Association of adverse childhood experiences with lipid profiles and atherogenic risk indices in a middle-to-older aged population

Abstract

Background

Adverse childhood experiences (ACE) have been associated with poor later life health outcomes, including cardiovascular disease (CVD). Limited research investigating potential underlying biological mechanisms linking ACE to CVD exists, particularly regarding lipid biomarkers.

Objectives

The aim of this study was to examine the associations between childhood adversity and unfavourable lipid profiles and derived atherogenic risk indices in a middle-to-older aged population.

Methods

This cross-sectional study includes 1820 participants from the Mitchelstown cohort (49% male) in Ireland. Participants' self-reported history of childhood adversity (overall and by subtypes household dysfunction, abuse and neglect) were assessed through a validated 10-item ACE questionnaire. Lipid profiles were determined and atherogenic risk indices including Castelli's Risk Index 1 and 2 (CRI-I and CRI-II), Atherogenic Coefficient (AC) and Atherogenic Index Plasma (AIP) were generated. Logistic regression analysed ACE associations with unfavourable lipid outcomes, controlling for potential confounders.

Results

ACE history (reported by 23% of sample), in particular childhood exposure to household dysfunction, was associated with later-life non-optimal TG and HDL-C concentrations and atherogenic risk indices CRI-II and AC in age and sex-adjusted models (all p < 0.05). In fully adjusted models, adults reporting ACE or exposure to household dysfunction were approximately twice as likely to have pro-atherogenic CRI-II relative to adults with no ACE (OR = 1.86, 95% CI: 1.19–2.92, p = 0.006 and OR = 2.19, 95% CI: 1.33–3.61, p = 0.002, respectively). Sex-stratified analysis demonstrated sex-specific associations.

Conclusions

This study provides evidence that ACEs are common among older adults in Ireland and are associated with unfavourable lipid profiles and derived atherogenic risk indices.

Highlights

• •Greater cardiovascular disease risk in adults with a history of childhood adversity.
• •Adverse childhood experiences are common; reported in almost a quarter of adults.
• •Unfavourable lipid profiles among adults exposed to adverse childhood events, with sex-specific associations.
• •Childhood adversity associated with non-optimal TG and HDL-C levels in adulthood.
• •Household dysfunction associated with pro-atherogenic lipid derived risk indices.

Abbreviations

ACE

Adverse Childhood Experience

CVD

Cardiovascular disease

CRI

Castelli Risk Index

AC

Atherogenic Coefficient

AIP

Atherogenic Risk Index

CRP

C-reactive protein

IL-6

Interleukin-6

LDL-C

Low-density Lipoprotein cholesterol

HDL-C

High-density lipoprotein cholesterol

GHQ

General Health Questionnaire

IPAQ

International Physical Activity Questionnaire

METs

Metabolic Equivalent of Tasks

FFQ

Food Frequency Questionnaire

VLDL-C

Very low-density lipoprotein cholesterol

TG

Triglyceride

IQR

Interquartile range

SD

Standard deviation

BMI

Body mass index

OR

Odds ratio

CI

Confidence interval

1. Introduction

Adverse childhood experiences (ACEs) include a wide range of events such as abuse, neglect and household dysfunction, that may have a serious impact on a child's physical or psychological well-being (Bartlett & Sacks, 2019). Furthermore, evidence suggests that ACEs may also have long-term effects and negatively influence later life health. Exposure to ACEs have been shown to be associated with depressive symptoms, multimorbidity and cardiovascular disease (CVD) (Cheong et al., 2017; Godoy et al., 2021; Sinnott et al., 2015). Much of the research regarding intermediate biomarkers of CVD risk in the context of ACE has focussed predominately on biomarkers of inflammation and glucose homeostasis (Campbell et al., 2018; Kerr et al., 2021). Results from a population-based sample from the Health and Retirement study (n = 11,198, mean age 69) suggest that those reporting at least one ACE have elevated high sensitivity c-reactive protein (CRP) concentrations (Lin et al., 2016). Childhood emotional abuse was associated with higher levels of interleukin-6 (IL-6) in midlife women (Nguyen & Thurston, 2020). Associations between childhood adversity and elevated fasting insulin levels and insulin resistance have also been observed (Campbell et al., 2018).

A casual role exists between high cholesterol concentrations and pathogenesis of CVD (Ference et al., 2017). An unfavourable lipid profile, defined by elevated plasma triglyceride and low-density lipoprotein (LDL) cholesterol and low high-density lipoprotein (HDL) cholesterol concentrations, are hallmarks of dyslipidaemia, an important risk factor for CVD (Kopin & Lowenstein, 2017). Atherogenic risk indices derived from the lipid measures include Castelli's Risk Index 1 and 2 (CRI-I, CRI-II) (Castelli et al., 1983), the Atherogenic Index of Plasma (AIP) (Liu et al., 2021) and the Atherogenic Coefficient (AC) (Sujatha & Kavitha, 2017). These indices capture alterations in the lipid profile.

Associations between social and cultural factors and health are well established (Antonovsky, 1967). More specifically, lower socioeconomic status has been linked with CVD risk factors and greater CVD risk (Schultz et al., 2018; Wang et al., 2023). Previous research has also shown the link between childhood adversity and socioeconomic factors including low parental education (Sedlak et al., 2010), as well as parental belief in corporal punishment (Russa & Rodriguez, 2010). Results from a Swedish national cohort study (n = 870,000) demonstrated associations of childhood household dysfunction with both poor health outcomes in later life and with socioeconomic disadvantage (Gauffin et al., 2016), highlighting the potential for dual strategies targeting both ACE and social inequalities. Furthermore, sex-specific associations of childhood abuse and household dysfunction and later life CVD risk factors have been reported (Aguayo et al., 2022).

