The Dunedin Study - DMHDRU

Publications

MRI

Is cardiovascular fitness associated with structural brain integrity in midlife? Evidence from a population-representative birth cohort study | 2020
d'Arbeloff, T., Cooke, M., Knodt, ... Show all » A. R., Sison, M., Melzer, T. R., Ireland, D., Poulton, R., Ramrakha, S., Moffitt, T. E., Caspi, A., Hariri, A. R. « Hide
Aging, 2020, 12(20), .
https://doi.org/10.18632/aging.104112
download pdfLink to full publication »
Our ref: RO744
Show abstract » Improving cardiovascular fitness may buffer against age-related cognitive decline and mitigate dementia risk by staving off brain atrophy. However, it is unclear if such effects reflect factors operating in childhood (neuroselection) or adulthood (neuroprotection). Using data from 807 members of the Dunedin Study, a population-representative birth cohort, we investigated associations between cardiovascular fitness and structural brain integrity at age 45, and the extent to which associations reflected possible neuroselection or neuroprotection by controlling for childhood IQ. Higher fitness, as indexed by VO2Max, was not associated with average cortical thickness, total surface area, or subcortical gray matter volume including the hippocampus. However, higher fitness was associated with thicker cortex in prefrontal and temporal regions as well as greater cerebellar gray matter volume. Higher fitness was also associated with decreased hippocampal fissure volume. These associations were unaffected by the inclusion of childhood IQ in analyses. In contrast, a higher rate of decline in cardiovascular fitness from 26 to 45 years was not robustly associated with structural brain integrity. Our findings are consistent with a neuroprotective account of adult cardiovascular fitness but suggest that effects are not uniformly observed across the brain and reflect contemporaneous fitness more so than decline over time
« Hide abstract

Association of childhood lead exposure with MRI measurements of structural brain integrity in midlife | 2020
Reuben, A. Elliott, M.L. Abraham, C. Broadbent, J. Houts, ... Show all » R.M. Ireland, D. Knodt, A.R. Poulton, R. Ramrakha, S. Hariri, A.R. Caspi, A. Moffitt, T.E. « Hide
JAMA, 2020, 324(19), 1970-1979.
https://doi.org/10.1001/jama.2020.19998
download pdfLink to full publication »
Our ref: RO742
Show abstract » Importance. Childhood lead exposure has been linked to disrupted brain development, but long-term consequences for structural brain integrity are unknown.

Objective. To test the hypothesis that childhood lead exposure is associated with magnetic resonance imaging (MRI) measurements of lower structural integrity of the brain in midlife.

Design, Setting, and Participants. The Dunedin Study followed a population-representative 1972-1973 birth cohort in New Zealand (N=564 analytic sample) to age 45 years (until April 2019).

Exposure. Childhood blood-lead levels measured at age 11 years.

Main Outcomes and Measures. Structural brain integrity at age 45 assessed via MRI (primary outcomes): gray matter (cortical thickness, surface area, hippocampal volume), white matter (white matter hyperintensities, fractional anisotropy [theoretical range: 0(diffusion is perfectly isotropic)-100(diffusion is perfectly anisotropic)]), and brainAGE, a composite index of the gap between chronological age and a machine-learning algorithm-estimated brain age (0 indicates a brain age equivalent to chronological age, positive and negative values represent an older and younger brain age, respectively). Age-45 cognitive function was assessed objectively via the Wechsler Adult Intelligence Scale–IV [IQ range, 40-160, standardized to mean(SD)=100(15)] and subjectively via informant and self-reports [z-score units, scale mean(SD)=0(1)].

