Forms & Other Documentation

Carotid Ultrasound

Cognitive Function Paper Tests

Scored Survey Forms

Raw Survey Forms

Protocols (with DOIs)

Scored Surveys

View Scoring Protocol

 

Cognitive Function

View Cognitive Function Protocol

 

Imaging (MRI)

 

    1. Anatomic Measures dx.doi.org/10.17504/protocols.io.n92ld5yw9v5b/v1
    2. Perfusion Arterial Spin Labeling dx.doi.org/10.17504/protocols.io.n92ld667xg5b/v1
    3. Diffusion Weighted Imaging dx.doi.org/10.17504/protocols.io.6qpvrkopolmk/v1
    4. White Matter Intensities dx.doi.org/10.17504/protocols.io.n2bvjebwngk5/v1
    5. Rest-State Resting-State Functional Magnetic Resonance Imaging dx.doi.org/10.17504/protocols.io.81wgbnp41gpk/v1

 

Labs & Biomarkers

 

    1. APOE4 genotyping [DOI]: dx.doi.org/10.17504/protocols.io.n92ldzo68v5b/v1
    2. Milliplex Assay [DOI]: dx.doi.org/10.17504/protocols.io.n2bvj8rkbgk5/v1
    3. AD Biomarkers [DOI]: dx.doi.org/10.17504/protocols.io.bp2l6dop5vqe/v1

Citation Policy
Data collection and sharing for the Precision Aging Network (PAN) is funded by the National Institute on Aging (National Institutes of Health Grant “Closing the Gap Between Cognitive Health Span and Human Lifespan” (U19AG065169)).
MindCrowd Data

MindCrowd® is a web-based cognitive testing platform hosted at The Translational Genomics Research Institute (TGen).  Initially funded through TGen and private donors in 2013 it grew to over 800,000 users across the globe.  MindCrowd® is Project 1 of the Precision Aging® Network grant and serves not only as an entryway to the Healthy Minds for Life (Project 2) study but also affords participants opportunities for other ancillary studies.

For purposes of sharing data, we are including complete case data on a set of demographics data (age, sex, race/ethnicity) and two scored cognitive function tests, the paired associates learning test and simple visual reaction time test (svRT or attention test).  We also include other user-provided health information that is included at the time the cognitive function tests were taken (note these variables may contain missing information). These data are complete through December 31, 2025.

A manuscript that describes the history of MindCrowd® and Precision Aging® Network more broadly can be found in a Gerontology editorial (https://doi.org/10.1159/000548488).

This data are included in a single data table as an Excel file through the research portal.

  Available (N=375,720)
Age 375,720 (100%)
Sex 375,720 (100%)
Race/Ethnicity 375,720 (100%)
Handedness 368,167 (98.0%)
Highest Level of Education Completed 374,712 (99.7%)
Attention Test (svRT) 375,289 (99.9%)
Paired Associates Test 375,412 (99.9%)
Number of Daily Medications 367,962 (97.9%)
No Chronic Conditions 287,576 (76.5%)
Depression/Anxiety 287,576 (76.5%)
Chronic Stress 287,576 (76.5%)
Stroke 287,576 (76.5%)
Alzheimers 287,576 (76.5%)
Brain Disease 287,576 (76.5%)
Loss of Consciousness 277812 (76.5%)
Memory Problems 287,576 (76.5%)
Seizures 287,576 (76.5%)
Parkinson’s Disease 287,576 (76.5%)
Sleep Problems 287,576 (76.5%)
Type I Diabetes 287,576 (76.5%)
Type II Diabetes 287,576 (76.5%)
Hypertension 287,576 (76.5%)
Heart Disease 287,576 (76.5%)
Smoking 287,576 (76.5%)
Cancer 287,576 (76.5%)

Data Dictionary

View MindCrowd Data Dictionary

Healthy Minds for Life Data

Healthy Minds for Life (HML) is the in-depth in-person longitudinal cohort project (Project 2) of the Precision Aging® Network. Data are collected from four sites: University of Arizona, Emory University, Johns Hopkins University and the University of Miami.

