Skip to main content
NCBI home page
Alcohol and Alcoholism (Oxford, Oxfordshire) logoLink to Alcohol and Alcoholism (Oxford, Oxfordshire)
. 2015 Nov 9;51(3):331–338. doi: 10.1093/alcalc/agv125

Adverse Childhood Experiences and Alcohol Consumption in Midlife and Early Old-Age

Jessica Pui Kei Leung 1,*, Annie Britton 1, Steven Bell 1
PMCID: PMC4830410  PMID: 26553290

Abstract

Aims

To examine the individual and cumulative effects of adverse childhood experiences (ACEs) on alcohol consumption in midlife and early old-age, and the role of ACEs in 10-year drinking trajectories across midlife.

Methods

Data were from the Whitehall II study, a longitudinal British civil service-based cohort study (N = 7870, 69.5% male). Multinomial logistic regression was used to examine the individual and cumulative effects of ACEs on weekly alcohol consumption. Mixed-effect multilevel modelling was used to explore the relationship between ACEs and change in alcohol consumption longitudinally.

Results

Participants who were exposed to parental arguments/fights in childhood were 1.24 (95% CI 1.06, 1.45) times more likely to drink at hazardous levels in midlife (mean age 56 years) after controlling for covariates and other ACEs. For each additional exposure to an ACE, the risk of hazardous drinking versus moderate drinking was increased by 1.12 (95% CI 1.03, 1.21) after adjusting for sex, age, adult socio-economic status, ethnicity and marital status. No associations between ACEs and increased risk of hazardous drinking in early old-age (mean age 66 years) were found. In longitudinal analyses, ACEs did not significantly influence 10-year drinking trajectories across midlife.

Conclusion

The effect of exposure to parental arguments on hazardous drinking persists into midlife.

INTRODUCTION

Adverse childhood experiences (ACEs) and alcohol misuse have been recognized globally as key public health issues (WHO, 2006, 2010).

ACEs refer to stressful events which an individual experiences in childhood, such as physical abuse, parental divorce or parental alcoholism (WHO, 2015). In the UK, a national household survey found that the prevalence of at least one ACE was 46.4% and 8.3% had experienced four or more ACEs (Bellis et al., 2014). Statistics on child maltreatment in the UK indicates that 1 in 14 children aged 11–17 (6.9%) report being physically abused by a parent or guardian during childhood (Radford et al., 2011). The number of children affected by parental divorce is significant. In 2012, there were 99,822 children who were aged under 16 when their parents divorced (Office for National Statistics, 2014). Furthermore, the number of children living with alcohol-dependent parents is largely unknown. Estimates suggest that ∼22% of children aged under 16 live with a hazardous drinker (Manning et al., 2009).

Childhood development has effects on health and well-being throughout life (Bartley, 2012). ACEs lead to poor health outcomes and health-threatening behaviours (Anda et al., 1999; Bair-Merritt et al., 2006; Shonkoff et al., 2012). Therefore, addressing the relationship between ACEs and alcohol use is important in designing and implementing prevention programmes and treating drinking problems at an individual and population level.

Previous studies have provided evidence on the pathways between childhood predictors and alcohol use, but most are based on adolescents and young adults. Participants with ACEs initiate alcohol use at an earlier age and are more likely to drink to cope with problems (Dube et al., 2006; Rothman et al., 2008). Adolescent drinking correlates with physical and sexual abuse, parental divorce, household substance abuse and mental illness of family members in childhood (Dube et al., 2006; Kristjansson et al., 2009; Shin et al., 2009). The relationship between ACEs and risk of problem drinking has been shown to persist into young adulthood (Hope et al., 1998; Kestilä et al., 2008; Thompson et al., 2008; Timko et al., 2008; Griffin and Amodeo, 2010).

Studies linking life-course determinants of alcohol misuse in middle-aged population are limited (Dubow et al., 2008). Most studies have tended to focus on parental drinking, parental divorce and physical abuse. Other ACEs such as childhood hospitalization, living in an orphanage and exposure to parental arguments have received relatively little attention (Spak et al., 1997; Mirsal et al., 2004; Kestilä et al., 2008). Additionally, the studies on alcohol use in midlife have tended to focus on a single type of ACE and have ignored the co-occurrence of various forms of ACEs. In fact, ACEs are highly interrelated, for example, children who have alcohol-dependent parents report having more parental divorce, physical abuse and family violence (Widom and Hiller-Sturmhöfel, 2001; Marshal, 2003; Gilbert et al., 2009). But few studies have examined the contribution of cumulative exposure to ACEs to the risk of hazardous drinking in midlife (Kauhanen et al., 2011). Moreover, there is a growing interest in the characteristics of those who abstain from alcohol. Some studies have found that abstainers have several psychosocial disadvantages, such as low education, not being married, depression (Bell et al., 2014), unemployment (Camacho et al., 1987; Naimi et al., 2005) and poorer health (Ng Fat and Shelton, 2012). However, the association between ACEs and abstinence is unknown. This topic should be explored further as ACEs might play a role in non-drinking behaviour. Furthermore, the impact of ACEs on midlife and later life alcohol consumption is poorly understood. A previous study found that there are long-term effects of exposure to childhood maltreatment on heavy drinking trajectories from adolescence to young adulthood (Shin et al., 2013). To our knowledge, the longitudinal relationship between ACEs and change in alcohol consumption in midlife has not been described.

The aim of this study was to examine the individual and cumulative effects of ACEs on weekly alcohol consumption in midlife and early old-age, and to investigate the role of ACEs in 10-year drinking trajectories from midlife through to early old-age.