Results. Of 1037 original participants, 997 were alive at age 45, of whom 564(57%) had received lead testing at age 11 years (302[54%] male) (median follow-up 34 years, IQR 33.7-34.7). Mean(SD) age-11 blood-lead level was 10.99(4.63) µg/dL. After adjusting for covariates, each 5µg/dL higher childhood blood-lead level was significantly associated with 1.19 cm2 smaller cortical surface area (95%CI:-2.35,-0.02, P=.05), 0.10 cm3 smaller hippocampal volume (95%CI:-0.17,-0.03, P=.006), lower global fractional anisotropy (b=-0.12, 95%CI:-0.24,-0.01, P=.04), and 0.77 years older brainAGE (95%CI:0.02, 1.51, P=.05). There were no statistically significant associations between blood-lead level and log-transformed white matter hyperintensity volume (b=0.05 log mm3, 95%CI:-0.02, 0.13, P=.17) or mean cortical thickness (b=-0.004 mm, 95%CI:-0.012, 0.004, P=.39). Each 5µg/dL higher childhood blood-lead level was significantly associated with a 2.07-point lower score (95%CI:-3.39,-0.74, P=.002) in age-45 IQ, and a 0.12-point higher score (95%CI:0.01, 0.23, P=.03) on informant-rated cognitive problems. There was no statistically significant association between childhood blood-lead levels and self-reported cognitive problems (b=-0.02 points, 95%CI: -0.10, 0.07, P=.68).

Conclusion and Relevance. In this longitudinal cohort study, with a median 34-year-follow-up, higher childhood blood-lead level was associated with differences in some MRI measures of brain structure that suggested lower structural brain integrity in mid-life. Because of the large number of statistical comparisons, some findings may represent type I error.

« Hide abstract

What is the test-retest reliability of common task-fMRI measures? New empirical evidence and a meta-analysis | 2020
Maxwell L. Elliott, Annchen R. Knodt, David Ireland, Meriwether L. Morris, Richie Poulton, ... Show all » Sandhya Ramrakha, Maria L. Sison, Terrie E. Moffitt, Avshalom Caspi, Ahmad R. Hariri « Hide
Psychological Science, 2020, 31(7), .
https://doi.org/10.1177%2F0956797620916786
download pdf Our ref: RO734
Show abstract » Identifying brain biomarkers of disease risk is a growing priority in neuroscience. The ability to identify meaningful biomarkers is limited by measurement reliability; unreliable measures are unsuitable for predicting clinical outcomes. Measuring brain activity using task functional MRI (fMRI) is a major focus of biomarker development; however, the reliability of task fMRI has not been systematically evaluated. We present converging evidence demonstrating poor reliability of task-fMRI measures. First, a meta-analysis of 90 experiments (N = 1,008) revealed poor overall reliability—mean intraclass correlation coefficient (ICC) = .397. Second, the test-retest reliabilities of activity in a priori regions of interest across 11 common fMRI tasks collected by the Human Connectome Project (N = 45) and the Dunedin Study (N = 20) were poor (ICCs = .067–.485). Collectively, these findings demonstrate that common task-fMRI measures are not currently suitable for brain biomarker discovery or for individual-differences research. We review how this state of affairs came to be and highlight avenues for improving task-fMRI reliability.
« Hide abstract

Pervasively Thinner Neocortex as a Transdiagnostic Feature of General Psychopathology | 2020
Adrienne L. Romer, Maxwell L. Elliott, Annchen R. Knodt, Maria L. Sison, David Ireland, ... Show all » Renate Houts, Sandhya Ramrakha, Richie Poulton, Ross Keenan, Tracy R. Melzer, Terrie E. Moffitt, Avshalom Caspi, Ahmad R. Hariri « Hide
Am J Psychiatry, 2020, .
https://doi.org/10.1176/appi.ajp.2020.19090934
download pdfLink to full publication »
Our ref: RO737
Show abstract » Objective:
Neuroimaging research has revealed that structural brain alterations are common across broad diagnostic families of disorders rather than specific to a single psychiatric disorder. Such overlap in the structural brain correlates of mental disorders mirrors already well-documented phenotypic comorbidity of psychiatric symptoms and diagnoses, which can be indexed by a general psychopathology or p factor. The authors hypothesized that if general psychopathology drives the convergence of structural alterations common across disorders, then 1) there should be few associations unique to any one diagnostic family of disorders, and 2) associations with the p factor should overlap with those for the broader diagnostic families.

Methods:
Analyses were conducted on structural MRI and psychopathology data collected from 861 members of the population-representative Dunedin Multidisciplinary Health and Development Study at age 45.

Results:
Study members with high scores across three broad diagnostic families of disorders (externalizing, internalizing, thought disorder) exhibited highly overlapping patterns of reduced global and widely distributed parcel-wise neocortical thickness. Study members with high p factor scores exhibited patterns of reduced global and parcel-wise neocortical thickness nearly identical to those associated with the three broad diagnostic families.

