– 0.567 1609 1.09 (0.70– 0.708 1.69) 1.74) 1.76) 1.70)   Cannabis 1609 1.47 (0.99– 0.059 1609 1.36 (0.90– 0.141 1609 1.35 (0.90– 0.147 1609 1.56 (1.02– 0.039 2.19) 2.03) 2.04) 2.37)   Multi-sub- 1609 1.22 (0.72– 0.457 1609 1.20 (0.71– 0.496 1609 1.22 (0.71– 0.468 1609 1.30 (0.73– 0.394 stance 2.07) 2.05) 2.08) 2.20)  Poor social reciprocity   Alcohol 1638 1.07 (0.80– 0.637 1638 1.05 (0.79– 0.730 1638 1.03 (0.77– 0.827 1638 1.10 (0.82– 0.544 1.43) 1.40) 1.38) 1.47)   Tobacco 1638 1.18 (0.86– 0.311 1638 1.20 (0.87– 0.262 1638 1.17 (0.85– 0.321 1638 1.14 (0.83 0.409 1.61) 1.64) 1.62) to1.58)   Cannabis 1638 1.28 (0.95– 0.102 1638 1.23 (0.90– 0.188 1638 1.19 (0.87– 0.267 1638 1.29 (0.94– 0.109 1.73) 1.66) 1.62) 1.77)   Multi-sub- 1638 1.04 (0.69– 0.867 1638 1.01 (0.67– 0.945 1638 1.00 (0.66– 0.984 1638 1.02 (0.67– 0.923 stance 1.59) 1.53) 1.51) 1.55) Poor non-verbal communication: ≥ seven total errors on the DANVA Poor social reciprocity: scoring yes on ≥ three from five sub questions on social reciprocity on the SCDC Poor social communication: total score of ≥ 16 on the SCDC Current alcohol use: ≥ 8 AUDIT Current tobacco use: use of tobacco is past 30 days Current cannabis use: use of cannabis in past 12 months Multi-substance users were classified as being current users of all three substances. (1) Adjusted for pre-birth/demographic characteristics: sex, parity, maternal social class, home ownership status, and maternal age. (2) Additionally adjusted for maternal substance use confounders: maternal cannabis use (offspring age 9) maternal smoking and binge drinking (offspring age 12). (3) Additionally adjusted for offspring confounders: IQ, peer problems, victimisation, borderline personality diagnosis Eur Child Adolesc Psychiatry (2018) 27:739–752 747 1 3 Table 3 Associations of poor childhood (age 7/8) identification of high and low intensity faces with adolescent current substance use (age 15) Unadjusted Adjustment 1 pre-birth/demographic Adjustment 2 maternal Adjustment 3 offspring N OR 95% CI p N OR 95% CI p N OR 95% CI p N OR 95% CI p Full case analysis  Poor identification of high intensity faces   Alcohol 2985 0.75 (0.62–0.91) 0.004 2398 0.78 (0.63–0.96) 0.017 1890 0.77 (0.60–0.98) 0.032 1567 0.70 (0.53–0.92) 0.010   Tobacco 2985 0.75 (0.60–0.93) 0.007 2398 0.76 (0.69–0.96) 0.020 1890 0.72 (0.55–0.96) 0.023 1567 0.63 (0.46–0.87) 0.004   Cannabis 2985 0.69 (0.56–0.85) 0.001 2398 0.73 (0.58–0.92) 0.009 1890 0.69 (0.53–0.91) 0.008 1567 0.63 (0.46–0.85) 0.003   Multi-substance 2985 0.62 (0.46–0.83) 0.002 2398 0.66 (0.48–0.92) 0.012 1890 0.71 (0.49–1.02) 0.064 1567 0.61 (0.39–0.93) 0.230  Poor identification of low intensity faces   Alcohol 2985 0.72 (0.60–0.87) 0.001 2398 0.68 (0.55–0.84) <0.001 1890 0.68 (0.53–0.85) 0.001 1567 0.67 (0.51–0.85) 0.002   Tobacco 2985 0.92 (0.75–1.12) 0.418 2398 0.82 (0.66–1.03) 0.091 1890 0.73 (0.56–0.96) 0.023 1567 0.67 (0.50–0.90) 0.009   Cannabis 2985 0.83 (0.68–1.01) 0.063 2398 0.78 (0.62–0.97) 0.027 1890 0.73 (0.57–0.95) 0.017 1567 0.68 (0.51–0.90) 0.008   Multi-substance 2985 0.86 (0.66–1.12) 0.254 2398 0.84 (0.62–1.13) 0.244 1890 0.79 (0.57–1.12) 0.185 1567 0.73 (0.49–1.07) 0.108 Complete case analysis  Poor identification of high intensity faces   Alcohol 1567 0.68 (0.53–0.89) 0.005 1567 0.67 (0.51–0.87) 0.003 1567 0.68 (0.52–0.89) 0.005 1567 0.70 (0.53–0.92) 0.010   Tobacco 1567 0.64 (0.47–0.88) 0.005 1567 0.64 (0.47–0.88) 0.005 1567 0.66 (0.48–.90) 0.009 1567 0.63 (0.46–0.87) 0.004   Cannabis 1567 0.63 (0.47–0.85) 0.002 1567 0.59 (0.44–0.80) 0.001 1567 0.61 (0.45–0.83) 0.001 1567 0.63 (0.46–0.85) 0.003   Multi-substance 1567 0.60 (0.40–0.92) 0.018 1567 0.58 (0.38–0.89) 0.012 1567 0.60 (0.40–0.93) 0.021 1567 0.61 (0.39–0.93) 0.230  Poor identification of low intensity faces   Alcohol 1567 0.66 (0.52–0.86) 0.002 1567 0.64 (0.50–0.83) 0.001 1567 0.65 (0.50–0.84) 0.001 1567 0.67 (0.51–0.85) 0.002   Tobacco 1567 0.70 (0.53–0.94) 0.018 1567 0.68 (0.51–0.92) 0.011 1567 0.69 (0.51–0.92) 0.013 1567 0.67 (0.50–0.90) 0.009   Cannabis 1567 0.67 (0.51–0.89) 0.006 1567 0.68 (0.51–0.90) 0.007 1567 0.66 (0.50–0.88) 0.005 1567 0.68 (0.51–0.90) 0.008   Multi-substance 1567 0.76 (0.52–1.10) 0.144 1567 0.72 (0.49–1.05) 0.088 1567 0.72 (0.49–1.06) 0.100 1567 0.73 (0.49–1.07) 0.108 Poor identification of high intensity faces: ≥ three errors on high intensity DANVA faces Poor identification of low intensity faces: ≥ five errors on low intensity DANVA faces Current alcohol use: ≥ eight AUDIT Current tobacco use: use of tobacco is past 30 days Current cannabis use: use of cannabis in past 12 months Multi-substance users were classified as being current users of all three substances. (1) Adjusted for pre-birth/demographic characteristics: sex, parity, maternal social class, home ownership status, and maternal age. (2) Additionally adjusted for maternal substance use confounders: maternal cannabis use (offspring age 9) maternal smoking and binge drinking (offspring age 12). (3) Additionally adjusted for offspring confounders: IQ, peer problems, victimisation, borderline personality diagnosis 748 Eur Child Adolesc Psychiatry (2018) 27:739–752 1 3 Table 4 Association of current substance use (age 15) predicting poor social cognition (age 18) Unadjusted Adjustment 1 pre-birth/demo- Adjustment 2 maternal Adjustment 3 offspring Adjustment 4 poor social cogni- graphic tion N OR 95% CI p N OR 95% CI p N OR 95% CI p N OR 95% CI p N OR 95% CI p Full case analysis  Alcohol   Social 3631 1.21 (0.94–1.57) 0.139 3089 1.45 (1.10–1.91) 0.009 2550 1.38 (1.00–1.88) 0.048 2002 1.49 (1.03–2.14) 0.033 1915 1.46 (0.99–2.14) 0.051 commu- nication   Social 3704 1.36 (1.12–1.65) 0.002 3147 1.50 (1.22–1.86) <0.001 2599 1.40 (1.11–1.78) 0.005 2041 1.59 (1.21–2.08) 0.001 1976 1.57 (1.18–2.09) 0.002 Reci- procity  Tobacco   Social 3662 2.05 (1.61–2.61) <0.001 3113 2.03 (1.55–2.68) <0.001 2570 2.02 (1.48–2.77) <0.001 2020 1.97 (1.37–2.84) <0.001 1933 1.95 (1.33–2.86) 0.001 commu- nication   Social 3736 1.97 (1.63–2.39) <0.001 3172 1.06 (1.66–2.55) <0.001 2620 2.02 (1.58–2.57) <0.001 2059 1.98 (1.49–2.62) <0.001 1994 1.92 (1.43–2.58) <0.001 Reci- procity  Cannabis   Social 3637 1.38 (1.07–1.77) 0.012 3095 1.31 (0.98–1.73) 0.064 2555 1.28 (0.39–1.77) 0.128 2009 1.32 (0.91–1.91) 0.146 1992 1.26 (0.85–1.86) 0.255 commu- nication   Social 3710 1.54 (1.27–1.86) <0.001 3153 1.54 (1.25–1.90) <0.001 2604 1.55 (1.22–1.96) <0.001 2048 1.57 (1.19–2.06) 0.001 1983 1.54 (1.16–2.05) 0.003 Reci- procity Complete case analysis  Alcohol   Social 1915 1.30 (0.91–1.87) 0.151 1915 1.38 (0.96–1.98) 0.087 1915 1.41 (0.98–2.04) 0.660 1915 1.49 (1.02–2.15) 0.037 1915 1.46 (0.99–2.14) 0.051 commu- nication   Social 1976 1.44 (1.10–1.88) 0.008 1976 1.53 (1.17–2.01) 0.002 1976 1.54 (1.17–2.02) 0.002 1976 1.61 (1.22–2.13) 0.001 1976 1.57 (1.18–2.09) 0.002 Reci- procity  Tobacco 1933 1.95 (1.36–2.78) <0.001 1933 1.94 (1.34–2.79) <0.001 1933 1.95 (1.34–2.82) <0.001 1933 1.94 (1.34–2.82) <0.001 1933 1.95 (1.33–2.86) 0.001   Social commu- nication   Social 1994 1.92 (1.46–2.53) <0.001 1994 1.99 (1.50–2.64) <0.001 1994 1.97 (1.49–2.62) <0.001 1994 2.00 (1.50–2.66) <0.001 1994 1.92 (1.43–2.58) <0.001 Reci- procity Eur Child Adolesc Psychiatry (2018) 27:739–752 749 and the subsequent desensitisation of CB receptors in the brain, requiring compensatory CB1 receptor activity else- where in the striatum [16–19]. Previous literature indicates strong familial bonds and open communication within fami- lies and schools may serve as a protective factor, or help to delay adolescent substance initiation [59–62]. However, in the other temporal direction (i.e., poor social cognition and subsequent substance use), there is currently little evidence. Our analyses help to rule out the possibility of reverse cau- sality, and strengthen our findings that substance use is asso- ciated with later impaired social cognition. Additionally, this analysis suggested that poor non-verbal communication may in fact be protective with respect to subsequent substance use. While this is clearly an area that warrants additional research and replication, one possible explanation for this finding is that adolescents with poor emotion recognition skills may less likely to have larger social groups [63, 64] and therefore less likely to engage in substance use due to less social inclusion [65–68]. Strengths of this study include a rich data set with multiple social cognitive and substance use variables col- lected at several time-points throughout the adolescence and early adulthood. This allows for the analysis of both temporal directions as well as examining different facets of social cognition. Additionally, a robust approach was taken to minimise confounding, using a range of possible con- founders from pre-birth throughout adolescence. There are also some limitations in our study to consider. First, some of our exposures were self-reported by the child (DANVA) while others were parent-completed (SCDC and DAWBA). Previous studies have indicated parental rating of offspring well-being to be more positive compared to self-report by offspring [69]. Similarly, the maternal-reported measure of SCDC taken when offspring were aged 17 may be captur - ing a breakdown in family communication or adolescent disobedience, as opposed to social cognition, due to the generally rebellious nature of the adolescent period. How- ever, a recent genome-wide association study conducted in ALSPAC found evidence of a genome-wide association of SCDC measures at age 17, suggesting there is a genetic architecture of social communication that can be reliably captured by the maternal SCDC measure [70]. Second, SCDC scores are known to remain constant across age groups [27], while studies have indicated DANVA scores to improve with age [26]. This is a potential problem if the ranking of scores across the population is not consist- ent; however, previous ALSPAC studies have indicated a test–retest reliability in the DANVA of 0.84 [71]. Third, as maternal data are collected frequently and are more extensive than partner data within ALSPAC, we only investigated the impact of maternal confounding. Fourth, our substance use outcomes are all reliant on self-report and we were not able to biochemically validate these 1 3 Table 4 (continued) Unadjusted Adjustment 1 pre-birth/demo- Adjustment 2 maternal Adjustment 3 offspring Adjustment 4 poor social cogni- graphic tion N OR 95% CI p N OR 95% CI p N OR 95% CI p N OR 95% CI p N OR 95% CI p Cannabis   Social 1922 1.16 (0.89–1.67) 0.440 1922 1.16 (0.80–1.68) 0.438 1922 1.20 (0.82–1.75) 0.344 1922 1.27 (0.86–1.85) 0.227 1992 1.26 (0.85–1.86) 0.255 commu- nication   Social 1983 1.44 (1.11–1.88) 0.007 1983 1.48 (1.13–1.93) 0.004 1983 1.48 (1.13–1.95) 0.005 1983 1.57 (1.19–2.06) 0.001 1983 1.54 (1.16–2.05) 0.003 Reci- procity Poor social reciprocity: scoring yes on ≥ three from five sub questions on social reciprocity on the SCDC Poor social communication: total score of ≥ 16 on the SCDC Current alcohol use: ≥ 20 drinks in past 6 months Current tobacco use: use of tobacco is past 30 days Current cannabis use: use of cannabis in past 12 months. (1) Adjusted for pre-birth/demographic characteristics: sex, parity, maternal social class, home ownership status, and maternal age. (2) Additionally adjusted for maternal substance use confounders: maternal cannabis use (offspring age 9) maternal smoking and binge drinking (offspring age 12). (3) Additionally adjusted for off- spring confounders: IQ, peer problems, victimisation, borderline personality diagnosis. (4) Additionally adjusted for poor SCDC or SCDC sub-scale (respectively) at offspring age 7 750 Eur Child Adolesc Psychiatry (2018) 27:739–752 and JH. All Authors approved the final version of the manuscript for responses. Additionally, we drew our outcomes from age submission. 