Abstract Body (Do not enter title and authors here): Background
People with congenital heart disease (CHD) face developmental and acquired risks to their neurocognitive health. Mechanisms and clinical-neuroimaging correlates are not well-defined. We report baseline neuroimaging findings, their associations with demographic/clinical factors, and cognitive performance, from a longitudinal study of brain health in adult CHD (ACHD).

Methods
Participants (n=99) aged >18y with moderate-severe complexity CHD were recruited from the ACHD referral centre for Western Canada. They underwent clinical review and had bloodwork, Holter, echocardiography MRI brain, cognitive testing (MoCA, NIH Toolbox [NIHTB]). Forward stepwise linear and logistic regression models were used to identify demographic/clinical factors that predicted MRI findings. Relationships between demographic/clinical factors with MRI findings, and MRI findings with cognition, were then explored with multivariable linear and logistic regression as appropriate.

Results
Median(IQR) age 35y (29-40); 42% female.(Figure 1) CHD complexity (21% severe) was independently associated with lower total brain volume (TBV). TBV independently predicted cognitive performance (B[95%CI] for MoCA change/100mm3: 2.73x10^-3 [3.0x10^-4 - 5.1 x 10^-3]; p=0.03; NIHTB 8.39x10^-3 [2.2x10^-3 - 1.46x10^-2]; p=0.009).
History of dyslipidemia was independently associated with white matter hyperintensity (WMH) volume but not presence of WMH. Cerebral microbleeds (CMB, 60% of participants) and lacunes (16%) were independently associated with history and number of cardiopulmonary bypass surgeries. None of WMH, CMB or lacunes predicted cognitive performance.

Conclusions
In this high-functioning cohort of mostly younger ACHD, neuroimaging abnormalities were common. TBV was independently associated with CHD severity. WMH were associated with dyslipidemia; CMB and lacunes with bypass. Only TBV predicted cognitive performance.
Acknowledging our modest cohort size with heterogenous CHD types, our results suggest that the pathophysiology impacting brain health reflects a combination of early- and later-life factors. Longitudinal studies may identify optimal preventative interventions and their timing; dyslipidemia may be a modifiable target.