METHODS

Data source

Data were drawn from phase 5 (1997–1999), phase 7 (2002–2004) and phase 9 (2007–2009) of the Whitehall II Study (Marmot and Brunner, 2005). The mean length of follow-up from phase 5 to 9 was 10.45 [Standard deviation (SD):0.58] years. Middle adulthood describes the age period of 40 to 60, and late adult transition is from 60 to 65 (Levinson, 1986). The mean age at phases 5 and 9 was 56 (SD: 6.04) and 66 (SD: 5.98) years, respectively. Phase 5 provides an estimate of alcohol consumption in midlife and acts as a baseline in the assessment of 10-year drinking trajectories from midlife to late adulthood. Phase 9 represents the period of early old-age. Phase 7 was within the transition period from midlife to early old-age and was used to provide additional information to capture the change in alcohol consumption during the transition period (Singer and Willett, 2003). The Whitehall II study therefore offers an important opportunity to examine the association between ACEs and alcohol consumption in both midlife and early old-age. The University College London ethics committee approved the study. Informed consent was obtained at baseline and renewed at each contact. Whitehall II data, protocols and other metadata are available to bona fide researchers for research purposes. Please refer to the Whitehall II data sharing policy at http://www.ucl.ac.uk/whitehallII/data-sharing.

Adverse childhood experiences

Six categories of ACE were studied: parental divorce, parental mental health/alcohol problems, physical abuse, hospitalization >4 weeks, living in an orphanage and exposure to parental arguments/fights. All questions about ACEs pertained to the participants' first 16 years of life and were assessed retrospectively at phase 5 through a self-completed questionnaire.

Alcohol consumption

Participants were asked to report the number of alcoholic drinks they had consumed in the last 7 days, in terms of spirits (measures), wine (glasses) and beer (pints). These were then converted into UK units of alcohol (in the UK, 1 unit of alcohol is equivalent to 8 g of ethanol) and summed to derive total weekly consumption. A conservative conversion was applied whereby a single measure of spirits and glass of wine was estimated to contain 1 unit of alcohol while a pint of beer was considered to contain 2. Categorical weekly alcohol consumption was used to examine the individual and cumulative effects of ACEs, while a continuous measure was used to estimate 10-year drinking trajectories in longitudinal analyses. Hazardous drinking was defined as consuming 21 or more units per week for men and 14 or more units for women (Department of Health, 1995). On the basis of weekly alcohol intake, participants were classified into three categories: past week non-drinkers (0 units), moderate drinkers (1–20 units for men, 1–13 units for women) and hazardous drinkers (21+ units for men and 14+ units for women). The reference group was moderate drinkers.

Possible confounders

Adjustments were made for a wide range of risk factors potentially associated with both alcohol consumption and ACEs. These included age, sex, ethnicity, smoking status, marital status and adult socio-economic position (SEP). Age was measured on the date of questionnaire completed at phases 5 and 9. Participants were classified into two ‘ethnic’ categories: white and non-white.

Smoking status

At phases 5, 7 and 9, participants were asked to report whether they currently smoked (cigarette/cigar), whether they previously smoked or whether they have never smoked (reference category). At Phases 5, 7 and 9, participants were classified into four categories: married/cohabiting (reference category), single, divorced and widowed.

Adult socio-economic position

Participants' civil service grade (or last recorded grade if the participant was no longer in the civil service at phase 5 and/or 9) defined their SEP as high (referent; administrative), intermediate (professional/executive) and low (clerical/support roles).

Statistical analysis

Multinomial logistic regression models were used to examine the associations between each category of ACE, as well as number of ACEs, and weekly alcohol consumption at phases 5 (midlife) and 9 (early old-age). Relative risk ratios (RRR) and 95% confidence interval (CI) for hazardous drinking and past week non-drinking in each category of ACE were estimated. For longitudinal analyses, mixed-effect multilevel modelling was used to investigate the effect of ACEs on drinking trajectories using data from phases 5, 7 and 9 (10-year interval; coefficients from these models reflect changes in alcohol consumption per unit increase in phase; ∼5 years). As the group of past week non-drinkers may bias the findings, a sensitivity analysis restricting these models to those who reported drinking at phase 5 was performed. To determine whether sex was a moderator in the relationship between ACEs and alcohol consumption, models were initially fitted with interactions between individual ACEs and sex. No significant interactions were observed and therefore analyses in this study were pooled with adjustment for sex. For each ACE, we present two models, one adjusted only for age and sex, and the other adjusted for all covariates listed above. In the multilevel models, we included an interaction between time and age to control for age-related changes in alcohol consumption (Britton et al., 2015). Age was centred on mean age at phase 5. Complete-case methods were used to analyse data. All statistical analyses were carried out using Stata 12.1 (StataCorp, Texas).

RESULTS

Participants' demographic characteristics

Participants' demographic characteristics at phases 5 and 9, and weekly alcohol consumption at phases 5, 7 and 9 are presented in Table 1. Nearly one-third of participants in this study were women. The mean age of participants at phases 5 and 9 was 56 (SD: 6.04; range from 45 to 69) and 66 (SD: 5.98; range from 55 to 80) years, respectively. Over 90% of participants were white; over 75% were married/cohabitated. A large proportion of participants were in high or intermediate SEP groups in adulthood. Mean weekly alcohol consumption declined from phase 5 to 9.

Table 1.