18, which provided us with a large sample size of individu- als whom had ever used substances. However, there were Compliance with ethical standards notably fewer individuals answering questions regarding frequency of use, which may have contributed to the low Role of funding source Nothing declared power for these analyses. Fifth, it is possible that our vari- able for multi-substance current use simply reflects current Conflict of interest On behalf of all authors, the corresponding au- cannabis use, since cannabis users typically also consume thor states that there is no conflict of interest. alcohol and tobacco [72]. Finally, there was evidence of Open Access This article is distributed under the terms of the differential loss to follow-up, as some children with high Creative Commons Attribution 4.0 International License (http://crea- SCDC scores, were slightly more likely to drop out of tivecommons.org/licenses/by/4.0/), which permits unrestricted use, the study before substance use and social cognition out- distribution, and reproduction in any medium, provided you give appro- come data was obtained. However, this does not necessar- priate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. ily imply selection bias in the association between social cognition and later substance use [73], and comparisons of full and complete cases display little change in results due to sample size. References Overall, we found differing patterns of relationships between social cognition and substance use behaviour 1. Organization WH (2016) WHO global report on trends in tobacco dependent on the specific social cognition examined and smoking 2000–2025. World Health Organization. http://www. temporal direction of association. While poor non-verbal who.int/tobacco/publications/surveillance/reportontrendstobac- communication in childhood appeared to have protective cosmoking/en/index3.html. Accessed 03 Feb 2017 2. Organization WH Cannabis. http://www.who.int/substance_abuse/ factors against later substance use, early adolescent sub- facts/cannabis/en/. Accessed 03 Feb 2017 stance use was associated with decreased social cognitive 3. Organization WH Global Information System on Alcohol and performance. This association would be worth pursuing for Health (GISAH). http://www.who.int/gho/alcohol/en/. Accessed replication in cohorts with similar or richer data on social 03 Feb 2017 4. Winstock ARB, M.; Ferris, J.; Maier, L. (2016) What we learned cognitive variables. Furthermore, the use of differing statisti- from GDS2016. https://www.globaldrugsurvey.com/wp-content/ cal analyses and methods will additionally help to strengthen uploads/2016/06/TASTER-KEY-FINDINGS-FROM-GDS2016. these findings. Given that causality cannot be inferred from pdf. Accessed 03 Feb 2017 observational data alone, future epidemiological studies 5. Curtin JJ, Patrick CJ, Lang AR, Cacioppo JT, Birbaume N (2001) Alcohol affects emotion through cognition. Psychol Sci investigating these associations should consider alternative 12(6):527–531. doi:10.1111/1467-9280.00397 statistical techniques. For example, studies can use genetic 6. Donadon MF, Osorio Fde L (2014) Recognition of facial expres- variants associated with substance use or social communi- sions by alcoholic patients: a systematic literature review. Neu- cation as proxies for these exposures within a Mendelian ropsychiatr Dis Treat 10:1655–1663. doi:10.2147/NDT.S65376 7. Townshend JM, Duka T (2003) Mixed emotions: alcoholics’ randomization framework. impairments in the recognition of specific emotional facial expres- sions. Neuropsychologia 41(7):773–782 Acknowledgements We are extremely grateful to all the families 8. Kornreich C, Blairy S, Philippot P, Hess U, Noel X, Streel E, Le who took part in this study, the midwives for their help in recruit- Bon O, Dan B, Pelc I, Verbanck P (2001) Deficits in recognition ing the participants, and the whole ALSPAC team, which includes of emotional facial expression are still present in alcoholics after interviewers, computer and laboratory technicians, clerical workers, mid- to long-term abstinence. J Stud Alcohol 62(4):533–542 research scientists, volunteers, managers, receptionists and nurses. The 9. Leventhal AM, Munafo M, Tidey JW, Sussman S, Monterosso JR, UK Medical Research Council and the Wellcome Trust (Grant Ref: Sun P, Kahler CW (2012) Anhedonia predicts altered processing 102215/2/13/2) and the University of Bristol provide core support for of happy faces in abstinent cigarette smokers. Psychopharmacol- ALSPAC. This publication is the work of the authors and MF, JH, and ogy 222(2):343–351. doi:10.1007/s00213-012-2649-5 MM will serve as guarantors for the contents of this paper. The research 10. Adams S, Attwood AS, Munafo MR (2014) Effects of nicotine was specifically funded by a University of Bristol postgraduate research and nicotine expectancy on attentional bias for emotional stimuli. scholarship awarded to MF, and by the Medical Research Council and Nicotine Tobacco Research. doi:10.1093/ntr/ntu219 the University of Bristol (MC_UU_12013/6). 11. Bayrakci A, Sert E, Zorlu N, Erol A, Saricicek A, Mete L (2015) MF and MM are members of the UK Centre for Tobacco and Alco- Facial emotion recognition deficits in abstinent cannabis depend- hol Studies, a UKCRC Public Health Research: Centre of Excellence. ent patients. Compr Psychiatry 58:160–164. doi:10.1016/j. Funding from the British Heart Foundation, Cancer Research UK, comppsych.2014.11.008 Economic and Social Research Council, Medical, Research Council 12. Hindocha C, Freeman TP, Schafer G, Gardener C, Das RK, Mor- and the National Institute for Health Research, under the auspices of gan CJ, Curran HV (2015) Acute effects of delta-9-tetrahydro - the UK Clinical Research Collaboration, is gratefully acknowledged. cannabinol, cannabidiol and their combination on facial emotion recognition: a randomised, double-blind, placebo-controlled study Author Contributions All authors designed the study. MEF con- ducted the analysis and drafted the manuscript with input from MRM 1 3 Eur Child Adolesc Psychiatry (2018) 27:739–752 751 in cannabis users. Eur Neuropsychopharmacol 25(3):325–334. 28. Skuse D, Warrington R, Bishop D, Chowdhury U, Lau J, Mandy doi:10.1016/j.euroneuro.2014.11.014 W, Place M (2004) The developmental, dimensional and diag- 13. Mitchell IJ, Beck SR, Boyal A, Edwards VR (2011) Theory of nostic interview (3di): a novel computerized assessment for mind deficits following acute alcohol intoxication. Eur Addict Res autism spectrum disorders. J Am Acad Child Adolesc Psychia- 17(3):164–168. doi:10.1159/000324871 try 43(5):548–558. doi:10.1097/00004583-200405000-00008 14. Nandrino JL, Gandolphe MC, Alexandre C, Kmiecik E, Yguel 29. Skuse DH, Mandy WPL, Scourfield J (2005) Measuring autistic J, Urso L (2014) Cognitive and affective theory of mind abili- traits: heritability, reliability and validity of the social and com- ties in alcohol-dependent patients: the role of autobiographi- munication disorders checklist. Brit J Psychiat 187:568–572. cal memory. Drug Alcohol Depend 143:65–73. doi:10.1016/j. doi:10.1192/bjp.187.6.568 drugalcdep.2014.07.010 30. Wallace JM Jr, Brown TN, Bachman JG, LaVeist TA (2003) 15. Bosco FM, Capozzi F, Colle L, Marostica P, Tirassa M (2014) The influence of race and religion on abstinence from alcohol, Theory of mind deficit in subjects with alcohol use disorder: an cigarettes and marijuana among adolescents. J Stud Alcohol analysis of mindreading processes. Alcohol Alcohol 49(3):299– 64(6):843–848 307. doi:10.1093/alcalc/agt148 31. Michalak L, Trocki K, Bond J (2007) Religion and alcohol in 16. Roser P, Lissek S, Tegenthoff M, Nicolas V, Juckel G, Brune M the U.S. National Alcohol Survey: how important is religion for (2012) Alterations of theory of mind network activation in chronic abstention and drinking? Drug Alcohol Depend 87(2–3):268– cannabis users. Schizophr Res 139(1–3):19–26. doi:10.1016/j. 280. doi:10.1016/j.drugalcdep.2006.07.013 schres.2012.05.020 32. Pardini D, White HR, Stouthamer-Loeber M (2007) Early ado- 17. Marjoram D, Job DE, Whalley HC, Gountouna VE, McIntosh lescent psychopathology as a predictor of alcohol use disorders AM, Simonotto E, Cunningham-Owens D, Johnstone EC, Lawrie by young adulthood. Drug Alcohol Depend 88(Suppl 1):S38– S (2006) A visual joke fMRI investigation into Theory of Mind S49. doi:10.1016/j.drugalcdep.2006.12.014 and enhanced risk of schizophrenia. Neuroimage 31(4):1850– 33. Evans EA, Grella CE, Washington DL, Upchurch DM (2017) 1858. doi:10.1016/j.neuroimage.2006.02.011 Gender and race/ethnic differences in the persistence of alcohol, 18. Romero J, Garcia-Palomero E, Castro JG, Garcia-Gil L, Ramos drug, and poly-substance use disorders. Drug Alcohol Depend JA, Fernandez-Ruiz JJ (1997) Effects of chronic exposure to 174:128–136. doi:10.1016/j.drugalcdep.2017.01.021 delta9-tetrahydrocannabinol on cannabinoid receptor binding