Therefore, it is important to understand the risk factors that may underlie associations between ACE and cardiometabolic health in adulthood, in both men and women. Thus far however, limited research has investigated the relationship between childhood adversity and later life dyslipidaemia and pro-atherogenic indices, in particular with sex-stratified analyses. This study aims to address these gaps in the knowledge base by examining the associations between ACE and its subtypes with lipid biomarkers and derived atherogenic risk indices in middle-to-older aged men and women.

2. Methods

2.1. Study population

The Cork and Kerry Diabetes and Heart Disease Study (Phase II—Mitchelstown Cohort) was a single-centre, cross-sectional study conducted between 2010 and 2011. Full details regarding study design, sampling methods, and data collection have been previously described (Kearney et al., 2013). In brief, the cohort includes 2047 middle-to-older aged men (49%) and women aged between 45 and 74 years. Stratified sampling was used to randomly selected patients attending the Livinghealth clinic in Mitchelstown, County Cork, Ireland. Ethics committee approval conforming to the Declaration of Helsinki was obtained from the Clinical Research Ethics Committee of University College Cork. A letter was sent by the general practitioner (GP) to all participants, with a reply slip indicating acceptance to the study. All participants provided signed informed consent to use their data for research purposes. The current analysis is based on 1820 participants with complete data available from the ACE questionnaire and lipid profiles.

2.2. Data collection and measures

All participants completed a general health questionnaire (GHQ) which provided data on demographic variables including age, sex, educational level, smoking status and alcohol consumption, prescription cholesterol-lowering medication use and CVD history. Information on frequency, duration, and intensity of physical activity was collected using the validated short-form International Physical Activity Questionnaire (IPAQ) (Craig et al., 2003). Diet was evaluated using a modified version of the self-completed European Prospective Investigation into Cancer and Nutrition (EPIC) food frequency questionnaire (FFQ) (Riboli et al., 1997), which has been validated extensively in several populations. The daily intake of energy and nutrients was computed from FFQ data using a tailored computer programme (FFQ Software Version 1.0; developed by the National Nutrition Surveillance Centre, School of Public Health, Physiotherapy and Sports Science, University College Dublin, Belfield, Dublin 4, Ireland), which linked frequency selections with the food equivalents in McCance and Widdowson Food Tables (Julia et al., 2014).

Anthropometric measurements were performed by trained researchers with reference to a standard operating procedures manual. Height was measured with a portable Seca Leicester height/length stadiometer (Seca, Birmingham, UK) and weight was measured using a portable electronic Tanita WB-100MA weighing scale (Tanita Corp, IL, USA). The weighing scale was placed on a firm flat surface and was calibrated weekly. Body mass index (BMI = weight (kg)/height(m)²) was calculated from measured weight and height.

2.2.1. Adverse childhood experience

Exposure to ACE was assessed using the 10-item ACE questionnaire, which is a validated instrument used to assess childhood adversity (Anda et al., 2010). This questionnaire addresses three categories: abuse (emotional, physical and sexual), neglect (emotional and physical) and household dysfunction (parental separation/divorce, violence against mother, substance abuse, mental illness and incarceration of household member). Participants were offered separate sealed envelopes to submit their responses to the ACE questionnaire during data collection. All ACE questions refer to the participant's first 18 years of life and require a binary response (yes/no). Overall ACE scores were calculated from responses. Responses also were categorised as any history of ACE (yes/no), with yes responses further classified by ACE subtype.

2.2.2. Lipid profiles and derived atherogenic risk indices

Study participants attended the clinic in the morning after an overnight fast (of at least 8 h) and blood samples were taken on arrival. Total cholesterol (total-C), low-density lipoprotein cholesterol (LDL-C), high-density lipoprotein cholesterol (HDL-C), very low-density lipoprotein cholesterol (VLDL-C), and triglyceride (TG) concentrations were measured by Cork University Hospital Biochemistry Laboratory on Olympus 5400 biochemistry analysers with Olympus reagents using standardised procedures and fresh samples (Olympus Diagnostica GmbH, Hamburg, Germany). Four atherogenic risk indices were generated from the lipid biomarkers. AIP was calculated as log (TG/HDL-C) (Nwagha et al., 2010). CRI-I was calculated as Total-C/HDL-C and CRI-II was calculated as LDL-C/HDL-C (Castelli et al., 1983). AC was calculated as (Total-C – HDL-C)/HDL-C (Sujatha & Kavitha, 2017). Established formulas and cut off points for the atherogenic risk indices (Bhardwaj et al., 2013) and each of the lipids are presented in Supplementary Table 1.