Demographic characteristics and weekly alcohol consumption of participants

Covariates Phase 5
 Phase 9
Men
 Women
 Total
 Men
 Women
 Total
Mean age (years, SD) 5473 55.69 (5.99) 2397 56.55 (6.11) 7870 55.95 (6.04) 4759 65.82 (5.90) 2002 66.43 (6.13) 6761 66.00 (5.98)
Ethnicity, N (%)
 White 5473 5116 (93.48) 2397 2070 (86.36) 7870 7186 (91.31) 4759 4472 (93.97) 2002 1746 (87.21) 6761 6218 (91.97)
 Non-white 357 (6.52) 327 (13.64) 684 (8.69) 287 (6.03) 256 (12.79) 543 (8.03)
Marital status, N (%)
 Married/cohabitated 4892 4156 (84.96) 2029 1269 (62.54) 6921 5425 (78.38) 4702 3917 (83.30) 1944 1082 (55.66) 6646 4999 (75.22)
 Single 438 (8.95) 380 (18.73) 818 (11.82) 386 (8.21) 365 (18.78) 751 (11.30)
 Divorce 231 (4.72) 234 (11.53) 465 (6.72) 240 (5.10) 235 (12.09) 475 (7.15)
 Widowed 67 (1.37) 146 (7.20) 213 (3.08) 159 (3.38) 262 (13.48) 421 (6.33)
Adult SEP, N (%)
 High (administrative) 5415 2776 (51.27) 2358 447 (18.96) 7773 3223 (41.46) 3102 1739 (56.06) 1258 318 (25.28) 4360 2057 (47.18)
 Intermediate (professional/executive) 2325 (42.94) 1071 (45.42) 3396 (43.69) 1210 (39.10) 594 (47.22) 1804 (41.38)
 Low (clerical/support) 314 (5.80) 840 (35.62) 1154 (14.85) 153 (4.93) 346 (27.50) 499 (11.44)
Smoking status, N (%)
 Current-smoker 5057 484 (9.57) 2141 283 (13.22) 7198 767 (10.66) 4480 273 (6.09) 1859 113 (6.08) 6339 386 (6.09)
 Ex-smoker 2204 (43.58) 700 (32.70) 2904 (40.34) 2235 (49.89) 750 (40.34) 2985 (47.09)
 Never-smoker 2369 (46.85) 1158 (54.09) 3527 (49.00) 1972 (44.02) 996 (53.58) 2968 (46.82)
Weekly alcohol consumption Phase 5 Phase 7 Phase 9
Men (N = 4979) Women (N = 2098) Total (N = 7077) Men (N = 4759) Women (N = 1981) Total (N = 6740) Men (N = 4612) Women (N = 1902) Total (N = 6514)
Mean alcohol units (SD) 16.16 (16.18) 7.14 (9.25) 13.49 (15.05) 14.25 (13.80) 6.04 (7.33) 11.84 (12.82) 12.23 (12.20) 5.40 (7.14) 10.24 (11.40)
Moderate drinkers, N (%)
(1–20 units for men, 1–13 units for women)		 2962 (59.49) 1140 (54.34) 4102 (57.96) 3044 (63.96) 1050 (53.00) 4094 (60.74) 3078 (66.74) 984 (51.74) 4062 (62.36)
Hazardous drinkers, N (%)
(21+ units for men, 14+ units for women)		 1443 (28.98) 371 (17.68) 1814 (25.63) 1148 (24.12) 327 (16.51) 1475 (21.88) 897 (19.45) 264 (13.88) 1161 (17.82)
Past week non-drinker (0 units) N (%) 574 (11.53) 587 (27.98) 1161 (16.41) 567 (11.91) 604 (30.49) 1171 (17.37) 637 (13.81) 654 (34.38) 1291 (19.82)
Open in a new tab

Phase 5: middle adulthood; Phase 7: transition period from middle adulthood to early old-age; Phase 9: early old-age.

Adverse childhood experiences

Women were more likely to report experiencing exposure to arguments/fights, parental divorce, parental mental health/alcohol problems and physical abuse (Table 2). The most common ACE across all participants was exposure to parental arguments/fights while the least common was living in an orphanage.

Table 2.

Prevalence of adverse childhood experiences

Category of ACEs Prevalence
Men
N (%)		 Women
N (%)		 Total
N (%)		 n
Exposure to parental arguments/fights 886 (18.36) 495 (24.37) 1381 (20.14) 6858
Parental divorce 189 (3.91) 108 (5.33) 297 (4.33) 6861
Parental mental health/alcohol problems 275 (5.70) 145 (7.15) 420 (6.13) 6855
Physical abuse 91 (1.89) 86 (4.24) 177 (2.58) 6853
Hospitalization >4 weeks 640 (13.24) 272 (13.35) 912 (13.28) 6870
Living in an orphanage 62 (1.29) 35 (1.73) 97 (1.42) 6840
ACE score
 0 3294 (68.88) 1222 (61.19) 4516 (66.62) 6779
 1 1089 (22.77) 559 (27.99) 1648 (24.31)
 2 300 (6.27) 166 (8.31) 466 (6.87)
 ≥3 99 (2.07) 50 (2.50) 149 (2.20)
Open in a new tab

ACEs: adverse childhood experiences.

Individual ACE and weekly alcohol consumption

Participants who experienced physical abuse in childhood were 1.43 (95% CI 1.01, 2.04) times more likely to be hazardous drinkers than moderate drinkers in midlife after adjusting for age and sex (Table 3). However, in the multivariable-adjusted model, the association was no longer statistically significant. Participants who were exposed to parental arguments/fights during childhood were 1.20 (95% CI 1.03, 1.38) times more likely to drink at a hazardous level in midlife than being moderate drinkers after adjusting for covariates. This association was robust in a sensitivity analysis adjusting for all other ACEs and covariates (data not shown).

Table 3.

Relationship between ACEs and weekly alcohol consumption after adjusting for covariates at phase 5

