# The Dallas Lifespan Brain Study

The Dallas Lifespan Brain Study (DLBS) is a longitudinal multi-modal neuroimaging study of the aging mind, which was initiated in 2008 (referred to as Wave 1). Participants returned for two additional waves of data collection with an approximate interval of 4-5 years between waves. The DLBS protocol encompasses various imaging modalities, including structural MRI, diffusion MRI, and functional MRI, as well as comprehensive cognitive and psychosocial assessments. Additionally, amyloid and tau PET imaging were performed as part of the study. This data set was among the pioneering long-duration longitudinal studies that investigated MRI metrics related to brain aging, PET biomarkers associated with Alzheimer's disease, and cognition across the entire adult lifespan (ranging from ages 20-90). Moreover, it was one of the earliest cognitive aging studies to include in vivo measurement of tauopathy using PET imaging with radiotracer AV-1451, also known as flortaucipir.

For detailed information on the data acquisition methods employed in the DLBS, please refer to our comprehensive document titled “Keys to the Kingdom”. Methodological references are also included at the end of this document.

Inquiries about this dataset should be directed to: [email protected].

This dataset is licensed under the [Creative Commons Zero (CC0) v1.0 License](https://creativecommons.org/publicdomain/zero/1.0/).


## Keys to the Kingdom

For more detailed information on all the data types mentioned below, please consult the “Keys to the Kingdom” file (The Dallas Lifespan Brain Study/DLBS KTTK.pdf). 

The “Keys to the Kingdom” is a master document that includes all the components of the Dallas Lifespan Brain Study. Within this document are listed the types of data that were collected, the source documents for each, how to access the study’s source documents and spreadsheets, and the coded item names and their abbreviations for each variable present in the spreadsheets. For neuroimaging data, detailed descriptions of the data processing procedures are also provided within the document.


## Individual Differences Measures

Individual differences measures for this sample are included in the participants.tsv and participants.json files. Key measures include age, sex, years of education, ethnicity, race, and the time interval between waves of data collection (in years).

Unavailable data points in participants.tsv are reported as "NA". Scores for participants that did not return for that wave are entered as "xx". 

## Cognitive Measures (/derivatives/cognition)

At each wave of data collection, a comprehensive cognitive battery was administered over two testing sessions. The cognitive measures were categorized into six distinct constructs:

1) *Speed of Processing*: Digit Comparison, WAIS-III Digit Symbol, NIH Toolbox Pattern Comparison Processing Speed Test 
	
2) *Working Memory*: CANTAB Spatial Working Memory, WAIS-III Letter Number Sequencing Task, Operation Span Task, NIH Toolbox List Sorting, CANTAB Delayed Matching to Sample Task, CANTAB Spatial Recognition Memory Task 	

4) *Episodic Memory*: Hopkins Verbal Learning Immediate & Delayed, CANTAB Verbal Recognition Memory, Woodcock-Johnson Memory for Names Immediate & Delayed, Wechsler Memory Scale Logical Memory, NIH Toolbox Picture Sequence Memory
	
5) *Reasoning*: Raven’s Matrices, ETS Letter Sets, CANTAB Stockings of Cambridge, Everyday Problem Solving 
	
6) *Vocabulary*: Educational Testing Service Advanced Vocabulary, Shipley Vocabulary, CANTAB Graded Naming Task, NIH Toolbox Oral Reading Recognition Test, NIH Toolbox Picture Vocabulary 
	
7) *Verbal Fluency*: Controlled Oral Word Association (FAS), Controlled Oral Association: Categories

Unavailable data points in the corresponding files are reported as "NA". Scores for participants that did not return for that wave are entered as "xx". 

Note: Construct numbering skips "3", as  a seventh construct (Executive Function) was included in the original design of the DLBS, but was dropped from the study due to persistent data collection and task availability issues. The partial data for this construct are not included in this dataset, but can be made available upon request.
	
## Surveys (/derivatives/surveys)

At each wave of data collection, a battery of surveys was administered either in-person following cognitive testing (Wave 1) or online typically within a week of cognitive testing (Waves 2-3). These surveys cover various domains related to physical health, mental health, and psychosocial aspects:

1) *Physical Health*: Fitness Survey, SF-36, Blood Pressure, NIH Toolbox Motor Assessment
	
2) *Mental Health*: Geriatric Depression Scale, Center for Epidemiological Studies-Depression (CESD), Alzheimer’s Disease Assessment Scale—Cognitive Subscale (ADAS-Cog)

3) *Psychosocial*: Martin and Park Environmental Demands (MPED) Questionnaire, Daily Activities Questionnaire, Lifetime Cognitive Activities, Need for Cognition Survey (NFC), Metamemory in Adulthood (MIA) Questionnaire, Self-Concept Clarity (SCC) Survey, Satisfaction with Life Scale, Revised Neuroticism-Extraversion-Openness Personality Inventory (NEO-PI-R), Big 5 Inventory, Personality Survey, NIH Toolbox Emotion Measures, Scale of Positive and Negative Experience (SPANE), Psychological Well-being (SWQ) 

Unavailable data points in the corresponding files are reported as "NA". Scores for participants that did not return for that wave are entered as "xx". 