2.2.3. Covariates

Categories of education included ‘some primary (not complete)’, ‘primary or equivalent’, ‘intermediate/group certificate or equivalent’, ‘leaving certificate or equivalent’, ‘diploma/certificate’, ‘primary university degree’ and ‘postgraduate/higher degree’. These were collapsed and recoded into a dichotomous variable: ‘primary education only’ (finished full-time education at age 13 years or younger) and ‘intermediate or higher’. Smoking status was categorised into never or former smoker and current smoker. Alcohol consumption was categorised into non-drinker (<1 unit/week) and drinker (>1 unit/week). Physical activity levels were classified as low or moderate-high, based on metabolic equivalents (METs) minutes per week in all activity types (Craig et al., 2003). Obesity was defined as a BMI ≥30 kg/m² and energy intake was measured in Kcals from FFQ responses. The presence of CVD was obtained from the GHQ by asking study participants if they had been diagnosed with any one of the following seven conditions: Heart Attack (including coronary thrombosis or myocardial infarction), Heart Failure, Angina, Aortic Aneurysm, Hardening of the Arteries, Stroke, or any other Heart Trouble. Subjects that indicated a diagnosis of any one of these conditions were classified as having CVD. Information on lipid-lowering medication was determined from a participant's response to a question in the GHQ regarding prescription medication use.

2.3. Statistical analysis

Descriptive characteristics were examined for all participants, stratified by sex and according to ACE history. These characteristics were described by counts and percentages for categorical variables and mean (SD) or median (IQR), as appropriate, for continuous variables. Pearson's chi-squared tests, independent t-tests and Mann-Whitney U tests were used to analyse differences between groups. Correlation analysis (Spearman's rho) investigating relationships between ACE scores and lipid biomarkers, stratified by sex, was performed. Logistic regression was conducted to examine unfavourable lipid and atherogenic risk index associations with ACE history and subtypes. Three models were run: unadjusted (model 1), adjusted for age (years) and sex (model 2) and fully adjusted (model 3), which additionally adjusted for BMI, energy intake, alcohol, smoking, physical activity, education and lipid-lowering medication use. Additional sensitivity analyses including sex-stratified analyses (Supplementary Tables 2 and 3) and analyses excluding those with a CVD diagnosis (Supplementary Tables 4 and 5) were conducted on fully adjusted models for lipids and atherogenic indices. A p-value (two-tailed) of less than 0.05 was considered to indicate statistical significance. To correct for the multiple testing performed Benjamini and Hochberg's correction for multiple testing was applied with a false discovery rate of 0.05 (Benjamini & Hochberg, 1995). Data analysis was conducted using SPSS version 27.0 (IBM Corporation, Armonk, NY, USA).

3. Results

3.1. Descriptive characteristics

Characteristics of the full study sample, and stratified by sex, are presented in Table 1. In total, 32.2% of participants were obese and 34.7% reported being on lipid-lowering medication. Over half of the study sample (51.8%) were current drinkers and just under half of participants (45.3%) had low levels of physical activity. ACE history was prevalent in 23% of participants. Household dysfunction was the most frequently reported subtype (14.6%), followed by abuse (12.6%) and neglect (6.6%).

Table 1.

Characteristics of the study sample for all participants and stratified by sex.

General Characteristics	Total	Males	Females	p
	n = 1820	n = 882	n = 938
Age, years (median [IQR])	59.5 (46.4–74)	59.4 (46.4–70.8)	59.6 (48.9–74)	0.847
BMI, kg/m2 (mean ± SD)	28.6 (±4.7)	29.2 (±4)	27.9 (±5)	<0.001
Obese (%)	585 (32.1)	321 (36.4)	264 (28.1)	<0.001
Cardiovascular disease (%)	193 (10.6)	128 (14.5)	65 (6.9)	<0.001
Lipid lowering medication (%)	632 (34.7)	301 (34.1)	331 (35.3)	0.603
Socioeconomic factors
Primary education only (%)	464 (25.5)	266 (30.2)	198 (21.1)	<0.001
Lifestyle factors
Energy intake, kcal (median [IQR])	1906.7 (1480.2–2436.8)	1913.9 (1481.5–2427.9)	1899.0 (1486.2–2435)	0.784
Current smoker (%)	260 (14.3)	124 (14.1)	136 (14.5)	0.83
Current drinker (%)	943 (51.8)	531 (60.2)	412 (43.9)	0.007
Low-level physical activity (%)	825 (45.3)	339 (38.4)	486 (51.8)	<0.001
Adverse Childhood Experiences
ACE history (%)	419 (23)	216 (24.5)	203 (21.6)	0.149
ACE abuse (%)	229 (12.6)	120 (13.6)	109 (11.6)	0.217
ACE neglect (%)	121 (6.6)	51 (5.8)	70 (7.5)	0.142
ACE household dysfunction (%)	265 (14.6)	140 (15.9)	125 (13.3)	0.125
Lipid biomarkers
Total-C, mmol/L (median [IQR])	5.3 (0–9.9)	5.2 (0–8.8)	5.4 (0–9.9)	<0.001
HDL-C, mmol/L (median [IQR])	1.4 (0–3)	1.3 (0–3)	1.6 (0–3)	<0.001
LDL-C, mmol/L (median [IQR])	3.2 (0–7.3)	3.1 (0–6.3)	3.2 (0–7.3)	0.007
VLDL-C, mmol/L (median [IQR])	0.6 (0–4.9)	0.6 (0–4.9)	0.5 (0–2.8)	<0.001
Triglycerides, mmol/L (median [IQR])	1.2 (0–10.7)	1.3 (0–10.7)	1.1 (0–6.1)	<0.001
AIP (median [IQR])	−0.1 (−1.6–2.6)	0.1 (−1.6–2.6)	−0.4 (−1.6–2)	<0.001
CRI-I (median [IQR])	3.7 (1.8–9.7)	4.0 (1.9–9.2)	3.4 (1.8–9.7)	<0.001
CRI-II (median [IQR])	2.2 (0–7.1)	2.5 (0–5.2)	2.1 (0.5–7.1)	<0.001
AC (median [IQR])	2.7 (0.8–8.7)	3.0 (0.9–8.2)	2.4 (0.8-8–7)	<0.001