ACEs No covariates added
 Sex + age
 Multivariable-adjusted (sex + age + adult SEP + ethnicity + marital status + smoking status)
N RRR (95% CI) N RRR (95% CI) N RRR (95% CI)
Hazardous drinking
 Exposure to parental arguments/fights 6775 1.22* (1.07; 1.41) 6775 1.25* (1.09; 1.44) 6476 1.20*(1.03; 1.38)
 Parental divorce 6779 1.01 (0.77; 1.34) 6779 1.05 (0.79; 1.38) 6480 0.97 (0.72; 1.31)
 Parental mental health/ alcohol problems 6772 1.00 (0.79; 1.27) 6772 1.00 (0.79; 1.27) 6472 0.95 (0.75; 1.22)
 Physical abuse 6771 1.35 (0.95; 1.92) 6771 1.43* (1.01; 2.04) 6474 1.37 (0.94; 1.98)
 Hospitalization >4 weeks 6788 1.02 (0.86; 1.20) 6788 1.04 (0.88; 1.23) 6487 1.05 (0.88; 1.24)
 Living in an orphanage 6759 0.72 (0.43; 1.22) 6759 0.76 (0.45; 1.28) 6465 0.75 (0.43; 1.32)
Past week non-drinkers
 Exposure to parental arguments/fights 6775 1.22* (1.04; 1.44) 6775 1.13 (0.96; 1.33) 6476 1.21* (1.01; 1.45)
 Parental divorce 6779 1.11 (0.81; 1.53) 6779 1.03 (0.74; 1.42) 6480 0.90 (0.64; 1.28)
 Parental mental health/ alcohol problems 6772 0.98 (0.74; 1.30) 6772 0.93 (0.70; 1.24) 6472 1.04 (0.77; 1.42)
 Physical abuse 6771 1.60* (1.08; 2.35) 6771 1.34 (0.90; 1.99) 6474 1.32 (0.87; 2.01)
 Hospitalization >4 weeks 6788 1.07 (0.89; 1.30) 6788 1.05 (0.86; 1.28) 6487 1.04 (0.84; 1.28)
 Living in an orphanage 6759 1.02 (0.59; 1.76) 6759 0.95 (0.54; 1.65) 6465 0.86 (0.47; 1.57)
Open in a new tab

Base outcome: moderate drinking.

Reference group for adult socio-economic position (SEP): high.

Reference group for smoking status: never-smoker.

RRR: relative risk ratios; CI: confidence interval; ACEs: adverse childhood experiences; Phase 5: middle adulthood.

*P < 0.05

Similarly, participants who experienced physical abuse in childhood were 1.60 (95% CI 1.08, 2.35) times more likely to report no consumption in past week in midlife, compared with moderate drinking. However, adjustment for covariates attenuated the effect to non-significant (RRR = 1.32; 95% CI 0.87, 2.01). Exposure to parental arguments/fights was associated with past week non-drinking in midlife (RRR = 1.22; 95% CI 1.04, 1.44). In the multivariable-adjusted model, participants who were exposed to parental arguments/fights during childhood were 1.21 (95% CI 1.01, 1.45) times more likely to not drink in the past week in midlife, compared with moderate drinking.

Associations between ACEs and hazardous drinking/past week non-drinkers at phase 9 were also examined (data not shown). Participants who lived in an orphanage during childhood had a lower risk of hazardous drinking in early old-age (RRR = 0.28; 95% CI 0.086, 0.93), after adjusting for covariates. However, in the multivariable-adjusted model, hazardous drinking in early old-age was not significantly associated with exposure to parental arguments/fights, parental divorce, parental mental health/alcohol problems, physical abuse and hospitalization >4 weeks. A significant association was observed between physical abuse in childhood and past week non-drinkers in early old-age. Participants who were exposed to physical abuse in childhood were more likely to be past week non-drinkers in early old-age, compared with moderate drinking (RRR = 1.73; 95% CI 1.18, 2.54). However, adjustment for covariates attenuated the effect (RRR = 0.95; 95% CI 0.53, 1.70).

Cumulative effect of ACEs on alcohol consumption

For each one category increase in the number of ACEs a participant was exposed to during childhood, the risk of hazardous drinking versus moderate drinking increased by 1.09 (95% CI 1.01, 1.18) (Table 4). The association remained significant after adjusting for sex and age (RRR = 1.10; 95% CI 1.02, 1.19). After further adjustment for adult SEP, ethnicity, marital status and smoking status, the association was no longer statistically significant (RRR = 1.08; 95% CI 0.99, 1.17).

Table 4.

Association between number of ACEs and weekly alcohol consumption at phase 5

Number of ACEs experienced
(0 as the reference group)		 No covariates added
(N:6699)		 Adjusted for Sex + Age
(N:6699)		 Multivariable-adjusted (sex + age + adult SEP + ethnicity + marital status + smoking status)
(N:6407)
RRR (95% CI) RRR (95% CI) RRR (95% CI)
Hazardous drinking
 1 1.18* (1.03; 1.35) 1.22* (1.07; 1.40) 1.17* (1.01–1.34)
 2 1.10 (0.87; 1.38) 1.12 (0.89; 1.41) 1.11 (0.88–1.41)
 ≥3 1.21 (0.83; 1.78) 1.24 (0.84; 1.82) 1.13 (0.76–1.67)
 Continuous ACE score 1.09* (1.01; 1.18) 1.10* (1.02; 1.19) 1.08 (0.99–1.17)
Past week non-drinkers
 1 1.15 (0.98; 1.35) 1.06 (0.90; 1.24) 1.11 (0.93–1.31)
 2 1.22 (0.94; 1.59) 1.12 (0.86; 1.45) 1.15 (0.87–1.53)
 ≥3 1.27 (0.82; 1.99) 1.19 (0.76; 1.88) 1.21 (0.74–1.97)
 Continuous ACE score 1.11* (1.01; 1.22) 1.06 (0.96; 1.16) 1.08 (0.98–1.19)
Open in a new tab

Base outcome: moderate drinking.

Reference group for adult socio-economic position (SEP): high.

Reference group for smoking status: never-smoker.

RRR: relative risk ratios; CI: confidence interval; ACEs: adverse childhood experiences; Phase 5: middle adulthood.

*P < 0.05.

At phase 9, for each additional exposure to an ACE, the risk of hazardous drinking versus moderate drinking increased by 1.07 (95% CI 0.97, 1.18) in early old-age. However, this association was not statistically significant (data not shown).