and 34. Skuse DH, Mandy W, Steer C, Miller LL, Goodman R, Law- mRNA levels in several rat brain regions. Brain Res Mol Brain rence K, Emond A, Golding J (2009) Social communication Res 46(1–2):100–108 competence and functional adaptation in a general population 19. Sim-Selley LJ (2003) Regulation of cannabinoid CB1 receptors of children: preliminary evidence for sex-by-verbal IQ differen- in the central nervous system by chronic cannabinoids. Crit Rev tial risk. J Am Acad Child Adolesc Psychiatry 48(2):128–137. Neurobiol 15(2):91–119 doi:10.1097/CHI.0b013e31819176b8 20. Hawkins JD, Catalano RF, Kosterman R, Abbott R, Hill KG 35. Argys LM, Rees DI, Averett SL, Witoonchart B (2006) Birth (1999) Preventing adolescent health-risk behaviors by strength- order and risky adolescent behavior. Econ Inq 44(2):215–233. ening protection during childhood. Arch Pediat Adol Med doi:10.1093/ei/cbj011 153(3):226–234 36. Warren J (1966) Birth order and social behaviour. Psychol Bull 21. Battistich V, Schaps E, Watson M, Solomon D (1996) Prevention 65(1):38–49 effects of the child development project: early findings from an 37. Buu A, DiPiazza C, Wang J, Puttler LI, Fitzgerald HE, Zucker ongoing multisite demonstration trial. J Adolescent Res 11(1):12– RA (2009) Parent, family, and neighborhood effects on the 35. doi:10.1177/0743554896111003 development of child substance use and other psychopathology 22. Rhee SH, Hewitt JK, Young SE, Corley RP, Crowley TJ, Stallings from preschool to the start of adulthood. J Stud Alcohol Drugs MC (2003) Genetic and environmental influences on substance 70(4):489–498 initiation, use, and problem use in adolescents. Arch Gen Psy- 38. Stone AL, Becker LG, Huber AM, Catalano RF (2012) Review chiatry 60(12):1256–1264. doi:10.1001/archpsyc.60.12.1256 of risk and protective factors of substance use and problem 23. King KM, Chassin L (2007) A prospective study of the effects of use in emerging adulthood. Addict Behav 37(7):747–775. age of initiation of alcohol and drug use on young adult substance doi:10.1016/j.addbeh.2012.02.014 dependence. J Stud Alcohol Drugs 68(2):256–265 39. Bradley RH, Corwyn RF (2002) Socioeconomic status and child 24. Agrawal A, Grant JD, Waldron M, Duncan AE, Scherrer JF, development. Annu Rev Psychol 53:371–399. doi:10.1146/ Lynskey MT, Madden PA, Bucholz KK, Heath AC (2006) Risk annurev.psych.53.100901.135233 for initiation of substance use as a function of age of onset of 40. Coyne CA, Langstrom N, Rickert ME, Lichtenstein P, cigarette, alcohol and cannabis use: findings in a Midwestern D’Onofrio BM (2013) Maternal age at first birth and offspring female twin cohort. Prev Med 43(2):125–128. doi:10.1016/j. criminality: using the children of twins design to test causal ypmed.2006.03.022 hypotheses. Dev Psychopathol 25(1):17–35. doi:10.1017/ 25. Boyd A, Golding J, Macleod J, Lawlor DA, Fraser A, Henderson S0954579412000879 J, Molloy L, Ness A, Ring S, Smith GD (2013) Cohort profile: the 41. Benjet C (2010) Childhood adversities of populations living ‘Children of the 90 s’-the index offspring of the Avon Longitudi- in low-income countries: prevalence, characteristics, and men- nal Study of Parents and Children. Int J Epidemiol 42(1):111–127. tal health consequences. Curr Opin Psychiatry 23(4):356–362. doi:10.1093/ije/dys064 doi:10.1097/YCO.0b013e32833ad79b 26. Nowicki S, Duke MP (1994) Individual-differences in the nonver - 42. Lopez-Castroman J (2014) Young maternal age and old pater- bal-communication of affect - the diagnostic-analysis of nonver - nal age induce similar risk of mental disorders in offspring. Evid bal accuracy scale. J Nonverbal Behav 18(1):9–35. doi:10.1007/ Based Ment Health 17(4):100. doi:10.1136/eb-2014-101861 Bf02169077 43. Arria AM, Mericle AA, Meyers K, Winters KC (2012) Paren- 27. Robinson EB, Munir K, Munafo MR, Hughes M, McCormick tal substance use impairment, parenting and substance use dis- MC, Koenen KC (2011) Stability of autistic traits in the general order risk. J Subst Abuse Treat 43(1):114–122. doi:10.1016/j. population: further evidence for a continuum of impairment. J jsat.2011.10.001 44. Chassin L, Pitts SC, Prost J (2002) Binge drinking trajectories Am Acad Child Adolesc Psychiatry 50(4):376–384. doi:10.1016/j. from adolescence to emerging adulthood in a high-risk sample: jaac.2011.01.005 1 3 752 Eur Child Adolesc Psychiatry (2018) 27:739–752 predictors and substance abuse outcomes. J Consult Clin Psychol substance use: the roles of family structure, function and gender. 70(1):67–78 Addiction 97(3):329–336 45. Trinidad DR, Johnson CA (2002) The association between 60. Kliewer W, Murrelle L (2007) Risk and protective factors for emotional intelligence and early adolescent tobacco and adolescent substance use: findings from a study in selected alcohol use. Pers Indiv Differ 32(1):95–105. doi:10.1016/ Central American countries. J Adolesc Health 40(5):448–455. S0191-8869(01)00008-3 doi:10.1016/j.jadohealth.2006.11.148 46. Lavin C (1996) The Wechsler intelligence scale for children third 61. Farrell AD, White KS (1998) Peer influences and drug use among edition and the Stanford–Binet Intelligence scale: fourth edition: urban adolescents: family structure and parent-adolescent relation- a preliminary study of validity. Psychol Rep 78(2):491–496 ship as protective factors. J Consult Clin Psychol 66(2):248–258 47. Tharp-Taylor S, Haviland A, D’Amico EJ (2009) Victimization 62. Spoth R, Redmond C, Shin C, Azevedo K (2004) Brief family from mental and physical bullying and substance use in early intervention effects on adolescent substance initiation: school- adolescence. Addict Behav 34(6–7):561–567. doi:10.1016/j. level growth curve analyses 6 years following baseline. J Consult addbeh.2009.03.012 Clin Psychol 72(3):535–542. doi:10.1037/0022-006X.72.3.535 48. Dodge KA, Lansford JE, Burks VS, Bates JE, Pettit GS, Fontaine 63. Leppanen JM, Hietanen JK (2001) Emotion recognition and R, Price JM (2003) Peer rejection and social information-process- social adjustment in school-aged girls and boys. Scand J Psychol ing factors in the development of aggressive behavior problems in 42(5):429–435 children. Child Dev 74(2):374–393 64. Barth JM, Bastiani A (1997) A longitudinal study of emotion rec- 49. Salmivalli C, Lagerspetz K, Bjorkqvist K, Osterman K, ognition and preschool children’s social behavior. Merrill Palmer Kaukiainen A (1996) Bullying as a group process: par- Q 43(1):107–128 ticipant roles and their relations to social status within 65. Urberg KA, Degirmencioglu SM, Pilgrim C (1997) Close friend the group. Aggress Behav 22(1):1–15. doi: 10.1002/ and group influence on adolescent cigarette smoking and alcohol (Sici)1098-2337(1996)22:1<1:Aid-Ab1>3.0.Co;2-T use. Dev Psychol 33(5):834–844 50. Wolke D, Woods S, Stanford K, Schulz H (2001) Bullying and vic- 66. McCrady BS (2004) To have but one true friend: implica- timization of primary school children in England and Germany: tions for practice of research on alcohol use disorders and prevalence and school factors. Br J Psychol 92(Pt 4):673–696 social network. Psychol Addict Behav 18(2):113–121. 51. Trull TJ, Sher KJ, Minks-Brown C, Durbin J, Burr R (2000) Bor- doi:10.1037/0893-164X.18.2.113 derline personality disorder and substance use disorders: a review 67. Shadur J, Hussong A (2014) Friendship intimacy, close friend and integration. Clin Psychol Rev 20(2):235–253 drug use, and self-medication in adolescence. J Soc Pers Relat 52. Roepke S, Vater A, Preissler S, Heekeren HR, Dziobek I (2013) 31(8):997–1018. doi:10.1177/0265407513516889 Social cognition in borderline personality disorder. Front Neuro- 68. Alfaro DLP, Ehlinger V, Spilka S, Ross J, Sentenac M, Godeau E sci-Switz 6. doi:ARTN 195 10.3389/fnins.2012.00195 (2017) Alcohol, tobacco and cannabis use: do students with mild- 53. Murphy SM, Faulkner DM, Farley LR (2014) The behaviour intellectual disability mimic students in the general population? of young children with social communication disorders during Res Dev Disabil 63:118–131. doi:10.1016/j.ridd.2016.10.009 dyadic interaction with peers. J Abnorm Child Psychol 42(2):277– 69. Waters E, Stewart-Brown S, Fitzpatrick R (2003) Agreement 289. doi:10.1007/s10802-013-9772-6 between adolescent self-report and parent reports of health and 54. Goodman R (2001) Psychometric properties of the strengths and well-being: results of an epidemiological study. Child Care Health difficulties questionnaire. J Am Acad Child Adolesc Psychiatry Dev 29(6):501–509 40(11):1337–1345. doi:10.1097/00004583-200111000-00015 70. St Pourcain B, Skuse DH, Mandy WP, Wang K, Hakonar- 55. Labouvie-Vief G, Blanchard-Fields F (1982) Cognitive ageing and son H, Timpson NJ, Evans DM, Kemp JP, Ring SM, McArdle psychological growth. Ageing Soc 2(2):183–209. doi:10.1017/ WL, Golding J, Smith GD (2014) Variability in the common S0144686X00009429Published genetic architecture of social-communication spectrum pheno- 56. Markou A, Kosten TR, Koob GF (1998) Neurobiological simi- types during childhood and adolescence. Mol Autism 5(1):18. larities in depression and drug dependence: a self-medication doi:10.1186/2040-2392-5-18 hypothesis. Neuropsychopharmacol 18(3):135–174. doi:10.1016/ 71. Barona M, Kothari R, Skuse D, Micali N (2015) Social communi- S0893-133x(97)00113-9 cation and emotion difficulties and second to fourth digit ratio in 57. Joormann J, Quinn ME (2014) Cognitive processes and emo- a large community-based sample. Mol Autism 6:68. doi:10.1186/ tion regulation in depression. Depress Anxiety 31(4):308–315. s13229-015-0063-7 doi:10.1002/da.22264 72. Raphael B, Wooding S, Stevens G, Connor J (2005) Comorbidity: 58. Joormann J, Stanton CH (2016) Examining emotion regulation cannabis and complexity. J Psychiatr Pract 11(3):161–176 in depression: a review and future directions. Behav Res Ther 73. Carter KN, Imlach-Gunasekara F, McKenzie SK, Blakely 86:35–49. doi:10.1016/j.brat.2016.07.007 T (2012) Differential loss of participants does not necessar- 59. McArdle P, Wiegersma A, Gilvarry E, Kolte B, McCarthy S, ily cause selection bias. Aust Nz J Publ Heal 36(3):218–222. Fitzgerald M, Brinkley A, Blom M, Stoeckel I, Pierolini A, doi:10.1111/j.1753-6405.2012.00867.x Michels I, Johnson R, Quensel S (2002) European adolescent 1 3 http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png European Child & Adolescent Psychiatry Springer Journals