Many of the data collected in this study were summarized in a manuscript in Frontiers in Aging Neuroscience (https://doi.org/10.3389/fnagi.2019.00128). Additionally, manuscript that describes the history and details of the HML cohort and how it fits into the Precision Aging® Network more broadly we recommend a Gerontology editorial on the topic (https://doi.org/10.1159/000548488).

With the exception of the full MindCrowd® dataset, all data tables described on this website describe data from HML participants.

Healthy Minds for Life Demographics (HML)

There are many health history, medications, demographic and biometrics data that have been collected both online (through MindCrowd) and in at in-person visits (Healthy Minds for Life).  Some demographics (like age) are different from these two demographics tables because of the time between online sign-up to MindCrowd and recruitment into Healthy Minds for Life, we suggest using Healthy Minds for Life demographics (age, sex, etc.) as these are the closest in time to other data collected in person.

  1. Healthy Minds for Life Demographics
  2. Healthy Minds for Life Eligibility
  3. Medications and Health Intake
  4. Mind Crowd Demographics (includes some health history variables)
  5. Biometrics Data

The data now include 115 participants with 2-year follow-up. In order to accommodate these data, a new variable “Visit” has been added to the data tables. Values of this new variable are either a value of 1 representing the baseline assessment, and a value of 2 representing the 2-year follow-up visit.  Additional visit values will appear in future releases as subsequent follow-up visits are completed.  Please see the help files for either the web interface or the relational database for programming notes on how to analyze data with this new variable.

  Baseline
(N=1063)		 2YR Follow-Up
(N=115)
Age1 1063 (100%) –
Sex1 1062 (99.9%) –
Race/Ethnicity1 1063 (100%) –
Handedness1 1055 (99.2%) –
Highest Level of Education Completed1 1063 (100%) –
Attention Test (svRT) 1063 (100%) 115 (100%)
Paired Associates Test 1063 (100%) 115 (100%)
Number of Daily Medications1 1049 (98.7%) –
Health Intake1 1063 (100%) –
MindCrowd Demographics (includes some health history variables)1 1017 (95.7%) –
Medications2 920 (86.5%) 79 (68.7%)
Biometrics Data 1063 (100%) 115 (100%)
Healthy Minds for Life Eligibility1 1063 (100%) –

1 Only collected at baseline.
2Medications are captured in Health Intake and Health & Medical survey tables until July 30, 2025. After July 30, 2025 complete medications are only captured in the Health & Medical survey table.

These data can be found in the HML Demographics, HML Eligibility, Health Intake, MindCrowd Demographics, and Biometrics

Data Dictionary

View the HML Demographics Data Dictionary

Note: The Health Intake data table contains information about medications. We created a variable “dsc_meds_label_1” which is a derived variable using “hif_meds_1” variable and “hif_meds_oth_1” variable that should be used when the “meds_label_1” is missing.

Carotid Ultrasound

Each of the 4 sites was equipped with a Terason uSmart 3300 NexGen Ultrasound system with a 15L4A MHz uSmart Linear Array Transducer or alternative ultrasound systems with a 10-15MHz linear array transducer and a pre-set vascular protocol.

The mean and maximum intima-media thickness (IMT) of the near and the far walls of both right and left common carotid, carotid bifurcation, and internal carotid arteries are measured in the plaque-free segments.  Doppler was used for measuring blood flow velocities in examination performed on the right and left common carotid, (CCA) internal (ICA) and external carotid arteries (ECA) for quick assessment of general extracranial hemodynamic status.

These data are located in the Carotid Ultrasound data table

  Available (N=1063)
Velocity Data 546 (51.36%)
Intima-Media Thickness (IMT) Data 363 (34.15%)

Standard Operating Procedures

Data Dictionary

View Carotid Ultrasound Data Dictionary

Cognitive Function

There are 4 paper scored tests, and 7 computer captured cognitive tests.  The data now include 115 participants with 2-year follow-up. In order to accommodate these data, a new variable “Visit” has been added to the data tables. Values of this new variable are either a value of 1 representing the baseline assessment, and a value of 2 representing the 2-year follow-up visit.  Additional visit values will appear in future releases as subsequent follow-up visits are completed.  Please see the help files for either the web interface or the relational database for programming notes on how to analyze data with this new variable.