Longitudinal effect of ACEs on drinking trajectories from phase 5 to 9

Both marital status and the interaction between sex and ACEs were not included in the final mixed-effect multilevel regression models as there was no evidence of any improvement in the model fit upon their inclusion according to likelihood ratio tests (Table 5). Alcohol consumption tended to decrease with increasing age. There was significant variation in both the slope and intercept of participants' drinking trajectories, indicating heterogeneity between participants' 10-year drinking trajectories. Participants who reported parental arguments/fights had a slower rate of decline in alcohol consumption than those with no reports (β = 0.039; 95% CI −0.34, 0.42) over 10-year interval, but this was not statistically significant (P = 0.84). Those who experienced hospitalization (β = −0.12; 95% CI −0.57, 0.33) or living in an orphanage (β = −0.15; 95% CI −1.45, 1.15) made greater decline in alcohol consumption than those with no ACEs over 10-year interval when all covariates were held constant. As participants who reported hospitalization or living in an orphanage during their childhood made greater reductions on average in alcohol consumption over 10-year interval, they may mask the cumulative effects of number of ACEs experienced which were associated with slower decline in alcohol consumption in mixed-effect multilevel regression modelling. When examining the cumulative effect of ACEs, these two ACEs were excluded in calculating the total number of ACE score. Those who experienced an increasing number of ACEs had a slower rate of decline in alcohol consumption than those who did not experience any ACEs (β = 0.10; 95% CI −0.14, 0.34) over the 10-year interval; however, this effect was not statistically significant (P = 0.42).

Table 5.

Fixed effects from multilevel analyses of the role of ACEs on drinking trajectories from phase 5 to 9

ACEs Unconditional model
 Multivariable-adjusted modela
N Slope Coefficient for interaction with slopeb N Slope Coefficient for interaction with slopeb
Exposure to parental arguments/fights 5744 −2.15 (−2.32; −1.98)** 0.038 (−0.34; 0.42) 5686 −2.19 (−2.36; −2.02)** 0.039 (−0.34; 0.42)
Parental divorce 5745 −2.18 (−2.33; −2.03)** 0.44 (−0.31; 1.20) 5686 −2.22 (−2.37; −2.06)** 0.45 (−0.31; 1.21)
Parental mental health/alcohol problem 5740 −2.18 (−2.33; −2.02)** 0.36 (−0.27; 0.99) 5680 −2.21 (−2.37; −2.05)** 0.31 (−0.32; 0.95)
Physical abuse 5740 −2.15 (−2.30; −2.00)** 0.084 (−0.90; 1.07) 5680 −2.19 (−2.34; −2.03)** 0.16 (−0.83; 1.15)
Hospitalization >4 weeks 5749 −2.12 (−2.28; −1.96)** −0.15 (−0.60; 0.30) 5689 −2.16 (−2.32; −2.00)** −0.12 (−0.57; 0.33)
Living in an orphanage 5726 −2.14 (−2.29; −1.99)** −0.33 (−1.63; 0.96) 5671 −2.1 8 (−2.33; −2.02)** −0.15 (−1.45; 1.15)
Number of ACEs experienced 5698 −2.18 (−2.35; −2.01)** 0.10 (−0.14; 0.34) 5641 −2.22 (−2.39; −2.05)** 0.10 (−0.14; 0.34)
Open in a new tab

Reference group for adult socio-economic position (SEP): high.

Reference group for smoking status: never-smoker.

ACEs: adverse childhood experiences; Phase 5: middle adulthood; Phase 9: early old-age.

aAdjusted for sex, age, adult SEP, ethnicity and smoking status.

bChange in alcohol consumption per increase in phase (∼5 years).

**P ≤ 0.001.

DISCUSSION

Among the six categories of ACEs examined in this study, only exposure to parental arguments/fights was associated with hazardous drinking in midlife, after controlling for a range of covariates. Exposure to parental arguments may act as a stressor to children, and children are more likely to have negative emotional reactions such as anger and aggression (Jenkins, 2000). At the same time, prolonged inter-parental conflict can influence parenting practices (Krishnakumar and Buehler, 2000). Parents who are in marital strife may spend less time monitoring or supervising their children (Cox et al., 2001). Parental factors such as parental monitoring and parental involvement are associated with delayed alcohol initiation and lower levels of alcohol use by offspring (Ryan et al., 2010).

Exposure to parental arguments/fights was also associated with both hazardous drinking and past week non-drinking in midlife. This finding might provide insights into the role of childhood adversity to alcohol abstinence as well as poorer health profiles among non-drinkers. Several studies have shown that abstainers have worse health profiles than moderate drinkers (Ng Fat and Shelton, 2012; Fekjær, 2013). Moderate consumption of alcohol is alleged to confer protective effects for multiple cardiovascular outcomes and all-cause mortality (Ronksley et al., 2011; Roerecke and Rehm, 2012; Movva and Figueredo, 2013). However, the underlying mechanism between abstinence and worse health profiles is still poorly understood (Fekjær, 2013). Non-drinkers may have poor health outcomes due to factors other than not drinking. Future research should examine the interrelationship between adverse childhood experiences, abstinence and poor health outcomes. In this study, the definition of non-drinkers was broad and based on the previous week only. It includes lifetime abstainers, ex-drinkers and occasional drinkers who did not drink in the week prior to participation. A more robust categorization of abstaining would be required in future studies.

A significant association between the continuous ACE score and risk of hazardous drinking in midlife was observed in this study. Exposure to multiple ACEs was associated with an increased risk of hazardous drinking in midlife. This result supports previous studies (Dube et al., 2002; Kauhanen et al., 2011; Bellis et al., 2014), but to our knowledge, this is the first study to examine the role of ACEs in drinking trajectories across midlife longitudinally. The decline in alcohol consumption with increasing age is consistent with other longitudinal studies (Moore et al., 2005; Karlamangla et al., 2006). Although ACEs play a role in heavy drinking trajectories from adolescence to young adulthood (Shin et al., 2013), this study found that ACEs did not impact on the rate of change in alcohol consumption from during a ∼10-year period from midlife to early old-age. The pathway from ACEs to adoption of drinking may be interrupted by resilience. Interaction between genetic factors, life experiences, personal characteristics and social environment over the time may explain the resilience from ACEs (Liem et al., 1997; Collishaw et al., 2007). Adult experiences provide important turning points for individuals to counter ACEs (Rutter, 2006). Through life experiences, maturation may reduce the impacts of ACEs among adults and develop more adaptive strategies to cope with stress (Draper et al., 2008).