## MRI Protocol

The MRI protocol included the following scans (duration in parentheses):

1. fMRI "Ventral Visual Task" (6min, 58 s) x 2 runs for each wave
2. fMRI "Words Task" (7 min, 56 s)
3. fMRI "Scenes Task" (5 min, 56 s) x 3 runs for each wave
4. MPRAGE (3 min, 57 s)
5. Resting-State (6min, 14 s), later scans included 2 runs for each wave
6. FLAIR (3 min, 40s) switched from 24 to 64 slices on 3/30/2009
7. Diffusion Tensor Imaging (3 min, 37 s)
8. Arterial Spin Labeling (ASL, 4 min, 13 s)
9. Hypercapnia (10 min)


## MRI Notes

Five dummy volumes were discarded at scan time to allow for T1 stabilization. Some participants had a second MPRAGE or FLAIR acquisition due to motion artifacts in their first scan. Participants with second MPRAGE run by wave: wave 1 (752), wave 2 (4868), wave 3 (208, 503, 557, 684, 688, 772, 1093, 1245)
Participants with second FLAIR run by wave: wave 1 (754, 779, 2872, 3693), wave 2 (196, 200, 2127, 2133), wave 3 (12, 934, 988, 1767, 2872, 4315)


## Structural MRI volume estimates (/derivatives/brainsummary)

For all three waves of DLBS data collection, MRI scans were processed cross-sectionally through Freesurfer ver. 5.3 with regional parcellations based on the Desikan-Killiany atlas. Extensively trained operators inspected the reconstructed white and grey matter surfaces and performed manual edits when necessary. Quality control was extensive. A second highly trained and independent group reviewed each parcellation for accuracy. A wide variety of derived Freesurfer measures, such as cortical thickness and grey matter volume, are included in the derivatives directory. The complete Freesufer-processed scan data, which are produced after running recon-all, are not included here and only extracted values are provided.


## Functional MRI Tasks

*fMRI "Scenes Task"*: This task was designed to assess how fMRI BOLD signal during the incidental encoding of images relates to performance in a subsequent recognition test. Using an incidental encoding paradigm, participants were presented with outdoor landscape scenes and had to determine whether there was water present in each scene by pressing a yes or no button. The task consisted of three runs with a random ordering of 96 outdoor scenes (32 scenes per run) presented in an event-related design. Each image was presented for 3s, and intertrial intervals (ITIs) were jittered between 4 and 14 seconds. The total scan time was 20 minutes. Responses were recorded using a fiber-optic button box held in the right hand. Visual stimuli were presented using E-prime software (Psychology Software Tools, Pittsburgh, PA, USA), projected through the back of the scanner and viewed though a mirror attached to the head coil. Approximately twenty minutes after the encoding trial, an off-line recognition task was administered outside the scanner. A total of 192 pictures were presented, consisting of 96 previously encountered target scenes and 96 closely matched lure scenes. Participants were instructed to make one of three judgements: 1. “high confidence remember” to indicate confident recognition of the exact picture ; 2) “low confidence remember” to indicate recognition with low confidence; 3. “new item” to indicate that the picture was not previously presented. The recognition task allowed participants to respond at their own pace with a maximum response time of 4 seconds for each trial. 

*fMRI "Words Task"*: This block-design fMRI task was designed to investigate patterns of fMRI BOLD signal during semantic judgements with varying levels of difficulty and ambiguity. Participants were shown a total of 128 words, presented in blocks of “easy” (64 non-ambiguous) and “hard” (64 ambiguous) trials based on the living/nonliving classification of the words. Upon viewing each word, participants made a living or non-living semantic judgment by pressing a button with their right index for “yes” (living) or middle finger for “no” (not living). Examples of easy (non-ambiguous) words included “walrus”, “truck”, “asphalt”. Examples of hard (ambiguous) words included “speaker”, “virus”, and “sponge”. The hard words were intended to elicit increased neural processing and greater task demands. During each stimulus presentation, the semantic decision response time was recorded. Each word was displayed for 2500 ms followed by a 500 ms fixation period (crosshair). The experiment consisted of 8 blocks of easy items and 8 blocks of hard items, with 8 words per block. The easy and hard blocks were presented in a pseudo-randomized order, and the words were randomly distributed among the easy and hard condition blocks. Additionally, three 24-second fixation blocks were included as a baseline for the scan session. The total scan time was 7.7 min, including a 6-second fixation interval before the first stimulus block. Responses were recorded using a fiber-optic button box held in the right hand. Visual stimuli were presented using E-prime software (Psychology Software Tools, Pittsburgh, PA, USA) projected through the back of the scanner and viewed through a mirror attached to the head coil. Subject-specific stimulus onset files for the words task were found to be inaccurate. Therefore, a study-wide stimulus onset asynchrony file is provided in this project’s root directory (task-Words_run-1_events.tsv). 

