Adverse childhood experiences and multimorbidity of internalising and cardiometabolic conditions in mid to older age
1Cardiff University
2University of Leeds
3University of Exeter
4-
*Corresponding author; Email: lkbenger98@gmail.comAbstract
Background
Multimorbidity of internalising and cardiometabolic conditions (ICM-MM) is the most common combination of mental and physical health conditions in older age. Few studies have examined the likelihood that individuals with adverse childhood experiences (ACEs) such as abuse or neglect will develop ICM-MM in mid to late adulthood, or gender disparities.
Methods
UK Biobank participants (n = 157,184, mean age 55.94, SD= 7.74; 68186 males and 88998 females) reported on ACEs as well as sociodemographic and lifestyle factors. Diagnoses of internalising conditions (depression and anxiety) and cardiometabolic conditions (hypertension, obesity, type 2 diabetes, dyslipidaemia and chronic kidney disease) were obtained through linked electronic healthcare records. Logistic regression models tested associations between ACEs and internalising conditions, cardiometabolic conditions and ICM-MM, accounting for gender differences and sociodemographic and lifestyle factors.
Results
ACEs were associated with all individual and multimorbid presentations. Stronger associations were found with internalising conditions (OR 1.84) and ICM-MM (OR range 1.73-2.15) than cardiometabolic conditions (OR range 1.08-1.44). Females were more likely to report most ACEs, but health risks following ACEs were similar for both genders. The associations remained when accounting for sociodemographic and lifestyle factors, including gender, age, socioeconomic status, ethnicity, diet, alcohol intake, smoking status and physical activity levels.
Conclusions
This is the first study to report associations between ACEs and the most common type of physical and mental health multimorbidity in mid to late adulthood. The results highlight the importance of early ACE intervention and trauma-informed healthcare.
Article notes
Competing Interest Statement
Michael J. Owen reports grants from Akrivia Health and Takeda Pharmaceuticals outside the submitted work.
Funding Statement
This work was funded by a PhD studentship to LKB from Health and Care Research Wales (MvdB; PH, MJO, FR, MMW; HS 22 04) and the Tackling Multimorbidity at Scale Strategic Priorities Fund programme (MvdB, PH, MJO, MMW, RP; MR/W014416/1) delivered by the Medical Research Council and the National Institute for Health Research in partnership with the Economic and Social Research Council and in collaboration with the Engineering and Physical Sciences Research Council.
Introduction
Multimorbidity refers to individuals experiencing two or more chronic conditions at the same time and has been identified as a major and increasing challenge to health care systems (Academy of Medical Sciences, 2018). It is particularly prevalent in the older population, being present in more than half of those aged over 60 years (Carlson & Yarns, 2023; Chowdhury, Das, Sunna, Beyene, & Hossain, 2023).
The most common physical and mental health multimorbidity cluster in older age is between internalising conditions (IC), such as anxiety or depression, and cardiometabolic conditions (CMC; preventable chronic diseases impacting the cardiovascular system and metabolic health). We will refer to this cluster of conditions as ICM-MM from here onwards. ICM-MM is associated with high levels of disability, increased complexity of disease management (S. W. Mercer, Gunn, Bower, Wyke, & Guthrie, 2012), high treatment burden for patients, poorer delivery of healthcare across fragmented health services (S. Mercer, Furler, Moffat, Fischbacher-Smith, & Sanci, 2016) and high healthcare costs (Soley-Bori et al., 2021; Stokes, Guthrie, Mercer, Rice, & Sutton, 2021).
CMCs such as type 2 diabetes (T2D), obesity, hypertension, chronic kidney disease (CKD) and dyslipidaemia are seen as precursors to cardiovascular disease and occur earlier in adulthood. Moreover, research indicates that risk of both IC and CMC start early in life, often before adulthood (Beesdo, Knappe, & Pine, 2009; Berenson et al., 1989; McGill Jr et al., 2000).
There is therefore a pressing need to understand the early life risk factors contributing to the development of ICM-MM in mid to late adulthood, as these insights may point to opportunities for prevention or early intervention.