Some potential limitations of this study should be considered when interpreting the results. First, the retrospective self-reported nature of ACEs raises the potential for recall bias. Participants may not be able to remember all the details in childhood fully or not willing to divulge the childhood adversities, resulting in those who truly had ACEs being misclassified as belonging to the unexposed group. Additionally, only six categories of ACE were offered in the questionnaire. Other important adverse experiences were not included, for example, being bullied by peers in childhood (Swahn et al., 2011; Topper et al., 2011). Some categories are very subjective, for example, exposure to parental arguments/fights, and some experiences are more adverse than others, for example, individuals who reported parental separation or divorce were likely to have suffered from physical abuse in childhood (Dong et al., 2004). This study did not capture the frequency or severity of ACEs which participants had experienced. Second, alcohol consumption could be either over- or under-reported by participants (Bellis et al., 2015). As both exposure (ACEs) and outcome (weekly alcohol consumption) were possibly underestimated, this may bias our findings towards the null hypothesis. Therefore, the strength of the relationships between ACEs and risk of hazardous drinking found in this study is likely to be conservative. Third, the use of data from the Whitehall II study may limit the generalizability of the findings from this study. It is not fully representative of the general population, primarily consisting of white-collar men of predominantly high-to-intermediate socio-economic position. The proportion of participants reporting two or more ACEs (9.1%) in the Whitehall II study was lower than that of population studies (23.7%) (Bellis et al., 2014). However, aetiological findings from the cohort have been shown to be consistent with those obtained from general population samples suggesting this bias may not be as substantial as previously thought (Batty et al., 2014). Bias may result from differential loss to follow-up in longitudinal studies (Hernán et al., 2004). Illness and death could lead participants to drop-out not only before phase 5 in the Whitehall II study (the baseline of this study) but even if participants took part at phase 5, drop-out might have occurred in subsequent phases. ACEs and hazardous drinking are associated with poor health outcomes and death (Room et al., 2005; Shonkoff et al., 2012). Participants with ACEs and/or hazardous drinking may be more likely to drop out, resulting in systematic differences between who remain and those who drop out. Loss to follow-up may increase the potential for selection bias in the remaining sample, and mask the effects of ACEs on hazardous drinking in early old-age and longitudinally. Furthermore, some covariates that were included as confounders in our models could conceivably be argued to be mediators of the association between ACEs and alcohol consumption, for example, experiencing ACEs could affect the socio-economic group or marital status a participant belongs to in adulthood, which in turn might affect drinking—so therefore controlling for these factors may represent over-adjustment (Schisterman et al., 2009). However, the age-/sex-adjusted estimates are similar to the multivariable-adjusted estimates, suggesting that these variables are unlikely to explain a substantial proportion of the association between ACEs and alcohol intake in midlife/old-age.

This study is one of few to examine the relationship between ACEs and hazardous drinking in midlife, and the first to explore the impacts of ACEs on change in alcohol consumption from midlife to early old-age. Target prevention and intervention programmes can help to reduce the occurrence of ACEs which may have favourable knock-on advantages in lowering the risk of hazardous drinking as well as negative health outcomes.

CONCLUSION

The impact of exposure to parental arguments on hazardous drinking persists into midlife, but not early old-age. A significant association between continuous ACE score and risk of hazardous drinking in midlife was also observed. Our findings highlight that ongoing efforts to prevent ACEs may also help to reduce hazardous drinking in midlife.

FUNDING

The Whitehall II study is supported by grants from the Medical Research Council [K013351], British Heart Foundation [RG/07/008/23674], Stroke Association, National Heart Lung and Blood Institute [HL036310] and National Institute on Aging [AG13196 and AG034454]. A.B. and S.B. are currently supported by the European Research Council (ERC-StG-2012-309337_AlcoholLifecourse, PI: Britton, http://www.ucl.ac.uk/alcohol-lifecourse) and UK Medical Research Council/Alcohol Research UK (MR/M006638/1). The remaining author has no financial relationships relevant to this article to disclose.

CONFLICT OF INTEREST STATEMENT

All authors have no conflicts of interest relevant to this article to disclose.

ACKNOWLEDGEMENTS

We thank all participants who have willingly shared their precious time to take part in the Whitehall II study, as well as all study group members.