Paper Tests

  1. Rey Auditory Verbal Learning Test (AVLT)
  2. Montreal Cognitive Assessment (MoCA)
  3. North American Adult Reading Test (NAART) Instructions
  4. North American Adult Reading Test (NAART) Answer Key
  5. Patient Health Questionnaire-9 (PHQ-9)

MindCrowd Memory and Attention Tests

  1. Paired Associates Test (summarized as total correct and age/race/education adjusted quartiles)
  2. Attention Test (summarized as median reaction time)

Scored Online Tests

  1. Verbal Paired Associate Learning (associative memory)
  2. Simple and Choice Reaction Time (information processing speed)

Three are executive functioning tasks

  1. Keep Track Task (updating/working memory)
  2. Flanker Task (inhibition)
  3. Letter Number Task (set shifting)
[Protocol: dx.doi.org/10.17504/protocols.io.14egnyqw6v5d/v1]

Participant level results from these tests can be found in data tables named by test names except for the Paired Associates and Attention Test summaries which can be found in the MindCrowd Demographics data table.  These tests will be available as raw data in the April 2026 data release.

  Baseline
(N=1063)		 2 YR Follow-Up
(N=115)
AVLT 1062 (99.91%) 115 (100%)
MoCA 1063 (100%) 115 (100%)
NAART1 1063 (100%) –
PHQ-9* 1061 (99.81%) 115 (100%)
Attention 1063 (100%) 115 (100%)
Flanker Task 898 (84.5%) 102 (88.7%)
Keep Track Task 813 (76.5%) 98 (85.2%)
Paired Associates 1063 (100%) 115 (100%)
Simple and Choice Reaction Time 908 (85.4%) 104 (90.4%)
Letter Number Task 828 (77.9%) 90 (78.3%)
Verbal Paired Associates Learning 859 (80.8%) 101 (87.8%)
*PHQ-9 is a depression screening tool
1Only collected at baseline

Data Dictionary

View Cognitive Function Data Dictionary

Imaging (MRI)

There are 7 data tables with processed MRI data.

  1. Anatomic Measures (using freesurfer)
  2. Perfusion Arterial Spin Labeling
  3. Diffusion Weighted Imaging
  4. Resting-State Functional Magnetic Resonance Imaging Cortical-Cerebellar
  5. Resting-State Functional Magnetic Resonance Imaging Whole Brain
  6. Resting-State Functional Magnetic Resonance Imaging Intra-Regional
  7. White Matter Hyperintensities
  Available (N=1063)
Anatomic Measures 700 (65.9%)
Perfusion Arterial Spin Labeling 671 (63.1%)
Diffusion Weighted Imaging 671 (63.1%)
Resting-State Functional Magnetic Resonance Imaging
Cortical-Cerebellar		 672 (63.2%)
Resting-State Functional Magnetic Resonance Imaging Whole
Brain		 672 (63.2%)
Resting-State Functional Magnetic Resonance Imaging
Intra-Regional		 672 (63.2%)
White Matter Hyperintensities 692 (65.1%)

The data tables for these data can be found here (in order): MRI Anatomic Measures; MRI Perfusion ASL; MRI Diffusion Imaging; Resting State fMRI corticocereb; Resting State fMRI wholebrain; Resting State fMRI; and White Matter Hyperintensities

Data Dictionary

View the Imaging (MRI) Data Dictionary

Protocols

There are 5 protocols associated with these data

  1. Anatomic Measures dx.doi.org/10.17504/protocols.io.n92ld5yw9v5b/v1
  2. Perfusion Arterial Spin Labeling dx.doi.org/10.17504/protocols.io.n92ld667xg5b/v1
  3. Diffusion Weighted Imaging dx.doi.org/10.17504/protocols.io.6qpvrkopolmk/v1
  4. White Matter Intensities dx.doi.org/10.17504/protocols.io.n2bvjebwngk5/v1
  5. Rest-State Resting-State Functional Magnetic Resonance Imaging dx.doi.org/10.17504/protocols.io.81wgbnp41gpk/v1
Labs & Biomarkers
  Baseline
(N=1063)		 2YR Follow-Up
(N=115)
AD Plasma Biomarkers 868 (81.66%) 31 (26.96%)
APOE4 Status1 1022 (96.14%) –
Cortisol1 855 (80.43%) –
Epigenetic Clocks 965 (90.78%) 103 (89.57%)
Millipore Cytokine Assay 1027 (96.61%) 104 (90.43%)
NULISA Seq CNS Disease Panel1 569 (53.53%) –
Polygenic Risk Scores1 1022 (96.14%) –
Sonora Quest Blood Chemistries2 1051 (98.87%) 114 (99.13%)
1Only collected at baseline.
2Blood chemistry values from samples collected after October 1, 2025 reflect a non-fasting state. Data collected prior to this date were obtained under fasting conditions.