Adverse childhood experiences (ACEs), such as abuse and neglect, have been associated with a wide range of mental and physical health conditions in adulthood (Felitti et al., 1998). Specifically, ACEs are associated with increased likelihood of ICs (Merrick et al., 2017) and CMCs (Suglia et al., 2018), including T2D (Hughes, Ford, & Bellis, 2020), obesity (Blissett, 2011), dyslipidaemia (Lin, Wang, Lu, Chen, & Guo, 2021), hypertension (Obi, McPherson, & Pollock, 2019) and CKD (Shi, Huang, & Jin, 2022). There are numerous studies on the association between ACEs and individual health outcomes but their impact on multimorbidity development in older age remains poorly understood, particularly for combinations of physical and mental health conditions. Exceptions are studies by Taylor and Demakakos (2024) and Hanlon et al. (2020), which have reported that experiencing a greater number of ACEs increases the risk of multimorbidity of mental and physical health conditions in older age, including ICs, and certain CMCs (i.e., diabetes, hypertension and CKD). However, these studies are based on self-reported conditions, which are liable to misreporting due to recall bias and limited health literacy (Sulieman et al., 2022) and need to be replicated using routinely recorded diagnoses from electronic health records (EHR). These studies also used counts of long-term conditions or disease categories, rather than combinations of specific mental and physical health conditions, to investigate multimorbidity. Identifying the precise conditions associated with ACEs would provide more meaningful information for both patients and clinicians, thereby facilitating the development of more targeted interventions.
There are well-documented gender differences in ACEs indicating that females are more likely to experience a higher frequency of ACEs and more complex ACE patterns than males (Haahr-Pedersen et al., 2020). Females are also more likely to experience most types of ACE, with the exception of physical abuse which has a similar or higher prevalence in males (Office for National Statistics, 2020a, 2020b). Furthermore, the risk of IC and CMC also tends to differ between the genders, with depression and anxiety more common in females (McLean, Asnaani, Litz, & Hofmann, 2011; Piccinelli & Wilkinson, 2000), whilst the prevalence of diagnosis of T2D, obesity, hypertension, dyslipidaemia and CKD differs between the genders in an age-dependent manner (Meloni et al., 2023; The World Health Organisation, 2023a, 2023b). It is therefore important that research into the associations between ACEs and ICM-MM examines gender differences.
Finally, a number of demographic and lifestyle factors, including lower socioeconomic status (SES), ethnic minority status, poor diet, smoking, high levels of alcohol use, and a sedentary lifestyle, are likely to influence the associations between ACEs, ICs and CMCs (Mersky, Choi, Lee, & Janczewski, 2021; Walsh, McCartney, Smith, & Armour, 2019). However, previous studies have not fully account for these factors (Hanlon et al., 2020; Taylor & Demakakos, 2024). Taking these factors into account will allow for better understanding of the relationships between ACES and individual condition and multimorbid presentations.
Our aims therefore were to examine:
- Whether individuals in mid to older age who have experienced a range of ACEs (e.g., emotional, physical and sexual abuse, and emotional and physical neglect) are more likely to develop IC, CMC or ICM-MM than those without such experiences;
- To what extent the associations between ACEs and IC, CMC or ICM-MM are impacted by demographic and lifestyle factors;
- Whether the associations between ACEs and IC, CMC or ICM-MM differ between females and males.
Methods
Sample
Data were collated from the UK Biobank cohort (Bycroft et al., 2018). The cohort contains information on around 500,000 people in the UK who were aged between 38-72 years when they were recruited between 2006 and 2010. Data were released under application number 79704 “Physical and mental health multimorbidity across the lifespan (LIfespaN multimorbidity research Collaborative: LINC)”. This work included participants with data on at least one ACE as well as diagnoses of IC and CMC from EHR data (n = 157,184).