Continuous data are presented as mean ± standard deviation, median (interquartile range), median (range) and categorical data as numbers (percentages).

p-values were determined from a Mann-Whitney U test, independent t-test or Pearson's chi-square. BMI Body Mass Index, ACE Adverse Childhood Experience, Total-C Total Cholesterol, HDL-C High-density lipoprotein cholesterol, LDL-C low-density lipoprotein cholesterol, VLDL-C Very low-density lipoprotein cholesterol.

Analysis following stratification by sex revealed that males were more likely to have a higher BMI, have CVD, be obese, have lower education levels (all p < 0.001) and were more likely to be drinkers (p = 0.007). Males also had a less favourable lipid profile, characterised by lower HDL-C and higher VLDL-C and TG concentrations (all p < 0.001). Not surprisingly, all four atherogenic risk indices were higher in males (all p < 0.001). Females were more likely to have low physical activity levels (p < 0.001) and had higher total-C (p < 0.001) and higher LDL-C concentrations (p = 0.007).

3.2. Descriptive characteristics according to ACE history

Table 2 presents study participants' characteristics, socioeconomic and lifestyle factors and lipid biomarkers according to ACE history. Almost one-quarter (23%) of study participants reported at least one ACE. Participants who experienced at least one ACE were younger (p < 0.001), had higher prevalence of CVD (p = 0.008) and were more likely to have completed education to primary level only (p = 0.004). Participants reporting an ACE history had less favourable lipid profiles characterised by lower HDL-C concentrations (p = 0.025) and higher, more pro-atherogenic CRI-I and AC (both p < 0.001) and AIP and CRI-II (p = 0.031 and p = 0.002 respectively).

Table 2.

Characteristics of the study sample stratified by ACE history.

	With ACE	Without ACE	p
General Characteristics	n = 419 (23%)	n = 1401 (77%)
Age, years (median [IQR])	57.5 (48.9–70.6)	60.1 (46.4–74.0)	<0.001
BMI, kg/m2 (mean ± SD)	28.9 (±4.9)	28.4 (4.6)	0.095
Obese (%)	146 (35.4)	439 (32.0)	0.208
Cardiovascular disease (%)	59 (14.1)	134 (9.6)	0.008
Lipid lowering medication (%)	145 (34.6)	487 (34.8)	0.953
Socioeconomic factors
Primary education only (%)	108 (25.8)	356 (25.4)	0.004
Lifestyle factors
Energy intake, kcal (median [IQR])	1881.8 (1506.4–2381.3)	1914.4 (1472.5–2442.8)	0.683
Current smoker (%)	63 (15)	197 (14.4)	0.634
Current drinker (%)	238 (56.8)	705 (59.3)	0.207
Low physical activity (%)	192 (45.8)	633 (45.2)	0.544
Lipid biomarkers
Total-C, mmol/L (median [IQR])	5.4 (0–9.9)	5.3 (0–9.8)	0.061
HDL-C, mmol/L (median [IQR])	1.4 (0–2.8)	1.4 (0–3)	0.025
LDL-C, mmol/L (median [IQR])	3.2 (0–7.2)	3.2 (0–7.3)	0.126
VLDL-C, mmol/L (median [IQR])	0.6 (0–4.9)	0.6 (0–3.3)	0.247
Triglycerides, mmol/L (median [IQR])	1.2 (0–10.7)	1.2 (0–7.1)	0.181
AIP (median [IQR])	−0.1 (−1.5–2.6)	−0.2 (−1.6–2.4)	0.031
CRI-I (median [IQR])	3.9 (2–9.7)	3.6 (1.8–8.2)	<0.001
CRI-II (median [IQR])	2.4 (0.1–7.1)	2.2 (0–5.2)	0.002
AC (median [IQR])	2.9 (1–8.7)	2.6 (0.8–7.2)	<0.001

Continuous data are presented as mean ± standard deviation, median (interquartile range), median (range) and categorical data as numbers (percentages).

p-values were determined from a Mann-Whitney U test, independent t-test or Pearson's chi-square. BMI Body Mass Index, ACE Adverse Childhood Experience, Total-C Total Cholesterol, HDL-C High-density lipoprotein cholesterol, LDL-C low-density lipoprotein cholesterol, VLDL-C Very low-density lipoprotein cholesterol.

3.3. Correlation analyses

The results of the Spearman's rho correlation analysis between the ACE score and lipid biomarkers for the full sample, and stratified by sex, are shown in Table 3. Overall, there were weak negative relationships observed between HDL-C concentrations and ACE scores r_s = -0.053, p < 0.025, which appeared to be driven by male participants (r_s = -0.067, p < 0.048) as no significant correlation was observed in the female participants. Positive weak correlations were observed between the ACE score and total and LDL-C concentrations in the female participants only (r_s = 0.070, p < 0.032 and r = 0.080), p < 0.014, for total-C and LDL-C respectively). Each of the four atherogenic risk indices were weakly correlated with the ACE score.

Table 3.