Longitudinal associations of social cognition and substance use in childhood and early adolescence: findings from the Avon Longitudinal Study of Parents and Children

Free
14 pages
Loading next page...
 
/lp/springer_journal/longitudinal-associations-of-social-cognition-and-substance-use-in-L1BzFjrxd6
Publisher
Springer Berlin Heidelberg
Copyright
Copyright © 2017 by The Author(s)
Subject
Medicine & Public Health; Psychiatry
ISSN
1018-8827
eISSN
1435-165X
D.O.I.
10.1007/s00787-017-1068-x
Publisher site
See Article on Publisher Site

Abstract

Eur Child Adolesc Psychiatry (2018) 27:739–752 https://doi.org/10.1007/s00787-017-1068-x ORIGINAL CONTRIBUTION Longitudinal associations of social cognition and substance use in childhood and early adolescence: findings from the Avon Longitudinal Study of Parents and Children 1,2,4 3 1,2 Meg E. Fluharty  · Jon Heron  · Marcus R. Munafò   Received: 1 April 2017 / Accepted: 12 October 2017 / Published online: 20 October 2017 © The Author(s) 2017. This article is an open access publication Abstract Substance use is associated with impaired social 1.92, 95% CI 1.43–2.58; cannabis: OR 1.54, 95% CI 1.16– cognition. Experimental studies have shown that acute 2.05). Overall, the relationship between social cognition and intoxication of alcohol, tobacco, and cannabis decreases substance use was different in each temporal direction. Poor the performance in non-verbal, social communication and non-verbal communication in childhood appeared protective theory of mind tasks. However, in epidemiological studies against later substance use, while adolescent substance use the temporal direction of this association has gone relatively was associated with decreased social cognitive performance. unstudied. We investigated both directions of association within an adolescent birth cohort: the association of social Keywords Social cognition · Substance use · cognition with subsequent substance use, and the association Adolescence · Epidemiology · ALSPAC of early substance use with subsequent social cognition. We used data from the Avon Longitudinal Study of Parents and Children, a UK birth cohort. Logistic regression indicated Introduction that poor childhood non-verbal communication was associ- ated with decreased odds of adolescent alcohol (OR 0.70, Alcohol, tobacco, and cannabis are the most commonly used 95% 0.54–0.91), tobacco (OR 0.62, 95% CI 0.47–0.83), and substances worldwide [1–3]. In 2016, the Global Drugs Sur- cannabis use (OR 0.62, 95% CI 0.46–0.83). Early adolescent vey found that 93% of respondents reported drinking, 60% substance use was associated with increased odds of poor smoking tobacco, and 63% using cannabis within the past social communication (alcohol: OR 1.46, 95% CI 0.99–2.14; 12 months [4]. Several studies have suggested that acute tobacco: OR 1.95, 95% CI 1.33–2.86) and poor social reci- administration of these substances, and/or prolonged use procity (alcohol: OR 1.57, 95% CI 1.18–2.09; tobacco: OR and abuse of these substances, is associated with deficits in social cognition (i.e., psychological processes involved in social interaction, comprising self-knowledge, perception of Electronic supplementary material The online version of this others, and motivational understanding). These deficits may article (doi:10.1007/s00787-017-1068-x) contains supplementary include social (i.e., pragmatic) or non-verbal (i.e., emotion material, which is available to authorized users. processing) communication, and/or Theory of Mind (ToM) * Meg E. Fluharty (i.e., the ability to attribute complex mental states to others) meg.fluharty@bristol.ac.uk processes, such as social reciprocity. 1 Studies indicate that alcohol, tobacco, and cannabis may MRC Integrative Epidemiology Unit (IEU) at the University of Bristol, Bristol, UK disrupt non-verbal communication: acute intoxication from 2 alcohol is associated with decreased reactivity to threat cues School of Experimental Psychology, UK Centre for Tobacco and Alcohol Studies, University of Bristol, Bristol, UK [5], while alcohol-dependent individuals display reduced 3 accuracy in judging sadness and disgust, and require greater School of Community and Social Medicine, University of Bristol, Bristol, UK emotional intensity to detect fear and anger [6]. These 4 impairments persist when alcohol-dependent individuals School of Experimental Psychology, University of Bristol, 12a Priory Road, Bristol BS8 1TU, UK are detoxified [ 7], and can be sustained up to 2 months into Vol.:(0123456789) 1 3 740 Eur Child Adolesc Psychiatry (2018) 27:739–752 sobriety [8]. In daily cigarette smokers, deficits become This study, conducted using data from the Avon Lon- apparent when individuals are tobacco-deprived. Acute gitudinal Study of Parents and Children (ALSPC), inves- withdrawal in smokers is associated with diminished pro- tigated the temporal associations between poor social cessing of happy faces relative to neutral faces [9], and dis- cognitive function (non-verbal communication, social rupted attentional bias to facial stimuli [10]. Additionally, communication, and social reciprocity) and substance use chronic cannabis use is associated with a reduced ability behaviours (current, frequent, and age of onset). We exam- to identify emotions, particularly negative emotions [11]. ined the association of poor childhood social cognition However, the acute effects of different cannabinoids are with subsequent adolescent substance use, and the asso- distinct, with ∆-9-tetrahydrocannabinol (THC) impairing ciation of early substance use behaviour with subsequent affect recognition, but cannabidiol (CBD) improving affect social cognition. We hypothesised that there would be recognition [12]. associations between poor social cognition and substance Experimental studies have also shown that acute intoxi- use in both temporal directions. cation with alcohol results in ToM deficits [13]. Alcohol- dependent individual display ToM deficits, as they have dif- ficulty identifying their own mental states and that of social partners [14, 15]. While chronic cannabis users display no Methods change in ToM task performance compared to healthy con- trols, when compared at the neuroanatomical level they show Participants differential network activation. Heavy cannabis users display less activation in the left parahippocampal gyrus, right pre- The Avon Longitudinal Study of Parents and Children cuneus and cuneus, but greater activation in the left cuneus (ALSPAC) is a prospective, population-based birth cohort and right anterior cingulate gyrus, suggesting changes at study that recruited 14,541 pregnant women resident in the physiological level [16]. This indicates an aberrant or Avon, UK, with expected delivery dates from April 1st greater activity of ToM network, and similar changes have 1991 to December 31st 1992 (http://www.alspac.bris. been observed in at-risk psychosis populations [16, 17]. ac.uk). Information has been collected on the participants Long-term cannabinoid exposure may result in changes and and their offspring from over 60 questionnaires and 9 functionality of the endocannabinoid system, and subsequent clinic assessments [25]. The study website contains details desensitisation of CB receptors may explain the compensa- of all the data that is available through a fully searchable tory elevated CB receptors elsewhere in the striatum [18] data dictionary (http://www.bris.ac.uk/alspac/research- noted in heavy cannabis users compared to controls [19]. ers/data-access/data-dictionar y/). The study included However, it remains unclear whether it is substance use 13,617 mother–offspring pairs from singleton live births itself causing these deficits, or whether these deficits lead who survived to at least 1  year; only these are consid- to substance use (for example, to enhance certain aspects ered here. Ethics approval for the study was obtained from of social cognition). One argument for the latter is that chil- the ALSPAC Ethics and Law Committee and the Local dren that have received social-cognitive interventions within Research Ethics Committee. schools and the home have lower rates of substance abuse in The analysis of the association between childhood social adolescence [20, 21]. It is also possible that the relationship cognition and subsequent substance use was restricted to between substance use and social cognition may be due to the offspring of parents who had completed the Social and shared risk factors (genetic or environmental). Communication Disorders Checklist (SCDC) (N = 3,007), The relationship between substance use and social cog- SCDC sub-scale (N = 3,058) at age 7, and/or offspring nition is therefore complex, as some deficits occur rapidly who had completed the Diagnostic Assessment of Non- with intoxication while others may arise only after longer Verbal Accuracy (DANVA) (N = 2,985) at age 8, and off- periods of use. Furthermore, despite evidence of associa- spring who had taken part in the substance use computer tions of poor social cognition with substance use, there has task at age 18 (N = 3,820). The analysis of the association been relatively little research into the temporal relationships between early adolescent substance and subsequent social between the two to date. As individuals are most likely to cognition was further restricted to the offspring who had experiment and initiate substance use during their adolescent taken part in the substance use computer task (N = 5,009) period [22–24], and several studies have suggested social at age 15, and offspring whose parents had completed the cognitive problems among hardened users, it is important Social and Communication Disorders Checklist (SCDC) to further understand whether substance use in early ado- (N = 5,506) at age 17. Flow diagrams (Figs. 1 and 2) dis- lescence is associated with later social cognitive deficits, play the final sample size for each temporal association or whether poor social cognition in childhood is associated analysis (see Supplementary Figure 1 for a longitudinal with later substance use. representation of assessments). 1 3 Eur Child Adolesc Psychiatry (2018) 27:739–752 741 Fig. 1 Flow diagram of final sample size in analysis of childhood social cognition (age 7/8) predicting adolescent substance use (age 18) Measures Social cognition Fig. 2 Flow diagram of final sample size in analysis of adolescent Non-verbal communication at age 8 was measured via com- substance (age 15) use predicting social cognition (age 18) puter session during a clinic visit using the faces subset of the DANVA [26]. This contains 24 photographs of children’s faces displaying an either high or low intensity version of current users of each respective substance. Additionally, age 18 measures of alcohol, tobacco and cannabis use were the following emotions: happy, sad, fear, or anger. Each photograph was displayed to the children for 2 s and they collected via a computer-based assessment during a clinic visit. Individuals were classified as users of each substance, responded as to what emotion they perceived. Scoring ≥ 7 total errors on the DANVA was coded as poor performance and a user of all three substances if appropriate. Individuals scoring ≥ 8 on the Alcohol Use Disorders Identic fi ation Test [26]. Social communication was measured by maternal com- pletion of SCDC at offspring age 7 and 17 via questionnaire, (AUDIT), smoking cigarettes in the past 30 days, or using cannabis in the past 12 months were classified as users of scoring ≥ 8 out of a possible of 24 was coded as poor perfor- mance [27]. Social reciprocity at age 7 and 17 was derived each respective substance. Due to widespread acceptance of alcohol use in the UK, the alcohol use variable was restricted from five questions on the SCDC that were specifically designed to measure social reciprocity [28, 29]. Responses to hazardous use on the AUDIT rather than an ever/never response, as never drinkers may differ in regards to other of yes to ≥ three questions were coded as poor performance. societal factors comparable to social drinkers (e.g., abstain- ers for religious reasons [30, 31], or individuals with high Substance use anxiety [32]). First, individuals using all three substances were additionally classified as multi-substance users, while Current use of alcohol, tobacco, and cannabis at age 15 was collected via computer session during a clinic visit. individuals using one to two substances were classified as non multi-substance users. Second, frequency of use was Individuals were classified as either current or non-users of each substance. Individuals reporting  ≥  20 drinks in categorised as either non-weekly or weekly use. Finally, age of onset was a categorical measure based on self-reported the past 6 months, smoking cigarettes in the past 30 days, or using cannabis in the past 12 months were classified as first use of each respective substance. 