Measures
Adverse Childhood Experiences
Following recruitment, 157,305 participants completed an online follow-up mental health questionnaire (MHQ), which queried ACEs, with responses from 157,184 participants used in analyses (see Supplementary Figure 1). The Childhood Trauma Screener (Grabe et al., 2012), was used to assess five ACE domains: emotional neglect, physical abuse, emotional abuse, sexual abuse and physical neglect. A binary variable was derived to indicate the presence of each ACE type following previous practice (Ho et al., 2020) (see Supplementary Table 1 for details and coding of ACEs). Responses of “Prefer not to say” were coded as missing data. Those who gave a response to at least one of the five questions were included in the analysis but anyone who answered “Prefer not to say” to all five questions was removed (n = 42) (see Supplementary Figure 1).
Demographic variables
Gender (female and male), age, SES and ethnicity, all collected at recruitment, were included in analyses. SES was measured using the Townsend Deprivation Index (TDI). A score of zero indicates the average material value of an area, positive values indicate high material deprivation and negative values indicate relative affluence. Ethnicity was categorised as white and non-white.
Lifestyle variables
Lifestyle variables (alcohol intake, diet, smoking status and physical activity) were taken from the touchscreen questionnaire completed at recruitment. Alcohol consumption was self-reported and coded as units per week (with >100 considered as missing data). A healthy diet score (ranging from ‘0’ least healthy to ‘7’ most healthy) was derived using data collected on consumption of 7 food groups (Hepsomali & Groeger, 2021; Wang et al., 2022). Smoker status was defined as never, previous or current. Physical activity was determined as total Metabolic Equivalent of Task (MET) in hours per week across all types of activity. Scores of over 168 hours per week were deemed implausible and coded as missing. See supplementary materials for more detail.
Statistical analysis
Statistical analyses were conducted using R (version 4.2.2).
We compared sociodemographic and lifestyle variables for individuals with and without ACEs (Table 1) for the full sample and separately for females and males. In the full sample, diet, alcohol intake, smoking status and physical activity levels were corrected for gender by including gender as a covariate in regression models. Associations with age and socioeconomic status were measured using t-tests. Associations with ethnicity were evaluated using a chi-squared test. Binomial logistic regression analysis regressing ACE on alcohol intake, smoking status and physical activity levels were used. The association between ACEs and diet was evaluated using ordinal logistic regression.
To address aim 1, binomial logistic regressions were conducted regressing 13 individual and multimorbid health presentations on each type of ACE (Table 2 and Supplementary Tables 4-8). Four binomial logistic regression models were tested: an unadjusted regression model (Model 1); followed by adjustment for demographic factors (age, gender, SES and ethnicity; Model 2), adjustment for lifestyle factors (diet, alcohol intake, smoking status and physical activity levels; Model 3) and adjustment for both demographic and lifestyle factors (Model 4). Models 2, 3 and 4 address Aim 2.
To examine whether the impact of the ACEs on health conditions differed by gender (aim 3), binomial logistic regressions were conducted regressing health conditions on each type of ACE in females and males separately. The significance of the difference in odds ratios (ORs) between genders was tested by adding an interaction term (gender * ACE) to the regression analyses in the combined sample (Table 3 and Supplementary Tables 9-13).
For Tables 2 and 3 (and Supplementary Tables 4-13) p-values are corrected for multiple testing using the Bonferroni correction, the adjusted alpha level after accounting for six ACE categories and 13 health conditions was 0.00064.
Results
Rates of ACEs in the total sample, and for females and males separately, are presented in Supplementary Table 3. 33.64% of the sample reported at least one ACE, and 5.41% three or more ACEs. The prevalences of emotional and physical neglect, and emotional and sexual abuse were significantly higher in females than males (all p-values < 2.2 x 10-16), but there was no difference for physical abuse. Emotional neglect was the most prevalent ACE for both females (23.15%) and males (21.67%), whilst physical neglect was the least prevalent ACE for both females (6.15%) and males (5.25%).