Conclusions:
A pattern of pervasively reduced neocortical thickness appears to be common across all forms of mental disorders and may represent a transdiagnostic feature of general psychopathology. As has been documented with regard to symptoms and diagnoses, the underlying brain structural correlates of mental disorders may not exhibit specificity, and the continued pursuit of such specific correlates may limit progress toward more effective strategies for etiological understanding, prevention, and intervention.

« Hide abstract

Associations between life-course-persistent antisocial behaviour and brain structure in a population-representative longitudinal birth cohort | 2020
Christina O Carlisi, Terrie E Moffitt, Annchen R Knodt, Honalee Harrington, David Ireland, ... Show all » Tracy R Melzer, Richie Poulton, Sandhya Ramrakha, Avshalom Caspi, Ahmad R Hariri, Essi Viding « Hide
Lancet Psychiatry , 2020, 7 245-53.
doi.org/10.1016/S2215-0366(20)30002-X
download pdfLink to full publication »
Our ref: RO731
Show abstract » Summary
Background Studies with behavioural and neuropsychological tests have supported the developmental taxonomy theory of antisocial behaviour, which specifies abnormal brain development as a fundamental aspect of life-coursepersistent antisocial behaviour, but no study has characterised features of brain structure associated with lifecourse-persistent versus adolescence-limited trajectories, as defined by prospective data. We aimed to determine whether life-course-persistent antisocial behaviour is associated with neurocognitive abnormalities by testing the hypothesis that it is also associated with brain structure abnormalities.
Methods
We used structural MRI data collected at 45 years of age from participants in the Dunedin Study, a populationrepresentative longitudinal birth cohort of 1037 individuals born between April 1, 1972, and March 31, 1973, in Dunedin, New Zealand, who were resident in the province and who participated in the first assessment at 3 years of age. Participants underwent MRI, and mean global cortical surface area and cortical thickness were extracted for each participant. Participants had been previously subtyped as exhibiting life-course-persistent, adolescence-limited, or no history of persistent antisocial behaviour (ie, a low trajectory group) based on informant-reported and self-reported conduct problems from the ages of 7 years to 26 years. Study personnel who processed the MRI images were masked to antisocial group membership. We used linear estimated ordinary least squares regressions to compare each antisocial trajectory group (life-course persistent and adolescence limited) with the low trajectory group to examine whether antisocial behaviour was related to abnormalities in mean global surface area and mean cortical thickness. Next, we used parcel-wise linear regressions to identify antisocial trajectory group differences in surface area and cortical thickness. All results were controlled for sex and false discovery rate corrected. Findings Data from 672 participants were analysed, and 80 (12%) were classified as having life-course-persistent antisocial behaviour, 151 (23%) as having adolescence-limited antisocial behaviour, and 441 (66%) as having low antisocial behaviour. Individuals on the life-course-persistent trajectory had a smaller mean surface area (standardised β=–0·18 [95% CI –0·24 to –0·11]; p<0·0001) and lower mean cortical thickness (standardised β=–0·10 [95% CI –0·19 to –0·02]; p=0·020) than did those in the low group. Compared with the low group, the life-course-persistent group had reduced surface area in 282 of 360 anatomically defined parcels and thinner cortex in 11 of 360 parcels encompassing circumscribed frontal and temporal regions associated with executive function, affect regulation, and motivation. Widespread differences in brain surface morphometry were not observed for the adolescence-limited group compared with either non-antisocial behaviour or life-course-persistent groups.
Interpretation
These analyses provide initial evidence that differences in brain surface morphometry are associated with life-course-persistent, but not adolescence-limited, antisocial behaviour. As such, the analyses are consistent with the developmental taxonomy theory of antisocial behaviour and highlight the importance of using prospective longitudinal data to define different patterns of antisocial behaviour development.