Currently we have 7 data tables containing labs and biomarkers:

  1. DNA methylation: Horvath, Altum, and DunedinPACE clocks
  2. TaqMan APOE4 genotyping for SNPs rs7412 and rs429358
  3. Polygenetic Risk Scores
  4. MILLIPLEX® Human Cytokine/Chemokine/Growth Factor Panel A – Immunology Multiplex Assay
  5. AD Biomarkers (Simoa Neuro 4-Plex E and Human pTau-181 Assays)
  6. Blood Chemistry (Sonora Quest Laboratory)
  7. NULISAseq Disease Panel 1

The data now include 115 participants with 2-year follow-up. In order to accommodate these data, a new variable “Visit” has been added to the data tables. Values of this new variable are either a value of 1 representing the baseline assessment, and a value of 2 representing the 2-year follow-up visit.  Additional visit values will appear in future releases as subsequent follow-up visits are completed.  Please see the help files for either the web interface or the relational database for programming notes on how to analyze data with this new variable.

Data tables include Epigenetic Clocks, APOE4, Polygenetic Risk Scores, Millipore Cytokine Assay, AD Plasma Biomarkers, Sonora Quest Blood Chemistries, and NULISAseq

Data Dictionary

View the Labs & Biomarkers Data Dictionary

Protocols

There are 3 protocols associated with these assays

  1. APOE4 genotyping [DOI]: dx.doi.org/10.17504/protocols.io.n92ldzo68v5b/v1
  2. Milliplex Assay [DOI]: dx.doi.org/10.17504/protocols.io.n2bvj8rkbgk5/v1
  3. AD Biomarkers [DOI]: dx.doi.org/10.17504/protocols.io.bp2l6dop5vqe/v1
Non-Invasive Technologies

These data are captured using non-invasive technology (sound, video, sweat and wearable devices). The processing of these data typically involves machine learning and novel techniques.

There are 4 categories of data

  1. Cortisol (palmer sweat Direct Analysis in Real Time Mass Spectrometry (DART-MS))
  2. Sensor/Wearable (Garmin)
  3. Frailty (Upper Extremity Function)
    These features are extracted from video, variables include Single and Dual Task Frailty (ST_FI and DT_FI, respectively) described here: https://doi.org/10.1159/000546227
  4. Language and Speech Task (LIWC-22)
    These features are extracted using LIWC-22 from transcripts of the two speech tasks. Transcripts were generated using Open AI’s Whisper, Model size medium. Files with a word count of below 100 were removed. In the Aging Well Task (speech task 1), transcripts were automatically processed to replace “aging well” with “aging-well” to avoid bias. (SOP attached below)
  5. Language and Speech Tasks (GeMAPS feature set)
    These features are extracted from the raw wav-files of the two speech tasks. Information about the openSMILE toolbox can be found here: https://audeering.github.io/opensmile/about.htmlMeans and Standard Deviation are calculated as: Arithmetic mean and coefficient of variation (standard deviation normalized by the arithmetic mean), information about how these are derived can be found here: 10.1109/TAFFC.2015.2457417
  Available (N=1063)
Language Speech Task (LIWC-22) 1039 (97.74%)
Language Speech Task (GeMAPS) 1041 (97.93%)
Garmin 576 (54.19%)
UEF Frailty 726 (68.3%)
Cortisol 847 (79.68%)

These data are found in the Cortisol, LabFront Garmin, UEF Frailty, and Language Speech Task data tables LIWC-22 and GeMAPS

Data Dictionary

View Wearables Data Dictionary

Surveys

Scored Surveys

There are a total of 15 surveys, with 7 scored instruments.  The data now include 115 participants with 2-year follow-up. In order to accommodate these data, a new variable “Visit” has been added to the data tables. Values of this new variable are either a value of 1 representing the baseline assessment, and a value of 2 representing the 2-year follow-up visit.  Additional visit values will appear in future releases as subsequent follow-up visits are completed.  Please see the help files for either the web interface or the relational database for programming notes on how to analyze data with this new variable.