Table 1 presents the sociodemographic and lifestyle characteristics of those who reported ACEs compared to those who did not. The findings are presented separately for females and males as well as combined (taking gender into account as a covariate for the lifestyle variables). Those with any ACE were younger than those without by approximately 8 months (t(106084) = 17.32, p < 2.2 x 10-16). Presence of any ACE was associated with lower socio-economic status (t(98187) = -34.42, p < 2.2 x 10-16)). There was a higher prevalence of people of non-white ethnic backgrounds in individuals with any ACE (4.54%) than without (1.99%) (χ² (1, N = 156629) = 809.22, p < 2.2 x 10-16). Compared to those who had never smoked, there was a higher risk of experiencing any ACE when the participant was a past or current smoker (past smoker OR = 1.37, p < 2.2 x 10-16; current smoker OR = 1.77, p < 2.2 x 10-16). Finally, those with any ACE had increased mean physical activity levels compared to those without ACEs, however the effect size and absolute difference in mean was minimal (combined sample OR = 1.001, p = 6.08 x 10-06). All the associations were present in both females and males, except for physical activity, which was only present in males (OR = 1.001, p = 8.51 x 10-06).
We next examined whether reporting any ACE was associated with risk of health presentations (Aim 1) (Table 2). We found associations with all individual and multimorbid presentations in Model 1 (unadjusted model). The associations were stronger between ACEs and any IC and any ICM-MM respectively, versus ACEs and any CMC. For example, the OR of being diagnosed with any IC after experiencing any ACE was 1.84, (p < 2.2 x 10-16), whereas for individual CMC conditions the ORs ranged from 1.08-1.44, with obesity showing the strongest association (OR = 1.44, p < 2.2 x 10-16). For any CMC, the OR was 1.14, (p = < 2.2 x 10-16). For pairwise ICM-MM combinations, ORs ranged from 1.73 (p < 2.2 x 10-16) for IC plus hypertension to 2.15 (p < 2.2 x 10-16) for IC plus obesity. Association with any ICM-MM (OR= 1.81, p < 2.2 x 10-16) was similar to that of any ACE and any IC, and broadly in line with multimorbid presentations.
We subsequently adjusted for demographic (Model 2) and lifestyle factors (Model 3) as well as both (Model 4) (aim 2). Similar associations to Model 1 were found in Models 2 and 4, whereas in Model 3, one difference was found: any ACE was no longer associated with CKD. A stepwise regression model indicated this reduction in odds ratio and loss of significance was attributable to the inclusion of smoking status (p = 2.66 x 10-3 : not statistically significant after multiple correction tests).
Unadjusted associations of individual ACEs and health conditions are reported in Supplementary Tables 4-8, Model 1 (aim 1). Physical neglect and physical abuse were associated with all individual and multimorbid presentations. Emotional neglect was associated with all conditions except for CKD, emotional abuse was associated with all conditions except for hypertension and CKD and sexual abuse was only associated with any IC, T2D, obesity and all multimorbid presentations. Supplementary Tables 4-8 also provide results for Model 2, 3 and 4 (Aim 2), and present that the associations remain largely unaffected by the inclusion of demographic and lifestyle factors. All individual ACEs were most strongly associated with IC plus obesity, except for sexual abuse which was most strongly associated with IC plus CKD (Model 4, OR = 1.73, p = 1.71 x 10-07). Generally, age and gender were key drivers of changes in statistical significance across models, whereby older and male participants were at higher risk of individual and multimorbid presentations.
Table 3 presents associations between ACEs and health conditions for females and males separately (aim 3). Presence of any ACE was associated with increased risk of all health presentations in both females and males. There was no evidence that the association between any ACE and any individual or multimorbid presentations varied by gender.
Results for all individual ACE types and health presentations (Supplementary Tables 9-13) indicated no gender differences, except for emotional neglect. Although both females and males have increased risk of any CMC after reporting emotional neglect, females have higher odds of any CMC diagnosis than males (female OR = 1.28, male OR = 1.11, p = 1.25 x 10-04).