Spearman correlations between ACE total score and lipid biomarkers in the full sample and stratified by sex.

Lipid biomarkers	Total		Males		Females
	rs	p	rs	p	rs	p
Total-C (mmol/L)	0.044	0.061	0.031	0.355	0.070	0.032
HDL-C (mmol/L)	−0.053	0.025	−0.067	0.048	−0.022	0.506
LDL- C (mmol/L)	0.036	0.126	−0.003	0.927	0.080	0.014
VLDL-C (mmol/L)	0.027	0.247	0.039	0.252	0.008	0.812
Triglycerides (mmol/L)	0.031	0.181	0.042	0.215	0.013	0.691
AIP	0.051	0.031	0.065	0.053	0.023	0.478
CRI-I	0.082	<0.001	0.078	0.02	0.071	0.03
CRI-II	0.075	0.001	0.057	0.09	0.077	0.018
AC	0.082	<0.001	0.078	0.02	0.071	0.03

C – Cholesterol.

3.4. Logistic regression analyses of ACE associations with lipids

Table 4 presents the analysis predicting the likelihood of unfavourable lipids based on ACE history and subtype. Childhood exposure to household dysfunction was associated with non-optimal HDL-C and VLDL-C concentrations in the crude model (OR = 1.43, 95% CI: 1.04–1.96, p = 0.029 and OR = 1.39, 95% CI: 1.05–1.84, p = 0.023, respectively). The association with HDL-C concentration persisted after adjustment for age and sex (OR = 1.51, 95% CI: 1.09–2.09, p = 0.012 respectively), but was attenuated in the fully adjusted model. No associations between ACE or ACE subtypes with unfavourable total-C or LDL-C concentrations were observed.

Table 4.

Logistic regression predicting the likelihood of non-optimal lipids based on ACE history and subtype.

Total-C	Model 1			Model 2			Model 3
	ORs	95% CI	p	ORs	95% CI	p	ORs	95% CI	p
ACE history	1.05	0.84–1.31	0.683	0.97	0.77–1.21	0.758	0.97	0.71–1.33	0.851
ACE abuse	0.90	0.68–1.19	0.456	0.83	0.62–1.10	0.193	0.72	0.48–1.06	0.093
ACE neglect	1.13	0.77–1.64	0.523	0.98	0.67–1.44	0.924	0.73	0.41–1.27	0.26
ACE household dysfunction	1.10	0.84–1.43	0.489	0.98	0.75–1.30	0.984	0.98	0.95–1.01	0.116
HDL-C
ACE history	1.26	0.96–1.67	0.098	1.32	1.00–1.74	0.053	1.35	0.88–1.97	0.185
ACE abuse	1.34	0.80–1.62	0.487	1.18	0.82–1.68	0.371	1.36	0.84–2.21	0.214
ACE neglect	1.12	0.70–1.79	0.649	1.17	0.73–1.88	0.523	1.26	0.62–2.55	0.527
ACE household dysfunctio	1.43	1.04–1.96	0.029	1.51	1.09–2.09	0.012	1.40	0.88–2.22	0.159
LDL-C
ACE history	1.16	0.93–1.45	0.178	1.10	0.87–1.36	0.46	1.10	0.81–1.49	0.528
ACE abuse	1.09	0.83–1.45	0.531	1.02	0.77–1.36	0.88	1.00	0.68–1.46	0.989
ACE neglect	1.23	0.85–1.78	0.268	1.12	0.77–1.63	0.551	0.74	0.42–1.27	0.272
ACE household dysfunction	1.09	0.84–1.42	0.531	0.99	0.76–1.30	0.954	1.12	0.78–1.59	0.545
VLDL-C
ACE history	1.21	0.95–1.54	0.118	1.17	0.92–1.50	0.209	0.97	0.70–1.34	0.85
ACE abuse	1.10	0.81–1.50	0.541	1.06	0.77–1.44	0.739	0.89	0.59–1.34	0.58
ACE neglect	1.16	0.77–1.74	0.472	1.22	0.81–1.85	0.348	0.94	0.52–1.69	0.832
ACE household dysfunction	1.39	1.05–1.84	0.023	1.33	0.99–1.77	0.055	1.076	0.74–1.57	0.706
TG
ACE history	1.49	1.06–2.07	0.02	1.41	1.00–1.98	0.050	1.08	0.68–1.73	0.743
ACE abuse	1.18	0.76–1.82	0.458	1.11	0.71–1.72	0.656	0.83	0.45–1.51	0.537
ACE neglect	1.93	1.17–3.17	0.01	2.08	1.25–3.48	0.005	1.68	0.79–3.59	0.178
ACE household dysfunction	1.64	1.12–2.39	0.011	1.53	1.03–2.26	0.035	1.28	0.76–2.15	0.363

Model 1 unadjusted, Model 2 adjusted for age and sex, Model 3 additionally adjusted for BMI, energy intake, alcohol intake, smoking status, physical activity, education and lipid-lowering medication. Reference values were no ACE and no ACE subtype within each comparison.

Exposure to an ACE, neglect and household dysfunction were associated with an increased likelihood of having unfavourable TG concentrations in the unadjusted analysis (OR = 1.49, 95% CI: 1.06–2.07, p = 0.020, OR = 1.93, 95% CI: 1.17–3.17, p = 0.010, OR = 1.64, 95% CI: 1.12–2.39, p = 0.011) respectively. Each of these associations persisted following adjustment for age and sex (OR = 1.41, 95% CI: 1.00–1.98, p = 0.050, OR = 2.08, 95% CI: 1.25–3.48, p = 0.005. OR = 1.53, 95% CI: 1.03–2.26, p = 0.035, respectively) but were attenuated in the fully adjusted models.