« Hide abstract

Brain-age in midlife is associated with accelerated biological aging and cognitive decline in a longitudinal birth cohort | 2019
Elliot, M.L. Belsky, D.W., Knodt, A.R., ... Show all » Ireland, D., Melzer, T.R., Poulton, R., Ramrakha, S., Caspi, A., Moffitt, T.E., Hariri, A.R « Hide
Molecular Psychiatry, 2019, .
https://doi.org/10.1038/s41380-019-0626-7
download pdf Our ref: RO733
Show abstract » An individual’s brainAGE is the difference between chronological age and age predicted from machine-learning models of brain-imaging data. BrainAGE has been proposed as a biomarker of age-related deterioration of the brain. Having an older brainAGE has been linked to Alzheimer’s, dementia, and mortality. However, these findings are largely based on crosssectional associations which can confuse age differences with cohort differences. To illuminate the validity of brainAGE as a biomarker of accelerated brain aging, a study is needed of a large cohort all born in the same year who nevertheless vary on brainAGE. In the Dunedin Study, a population-representative 1972–73 birth cohort, we measured brainAGE at age 45 years, as well as the pace of biological aging and cognitive decline in longitudinal data from childhood to midlife (N = 869). In this cohort, all chronological age 45 years, brainAGE was measured reliably (ICC = 0.81) and ranged from 24 to 72 years. Those with older midlife brainAGEs tended to have poorer cognitive function in both adulthood and childhood, as well as impaired brain health at age 3. Furthermore, those with older brainAGEs had an accelerated pace of biological aging, older facial appearance, and early signs of cognitive decline from childhood to midlife. These findings help to validate brainAGE as a potential surrogate biomarker for midlife intervention studies that seek to measure dementia-prevention efforts in midlife. However, the findings also caution against the assumption that brainAGE scores represent only age-related deterioration of the brain as they may also index central nervous system variation present since childhood.
« Hide abstract

White matter hyperintensities are common in midlife and already associated with cognitive decline | 2019
Tracy d’Arbeloff, Maxwell L. Elliott, Annchen R. Knodt, Tracy R. Melzer, Ross Keenan, ... Show all » David Ireland, Sandhya Ramrakha, Richie Poulton, Tim Anderson, Avshalom Caspi, Terrie E. Moffitt, Ahmad R. Hariri « Hide
Brain Communications, 2019, 1(1), .
https://doi.org/10.1093/braincomms/fcz041
download pdf Our ref: RO728
Show abstract » White matter hyperintensities proliferate as the brain ages and are associated with increased risk for cognitive decline as well as Alzheimer’s disease and related dementias. As such, white matter hyperintensities have been targeted as a surrogate biomarker in intervention trials with older adults. However, it is unclear at what stage of aging white matter hyperintensities begin to relate to cognition and if they may be a viable target for early prevention. In the Dunedin Study, a population-representative cohort followed since birth, we measured white matter hyperintensities in 843 45-year-old participants using T2-weighted magnetic resonance imaging and we assessed cognitive decline from childhood to midlife. We found that white matter hyperintensities were common at age 45 and that white matter hyperintensity volume was modestly associated with both lower childhood (ß = −0.08, P = 0.013) and adult IQ (ß=−0.15, P < 0.001). Moreover, white matter hyperintensity volume was associated with greater cognitive decline from childhood to midlife (ß=−0.09, P < 0.001). Our results demonstrate that a link between white matter hyperintensities and early signs of cognitive decline is detectable decades before clinical symptoms of dementia emerge. Thus, white matter hyperintensities may be a useful surrogate biomarker for identifying individuals in midlife at risk for future accelerated cognitive decline and selecting participants for dementia prevention trials.
« Hide abstract

Replicability of structural brain alterations associated with general psychopathology: evidence from a population-representative birth cohort | 2019
Romer, A. L., Knodt, A. R., Sison, ... Show all » M., Ireland, D., Ramrakha, S., Poulton, R., Keenan, R., Melzer, T. R., Moffitt, T. E., Caspi, A., Hariri, A. R. « Hide
Molecular Psychiatry, 2019, .
https://doi.org/10.1038/s41380-019-0621-z
download pdf Our ref: RO726
Show abstract » Transdiagnostic research has identified a general psychopathology factor – often called the ‘p’
factor – that accounts for shared liability to internalizing, externalizing, and thought disorders in
diverse samples. It has been argued that the p factor may reflect dysfunctional thinking present
in serious mental illness. In support of this, we previously used a theory-free, data-driven
multimodal neuroimaging approach to find that higher p factor scores are associated with
structural deficits within a cerebello-thalamo-cortical circuit (CTCC) and visual association
cortex, both of which are important for monitoring and coordinating information processing in
the service of executive control. Here we attempt to replicate these associations by conducting
region-of-interest analyses of the CTCC and visual association cortex using data from 831
members of the Dunedin Multidisciplinary Health and Development Study, a five-decade
longitudinal study of a population-representative birth cohort now 45 years old. We further
sought to replicate a more recent report that p factor scores can be predicted by patterns of
distributed cerebellar morphology as estimated through independent component analysis. We
successfully replicated associations between higher p factor scores and both reduced grey
matter volume of the visual association cortex and fractional anisotropy of pontine white
matter pathways within the CTCC. In contrast, we failed to replicate prior associations between
cerebellar structure and p factor scores. Collectively, our findings encourage further focus on
the CTCC and visual association cortex as core neural substrates and potential biomarkers of
transdiagnostic risk for mental illness.