Discussion
This is the first study to examine the impact of childhood adversity on the multimorbid presentation of IC and a range of CMCs, which confer risk for later cardiovascular disease. We studied the associations between a range of ACEs and diagnoses of any IC, T2D, obesity, hypertension, dyslipidaemia, CKD, any CMC and their multimorbid combinations in a large population-based cohort of mid to older age people. We found that experiences of a range of different ACEs increased the risk of all individual and multimorbid presentations, with similar findings for females and males. Furthermore, the findings generally remained the same after adjustment for demographic and lifestyle variables. These findings highlight the impact of ACEs on complex and difficult to treat chronic health presentations in older age and have implications for prevention and early intervention efforts.
The presence of any ACE increased the risk of any IC by 75%, any CMC by 18% and any ICM-MM by 73% even when accounting for demographic and lifestyle factors. This work builds on prior studies by extending the investigation to an older population and shifting the focus from counts of long-term conditions or disease categories to specific combinations of ICs and CMCs. Importantly, the current study used formal diagnoses recorded in EHRs. These results align, however, with previous research reporting that ACEs are associated with increased likelihood of IC (Merrick et al., 2017), CMC (Suglia et al., 2018) and mental and physical health multimorbidity including ICs and CMCs (Hanlon et al., 2020; Taylor & Demakakos, 2024). The results suggest that ACEs have a stronger relationship with IC and multimorbid presentations than individual CMCs, confirming and extending findings by Hughes et al. (2017). The same is true across all individual ACE types.
The presence of any ACE was most strongly associated with the combination of IC and obesity which is consistent with previous studies finding IC and obesity as comorbid outcomes of ACEs (Davies et al., 2024; Taylor & Demakakos, 2024). Obesity is a trait with complex multifactorial aetiologies, therefore associations between IC and obesity are likely to be complicated and nuanced (Masood & Moorthy, 2023). Possible theories range from the notion that ACEs can motivate people to gain weight to change their body from the one they were abused in (Felitti et al., 1998) and a probable casual bidirectional relationship between BMI and depression (Badillo, Khatib, Kahar, & Khanna, 2022; Steptoe & Frank, 2023).
The patterns of associations tended to be similar across the five different ACEs. All individual ACEs were significantly associated with any IC and all multimorbid combinations of IC and CMC. IC plus obesity, IC plus T2D and IC plus CKD were most strongly associated with individual ACEs, suggesting those who have experienced ACEs might be particularly vulnerable to these conditions as they age. One theory for this association is that chronic stress, which often accompanies adverse childhood experiences, can lead to allostatic load and constant arousal of hypothalamic-pituitary-adrenal (HPA) axis (Kalmakis, Meyer, Chiodo, & Leung, 2015). This in turn may disrupt the cortisol levels in the body, which can be a biomarker of anxiety and depression (Brindle, Pearson, & Ginty, 2022). Disruption of cortisol regulation can also increase food intake and cravings for sugar and salt, giving rise to excessive consumption as a coping mechanism with consequences for physical health (Epel, Lapidus, McEwen, & Brownell, 2001; Jackson, Kirschbaum, & Steptoe, 2017; Testa et al., 2024). Furthermore, both anxiety and depression have been directly related to overeating and binge eating as a form of coping with poor mental health (Torres & Nowson, 2007).
Reporting any ACE increased the likelihood of a diagnosis of IC, CMC and ICM-MM in both females and males. One finding indicated a sex difference: females who had experienced emotional neglect had greater risk of any CMC than males. However, overall, this study highlights that both genders have similar risks of poor health in mid to late adulthood after experiencing ACEs, including ICs. This novel finding adds to a literature that has predominantly focused on females or combined samples.
Those with an ACE were, on average, younger, living in more economically deprived areas and less likely to be of European ancestry, highlighting that the experience of ACEs is socially and culturally patterned as confirmed by previous studies (Merrick, 2019; Metzler, Merrick, Klevens, Ports, & Ford, 2017). Research in populations with a more diverse range of ethnic backgrounds is needed to further establish the relationship between ACEs and ethnicity.