Supplementary Table 2 presents results of the fully adjusted sex-stratified analysis predicting the likelihood of unfavourable lipids based on ACE and subtype. Exposure to childhood neglect was associated with non-optimal TG concentrations in males (OR = 2.63, 95% CI: 1.15–5.99, p = 0.022). No significant associations of ACE exposure with unfavourable lipids were observed among females.

3.5. Logistic regression analyses of ACE associations with atherogenic risk indices

Table 5 presents results of the analysis predicting the likelihood of atherogenic indices based on ACE history and subtype. Exposure to ACE was associated with an increased likelihood of unfavourable CRI-II and AC in crude models (OR = 1.60, 95% CI: 1.15–2.30, p = 0.005 and OR = 1.44, 95% CI: 1.15–1.81, p = 0.002, respectively). Examination of ACE subtypes revealed associations of household dysfunction, but not abuse or neglect, with pro-atherogenic AC and CRI-II in the crude and age and sex adjusted models. The associations with CRI-II persisted in the fully adjusted models; those reporting ACE, particularly childhood exposure to household dysfunction, were approximately twice as likely to have unfavourable CRI-II relative to those not exposed (OR = 1.86, 95% CI: 1.19–2.92, p = 0.006 and OR = 2.19, 95% CI: 1.33–3.61, p = 0.002, respectively). However, following adjustment for multiple testing none of the associations persisted.

Table 5.

Logistic regression predicting likelihood of unfavourable atherogenic risk indices based on ACE history and subtype.

AIP	Model 1			Model 2			Model 3
	ORs	95% CI	p	ORs	95% CI	p	ORs	95% CI	p
ACE history	1.12	0.89–1.40	0.348	1.10	0.87–1.40	0.431	0.91	0.65–1.28	0.593
ACE abuse	0.92	0.69–1.24	0.594	0.89	0.66–1.22	0.471	0.65	0.42–1.00	0.048
ACE neglect	0.98	0.66–1.44	0.898	1.10	0.73–1.66	0.645	0.73	0.39–1.37	0.323
ACE household dysfunction	1.31	1.00–1.71	0.052	1.30	0.98–1.72	0.073	1.12	0.76–1.66	0.577
AC
ACE history	1.44	1.15–1.81	0.002	0.98	0.96–1.00	0.012	1.26	0.91–1.75	0.17
ACE abuse	1.13	0.85–1.52	0.4	1.03	0.76–1.40	0.867	0.94	0.63–1.43	0.785
ACE neglect	1.18	0.80–1.77	0.395	1.29	0.86–1.94	0.227	1.10	0.59–1.99	0.796
ACE household dysfunction	1.67	1.28–2.19	<0.001	1.55	1.16–2.06	0.003	1.41	0.96–2.07	0.083
CRI-I
ACE history	1.22	0.96–1.56	0.107	1.17	0.91–1.49	0.225	1.02	0.74–1.41	0.889
ACE abuse	1.08	0.79–1.47	0.627	1.03	0.75–1.40	0.877	0.81	0.54–1.21	0.297
ACE neglec	1.28	0.84–1.96	0.247	1.22	0.80–1.86	0.363	1.20	0.67–2.18	0.538
ACE household dysfunction	1.41	1.04–1.90	0.027	1.32	0.97–1.79	0.075	1.33	0.90–1.98	0.158
CRI-II
ACE history	1.60	1.15–2.30	0.005	1.46	1.04–2.05	0.028	1.86	1.19–2.92	0.006
ACE abuse	1.46	0.96–2.20	0.075	1.33	0.87–2.03	0.182	1.40	0.80–2.44	0.242
ACE neglect	1.26	0.72–2.21	0.425	1.27	0.71–2.26	0.419	0.98	0.40–2.40	0.959
ACE household dysfunction	1.77	1.22–2.58	0.003	1.58	1.08–2.33	0.019	2.19	1.33–3.61	0.002

Model 1 unadjusted, Model 2 adjusted for age and sex, Model 3 additionally adjusted BMI, energy intake, alcohol intake, smoking status, physical activity, education and lipid-lowering medication.

Reference values were no ACE and no ACE subtype within each comparison.

Supplementary Table 3 presents the fully adjusted sex-stratified analysis predicting the likelihood of non-optimal atherogenic risk indices based on ACE and subtype. Exposure to childhood abuse in males was associated with AIP (OR = 0.53, 95% CI: 0.32–0.90, p = 0.019). ACE history, abuse and household dysfunction were associated with CRI-II in females (OR = 5.45, 95% CI: 2.09–14.21, p < 0.001, OR = 3.63, 95% CI: 1.36–9.72, p = 0.010 and OR = 4.60, 95% CI: 1.74–12.13, p < 0.002, respectively).

Additional sensitivity analyses restricted to adults without a CVD diagnosis revealed that exposure to a history of ACE was associated with higher likelihood of non-optimal HDL-C concentrations (OR = 1.63, 95% CI:1.06–2.51, p = 0.026) (Supplementary Table 4). ACE history, particularly exposure to household dysfunction were associated with unfavourable CRI-II (OR = 1.80, CI: 1.13–2.86, p = 0.014 and OR = 2.09, CI: 1.24–3.52, p = 0.006, respectively) (Supplementary Table 5).