1 3 742 Eur Child Adolesc Psychiatry (2018) 27:739–752 Confounders Secondary analysis Based on the literature, risk factors for poor social cog- Additionally, a secondary analysis was conducted after nition and substance use were considered as potential initial investigation of the DANVA exposure results. This confounders. These included: (1) pre-birth/demographic followed the same statistical procedure as above but inves- confounders (sex [33, 34], parity [35, 36] and socioeco- tigated response accuracy to individual emotions (happy, nomic measures [37–42] including maternal social class, sad, fear anger) and level of affect intensity (low to high) of maternal education status, maternal home ownership sta- emotions as opposed to task accuracy as a whole. tus, and maternal age) as measured by baseline maternal questionnaire; (2) maternal substance use [43, 44] con- founders (maternal cannabis use at offspring age 9, mater - Results nal binge drinking and smoking at offspring age 12) col- lected via maternal questionnaire at offspring ages 9 and/ Characteristics of participants or 12; (3) childhood confounders (IQ [34, 45] measured by the Wechsler Intelligence Scale for Children-III [46], Data were available on N = 3058 participants for the analy- victimisation [47–49] measured by a modified version sis of childhood social cognition with subsequent substance of the Bullying and Friendship Interview Schedule [50], use, and N = 3613 for the analysis of early adolescent sub- borderline personality [51, 52] measured via interview, stance use with subsequent social cognition. Characteris- and peer problems [53] measured via interview, and The tics of these participants are shown in Table 1. Confounder Strengths and Difficulties Questionnaire [54]) all collected characteristics and associations with each outcome are pre- via clinic assessment at age 8 or maternal questionnaire. sented in Supplementary Table S1. The results presented Additionally, for the analysis of early substance use and below are from the fully adjusted models. Unadjusted and subsequent social cognition, confounders included (4) pre- partially adjusted models are presented in Supplementary vious incidence of poor social cognition (age 7 SCDC and Tables (S2–S4). In general, sex-stratified analyses did not SCDC sub-scale scores, as described above). indicate any clear differences in the strength of association observed for males and females separately. The results are therefore presented unstratified, except where indicated, with Statistical Analysis sex-stratified analyses presented in Supplementary Tables S5–S8. First, we examined the association of social cognition at age 7/8 (exposure) with subsequent substance use behaviour at age 18 (outcome). Next, we examined the Association of childhood social cognition (age 7/8) association of early substance use behaviours at age 15 with adolescent substance use (age 18) (exposure) with subsequent social cognition at age 17 (out- come). We assessed both temporal relationships before and Non‑verbal communication after adjustment for covariates using logistic regression. We examined the impact of confounding by comparing Poor non-verbal communication was associated with mod- unadjusted results with those adjusted for pre-birth/demo- erately decreased odds of alcohol (fully adjusted OR 0.70, graphics confounders (model 1), and then additionally and 95% CI 0.54–0.91, P = 0.007), tobacco (fully adjusted OR cumulatively maternal substance use (model 2), childhood 0.62, 95% CI 0.47–0.83, P = 0.001), and cannabis use (fully confounders (model 3), and (for the association of early adjusted OR 0.62, 95% CI 0.46–0.83, P = 0.001). These adolescent substance use with subsequent social cognition) results are shown in Table  2. No clear evidence of asso- history of social cognition at age 7/8 (model 4). Finally, ciation was observed for age of onset, or frequency of use we ran a second set of confounder-adjusted analyses only (non-weekly/weekly) at age 18 (see Supplementary Tables including the complete cases from model 3 (for the asso- S2–S3). ciation of childhood social cognition with subsequent sub- stance use) or 4 (for the association of early adolescent Social communication and social reciprocity substance use with subsequent social cognition). Both analyses were conducted unstratified and stratified by sex. There was no clear evidence of an association of either Each analysis was conducted in full (total sample) and poor social communication or social reciprocity with alco- complete cases (sample restricted to data available at both hol, tobacco, cannabis, or all substance use. These results time-points). Analyses were conducted in Stata version 13 are shown in Table 2. Additionally, no clear evidence of (Stata Corp LP, College Station TX USA). 1 3 Eur Child Adolesc Psychiatry (2018) 27:739–752 743 Table 1 Characteristics of participants N Normal Poor a,b,c Childhood social cognitive ability (age 7/8)  Social communication 7907 90% (7138) 10% (6814)  Social reciprocity 8058 84% (6757) 16% (1301)  Non-verbal communication 6814 78% (5290) 22% (1524) N Normal Poor a,b Adolescent social cognitive ability (age 18)  Social communication 5468 88% (4833) 12% (4300)  Social reciprocity 5571 77% (4300) 23% (1271) Current use N No Yes d,e,f Early adolescent substance use (age 15)  Cannabis 5048 81% (4064) 19% (984)  Tobacco 5107 83% (4214) 17% (893)  Alcohol 5051 81% (4077) 19% (974) Current use Frequency N No Yes N ≥ Weekly < Weekly d,f,g,h,i Late adolescent substance use (age 18)  Cannabis 3820 70% (2656) 30% (1164) 1187 85% (1014) 15% (173)  Tobacco 3820 71% (2702) 29% (1118) 1181 61% (716) 39% (465)  Alcohol 3820 57% (2196) 43% (1624) 3886 74% (2874) 25% (1012)  Multi-substance 3820 86% (3268) 14% (552) Age N Cannabis Tobacco Alcohol Age of first substance  Six 1443 0% (0) 0.10% (1) 0.20% (3)  Seven 1443 0% (0) 0.14% (2) 0.69% (10)  Eight 1443 0.10% (1) 0.30% (4) 0.90% (13)  Nine 1443 0.14% (2) 0.50% (7) 1% (21)  Ten 1443 0.14% (2) 2% (21) 6% (81)  Eleven 1443 1% (16) 5% (70) 7% (96)  Twelve 1443 4% (51) 11% (160) 17% (250)  Thirteen 1443 9% (133) 17% (246) 23% (335)  Fourteen 1443 17% (246) 21% (307) 25% (354)  Fifteen 1443 24% (345) 20% (293) 15% (212)  Sixteen 1443 31% (447) 17% (242) 4% (60)  Seventeen 1443 12% (447) 6% (81%) 0.50% (8)  Eighteen 1443 1% (19) 0.60% (8) 0% (0)  Nineteen 1443 0.14% (2) 0.10% (1) 0% (0) Poor social communication: total score of ≥ 16 on the SCDC Poor social reciprocity: scoring yes on ≥ 3 from 5 sub questions on social reciprocity on the SCDC Poor non-verbal communication: ≥ 7 total errors on the DANVA Current tobacco use: use of tobacco is past 30 days Current alcohol use: ≥ 20 drinks in past 6 months Current cannabis use: use of cannabis in past 12 months Current alcohol use: ≥ 8 AUDIT Multi-substance users were classified as being current users of all three substances Frequency of use: measure of less or more than weekly use Age of first use: categorical age of first use as measured by computerised interview 1 3 744 Eur Child Adolesc Psychiatry (2018) 27:739–752 association was observed for age of onset, or frequency of Discussion use (non-weekly/weekly) at age 18 (see Supplementary Tables S2–S3). Our results indicate that, in this cohort, poor non-verbal communication at age 8 is associated with decreased alcohol, tobacco, and cannabis use. Adjustment for pre- Secondary analyses birth/demographic, maternal, and childhood confounders strengthened the associations for tobacco and cannabis use, To further investigate the association of non-verbal com- but weakened the associations for alcohol. We analysed munication and current substance use, we investigated individual emotions within the DANVA to identify whether the DANVA by individual emotion and intensity. There sensitivity to specific emotions were driving this association. was no clear pattern of association across the individual No pattern of association was found for individual emotions, emotions (see Supplementary Table S4). However, indi- although poor identification of both low and high intensity of viduals displaying reduced ability to identify emotions in emotional expression was associated with alcohol, tobacco, general, as demonstrated by poor identification of both cannabis, and all substance use. Adjustment for confound- ‘low’ and ‘high’ intensity emotionally expressive faces, ers strengthened the associations for alcohol, tobacco, and had decreased odds of substance use onset, similar to the cannabis, but weakened the association for all substance use. results seen above. Poor identification of low and high Interestingly, poor non-verbal communication appeared to intensity faces was associated with decreased odds of be protective against later substance use; thus the deficits in alcohol, tobacco, and cannabis use, and this was robust to non-verbal communication previously reported in substance adjustment (see Table 3 for details). users are more likely to be the outcome of prolonged use [6–8, 10, 11], as opposed to reflecting self-medication of these deficits. In the opposite temporal direction, our results Association of early adolescent substance use (age 15) indicate that current alcohol, tobacco, and/or cannabis use with later social cognition (age 18) at age 15 is associated with poor social communication and social reciprocity at 17. In all cases, adjustment for pre-birth, Social communication maternal, childhood, or previous indication of poor social cognition (age 7) did not substantially alter these associa- Increased odds of poor social communication was associ- tions. As both analyses adjust for previous indication of ated with earlier adolescent alcohol (fully adjusted OR poor social cognition prior to the onset of any substance 1.46, 95% CI 0.99–2.14, P = 0.051), and tobacco use (fully use (age 7), this suggests that being a current user of alco- adjusted OR 1.95, 95% CI 1.33–2.86, P = 0.001). There hol, tobacco, and cannabis may have a substantial impact on was no clear evidence of an association of poor social social cognitive abilities. communication with earlier cannabis use. These results Generally, these analyses suggest that social cognitive are shown in Table 4. In stratified analyses, associations deficits may result from the initiation and/or regular use of were slightly stronger for males, with respect to tobacco these substances. While previous literature has suggested outcomes. these social cognitive deficits can arise during periods of acute intoxication [5, 12] or withdrawal [10], our results suggest these deficits remain present over longer periods of Social reciprocity time among users. Alcohol dependence has been associated with impaired semantic memory (i.e., deficits general knowl- Increased odds of poor social reciprocity was associated edge accumulated through personal experience). As seman- with earlier adolescent alcohol (fully adjusted OR 1.57, tic memory may be necessary for the maintenance of social 95% CI 1.18–2.09, P = 0.002), tobacco (fully adjusted OR networks [55], this may subsequently lead to more specific 1.92, 95% CI 1.43–2.58, P = < 0.001), and cannabis use social cognitive