Research has indicated that habitual lifestyle health risk behaviours such as smoking, alcohol use, poor diet and sedentary lifestyle could account for the enduring impact of ACEs on physical and mental health (Bellis, Lowey, Leckenby, Hughes, & Harrison, 2014). In this study, people with ACEs were more likely to be current or past smokers, confirming that ACEs may increase engagement in health risk behaviours. We also found associations between higher physical activity levels and ACEs; however, the effect sizes were small and including physical activity levels as a covariate in analysis had little impact on the findings. Alcohol intake was not associated with ACEs, potentially indicative of the age of UK Biobank participants. Levels of alcohol intake decrease in later adulthood (Britton, Ben-Shlomo, Benzeval, Kuh, & Bell, 2015) and it may be possible that stronger associations would be present in younger cohorts (Zhen-Duan, Colombo, Cruz-Gonzalez, Hoyos, & Alvarez, 2023). We did not find evidence that diet was associated with ACEs.
Strengths and Limitations
Important strengths of this study include the large sample size, use of linked EHR data allowing for formal diagnoses of health conditions and the availability of key covariates such as lifestyle variables. However, there are limitations that must be acknowledged. It is possible there may be reporting biases such that individuals who are most affected by their trauma may have responded “Prefer not to say” to the CTS. We coded these responses as missing and thus these findings may represent an underestimate of the impact of ACEs on mental and physical health in older age.
Compared to the general population, UK Biobank participants are more affluent (Fry et al., 2017). In this study, although those with ACEs were more deprived than those without, the whole sample was still less materially deprived relative to the UK national average. UK Biobank participants are also less likely to have ethnic minority status, or to be obese, smoke, drink alcohol daily and have fewer self-reported health conditions (Fry et al., 2017). Those that continued to participate in the follow-up MHQ (which included the ACEs) were better educated, healthier, of a higher socioeconomic status and had lower rates of smoking than those who did not (Davis et al., 2020). Importantly, despite these biases, associations between ACEs and health conditions remained, suggesting that in the general population, where there may be increased vulnerability to ACEs and poor later health, these associations may be stronger.
The use of EHRs ensured the results were not biased by self-report, such as social desirability bias or recall bias. However, EHRs were designed to support clinical practice rather than research and therefore have limitations such as selective recording of conditions which may produce incomplete or misrepresentative data. Mental health diagnoses, particularly, are poorly captured in EHRs (Madden, Lakoma, Rusinak, Lu, & Soumerai, 2016). Furthermore, UK Biobank has incomplete coverage of primary care records and many of the conditions examined are managed in primary care, suggesting that the recorded frequencies of diagnoses may be lower than the true frequencies.
The ACEs studied in this work are based on retrospective reports and may be subject to recall bias such that those experiencing trauma may find it difficult to access autobiographical memories (Moore & Zoellner, 2007) and that those who are currently euthymic may find it difficult to recall negative memories (Colman et al., 2016). Furthermore, there are differences in associations of ACEs and psychopathology depending on whether the ACE is assessed retrospectively or prospectively, which highlights that subjective memory is potentially important in explaining associations between ACEs and IC in this study (Danese & Widom, 2020).
Finally, the fact that stronger associations were found with IC and multimorbid presentations than individual CMCs raises the question of whether there are direct pathways from ACEs to IC and CMC respectively, or whether ACEs increase the risk of ICs which in turn increase the risk of CMCs. Investigation into this was beyond the scope of the study because it would require conditioning on those who had any IC thereby creating risk of collider bias and spurious associations (Holmberg & Andersen, 2022).
Clinical implications
The results of this study amplify recent claims that ACEs should be treated as a public health issue (Hughes et al., 2017). Those who have experienced ACEs may prefer to be aware of the risk of developing IC, CMC and ICM-MM in older age, allowing for self-advocacy in clinical settings and motivating the adoption of healthy lifestyle habits throughout life to support later wellbeing. Importantly, risk prediction within the patient and public domain should be handled with care, due to the risk of stigmatisation and self-fulfilling prophecy (Stephens et al., 2023). Males may experience additional barriers to service-seeking after trauma such as fear of disclosure and challenge to masculinity (Denhard et al., 2024). Therefore, an emphasis should be placed on motivating males to actively seek similar support after trauma as this study highlights that both females and males have similar risk of experiencing IC after ACEs. This research presents evidence that ACEs contribute to multimorbidity, a major driver of costs incurred by the public health services worldwide. At the population level, faster detection and intervention of ACEs by public health services may reduce cases of multimorbidity and allow for a decrease in healthcare costs. Moreover, joining up public health services, creating one integrated system of care, could allow for more effective information sharing, giving health care providers context for better decision making and individualised care.