*fMRI "Ventral Visual Task" ("vv")*:  This task was designed to assess age-related changes in dedifferentiation, including the neural selectivity to various visual stimuli. Participants viewed images from seven categories: human faces, primate faces, domestic cat faces, wild cat faces, houses, chairs, and phase-scrambled control stimuli. Each category consisted of 64 gray-scale photographs (400 pixels wide x 300 pixels tall), except for domestic and wild cat faces, which had 32 photographs each. The images of human faces were sourced from the face library of Minear and Park (2004) and included individuals of different ages, races, and genders. All face images depicted individuals facing forward with neutral expressions. Animal photographs (primate, domestic and wild cats) were obtained from the internet and cropped to ensure clear visibility of the animal’s face. Similar to human faces, only animal images with front-facing views and neutral expressions were selected. Houses were photographed from various locations across the United States, and photographs of chairs were sourced from furniture websites. To create control stimuli, phase-scrambled images were generated by scrambling the phase information in all the experimental stimuli. This process preserved the spatial frequency information while rendering the visual content meaningless. The images were presented in 16-second blocks, with each image displayed for 2 seconds. A total of 8 blocks were presented for each category, distributed across two separate runs with four blocks per run. The order of the blocks was randomized for each participant to minimize potential order effects. The stimuli were presented using E-prime software (Psychology Software Tools) and displayed through a back-projection system. 

Note: Slice timing info for these three tasks is not currently provided.


## PET data (/derivatives/brainsummary)

*PET-Amyloid Data*: The Dallas Lifespan Brain Study (DLBS), initiated in 2008, employed new in vivo imaging techniques to assess the presence of AD pathology, aiming to investigate the aging process and cognition. One such method scanned for beta-amyloid using the radioligand 18F-AV-45, also known as florbetapir. The corresponding data set includes standardized uptake value ratios (SUVRs) for eight key regions that were averaged to form a global SUVR value as well as PET counts for a wide range of Freesurfer regions. The eight regions included: anterior cingulate, posterior cingulate, lateral prefrontal, orbitofrontal, precuneus, lateral parietal, lateral occipital, and lateral temporal cortices. Whole cerebellum was used as the reference region when computing SUVRs. For further details on the processing of PET measures, please refer to the Key to the Kingdom (KTTK) file for additional processing information for PET measures. 

*PET-Tau Data*: At waves 2 and 3, scanning of tau burden was conducted using the radioligand 18F-AV-1451, also known as flortaucipir. The corresponding data set includes SUVRs for a temporal meta region presented by Jack et al., 2018, which includes inferior temporal, middle temporal, entorhinal, parahippocampus, fusiform, and amygdala. This region was selected due to its sensitivity in detecting tau accumulation in otherwise healthy aging. Whole cerebellum was used as the reference region when computing SUVRs.

Reference: Jack, C.R., Wiste, H.J., Schwarz, C.G., Lowe, V.J., Senjem, M.L., Vemuri, P., . . .Petersen, R.C. (2018). Longitudinal tau PET in ageing and Alzheimer’s disease. Brain, 141, 1517-1528. 


## Additional Information: Participant Eligibility

This sample was recruited in two cohorts, with the second cohort implementing modified inclusion/exclusion criteria to ensure better representation of the Dallas community. At waves 2 & 3, the cutoff score for the Mini-Mental State Examination was lowered to allow for potential cognitive decline to be captured.

Inclusion and Exclusion Criteria for the first cohort of 315 individuals at Wave 1:
Participants had to be right-handed (not ambidextrous), between 20-89 years old at Wave 1, have at least 10th grade education, and be fluent in English. Exclusion criteria included: a Mini-Mental State Examination score below 26, major psychiatric or neurological disorder, chemotherapy in past 5 years, coronary bypass, history of substance abuse, history of central nervous system disease or brain injury, history of disorders of the immune, kidney or liver disorder, low glomerular filtration rate (GFR), history of pulmonary disease (i.e., COPD, asthma), corrected vision poorer than 20/30, insulin-dependent diabetic, taking sedatives/benzodiazepines/anti-psychotics, self-reported caffeine consumption exceeding 6 cups of coffee or 12 soft drinks per day, self-reported cholesterol of at least 200, self-reported blood pressure above 160/90, BMI greater than or equal to 35, recreational drug use in past 6 months, conditions that would contra-indicate MRI.