4. Discussion

This study investigated ACE and ACE subtype associations with lipids and derived atherogenic risk indices in middle-older-aged adults. ACE history, in particular childhood exposure to household dysfunction, was associated with non-optimal HDL-C and TG concentrations and unfavourable atherogenic risk indices CRI-II and AC. The ACE associations with CRI-II persisted in the fully adjusted model and appeared to be driven by a history of household dysfunction. Adults reporting ACE or a history of childhood exposure to household dysfunction were approximately twice as likely to have a more pro-atherogenic CRI-II compared to adults without ACE. Sex-stratified analysis provided evidence of sex-specific associations. Overall, these results provide new insights into the relationship between ACE and lipid biomarkers of CVD risk in males and females and highlight a potential mechanism underlying the greater CVD risk reported in adults with an ACE history.

Atherogenic risk indices, derived from lipid measures, have been suggested as significant indicators of cardiometabolic health status and CVD risk (Olamoyegun et al., 2016). Higher AIP levels have been independently associated with diabetes, hypertension, rapid plaque progression, which is strongly linked with greater future CVD risk (Khosravi et al., 2022), plaque instability and with coronary artery disease (CAD), beyond traditional CVD risk factors (Cai et al., 2017; Dobiasova & Frohlich, 2001; Onat et al., 2010; Won et al., 2021; Wu et al., 2018). Recent evidence suggests associations of AIP, CRI-I and CRI-II with arterial stiffness (Wang et al., 2022), which may be a consequence of the pathological changes that occur in the arterial wall during progression of atherosclerosis (Hansen & Taylor, 2016). Results from the Avon Longitudinal Study of Parents and Children (ALSPAC) (n = 3612 middle-aged women) revealed that exposure to childhood psychosocial adversity was associated with lower arterial distensibility (Anderson et al., 2018). Results from longitudinal follow-up of several other birth cohorts have demonstrated consistent associations of childhood adversity with outcomes of poorer lipid profiles and increased inflammation (Danese et al., 2008; Li et al., 2019; Poulton et al., 2015). However limited investigation of atherogenic indices in the context of childhood adversity and later life cardiometabolic health has been conducted to date.

In the current study, exposure to ACE history, in particular childhood exposure to household dysfunction, was associated with non-optimal HDL-C and TG concentrations, and pro-atherogenic CRI-II. Furthermore, in analyses restricted to participants without a CVD diagnosis, the associations between ACE history and household dysfunction with unfavourable CRI-II persisted, suggesting the robustness of this relationship. Additionally, an association between ACE history and non-optimal HDL-C concentrations was observed. These findings are consistent with previous research which identified an association between exposure to childhood abuse and pro-atherogenic CRI-II in an observational study (n = 452) (Spann et al., 2014).

Regarding lipid profiles, in keeping with the results of the present study, associations between childhood neglect were previously observed with unfavourable TG concentrations (Li et al., 2016; Miller & Lacey, 2022; Péterfalvi et al., 2019). It is interesting to note that these studies additionally reported associations between abuse and parental offending with TG concentrations. Associations between childhood abuse and later life non-optimal HDL-C levels have been reported (Soares et al., 2021; Spann et al., 2014). Conversely, exposure to neglect and physical abuse in childhood was positively associated with HDL-C concentrations in the 1958 British birth cohort study (n = 9000) (Li et al., 2016). Findings from the 1958 British birth cohort also showed that exposure to parental divorce/separation, neglect and psychological abuse was associated with lower HDL-C concentrations (n = 8511) (Miller & Lacey, 2022). It should be noted that none of the aforementioned studies investigated household dysfunction, highlighting the value of broader examination of ACE subtypes. However, in the CARDIA study a ‘risky’ family environment, including ‘unaffectionate family interactions and conflict’, which may be considered as household dysfunction, was associated with HDL-C concentrations (Loucks et al., 2011).

In the present study, exposure to household dysfunction was also associated with unfavourable CRI-I and non-optimal VLDL-C concentrations in crude models, however these associations were attenuated upon adjustment. Only one previous small study (n = 42) reported positive associations between ACE exposure, in particular childhood neglect, and CRI-I (Péterfalvi et al., 2019). To the best of our knowledge, no other study has examined VLDL-C as a biomarker for CVD in the context of ACE. We did not identify any relationships between ACE exposure with either total-C or LDL-C concentrations. These results are consistent with findings from the CARDIA study regarding exposure to childhood maltreatment and total-C concentrations (Loucks et al., 2011). In contrast, results from investigation of the 1958 British Birth cohort suggested that exposure to physical abuse was associated with LDL-C concentrations (Li et al., 2016).

We also did not observe any associations between ACE or ACE subtypes with AIP. Results from a comparative study revealed that AIP had the strongest correlation of determination in CAD patients, leading the authors to conclude that AIP is the most sensitive marker of CVD risk compared to CRI-I, CRI-II and AC (Bhardwaj et al., 2013). Unlike the other atherogenic indices, AIP is calculated with TG concentrations which may partly account for the observed differences.