deficits [ 14]. Prolonged nicotine exposure (fully adjusted OR 1.54, 95% CI 1.16–2.05, P = 0.003). may dysregulate the hypothalamic–pituitary–adrenal system, These results are shown in Table 4. In stratified analyses, resulting in the hypersecretions of cortisol and alterations in associations were slightly stronger for males, with respect the activity of the associated monoamine neurotransmitter to tobacco outcomes (see Supplementary Tables S8). system, which contributes to stress-regulation [56]. This may result in individuals being more susceptible to environmen- tal stressors and associated difficulties with affect and emo- tional regulation [57, 58]. Finally, evidence from imaging studies indicate neuroanatomical changes in heavy cannabis uses associated with prolonged endocannabinoid exposure, 1 3 Eur Child Adolesc Psychiatry (2018) 27:739–752 745 1 3 Table 2 Associations of poor childhood social cognition (age 7/8) with adolescent current substance use (age 18) Unadjusted Adjustment 1 pre-birth/demographic Adjustment 2 maternal Adjustment 3 offspring N OR 95% CI p N OR 95% CI p N OR 95% CI p N OR 95% CI p Full case analysis  Poor non-verbal communication   Alcohol 2985 0.74 (0.62– 0.001 2395 0.74 (0.61– 0.004 1890 0.77 (0.62– 0.027 1567 0.70 (0.54– 0.007 0.89) 0.91) 0.97) 0.91)   Tobacco 2985 0.83 (0.68– 0.059 2395 0.77 (0.62– 0.024 1890 0.70 (0.54– 0.009 1567 0.62 (0.47– 0.001 1.01) 0.97) 0.92) 0.83)   Cannabis 2985 0.72 (0.59– 0.001 2395 0.75 (0.60– 0.009 1890 0.72 (0.56– 0.010 1567 0.62 (0.46– 0.001 0.88) 0.93) 0.92) 0.83)   Multi-sub- 2985 0.73 (0.56– 0.024 2395 0.75 (0.56– 0.060 1890 0.80 (0.57– 0.191 1567 0.67 (0.46– 0.047 stance 0.96) 1.01) 1.12) 1.00)  Poor social communication   Alcohol 3007 1.33 (1.01– 0.045 2543 1.27 (0.93– 0.126 2025 1.29 (0.91– 0.155 1609 1.42 (0.95– 0.089 1.77) 1.73) 1.84) 2.12)   Tobacco 3007 1.39 (1.03– 0.030 2543 1.32 (0.95– 0.100 2025 1.09 (0.74– 0.667 1609 1.09 (0.70– 0.708 1.87) 1.82) 1.61) 1.70)   Cannabis 3007 1.57 (1.17– 0.002 2543 1.38 (1.00– 0.047 2025 1.32 (0.91– 0.146 1609 1.56 (1.02– 0.039 2.09) 1.91) 1.91) 2.37)   Multi-sub- 3007 1.55 (1.08– 0.017 2543 1.27 (0.93– 0.126 2025 1.18 (0.73– 0.508 1609 1.30 (0.73– 0.394 stance 2.22) 1.73) 1.90) 2.20)  Poor social reciprocity   Alcohol 3058 1.11 (0.91– 0.305 2586 1.12 (0.90– 0.304 2061 1.06 (0.82– 0.640 1638 1.10 (0.82– 0.544 1.37) 1.41) 1.38) 1.47)   Tobacco 3058 1.23 (0.99– 0.064 2586 1.21 (0.95– 0.120 2061 1.17 (0.89– 0.263 1638 1.14 (0.83 0.409 1.53) 1.54) 1.56) to1.58)   Cannabis 3058 1.27 (1.02– 0.029 2586 1.28 (1.01– 0.043 2061 1.29 (0.98– 0.067 1638 1.29 (0.94– 0.109 1.58) 1.62) 1.69) 1.77)   Multi-sub- 3058 1.24 (0.94– 0.126 2586 1.19 (0.88– 0.256 2061 1.09 (0.76– 0.641 1638 1.02 (0.67– 0.923 stance 1.64) 1.62) 1.56) 1.55) Complete case analysis  Poor non-verbal communication   Alcohol 1567 0.68 (0.53– 0.003 1567 0.66 (0.52– 0.002 1567 0.58 (0.44– <0.001 1567 0.70 (0.54– 0.007 0.88) 0.86) 0.78) 0.91)   Tobacco 1567 0.66 (0.49– 0.004 1567 0.64 (0.48– 0.003 1567 0.65 (0.48– 0.004 1567 0.62 (0.47– 0.001 0.88) 0.86) 0.87) 0.83) 746 Eur Child Adolesc Psychiatry (2018) 27:739–752 1 3 Table 2 (continued) Unadjusted Adjustment 1 pre-birth/demographic Adjustment 2 maternal Adjustment 3 offspring N OR 95% CI p N OR 95% CI p N OR 95% CI p N OR 95% CI p   Cannabis 1567 0.63 (0.48– 0.001 1567 0.58 (0.44– <0.001 1567 0.58 (0.44– <0.001 1567 0.62 (0.46– 0.001 0.83) 0.78) 0.78) 0.83)   Multi-sub- 1567 0.68 (0.47– 0.050 1567 0.64 (0.44– 0.025 1567 0.66 (0.45– 0.035 1567 0.67 (0.46– 0.047 stance 1.00) 0.95) 0.97) 1.00)  Poor social communication   Alcohol 1609 1.28 (0.87– 0.212 1609 1.26 (0.85– 0.245 1609 1.27 (0.96– 0.228 1609 1.42 (0.95– 0.089 1.89) 1.86) 1.89) 2.12)   Tobacco 1609 1.10 (0.72– 0.648 1609 1.13 (0.73– 0.581 1609 1.14 (0.73– 0.567 1609 1.09 (0.70– 0.708 1.69) 1.74) 1.76) 1.70)   Cannabis 1609 1.47 (0.99– 0.059 1609 1.36 (0.90– 0.141 1609 1.35 (0.90– 0.147 1609 1.56 (1.02– 0.039 2.19) 2.03) 2.04) 2.37)   Multi-sub- 1609 1.22 (0.72– 0.457 1609 1.20 (0.71– 0.496 1609 1.22 (0.71– 0.468 1609 1.30 (0.73– 0.394 stance 2.07) 2.05) 2.08) 2.20)  Poor social reciprocity   Alcohol 1638 1.07 (0.80– 0.637 1638 1.05 (0.79– 0.730 1638 1.03 (0.77– 0.827 1638 1.10 (0.82– 0.544 1.43) 1.40) 1.38) 1.47)   Tobacco 1638 1.18 (0.86– 0.311 1638 1.20 (0.87– 0.262 1638 1.17 (0.85– 0.321 1638 1.14 (0.83 0.409 1.61) 1.64) 1.62) to1.58)   Cannabis 1638 1.28 (0.95– 0.102 1638 1.23 (0.90– 0.188 1638 1.19 (0.87– 0.267 1638 1.29 (0.94– 0.109 1.73) 1.66) 1.62) 1.77)   Multi-sub- 1638 1.04 (0.69– 0.867 1638 1.01 (0.67– 0.945 1638 1.00 (0.66– 0.984 1638 1.02 (0.67– 0.923 stance 1.59) 1.53) 1.51) 1.55) Poor non-verbal communication: ≥ seven total errors on the DANVA Poor social reciprocity: scoring yes on ≥ three from five sub questions on social reciprocity on the SCDC Poor social communication: total score of ≥ 16 on the SCDC Current alcohol use: ≥ 8 AUDIT Current tobacco use: use of tobacco is past 30 days Current cannabis use: use of cannabis in past 12 months Multi-substance users were classified as being current users of all three substances. (1) Adjusted for pre-birth/demographic characteristics: sex, parity, maternal social class, home ownership status, and maternal age. (2) Additionally adjusted for maternal substance use confounders: maternal cannabis use (offspring age 9) maternal smoking and binge drinking (offspring age 12). (3) Additionally adjusted for offspring confounders: IQ, peer problems, victimisation, borderline personality diagnosis Eur Child Adolesc Psychiatry (2018) 27:739–752 747 1 3 Table 3 Associations of poor childhood (age 7/8) identification of high and low intensity faces with adolescent current substance use (age 15) Unadjusted Adjustment 1 pre-birth/demographic Adjustment 2 maternal Adjustment 3 offspring N OR 95% CI p N OR 95% CI p N OR 95% CI p N OR 95% CI p Full case analysis  Poor identification of high intensity faces   Alcohol 2985 0.75 (0.62–0.91) 0.004 2398 0.78 (0.63–0.96) 0.017 1890 0.77 (0.60–0.98) 0.032 1567 0.70 (0.53–0.92) 0.010   Tobacco 2985 0.75 (0.60–0.93) 0.007 2398 0.76 (0.69–0.96) 0.020 1890 0.72 (0.55–0.96) 0.023 1567 0.63 (0.46–0.87) 0.004   Cannabis 2985 0.69 (0.56–0.85) 0.001 2398 0.73 (0.58–0.92) 0.009 1890 0.69 (0.53–0.91) 0.008 1567 0.63 (0.46–0.85) 0.003   Multi-substance 2985 0.62 (0.46–0.83) 0.002 2398 0.66 (0.48–0.92) 0.012 1890 0.71 (0.49–1.02) 0.064 1567 0.61 (0.39–0.93) 0.230  Poor identification of low intensity faces   Alcohol 2985 0.72 (0.60–0.87) 0.001 2398 0.68 (0.55–0.84) <0.001 1890 0.68 (0.53–0.85) 0.001 1567 0.67 (0.51–0.85) 0.002   Tobacco 2985 0.92 (0.75–1.12) 0.418 2398 0.82 (0.66–1.03) 0.091 1890 0.73 (0.56–0.96) 0.023 1567 0.67 (0.50–0.90) 0.009   Cannabis 2985 0.83 (0.68–1.01) 0.063 2398 0.78 (0.62–0.97) 0.027 1890 0.73 (0.57–0.95) 0.017 1567 0.68 (0.51–0.90) 0.008   Multi-substance 2985 0.86 (0.66–1.12) 0.254 2398 0.84 (0.62–1.13) 0.244 1890 0.79 (0.57–1.12) 0.185 1567 0.73 (0.49–1.07) 0.108 Complete case analysis  Poor identification of high intensity faces   Alcohol 1567 0.68 (0.53–0.89) 0.005 1567 0.67 (0.51–0.87) 0.003 1567 0.68 (0.52–0.89) 0.005 1567 0.70 (0.53–0.92) 0.010   Tobacco 1567 0.64 (0.47–0.88) 0.005 1567 0.64 (0.47–0.88) 0.005 1567 0.66 (0.48–.90) 0.009 1567 0.63 (0.46–0.87) 0.004   Cannabis 1567 0.63 (0.47–0.85) 0.002 1567 0.59 (0.44–0.80) 0.001 1567 0.61 (0.45–0.83) 0.001 1567 0.63 (0.46–0.85) 0.003   Multi-substance 1567 0.60 (0.40–0.92) 0.018 1567 0.58 (0.38–0.89) 0.012 1567 0.60 (0.40–0.93) 0.021 1567 0.61 (0.39–0.93) 0.230  Poor identification of low intensity faces   Alcohol 1567 0.66 (0.52–0.86) 0.002 1567 0.64 (0.50–0.83) 0.001 1567 0.65 (0.50–0.84) 0.001 1567 0.67 (0.51–0.85) 0.002   Tobacco 1567 0.70 (0.53–0.94) 0.018 1567 0.68 (0.51–0.92) 0.011 1567 0.69 (0.51–0.92) 0.013 1567 0.67 (0.50–0.90) 0.009   Cannabis 1567 0.67 (0.51–0.89) 0.006 1567 0.68 (0.51–0.90) 0.007 1567 0.66 (0.50–0.88) 0.005 1567 0.68 (0.51–0.90) 0.008   Multi-substance 1567 0.76 (0.52–1.10) 0.144 1567 0.72 (0.49–1.05) 0.088 1567 0.72 (0.49–1.06) 0.100 1567 0.73 (0.49–1.07) 0.108 Poor identification of high intensity faces: ≥ three errors on high intensity DANVA faces Poor identification of low intensity faces: ≥ five errors on low intensity DANVA faces Current alcohol use: ≥ eight AUDIT Current tobacco use: use of tobacco is past 30 days Current cannabis use: use of cannabis in past 12 months Multi-substance users were classified as being current users of all three substances. (1) Adjusted for pre-birth/demographic characteristics: sex, parity, maternal social class, home ownership status, and maternal age. (2) Additionally adjusted for maternal substance use confounders: maternal cannabis use (offspring age 9) maternal smoking and binge drinking (offspring age 12). (3) Additionally adjusted for offspring confounders: IQ, peer problems, victimisation, borderline personality diagnosis 748 Eur Child Adolesc Psychiatry (2018) 27:739–752 1 3 Table 4 Association of current substance use (age 15) predicting poor social cognition (age 18) Unadjusted Adjustment 1 pre-birth/demo- Adjustment 2 maternal Adjustment 3 offspring Adjustment 4 poor social cogni- graphic tion N OR 95% CI p N OR 95% CI p N OR 95% CI p N OR 95% CI p N OR 95% CI p Full case analysis  Alcohol   Social 3631 1.21 (0.94–1.57) 0.139 3089 1.45 (1.10–1.91) 0.009 2550 1.38 (1.00–1.88) 0.048 2002 1.49 (1.03–2.14) 0.033 1915 1.46 (0.99–2.14) 0.051 commu- nication   Social 3704 1.36 (1.12–1.65) 0.002 3147 1.50 (1.22–1.86) <0.001 2599 1.40 (1.11–1.78) 0.005 2041 1.59 (1.21–2.08) 0.001 1976 1.57 (1.18–2.09) 0.002 Reci- procity  Tobacco   Social 3662 2.05 (1.61–2.61) <0.001 3113 2.03 (1.55–2.68) <0.001 2570 2.02 (1.48–2.77) <0.001 2020 1.97 (1.37–2.84) <0.001 1933 1.95 (1.33–2.86) 0.001 commu- nication   Social 3736 1.97 (1.63–2.39) <0.001 3172 1.06 (1.66–2.55) <0.001 2620 2.02 (1.58–2.57) <0.001 2059 1.98 (1.49–2.62) <0.001 1994 1.92 (1.43–2.58) <0.001 Reci- procity  Cannabis   Social 3637 1.38 (1.07–1.77) 0.012 3095 1.31 (0.98–1.73) 0.064 2555 1.28 (0.39–1.77) 0.128 2009 1.32 (0.91–1.91) 0.146 1992 1.26 (0.85–1.86) 0.255 commu- nication   Social 3710 1.54 (1.27–1.86) <0.001 3153 1.54 (1.25–1.90) <0.001 2604 1.55 (1.22–1.96) <0.001 2048 1.57 (1.19–2.06) 0.001 1983 1.54 (1.16–2.05) 0.003 Reci- procity Complete case analysis  Alcohol   Social 1915 1.30 (0.91–1.87) 0.151 1915 1.38 (0.96–1.98) 0.087 1915 1.41 (0.98–2.04) 0.660 1915 1.49 (1.02–2.15) 0.037 1915 1.46 (0.99–2.14) 0.051 commu- nication   Social 1976 1.44 (1.10–1.88) 0.008 1976 1.53 (1.17–2.01) 0.002 1976 1.54 (1.17–2.02) 0.002 1976 1.61 (1.22–2.13) 0.001 1976 1.57 (1.18–2.09) 0.002 Reci- procity  Tobacco 1933 1.95 (1.36–2.78) <0.001 1933 1.94 (1.34–2.79) <0.001 1933 1.95 (1.34–2.82) <0.001 1933 1.94 (1.34–2.82) <0.001 1933 1.95 (1.33–2.86) 0.001   Social commu- nication   Social 1994 1.92 (1.46–2.53) <0.001 1994 1.99 (1.50–2.64) <0.001 1994 1.97 (1.49–2.62) <0.001 1994 2.00 (1.50–2.66) <0.001 1994 1.92 (1.43–2.58) <0.001 Reci- procity Eur Child Adolesc Psychiatry (2018) 27:739–752 749 and the subsequent desensitisation of CB receptors in the brain, requiring compensatory CB1 receptor activity else- where in the striatum [16–19]. Previous literature indicates strong familial bonds and open communication within fami- lies and schools may serve as a protective factor, or help to delay adolescent substance initiation [59–62]. However, in the other temporal direction (i.e., poor social cognition and subsequent substance use), there is currently little evidence. Our analyses help to rule out the possibility of reverse cau- sality, and strengthen our findings that substance use is asso- ciated with later impaired social cognition. Additionally, this analysis suggested that poor non-verbal communication may in fact be protective with respect to subsequent substance use. While this is clearly an area that warrants additional research and replication, one possible explanation for this finding is that adolescents with poor emotion recognition skills may less likely to have larger social groups [63, 64] and therefore less likely to engage in