  Baseline
(N=1063)		 2YR Follow-Up
(N=115)
Anxiety 884 (83.2%) 79 (68.7%)
Stress 896 (84.3%) 80 (69.6%)
QPAR 875 (82.3%) 78 (67.8%)
Sleep 902 (84.9%) 80 (69.6%)
Social Stressors 868 (81.7%) 78 (67.8%)
Social Support 885 (83.3%) 80 (69.6%)
SWLS 880 (82.8%) 77 (67%)

Data collection forms for the scored surveys are included in the Scoring Protocol.

Scored Survey Forms

[Scoring Protocol here DOI: dx.doi.org/10.17504/protocols.io.q26g7nwy9lwz/v1]

These surveys are included in the HML Scored Survey data table and raw data for these surveys can be found in the data tables named like: Perceived Stress Raw Data

The visit variable is present in all 7 of the survey raw data tables and the HML Scored Surveys data table.

Data Dictionary (Scored Surveys)

View the Scored Surveys Data Dictionary

Remaining surveys are participant reported and can be found in the data tables with the following names:

  • Socioeconomic Survey
  • Subjective English Survey
  • Health Medical Survey
  • Family History Survey
  • Diet Survey
  • Covid Survey
  • Brain Disease Survey
  • ADL Survey

The visit variable is present in the ADL Survey, Diet Survey, and Health Medical Survey data tables but not the Brain Disease Survey, Covid Survey, Family History Survey, Subjective English Survey or Socioeconomic Survey data tables.

Raw Surveys

  Baseline
(N=1063)		 2YR Follow-Up
(N=115)
ADL 890 (83.7%) 77 (67%)
Anxiety 884 (83.2%) 79 (68.7%)
Brain Disease1 908 (85.4%) –
Covid1 915 (86.1%) –
Diet 901 (84.8%) 81 (70.4%)
FHAD1 910 (85.6%) –
Health Medical 920 (86.5%) 79 (68.7%)
QPAR 875 (82.3%) 78 (67.8)
SES1 912 (85.8%) –
SWLS 880 (82.8%) 77 (67%)
Sleep 902 (84.9%) 80 (69.6%)
Social Stressors 868 (81.7%) 78 (67.8%)
Social Support 885 (83.3%) 80 (69.6%)
Stress 896 (84.3%) 80 (69.6%)
Subjective English1 908 (85.4%) –

Data Dictionary (Raw Surveys)

View the Raw Surveys Data Dictionary

Survey Forms (Raw Data)

Forms & Other Documentation

Carotid Ultrasound

Cognitive Function Paper Tests

Scored Survey Forms

Raw Survey Forms

Protocols (with DOIs)

Scored Surveys

View Scoring Protocol

 

Cognitive Function

View Cognitive Function Protocol

 

Imaging (MRI)

 

    1. Anatomic Measures dx.doi.org/10.17504/protocols.io.n92ld5yw9v5b/v1
    2. Perfusion Arterial Spin Labeling dx.doi.org/10.17504/protocols.io.n92ld667xg5b/v1
    3. Diffusion Weighted Imaging dx.doi.org/10.17504/protocols.io.6qpvrkopolmk/v1
    4. White Matter Intensities dx.doi.org/10.17504/protocols.io.n2bvjebwngk5/v1
    5. Rest-State Resting-State Functional Magnetic Resonance Imaging dx.doi.org/10.17504/protocols.io.81wgbnp41gpk/v1

 

Labs & Biomarkers

 

    1. APOE4 genotyping [DOI]: dx.doi.org/10.17504/protocols.io.n92ldzo68v5b/v1
    2. Milliplex Assay [DOI]: dx.doi.org/10.17504/protocols.io.n2bvj8rkbgk5/v1
    3. AD Biomarkers [DOI]: dx.doi.org/10.17504/protocols.io.bp2l6dop5vqe/v1