Inclusion and Exclusion Criteria for the second cohort of 149 individuals at Wave 1 ("new cohort"):
Participants had to be enrolled in the Amyloid PET Scan Study (IRB 072010-219), be right-handed (not ambidextrous), between age 50-89 at Wave 1, fluent in English, have at least a 9th grade education, and have at least 15 years of education if they had a health condition such as hypertension, diabetes, cancer, or heart surgery. Additionally, participants were required to score 50% or greater on an abbreviated form of a standardized test of reading proficiency (i.e., North American Adult Reading Test; NAART). Exclusion criteria for this cohort included: a Mini-Mental State Examination score below 26, having 15 or more years of education if participant had no history of chronic illness (likely “super ager”), a major psychiatric or neurological disorder within the last 3 years, cancer treatment or surgery within the last year, coronary bypass within the last year, history of substance abuse, recreational drug use in past six months, history of central nervous systems disease or brain injury, corrected vision poorer than 20/30, taking sedatives/benzodiazepines/anti-psychotics, unable to undergo MRI scanning due to weight concerns or other contraindications.

Inclusion and Exclusion Criteria for waves 2 and 3:
Participants were required to complete testing procedures for this study within a re-test interval of 2.5-11 years and not have the following exclusionary criteria: a Mini-Mental State Examination score below 22, taking sedatives/benzodiazepines/anti-psychotics, currently undergoing chemotherapy or radiation for cancer, a new history of substance abuse, recreational drug use in the past six months, central nervous system disease or brain injury that would preclude participation in this study, psychiatric or neurological disorder that would preclude participation in this study.

## Methodological references:

For detailed information on the data acquisition methods employed in the DLBS, please refer to the relevant studies cited below, as well as our comprehensive document titled “Keys to the Kingdom”

**Cognitive battery**: McDonough, I.M., Bischof, G.N., Kennedy, K.M., Rodrigue, K.M., Farrell, M.E., & Park, D.C. (2016). Discrepancies between fluid and crystallized ability in healthy adults: a behavioral marker of preclinical Alzheimer's disease. Neurobiology of aging, 46, 68-75. https://doi.org/10.1016/j.neurobiolaging.2016.06.011

**fMRI semantic judgments (words task)**: Kennedy, K.M., Rodrigue, K.M., Bischof, G.N., Hebrank, A.C., Reuter-Lorenz, P.A., Park, D.C., 2015. Age trajectories of functional activation under conditions of low and high processing demands: An adult lifespan fMRI study of the aging brain. NeuroImage 104, 21–34. https://doi.org/10.1016/j.neuroimage.2014.09.056

**fMRI episodic memory (scenes task)**: Chen, X., Farrell, M.E., Rundle, M.M., Chan, M.Y., Moore, W., Wig, G.S., & Park, D.C. (2021). The relationship of functional hippocampal activity, amyloid deposition, and longitudinal memory decline to memory complaints in cognitively healthy older adults. Neurobiology of aging, 105, 318-326. https://doi.org/10.1016/j.neurobiolaging.2021.04.020

**fMRI dedifferentiation (ventral visual task)**: Park, J., Carp, J., Kennedy, K.M., Rodrigue, K.M., Bischof, G.N., Huang, C.M., ... & Park, D.C. (2012). Neural broadening or neural attenuation? Investigating age-related dedifferentiation in the face network in a large lifespan sample. Journal of Neuroscience, 32(6), 2154-2158. https://doi.org/10.1523/JNEUROSCI.4494-11.2012

**Resting-state imaging**: Chan, M.Y., Park, D.C., Savalia, N.K., Petersen, S.E., & Wig, G.S. (2014). Decreased segregation of brain systems across the healthy adult lifespan. Proceedings of the National Academy of Sciences, 111(46), E4997-E5006. https://doi.org/10.1073/pnas.1415122111

**Arterial spin labeling imaging**: Thomas, B.P., Liu, P., Park, D.C., Van Osch, M.J., & Lu, H. (2014). Cerebrovascular reactivity in the brain white matter: magnitude, temporal characteristics, and age effects. Journal of Cerebral Blood Flow & Metabolism, 34(2), 242-247. https://doi.org/10.1038/jcbfm.2013.194

**PET-amyloid imaging**: Rodrigue, K.M., Kennedy, K.M., Devous, M.D., Rieck, J.R., Hebrank, A.C., Diaz-Arrastia, R., ... & Park, D.C. (2012). β-Amyloid burden in healthy aging: regional distribution and cognitive consequences. Neurology, 78(6), 387-395. https://doi.org/10.1212/WNL.0b013e318245d295

Farrell, M.E., Chen, X., Rundle, M.M., Chan, M.Y., Wig, G.S., & Park, D.C. (2018). Regional amyloid accumulation and cognitive decline in initially amyloid-negative adults. Neurology, 91(19), e1809-e1821. https://doi.org/10.1212/WNL.0000000000006469