substance use due to less social inclusion [65–68]. Strengths of this study include a rich data set with multiple social cognitive and substance use variables col- lected at several time-points throughout the adolescence and early adulthood. This allows for the analysis of both temporal directions as well as examining different facets of social cognition. Additionally, a robust approach was taken to minimise confounding, using a range of possible con- founders from pre-birth throughout adolescence. There are also some limitations in our study to consider. First, some of our exposures were self-reported by the child (DANVA) while others were parent-completed (SCDC and DAWBA). Previous studies have indicated parental rating of offspring well-being to be more positive compared to self-report by offspring [69]. Similarly, the maternal-reported measure of SCDC taken when offspring were aged 17 may be captur - ing a breakdown in family communication or adolescent disobedience, as opposed to social cognition, due to the generally rebellious nature of the adolescent period. How- ever, a recent genome-wide association study conducted in ALSPAC found evidence of a genome-wide association of SCDC measures at age 17, suggesting there is a genetic architecture of social communication that can be reliably captured by the maternal SCDC measure [70]. Second, SCDC scores are known to remain constant across age groups [27], while studies have indicated DANVA scores to improve with age [26]. This is a potential problem if the ranking of scores across the population is not consist- ent; however, previous ALSPAC studies have indicated a test–retest reliability in the DANVA of 0.84 [71]. Third, as maternal data are collected frequently and are more extensive than partner data within ALSPAC, we only investigated the impact of maternal confounding. Fourth, our substance use outcomes are all reliant on self-report and we were not able to biochemically validate these 1 3 Table 4 (continued) Unadjusted Adjustment 1 pre-birth/demo- Adjustment 2 maternal Adjustment 3 offspring Adjustment 4 poor social cogni- graphic tion N OR 95% CI p N OR 95% CI p N OR 95% CI p N OR 95% CI p N OR 95% CI p Cannabis   Social 1922 1.16 (0.89–1.67) 0.440 1922 1.16 (0.80–1.68) 0.438 1922 1.20 (0.82–1.75) 0.344 1922 1.27 (0.86–1.85) 0.227 1992 1.26 (0.85–1.86) 0.255 commu- nication   Social 1983 1.44 (1.11–1.88) 0.007 1983 1.48 (1.13–1.93) 0.004 1983 1.48 (1.13–1.95) 0.005 1983 1.57 (1.19–2.06) 0.001 1983 1.54 (1.16–2.05) 0.003 Reci- procity Poor social reciprocity: scoring yes on ≥ three from five sub questions on social reciprocity on the SCDC Poor social communication: total score of ≥ 16 on the SCDC Current alcohol use: ≥ 20 drinks in past 6 months Current tobacco use: use of tobacco is past 30 days Current cannabis use: use of cannabis in past 12 months. (1) Adjusted for pre-birth/demographic characteristics: sex, parity, maternal social class, home ownership status, and maternal age. (2) Additionally adjusted for maternal substance use confounders: maternal cannabis use (offspring age 9) maternal smoking and binge drinking (offspring age 12). (3) Additionally adjusted for off- spring confounders: IQ, peer problems, victimisation, borderline personality diagnosis. (4) Additionally adjusted for poor SCDC or SCDC sub-scale (respectively) at offspring age 7 750 Eur Child Adolesc Psychiatry (2018) 27:739–752 and JH. All Authors approved the final version of the manuscript for responses. Additionally, we drew our outcomes from age submission. 18, which provided us with a large sample size of individu- als whom had ever used substances. However, there were Compliance with ethical standards notably fewer individuals answering questions regarding frequency of use, which may have contributed to the low Role of funding source Nothing declared power for these analyses. Fifth, it is possible that our vari- able for multi-substance current use simply reflects current Conflict of interest On behalf of all authors, the corresponding au- cannabis use, since cannabis users typically also consume thor states that there is no conflict of interest. alcohol and tobacco [72]. Finally, there was evidence of Open Access This article is distributed under the terms of the differential loss to follow-up, as some children with high Creative Commons Attribution 4.0 International License (http://crea- SCDC scores, were slightly more likely to drop out of tivecommons.org/licenses/by/4.0/), which permits unrestricted use, the study before substance use and social cognition out- distribution, and reproduction in any medium, provided you give appro- come data was obtained. However, this does not necessar- priate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. ily imply selection bias in the association between social cognition and later substance use [73], and comparisons of full and complete cases display little change in results due to sample size. References Overall, we found differing patterns of relationships between social cognition and substance use behaviour 1. Organization WH (2016) WHO global report on trends in tobacco dependent on the specific social cognition examined and smoking 2000–2025. World Health Organization. http://www. temporal direction of association. While poor non-verbal who.int/tobacco/publications/surveillance/reportontrendstobac- communication in childhood appeared to have protective cosmoking/en/index3.html. Accessed 03 Feb 2017 2. Organization WH Cannabis. http://www.who.int/substance_abuse/ factors against later substance use, early adolescent sub- facts/cannabis/en/. Accessed 03 Feb 2017 stance use was associated with decreased social cognitive 3. Organization WH Global Information System on Alcohol and performance. This association would be worth pursuing for Health (GISAH). http://www.who.int/gho/alcohol/en/. Accessed replication in cohorts with similar or richer data on social 03 Feb 2017 4. Winstock ARB, M.; Ferris, J.; Maier, L. (2016) What we learned cognitive variables. Furthermore, the use of differing statisti- from GDS2016. https://www.globaldrugsurvey.com/wp-content/ cal analyses and methods will additionally help to strengthen uploads/2016/06/TASTER-KEY-FINDINGS-FROM-GDS2016. these findings. Given that causality cannot be inferred from pdf. Accessed 03 Feb 2017 observational data alone, future epidemiological studies 5. Curtin JJ, Patrick CJ, Lang AR, Cacioppo JT, Birbaume N (2001) Alcohol affects emotion through cognition. Psychol Sci investigating these associations should consider alternative 12(6):527–531. doi:10.1111/1467-9280.00397 statistical techniques. For example, studies can use genetic 6. Donadon MF, Osorio Fde L (2014) Recognition of facial expres- variants associated with substance use or social communi- sions by alcoholic patients: a systematic literature review. Neu- cation as proxies for these exposures within a Mendelian ropsychiatr Dis Treat 10:1655–1663. doi:10.2147/NDT.S65376 7. Townshend JM, Duka T (2003) Mixed emotions: alcoholics’ randomization framework. impairments in the recognition of specific emotional facial expres- sions. Neuropsychologia 41(7):773–782 Acknowledgements We are extremely grateful to all the families 8. Kornreich C, Blairy S, Philippot P, Hess U, Noel X, Streel E, Le who took part in this study, the midwives for their help in recruit- Bon O, Dan B, Pelc I, Verbanck P (2001) Deficits in recognition ing the participants, and the whole ALSPAC team, which includes of emotional facial expression are still present in alcoholics after interviewers, computer and laboratory technicians, clerical workers, mid- to long-term abstinence. J Stud Alcohol 62(4):533–542 research scientists, volunteers, managers, receptionists and nurses. The 9. Leventhal AM, Munafo M, Tidey JW, Sussman S, Monterosso JR, UK Medical Research Council and the Wellcome Trust (Grant Ref: Sun P, Kahler CW (2012) Anhedonia predicts altered processing 102215/2/13/2) and the University of Bristol provide core support for of happy faces in abstinent cigarette smokers. Psychopharmacol- ALSPAC. This publication is the work of the authors and MF, JH, and ogy 222(2):343–351. doi:10.1007/s00213-012-2649-5 MM will serve as guarantors for the contents of this paper. The research 10. Adams S, Attwood AS, Munafo MR (2014) Effects of nicotine was specifically funded by a University of Bristol postgraduate research and nicotine expectancy on attentional bias for emotional stimuli. scholarship awarded to MF, and by the Medical Research Council and Nicotine Tobacco Research. doi:10.1093/ntr/ntu219 the University of Bristol (MC_UU_12013/6). 11. Bayrakci A, Sert E, Zorlu N, Erol A, Saricicek A, Mete L (2015) MF and MM are members of the UK Centre for Tobacco and Alco- Facial emotion recognition deficits in abstinent cannabis depend- hol Studies, a UKCRC Public Health Research: Centre of Excellence. ent patients. Compr Psychiatry 58:160–164. doi:10.1016/j. Funding from the British Heart Foundation, Cancer Research UK, comppsych.2014.11.008 Economic and Social Research Council, Medical, Research Council 12. Hindocha C, Freeman TP, Schafer G, Gardener C, Das RK, Mor- and the National Institute for Health Research, under the auspices of gan CJ, Curran HV (2015) Acute effects of delta-9-tetrahydro - the UK Clinical Research Collaboration, is gratefully acknowledged. cannabinol, cannabidiol and their combination on facial emotion recognition: a randomised, double-blind, placebo-controlled study Author Contributions All authors designed the study. MEF con- ducted the analysis and drafted the manuscript with input from MRM 1 3 Eur Child Adolesc Psychiatry (2018) 27:739–752 751 in cannabis users. Eur Neuropsychopharmacol 25(3):325–334. 28. Skuse D, Warrington R, Bishop D, Chowdhury U, Lau J, Mandy doi:10.1016/j.euroneuro.2014.11.014 W, Place M (2004) The developmental, dimensional and diag- 13. Mitchell IJ, Beck SR, Boyal A, Edwards VR (2011) Theory of nostic interview (3di): a novel computerized assessment for mind deficits following acute alcohol intoxication. Eur Addict Res autism spectrum disorders. J Am Acad Child Adolesc Psychia- 17(3):164–168. doi:10.1159/000324871 try 43(5):548–558. doi:10.1097/00004583-200405000-00008 14. Nandrino JL, Gandolphe MC, Alexandre C, Kmiecik E, Yguel 29. Skuse DH, Mandy WPL, Scourfield J (2005) Measuring autistic J, Urso L (2014) Cognitive and affective theory of mind abili- traits: heritability, reliability and validity of the social and com- ties in alcohol-dependent patients: the role of autobiographi- munication disorders checklist. Brit J Psychiat 187:568–572. cal memory. Drug Alcohol Depend 143:65–73. doi:10.1016/j. doi:10.1192/bjp.187.6.568 drugalcdep.2014.07.010 30. Wallace JM Jr, Brown TN, Bachman JG, LaVeist TA (2003) 15. Bosco FM, Capozzi F, Colle L, Marostica P, Tirassa M (2014) The influence of race and religion on abstinence from alcohol, Theory of mind deficit in subjects with alcohol use disorder: an cigarettes and marijuana among adolescents. J Stud Alcohol analysis of mindreading processes. Alcohol Alcohol 49(3):299– 64(6):843–848 307. doi:10.1093/alcalc/agt148 31. Michalak L, Trocki K, Bond J (2007) Religion and alcohol in 16. Roser P, Lissek S, Tegenthoff M, Nicolas V, Juckel G, Brune M the U.S. National Alcohol Survey: how important is religion for (2012) Alterations of theory of mind network activation in chronic abstention and drinking? Drug Alcohol Depend 87(2–3):268– cannabis users. Schizophr Res 139(1–3):19–26. doi:10.1016/j. 280. doi:10.1016/j.drugalcdep.2006.07.013 schres.2012.05.020 32. Pardini D, White HR, Stouthamer-Loeber M (2007) Early ado- 17. Marjoram D, Job DE, Whalley HC, Gountouna VE, McIntosh lescent psychopathology as a predictor of alcohol use disorders AM, Simonotto E, Cunningham-Owens D, Johnstone EC, Lawrie by young adulthood. Drug Alcohol Depend 88(Suppl 1):S38– S (2006) A visual joke fMRI investigation into Theory of Mind S49. doi:10.1016/j.drugalcdep.2006.12.014 and enhanced risk of schizophrenia. Neuroimage 31(4):1850– 33. Evans EA, Grella