REFERENCES

  1. Anda RF, Croft JB, Felitti VJ et al. (1999) Adverse childhood experiences and smoking during adolescence and adulthood. JAMA 282:1652–8. [DOI] [PubMed] [Google Scholar]
  2. Bair-Merritt MH, Blackstone M, Feudtner C (2006) Physical health outcomes of childhood exposure to intimate partner violence: a systematic review. Pediatrics 117:e278–90. [DOI] [PubMed] [Google Scholar]
  3. Bartley M. (2012) Life Gets Under Your Skin. London, UK: UCL Research Department of Epidemiology and Public Health. [Google Scholar]
  4. Batty GD, Shipley M, Tabák A et al. (2014) Generalizability of occupational cohort study findings. Epidemiology 25:932–3. [DOI] [PubMed] [Google Scholar]
  5. Bell S, Britton A, Kubinova R et al. (2014) Drinking pattern, abstention and problem drinking as risk factors for depressive symptoms: evidence from three urban Eastern European populations. PLoS One 9:e104384. [DOI] [PMC free article] [PubMed] [Google Scholar]
  6. Bellis MA, Hughes K, Leckenby N et al. (2014) National household survey of adverse childhood experiences and their relationship with resilience to health-harming behaviors in England. BMC Med 12:72. [DOI] [PMC free article] [PubMed] [Google Scholar]
  7. Bellis MA, Hughes K, Jones L et al. (2015) Holidays, celebrations, and commiserations: measuring drinking during feasting and fasting to improve national and individual estimates of alcohol consumption. BMC Med 13:113. [DOI] [PMC free article] [PubMed] [Google Scholar]
  8. Britton A, Ben-Shlomo Y, Benzeval M et al. (2015) Life course trajectories of alcohol consumption in the United Kingdom using longitudinal data from nine cohort studies. BMC Med 13:47. [DOI] [PMC free article] [PubMed] [Google Scholar]
  9. Camacho TC, Kaplan GA, Cohen RD (1987) Alcohol consumption and mortality in Alameda County. J Chronic Dis 40:229–36. [DOI] [PubMed] [Google Scholar]
  10. Collishaw S, Pickles A, Messer J et al. (2007) Resilience to adult psychopathology following childhood maltreatment: evidence from a community sample. Child Abuse Negl 31:211–29. [DOI] [PubMed] [Google Scholar]
  11. Cox MJ, Paley B, Harter K (2001) Interparental conflict and parent-child relationships. In Grych JH, Fincham FD (eds). Interparental Conflict and Child Development: Theory, Research and Applications. New York, USA: Cambridge University Press, 249–72. [Google Scholar]
  12. Department of Health. (1995) Sensible Drinking: Report of an Inter-departmental Working Group. London, UK: Department of Health. [Google Scholar]
  13. Dong M, Anda RF, Felitti VJ et al. (2004) The interrelatedness of multiple forms of childhood abuse, neglect, and household dysfunction. Child Abuse Negl 28:771–84. [DOI] [PubMed] [Google Scholar]
  14. Draper B, Pfaff JJ, Pirkis J et al. (2008) Long-term effects of childhood abuse on the quality of life and health of older people: results from the Depression and Early Prevention of Suicide in General Practice Project. J Am Geriatr Soc 56:262–71. [DOI] [PubMed] [Google Scholar]
  15. Dube SR, Anda RF, Felitti VJ et al. (2002) Adverse childhood experiences and personal alcohol abuse as an adult. Addict Behav 27:713–25. [DOI] [PubMed] [Google Scholar]
  16. Dube SR, Miller JW, Brown DW et al. (2006) Adverse childhood experiences and the association with ever using alcohol and initiating alcohol use during adolescence. J Adolesc Health 38:444.e1–10. [DOI] [PubMed] [Google Scholar]
  17. Dubow EF, Boxer P, Huesmann LR (2008) Childhood and adolescent predictors of early and middle adulthood alcohol use and problem drinking: the Columbia County Longitudinal Study. Addiction 103:36–47. [DOI] [PubMed] [Google Scholar]
  18. Fekjær HO. (2013) Alcohol—a universal preventive agent? A critical analysis. Addiction 108:2051–7. [DOI] [PubMed] [Google Scholar]
  19. Gilbert R, Widom CS, Browne K et al. (2009) Burden and consequences of child maltreatment in high-income countries. Lancet 373:68–81. [DOI] [PubMed] [Google Scholar]
  20. Griffin ML, Amodeo M (2010) Predicting long-term outcomes for women physically abused in childhood: contribution of abuse severity versus family environment. Child Abuse Negl 34:724–33. [DOI] [PubMed] [Google Scholar]
  21. Hernán MA, Hernández-Díaz S, Robins JM (2004) A structural approach to selection bias. Epidemiology 15:615–25. [DOI] [PubMed] [Google Scholar]
  22. Hope S, Power C, Rodgers B (1998) The relationship between parental separation in childhood and problem drinking in adulthood. Addiction 93:505–14. [DOI] [PubMed] [Google Scholar]
  23. Jenkins JM. (2000) Marital conflict and children's emotions: the development of an anger organization. J Marriage Fam 62:723–36. [Google Scholar]
  24. Karlamangla A, Zhou K, Reuben D et al. (2006) Longitudinal trajectories of heavy drinking in adults in the United States of America. Addiction 101:91–9. [DOI] [PubMed] [Google Scholar]
  25. Kauhanen L, Leino J, Lakka HM et al. (2011) Adverse childhood experiences and risk of binge drinking and drunkenness in middle-aged Finnish men. Adv Prev Med 2011:478741. [DOI] [PMC free article] [PubMed] [Google Scholar]
  26. Kestilä L, Martelin T, Rahkonen O et al. (2008) Childhood and current determinants of heavy drinking in early adulthood. Alcohol Alcohol 43:460–9. [DOI] [PubMed] [Google Scholar]
  27. Krishnakumar A, Buehler C (2000) Interparental conflict and parenting behaviors: a meta-analytic review. Fam Relat 49:25–44. [Google Scholar]
  28. Kristjansson AL, Sigfusdottir ID, Allegrante JP et al. (2009) Parental divorce and adolescent cigarette smoking and alcohol use: assessing the importance of family conflict. Acta Paediatr 98:537–42. [DOI] [PubMed] [Google Scholar]
  29. Levinson DJ. (1986) A conception of adult development. Am Psychol 41:3–13. [Google Scholar]
  30. Liem JH, James JB, O'Toole JG et al. (1997) Assessing resilience in adults with histories of childhood sexual abuse. Am J Orthopsychiatry 67:594–606. [DOI] [PubMed] [Google Scholar]
  31. Manning V, Best DW, Faulkner N et al. (2009) New estimates of the number of children living with substance misusing parents: results from UK national household surveys. BMC Public Health 9:377. [DOI] [PMC free article] [PubMed] [Google Scholar]