Differences in ACE exposure between males and females have been demonstrated in other studies (Almuneef et al., 2017; Jones et al., 2022). Sex-specific associations of childhood abuse and household dysfunction and later life CVD risk factors have also been reported (Aguayo et al., 2022). Our sex-stratified analysis revealed some sex-specific associations. Childhood neglect and abuse were associated with non-optimal TG concentrations and with AIP among males. These findings are in contrast with observations from previous research (Li et al., 2016; Miller & Lacey, 2022; Péterfalvi et al., 2019). Among females, ACE history, household dysfunction and abuse were associated with CRI-II. These findings are consistent with the literature; in an observational study females exposed to childhood abuse had greater associations with non-optimal CRI-II compared to males (Spann et al., 2014). Furthermore, greater risk of hyperlipidaemia was observed among white women who experienced abuse and household dysfunction (HR = 3.61; 95% CI: 1.62–8.05) or households with low levels of organisation (HR = 2.05; 95% CI: 1.25–3.36) compared with white women who experienced abuse but lived in well-organised households (HR = 0.66; 95% CI: 0.41–1.06) (Aguayo et al., 2022).

Household dysfunction was prevalent in the Mitchelstown cohort and comprises a greater range of sub-categories, compared to neglect and abuse. Sub-categories of household dysfunction can include adults with alcohol and drug use problems and separation of parents (Prevention & Early Intervention Network, 2019). Data from the Irish Central Statistics Office demonstrate a 51.5% increase in marital separation over a twenty year period from 1996 to 2016 (Central Statistic Office, 2016). Ireland has been ranked 9th among OECD countries in terms of alcohol consumption and 8th in the world for binge drinking (Health research Board, 2021). Alcohol consumption and/or parental separation may account for the impact of household dysfunction in this population. It has been postulated that ACE may directly influence disease risk through an unhealthy lifestyle, including poor diet, smoking, heavy alcohol consumption and sedentary behaviour (Yang et al., 2022), which tends to follow a social gradient. It is interesting to note that in the current study none of these lifestyle factors were different between study participants with and without an ACE history. However, among those reporting ACE, over one third had a medical card, a proxy of socioeconomic status, and a quarter had attained primary education only, suggesting some socioeconomic disadvantage in those who experienced childhood adversity. It was demonstrated in ALSPAC that ACE was associated with lower education attainment (Houtepen et al., 2020) and results from a British birth cohort revealed early life social disadvantage associations with premature mortality (Rogers et al., 2021). Examination of adverse childhood psychosocial experiences in the Dunedin Multidisciplinary Health and Development Study (n = 1037) revealed associations of social isolation, maltreatment and socioeconomic disadvantage with a clustering of metabolic risk factors including low HDL-C concentrations (Danese et al., 2009). Experience of socioeconomic disadvantage in childhood was also associated with elevated age-related disease risks in adulthood. Collectively, these data highlight the importance of examining ACE and ACE subtypes with regard to socio-cultural background and context. This could facilitate a more nuanced understanding of subgroups at greatest risk, with a view to informing intervention and treatment strategies and programmes.

A limitation of this study is that ACE scores were used to classify the history of ACE exposure as a binary variable. Therefore, the exposed group includes participants with ACE scores ranging from 1 to 8. This approach presumes equivalence of ACE scores. There is evidence of a dose-response effect of ACEs with those reporting 3+ at much higher risk of adverse later life health compared with those reporting 1 or 2 which leads to potential for masking the effect of variation in ACE exposure (Chang et al., 2019; Felitti et al., 1998). There is also evidence that males and females have distinct patterns of ACE which may differentially influence later life social, emotional and mental health outcomes (Haahr-Pedersen et al., 2020). Nonetheless, previous research has classified ACE exposure in the same way (Cheong et al., 2017; Sinnott et al., 2015) and it should be noted that very few participants had higher ACE scores (e.g. 7 or 8). The use of self-reported questionnaires, such as the ACE questionnaire, is subject to potential inaccuracies and recall and reporting bias. Lastly, although we included a wide range of potential confounders, residual confounding arising from imprecise measurement of variables, or indeed unmeasured confounders, should also be considered.

In conclusion, the results of this study highlight that a history of childhood adversity is prevalent in adults in Ireland and may be associated with non-optimal lipid profiles in later life. ACE history and childhood exposure to household dysfunction were found to be associated with an increased likelihood of having unfavourable TG and HDL-C concentrations and pro-atherogenic CRI-II. Sex-stratified analysis further demonstrated sex-specific associations. Further research investigating the relationship between ACE and CVD biomarkers in men and women according to socioeconomic status is warranted.

Source of support

This work was supported by a research grant from the Irish Health Research Board (reference: HRC/2007/13). The funders had no role in the study design, data collection and analysis, decision to publish or preparation of the manuscript.

Ethical statement

This work is based on secondary data analysis of the Cork and Kerry Diabetes and Heart Disease Study (Phase II—Mitchelstown Cohort). Ethics committee approval conforming to the Declaration of Helsinki was obtained from the Clinical Research Ethics Committee of University College Cork.

CRediT author statement

EOL, CMP and IJP: Conceptualisation, EOL and CMP: Investigation, SRM, IJP and CMP: Resources, IJP and CMP: Project administration, EOL: Formal analysis, CMP: Supervision, IJP and CMP: Funding acquisition, EOL and CMP: Writing – Original Draft, EOL, SRM, IJP and CMP: Writing – Review and Editing.

Declaration of competing interest

None declared.

Appendix A. Supplementary data

The following is the Supplementary data to this article.

Data availability

Data will be made available on request.