CE, Washington DL, Upchurch DM (2017) 1858. doi:10.1016/j.neuroimage.2006.02.011 Gender and race/ethnic differences in the persistence of alcohol, 18. Romero J, Garcia-Palomero E, Castro JG, Garcia-Gil L, Ramos drug, and poly-substance use disorders. Drug Alcohol Depend JA, Fernandez-Ruiz JJ (1997) Effects of chronic exposure to 174:128–136. doi:10.1016/j.drugalcdep.2017.01.021 delta9-tetrahydrocannabinol on cannabinoid receptor binding and 34. Skuse DH, Mandy W, Steer C, Miller LL, Goodman R, Law- mRNA levels in several rat brain regions. Brain Res Mol Brain rence K, Emond A, Golding J (2009) Social communication Res 46(1–2):100–108 competence and functional adaptation in a general population 19. Sim-Selley LJ (2003) Regulation of cannabinoid CB1 receptors of children: preliminary evidence for sex-by-verbal IQ differen- in the central nervous system by chronic cannabinoids. Crit Rev tial risk. J Am Acad Child Adolesc Psychiatry 48(2):128–137. Neurobiol 15(2):91–119 doi:10.1097/CHI.0b013e31819176b8 20. Hawkins JD, Catalano RF, Kosterman R, Abbott R, Hill KG 35. Argys LM, Rees DI, Averett SL, Witoonchart B (2006) Birth (1999) Preventing adolescent health-risk behaviors by strength- order and risky adolescent behavior. Econ Inq 44(2):215–233. ening protection during childhood. Arch Pediat Adol Med doi:10.1093/ei/cbj011 153(3):226–234 36. Warren J (1966) Birth order and social behaviour. Psychol Bull 21. Battistich V, Schaps E, Watson M, Solomon D (1996) Prevention 65(1):38–49 effects of the child development project: early findings from an 37. Buu A, DiPiazza C, Wang J, Puttler LI, Fitzgerald HE, Zucker ongoing multisite demonstration trial. J Adolescent Res 11(1):12– RA (2009) Parent, family, and neighborhood effects on the 35. doi:10.1177/0743554896111003 development of child substance use and other psychopathology 22. Rhee SH, Hewitt JK, Young SE, Corley RP, Crowley TJ, Stallings from preschool to the start of adulthood. J Stud Alcohol Drugs MC (2003) Genetic and environmental influences on substance 70(4):489–498 initiation, use, and problem use in adolescents. Arch Gen Psy- 38. Stone AL, Becker LG, Huber AM, Catalano RF (2012) Review chiatry 60(12):1256–1264. doi:10.1001/archpsyc.60.12.1256 of risk and protective factors of substance use and problem 23. King KM, Chassin L (2007) A prospective study of the effects of use in emerging adulthood. Addict Behav 37(7):747–775. age of initiation of alcohol and drug use on young adult substance doi:10.1016/j.addbeh.2012.02.014 dependence. J Stud Alcohol Drugs 68(2):256–265 39. Bradley RH, Corwyn RF (2002) Socioeconomic status and child 24. Agrawal A, Grant JD, Waldron M, Duncan AE, Scherrer JF, development. Annu Rev Psychol 53:371–399. doi:10.1146/ Lynskey MT, Madden PA, Bucholz KK, Heath AC (2006) Risk annurev.psych.53.100901.135233 for initiation of substance use as a function of age of onset of 40. Coyne CA, Langstrom N, Rickert ME, Lichtenstein P, cigarette, alcohol and cannabis use: findings in a Midwestern D’Onofrio BM (2013) Maternal age at first birth and offspring female twin cohort. Prev Med 43(2):125–128. doi:10.1016/j. criminality: using the children of twins design to test causal ypmed.2006.03.022 hypotheses. Dev Psychopathol 25(1):17–35. doi:10.1017/ 25. Boyd A, Golding J, Macleod J, Lawlor DA, Fraser A, Henderson S0954579412000879 J, Molloy L, Ness A, Ring S, Smith GD (2013) Cohort profile: the 41. Benjet C (2010) Childhood adversities of populations living ‘Children of the 90 s’-the index offspring of the Avon Longitudi- in low-income countries: prevalence, characteristics, and men- nal Study of Parents and Children. Int J Epidemiol 42(1):111–127. tal health consequences. Curr Opin Psychiatry 23(4):356–362. doi:10.1093/ije/dys064 doi:10.1097/YCO.0b013e32833ad79b 26. Nowicki S, Duke MP (1994) Individual-differences in the nonver - 42. Lopez-Castroman J (2014) Young maternal age and old pater- bal-communication of affect - the diagnostic-analysis of nonver - nal age induce similar risk of mental disorders in offspring. Evid bal accuracy scale. J Nonverbal Behav 18(1):9–35. doi:10.1007/ Based Ment Health 17(4):100. doi:10.1136/eb-2014-101861 Bf02169077 43. Arria AM, Mericle AA, Meyers K, Winters KC (2012) Paren- 27. Robinson EB, Munir K, Munafo MR, Hughes M, McCormick tal substance use impairment, parenting and substance use dis- MC, Koenen KC (2011) Stability of autistic traits in the general order risk. J Subst Abuse Treat 43(1):114–122. doi:10.1016/j. population: further evidence for a continuum of impairment. J jsat.2011.10.001 44. Chassin L, Pitts SC, Prost J (2002) Binge drinking trajectories Am Acad Child Adolesc Psychiatry 50(4):376–384. doi:10.1016/j. from adolescence to emerging adulthood in a high-risk sample: jaac.2011.01.005 1 3 752 Eur Child Adolesc Psychiatry (2018) 27:739–752 predictors and substance abuse outcomes. J Consult Clin Psychol substance use: the roles of family structure, function and gender. 70(1):67–78 Addiction 97(3):329–336 45. Trinidad DR, Johnson CA (2002) The association between 60. Kliewer W, Murrelle L (2007) Risk and protective factors for emotional intelligence and early adolescent tobacco and adolescent substance use: findings from a study in selected alcohol use. Pers Indiv Differ 32(1):95–105. doi:10.1016/ Central American countries. J Adolesc Health 40(5):448–455. S0191-8869(01)00008-3 doi:10.1016/j.jadohealth.2006.11.148 46. Lavin C (1996) The Wechsler intelligence scale for children third 61. Farrell AD, White KS (1998) Peer influences and drug use among edition and the Stanford–Binet Intelligence scale: fourth edition: urban adolescents: family structure and parent-adolescent relation- a preliminary study of validity. Psychol Rep 78(2):491–496 ship as protective factors. J Consult Clin Psychol 66(2):248–258 47. Tharp-Taylor S, Haviland A, D’Amico EJ (2009) Victimization 62. Spoth R, Redmond C, Shin C, Azevedo K (2004) Brief family from mental and physical bullying and substance use in early intervention effects on adolescent substance initiation: school- adolescence. Addict Behav 34(6–7):561–567. doi:10.1016/j. level growth curve analyses 6 years following baseline. J Consult addbeh.2009.03.012 Clin Psychol 72(3):535–542. doi:10.1037/0022-006X.72.3.535 48. Dodge KA, Lansford JE, Burks VS, Bates JE, Pettit GS, Fontaine 63. Leppanen JM, Hietanen JK (2001) Emotion recognition and R, Price JM (2003) Peer rejection and social information-process- social adjustment in school-aged girls and boys. Scand J Psychol ing factors in the development of aggressive behavior problems in 42(5):429–435 children. Child Dev 74(2):374–393 64. Barth JM, Bastiani A (1997) A longitudinal study of emotion rec- 49. Salmivalli C, Lagerspetz K, Bjorkqvist K, Osterman K, ognition and preschool children’s social behavior. Merrill Palmer Kaukiainen A (1996) Bullying as a group process: par- Q 43(1):107–128 ticipant roles and their relations to social status within 65. Urberg KA, Degirmencioglu SM, Pilgrim C (1997) Close friend the group. Aggress Behav 22(1):1–15. doi: 10.1002/ and group influence on adolescent cigarette smoking and alcohol (Sici)1098-2337(1996)22:1<1:Aid-Ab1>3.0.Co;2-T use. Dev Psychol 33(5):834–844 50. Wolke D, Woods S, Stanford K, Schulz H (2001) Bullying and vic- 66. McCrady BS (2004) To have but one true friend: implica- timization of primary school children in England and Germany: tions for practice of research on alcohol use disorders and prevalence and school factors. Br J Psychol 92(Pt 4):673–696 social network. Psychol Addict Behav 18(2):113–121. 51. Trull TJ, Sher KJ, Minks-Brown C, Durbin J, Burr R (2000) Bor- doi:10.1037/0893-164X.18.2.113 derline personality disorder and substance use disorders: a review 67. Shadur J, Hussong A (2014) Friendship intimacy, close friend and integration. Clin Psychol Rev 20(2):235–253 drug use, and self-medication in adolescence. J Soc Pers Relat 52. Roepke S, Vater A, Preissler S, Heekeren HR, Dziobek I (2013) 31(8):997–1018. doi:10.1177/0265407513516889 Social cognition in borderline personality disorder. Front Neuro- 68. Alfaro DLP, Ehlinger V, Spilka S, Ross J, Sentenac M, Godeau E sci-Switz 6. doi:ARTN 195 10.3389/fnins.2012.00195 (2017) Alcohol, tobacco and cannabis use: do students with mild- 53. Murphy SM, Faulkner DM, Farley LR (2014) The behaviour intellectual disability mimic students in the general population? of young children with social communication disorders during Res Dev Disabil 63:118–131. doi:10.1016/j.ridd.2016.10.009 dyadic interaction with peers. J Abnorm Child Psychol 42(2):277– 69. Waters E, Stewart-Brown S, Fitzpatrick R (2003) Agreement 289. doi:10.1007/s10802-013-9772-6 between adolescent self-report and parent reports of health and 54. Goodman R (2001) Psychometric properties of the strengths and well-being: results of an epidemiological study. Child Care Health difficulties questionnaire. J Am Acad Child Adolesc Psychiatry Dev 29(6):501–509 40(11):1337–1345. doi:10.1097/00004583-200111000-00015 70. St Pourcain B, Skuse DH, Mandy WP, Wang K, Hakonar- 55. Labouvie-Vief G, Blanchard-Fields F (1982) Cognitive ageing and son H, Timpson NJ, Evans DM, Kemp JP, Ring SM, McArdle psychological growth. Ageing Soc 2(2):183–209. doi:10.1017/ WL, Golding J, Smith GD (2014) Variability in the common S0144686X00009429Published genetic architecture of social-communication spectrum pheno- 56. Markou A, Kosten TR, Koob GF (1998) Neurobiological simi- types during childhood and adolescence. Mol Autism 5(1):18. larities in depression and drug dependence: a self-medication doi:10.1186/2040-2392-5-18 hypothesis. Neuropsychopharmacol 18(3):135–174. doi:10.1016/ 71. Barona M, Kothari R, Skuse D, Micali N (2015) Social communi- S0893-133x(97)00113-9 cation and emotion difficulties and second to fourth digit ratio in 57. Joormann J, Quinn ME (2014) Cognitive processes and emo- a large community-based sample. Mol Autism 6:68. doi:10.1186/ tion regulation in depression. Depress Anxiety 31(4):308–315. s13229-015-0063-7 doi:10.1002/da.22264 72. Raphael B, Wooding S, Stevens G, Connor J (2005) Comorbidity: 58. Joormann J, Stanton CH (2016) Examining emotion regulation cannabis and complexity. J Psychiatr Pract 11(3):161–176 in depression: a review and future directions. Behav Res Ther 73. Carter KN, Imlach-Gunasekara F, McKenzie SK, Blakely 86:35–49. doi:10.1016/j.brat.2016.07.007 T (2012) Differential loss of participants does not necessar- 59. McArdle P, Wiegersma A, Gilvarry E, Kolte B, McCarthy S, ily cause selection bias. Aust Nz J Publ Heal 36(3):218–222. Fitzgerald M, Brinkley A, Blom M, Stoeckel I, Pierolini A, doi:10.1111/j.1753-6405.2012.00867.x Michels I, Johnson R, Quensel S (2002) European adolescent 1 3

Journal

European Child & Adolescent PsychiatrySpringer Journals

Published: Oct 20, 2017

References

You’re reading a free preview. Subscribe to read the entire article.


DeepDyve is your
personal research library

It’s your single place to instantly
discover and read the research
that matters to you.

Enjoy affordable access to
over 18 million articles from more than
15,000 peer-reviewed journals.

All for just $49/month

Explore the DeepDyve Library

Search

Query the DeepDyve database, plus search all of PubMed and Google Scholar seamlessly

Organize

Save any article or search result from DeepDyve, PubMed, and Google Scholar... all in one place.

Access

Get unlimited, online access to over 18 million full-text articles from more than 15,000 scientific journals.

Your journals are on DeepDyve

Read from thousands of the leading scholarly journals from SpringerNature, Elsevier, Wiley-Blackwell, Oxford University Press and more.

All the latest content is available, no embargo periods.

See the journals in your area

DeepDyve

Freelancer

DeepDyve

Pro

Price

FREE

$49/month
$360/year

Save searches from
Google Scholar,
PubMed

Create lists to
organize your research

Export lists, citations

Read DeepDyve articles

Abstract access only

Unlimited access to over
18 million full-text articles

Print

20 pages / month

PDF Discount

20% off