The implications of the presence of ACEs for clinical practice remain unclear. The literature on the implementation of routine ACE enquiry in primary care has highlighted a number of caveats such as an over-reliance on crude ACE scores, which do not account for the subjective experience of different ACE types, and a lack of training on safely implementing the enquiry in primary care (Anda, Porter, & Brown, 2020; Gentry & Paterson, 2022; Hardcastle & Bellis, 2018; Loveday et al., 2022; McLennan, Gonzalez, MacMillan, & Afifi, 2024). A focus on strengthening patient-clinician interactions through connection, longer appointment times in which trust can be built and training on safely discussing trauma, may be more beneficial (Jones, Merrick, & Houry, 2020; Zulman et al., 2020).
Finally, it should be acknowledged that a substantial number of people go on to healthy and fulfilling adulthoods after having experienced ACEs, and their lived experience could be drawn upon to develop interventions that improve later health outcomes. For example, there is a growing body of literature on the concept of “continuing adversity” (those who experience prolonged adversity after childhood have worse outcomes compared to those whose trauma is confined to childhood (Horwitz, Widom, McLaughlin, & White, 2001)), and the importance of improving resilience after experiencing ACEs to mediate poor health outcomes in childhood and adulthood (Haczkewicz et al., 2024).
Conclusion
This study shows that ACEs impact IC, CMC and ICM-MM in older adulthood and demographic and lifestyle factors do not change these associations. Males are at similar risk as females of poor health in adulthood following ACEs. The importance of early detection and intervention to prevent ACEs and the use of trauma-informed care cannot be understated.
Supporting information
Data Availability
All data produced are available online at https://www.ukbiobank.ac.uk/use-our-data/
Acknowledgements
This research has been conducted using the UK Biobank Resource under application number 79704. Full list of LINC members: Marianne B. M. van den Bree, George Kirov, Michael J. Owen, James T. R. Walters, Peter A. Holmans, Jane Lynch, Ioanna K. Katzourou, Lowri O’Donovan (Cardiff University, UK). David A. van Heel, Sarah Finer, Daniel Stow (Queen Mary University of London, UK). Golam M. Khandaker, Nicholas J. Timpson, John A. A. MacLeod, Julie P. Clayton, Ruby S. M. Tsang, Jane Sprackman, Shahid Khan (University of Bristol, UK). Inês Barroso, Rupert A. Payne (University of Exeter, UK). Mark Mon-Williams, Megan L. Wood (University of Leeds, UK). Hilary C. Martin (Wellcome Sanger Institute, UK). Thomas Werge, Andrés Ingason (Institute of Biological Psychiatry, Denmark).
We would like to acknowledge the LIfespaN multimorbidity research Collaborative (LINC) patient and public involvement group for their input on this research.
Financial contributions
This work was funded by a PhD studentship to LKB from Health and Care Research Wales (MvdB; PH, MJO, FR, MMW; HS 22 04) and the Tackling Multimorbidity at Scale Strategic Priorities Fund programme (MvdB, PH, MJO, MMW, RP; MR/W014416/1) delivered by the Medical Research Council and the National Institute for Health Research in partnership with the Economic and Social Research Council and in collaboration with the Engineering and Physical Sciences Research Council.
Ethical standards
The authors assert that all procedures contributing to this work comply with the ethical standards of the relevant national and institutional committees on human experimentation and with the Helsinki Declaration of 1975, as revised in 2008.
Competing interests
Michael J. Owen reports grants from Akrivia Health and Takeda Pharmaceuticals outside the submitted work.