  32. Marmot M, Brunner E (2005) Cohort profile: the Whitehall II study. Int J Epidemiol 34:251–6. [DOI] [PubMed] [Google Scholar]
  33. Marshal MP. (2003) For better or for worse? The effects of alcohol use on marital functioning. Clin Psychol Rev 23:959–97. [DOI] [PMC free article] [PubMed] [Google Scholar]
  34. Mirsal H, Kalyoncu A, Pektaş Ö et al. (2004) Childhood trauma in alcoholics. Alcohol Alcohol 39:126–9. [DOI] [PubMed] [Google Scholar]
  35. Moore AA, Gould R, Reuben DB et al. (2005) Longitudinal patterns and predictors of alcohol consumption in the United States. Am J Public Health 95:458–64. [DOI] [PMC free article] [PubMed] [Google Scholar]
  36. Movva R, Figueredo VM (2013) Alcohol and the heart: to abstain or not to abstain? Int J Cardiol 164:267–76. [DOI] [PubMed] [Google Scholar]
  37. Naimi TS, Brown DW, Brewer RD et al. (2005) Cardiovascular risk factors and confounders among nondrinking and moderate-drinking U.S. adults. Am J Prev Med 28:369–73. [DOI] [PubMed] [Google Scholar]
  38. Ng Fat L, Shelton N (2012) Associations between self-reported illness and non-drinking in young adults. Addiction 107:1612–20. [DOI] [PubMed] [Google Scholar]
  39. Office for National Statistics. (2014) Divorces in England and Wales, 2012. London, UK: Office for National Statistics. [Google Scholar]
  40. Radford L, Corral S, Bradley C et al. (2011) Child Abuse and Neglect in the UK Today. London, UK: National Society for the Prevention of Cruelty to Children. [Google Scholar]
  41. Roerecke M, Rehm J (2012) The cardioprotective association of average alcohol consumption and ischaemic heart disease: a systematic review and meta-analysis. Addiction 107:1246–60. [DOI] [PMC free article] [PubMed] [Google Scholar]
  42. Ronksley PE, Brien SE, Turner BJ et al. (2011) Association of alcohol consumption with selected cardiovascular disease outcomes: a systematic review and meta-analysis. Br Med J 342:d671. [DOI] [PMC free article] [PubMed] [Google Scholar]
  43. Room R, Babor T, Rehm J (2005) Alcohol and public health. Lancet 365:519–30. [DOI] [PubMed] [Google Scholar]
  44. Rothman EF, Edwards EM, Heeren T et al. (2008) Adverse childhood experiences predict earlier age of drinking onset: results from a representative US sample of current or former drinkers. Pediatrics 122:e298–304. [DOI] [PubMed] [Google Scholar]
  45. Rutter M. (2006) The promotion of resilience in the face of adversity. In Clarke-Stewart A, Dunn J (eds). Families Count: Effects on Child and Adolescent Development. New York, NY: Cambridge University Press, 26–52. [Google Scholar]
  46. Ryan SM, Jorm AF, Lubman DI (2010) Parenting factors associated with reduced adolescent alcohol use: a systematic review of longitudinal studies. Aust N Z J Psychiatry 44:774–83. [DOI] [PubMed] [Google Scholar]
  47. Schisterman EF, Cole SR, Platt RW (2009) Overadjustment bias and unnecessary adjustment in epidemiologic studies. Epidemiology 20:488–95. [DOI] [PMC free article] [PubMed] [Google Scholar]
  48. Shin SH, Edwards EM, Heeren T (2009) Child abuse and neglect: relations to adolescent binge drinking in the national longitudinal study of Adolescent Health (AddHealth) Study. Addict Behav 34:277–80. [DOI] [PMC free article] [PubMed] [Google Scholar]
  49. Shin SH, Miller DP, Teicher MH (2013) Exposure to childhood neglect and physical abuse and developmental trajectories of heavy episodic drinking from early adolescence into young adulthood. Drug Alcohol Depend 127:31–8. [DOI] [PMC free article] [PubMed] [Google Scholar]
  50. Shonkoff JP, Garner AS, Committee on Psychosocial Aspects of Child and Family Health et al. (2012) The lifelong effects of early childhood adversity and toxic stress. Pediatrics 129:e232–46. [DOI] [PubMed] [Google Scholar]
  51. Singer JD, Willett JB (2003) Applied Longitudinal Data Analysis: Modeling Change and Event Occurrence. New York, USA: Oxford University Press. [Google Scholar]
  52. Spak L, Spak F, Allebeck P (1997) Factors in childhood and youth predicting alcohol dependence and abuse in Swedish women: findings from a general population study. Alcohol Alcohol 32:267–74. [DOI] [PubMed] [Google Scholar]
  53. Swahn MH, Topalli V, Ali B et al. (2011) Pre-teen alcohol use as a risk factor for victimization and perpetration of bullying among middle and high school students in Georgia. West J Emerg Med 12:305–9. [PMC free article] [PubMed] [Google Scholar]
  54. Thompson RG, Lizardi D, Keyes KM et al. (2008) Childhood or adolescent parental divorce/separation, parental history of alcohol problems, and offspring lifetime alcohol dependence. Drug Alcohol Depend 98:264–9. [DOI] [PMC free article] [PubMed] [Google Scholar]
  55. Timko C, Sutkowi A, Pavao J et al. (2008) Women's childhood and adult adverse experiences, mental health, and binge drinking: the California Women's Health Survey. Subst Abuse Treat Prev Policy 3:15. [DOI] [PMC free article] [PubMed] [Google Scholar]
  56. Topper LR, Castellanos-Ryan N, Mackie C et al. (2011) Adolescent bullying victimisation and alcohol-related problem behaviour mediated by coping drinking motives over a 12 month period. Addict Behav 36:6–13. [DOI] [PubMed] [Google Scholar]
  57. WHO. (2006) Preventing Child Maltreatment: A Guide to Taking Action and Generating Evidence. Geneva, Switzerland: World Health Organization. [Google Scholar]
  58. WHO. (2010) Global Strategy to Reduce the Harmful Use of Alcohol. Geneva, Switzerland: World Health Organization. [Google Scholar]
  59. WHO. (2015) Adverse Childhood Experiences International Questionnaire (ACE-IQ) as at 2015. http://www.who.int/violence_injury_prevention/violence/activities/adverse_childhood_experiences/en/ (29 September 2015, date last accessed).
  60. Widom CS, Hiller-Sturmhöfel S (2001) Alcohol abuse as a risk factor for and consequence of child abuse. Alcohol Res Health 25:52–7. [PMC free article] [PubMed] [Google Scholar]

Articles from Alcohol and Alcoholism (Oxford, Oxfordshire) are provided here courtesy of Oxford University Press

RESOURCES