« Hide abstract

Association of Neurocognitive and Physical Function With Gait Speed in Midlife | 2019
Line Jee Hartmann Rasmussen, Avshalom Caspi, Antony Ambler, Jonathan M. Broadbent, Harvey J. Cohen, ... Show all » Tracy d’Arbeloff, Maxwell Elliott, Robert J. Hancox, HonaLee Harrington, Sean Hogan, Renate Houts, David Ireland, Annchen R. Knodt, Kim Meredith-Jones, Miriam C. Morey, Lynda Morrison, Richie Poulton, Sandhya Ramrakha, Leah Richmond-Rakerd, Maria L. Sison, Kate Sneddon, W. Murray Thomson, Ahmad R. Hariri, Terrie E. Moffitt « Hide
JAMA Network Open, 2019, .
10.1001/jamanetworkopen.2019.13123
download pdfLink to full publication »
Our ref: RO725
Show abstract » IMPORTANCE Gait speed is a well-known indicator of risk of functional decline and mortality in older adults, but little is known about the factors associated with gait speed earlier in life.
OBJECTIVES To test the hypothesis that slow gait speed reflects accelerated biological aging at midlife, as well as poor neurocognitive functioning in childhood and cognitive decline from childhood to midlife.

DESIGN, SETTING, AND PARTICIPANTS This cohort study uses data from the Dunedin Multidisciplinary Health and Development Study, a population-based study of a representative 1972
to 1973 birth cohort in New Zealand that observed participants to age 45 years (until April 2019). Data analysis was performed from April to June 2019.

EXPOSURES Childhood neurocognitive functions and accelerated aging, brain structure, and concurrent physical and cognitive functions in adulthood.

MAIN OUTCOMES AND MEASURES Gait speed at age 45 years, measured under 3 walking conditions: usual, dual task, and maximum gait speeds.

RESULTS Of the 1037 original participants (91% of eligible births; 535 [51.6%] male), 997 were alive at age 45 years, of whom 904 (90.7%) had gait speed measured (455 [50.3%] male; 93%white). The mean (SD) gait speeds were 1.30 (0.17) m/s for usual gait, 1.16 (0.23) m/s for dual task gait, and 1.99 (0.29) m/s for maximum gait. Adults with more physical limitations (standardized regression coefficient [β], −0.27; 95%CI, −0.34 to −0.21; P < .001), poorer physical functions (ie, weak grip strength [β, 0.36; 95%CI, 0.25 to 0.46], poor balance [β, 0.28; 95%CI, 0.21 to 0.34], poor visualmotor coordination [β, 0.24; 95%CI, 0.17 to 0.30], and poor performance on the chair-stand [β, 0.34; 95%CI, 0.27 to 0.40] or 2-minute step tests [β, 0.33; 95%CI, 0.27 to 0.39]; all P < .001), accelerated biological aging across multiple organ systems (β, −0.33; 95%CI, −0.40 to −0.27; P < .001), older facial appearance (β, −0.25; 95%CI, −0.31 to −0.18; P < .001), smaller brain volume (β, 0.15; 95%CI, 0.06 to 0.23; P < .001), more cortical thinning (β, 0.09; 95%CI, 0.02 to 0.16; P = .01), smaller cortical surface area (β, 0.13; 95%CI, 0.04 to 0.21; P = .003), and more white matter hyperintensities (β, −0.09; 95%CI, −0.15 to −0.02; P = .01) had slower gait speed. Participants with lower IQ in midlife (β, 0.38; 95%CI, 0.32 to 0.44; P < .001) and participants who exhibited cognitive
decline from childhood to adulthood (β, 0.10; 95%CI, 0.04 to 0.17; P < .001) had slower gait at age 45 years. Those with poor neurocognitive functioning as early as age 3 years had slower gait in midlife (β, 0.26; 95%CI, 0.20 to 0.32; P < .001).

« Hide abstract

General functional connectivity: Shared features of resting-state and task fMRI drive reliable and heritable individual differences in functional brain networks | 2019
Maxwell L. Elliott, Annchen R. Knodt, Megan Cooke, M. Justin Kim, Tracy R. Melzer, ... Show all » Ross Keenan, David Ireland, Sandhya Ramrakha, Richie Poulton, Avshalom Caspi, Terrie E. Moffitt, Ahmad R. Hariri « Hide
NeuroImage, 2019, 189 516-532.
https://doi.org/10.1016/j.neuroimage.2019.01.068
download pdfLink to full publication »
Our ref: RO719
Show abstract » Intrinsic connectivity, measured using resting-state fMRI, has emerged as a fundamental tool in the study of the human brain. However, due to practical limitations,
many studies do not collect enough resting-state data to generate reliable measures of intrinsic connectivity necessary for studying individual differences. Here we
present general functional connectivity (GFC) as a method for leveraging shared features across resting-state and task fMRI and demonstrate in the Human Connectome
Project and the Dunedin Study that GFC offers better test-retest reliability than intrinsic connectivity estimated from the same amount of resting-state data
alone. Furthermore, at equivalent scan lengths, GFC displayed higher estimates of heritability than resting-state functional connectivity. We also found that predictions
of cognitive ability from GFC generalized across datasets, performing as well or better than resting-state or task data alone. Collectively, our work suggests that GFC can improve the reliability of intrinsic connectivity estimates in existing datasets and, subsequently, the opportunity to identify meaningful correlates of individual differences in behavior. Given that task and resting-state data are often collected together, many researchers can immediately derive more reliable measures of intrinsic connectivity through the adoption of GFC rather than solely using resting-state data. Moreover, by better capturing heritable variation in intrinsic connectivity, GFC represents a novel endophenotype with broad applications in clinical neuroscience and biomarker discovery.

« Hide abstract

A Polygenic Score for Higher Educational Attainment is Associated with Larger Brains | 2019
Elliott, M. L. Belsky, D. W. Anderson, K. Corcoran, D. L. Ge, ... Show all » T. Knodt, A. Prinz, J. A. Sugden, K. Williams, B. Ireland, D. Poulton, R. Caspi, A. Holmes, A. Moffitt, T.E. Hariri, A. R. « Hide
Cerebal Cortex, 2019, 1-9.
https://doi.org/10.1093/cercor/bhy219
download pdf Our ref: RO706
Show abstract » People who score higher on intelligence tests tend to have larger brains. Twin studies suggest the same genetic factors influence both brain size and intelligence. This has led to the hypothesis that genetics influence intelligence partly by contributing to development of larger brains. We tested this hypothesis with molecular genetic data using discoveries from a genome-wide association study (GWAS) of educational attainment, a correlate of intelligence. We analyzed genetic, brain imaging, and cognitive test data from the UK Biobank, the Dunedin Study, the Brain Genomics Superstruct Project (GSP), and the Duke Neurogenetics Study (DNS) (combined N=8,271). We measured genetics using polygenic scores based on published GWAS. We conducted meta-analysis to test associations among participants' genetics, total brain volume (i.e., brain size), and cognitive test performance. Consistent with previous findings, participants with higher polygenic scores achieved higher scores on cognitive tests, as did participants with larger brains. Participants with higher polygenic scores also had larger brains. We found some evidence that brain size partly mediated associations between participants' education polygenic scores and their cognitive test performance. Effect-sizes were larger in the population-based UK Biobank and Dunedin samples than in the GSP and DNS samples. Sensitivity analysis suggested this effect-size difference partly reflected restricted range of cognitive performance in the GSP and DNS samples. Recruitment and retention of population-representative samples should be a priority for neuroscience research. Findings suggest promise for studies integrating GWAS discoveries with brain imaging data to understand neurobiology linking genetics with individual differences in cognitive performance.
« Hide abstract