Research avenues that don't exist yet

CDC / National Center for Chronic Disease Prevention and Health Promotion

Annual telephone survey of health-related risk behaviors, chronic conditions (including a cognitive decline/caregiver module), and preventive service use.

• Geography: National, state, and (via SMART) select city/county
• Years: 1984-present (cognitive decline module varies by state/year)

Why it would help: State-level cognitive-decline and caregiver-burden self-report data, and the underlying source for most of the pct_diabetes/obesity/smoking/etc columns in panel_state.csv.

cdc.gov

CDC / NCHS

Ad-hoc mortality query system; underlying/multiple cause-of-death counts and rates by ICD-10 code (G30 = Alzheimer's), year, age, sex, race, state, county.

• Geography: National, state, county
• Years: 1968-present (detailed cause-of-death 1999-present)

Why it would help: The authoritative source for filling the mortality gaps left blank in timeseries_national.csv, and for cross-checking the Facts & Figures state mortality table.

wonder.cdc.gov

CDC / NCHS

Combined interview and physical exam survey with labs, biospecimens, and (in later cycles) cognitive testing modules.

• Geography: National only (not state-representative)
• Years: 1960s (initial cycles), continuous since 1999; most recent public releases through 2021-2023

Why it would help: Biomarker/lab and cognitive-screen data for national-level covariates (e.g., cardiometabolic risk factors) against national AD trends.

cdc.gov

Johns Hopkins Bloomberg School of Public Health, funded by NIA

Annual in-person interviews of Medicare beneficiaries 65+ on disability, cognitive impairment, dementia status, and late-life trajectories, linkable to Medicare claims.

• Geography: Nationally representative sample (respondent-level, not state-representative)
• Years: 2011-present (annual rounds)

Why it would help: Individual-level dementia/disability trajectories linkable to Medicare claims -- strong for cost/utilization correlation work.

nhats.org

Multi-site consortium, coordinated via LONI/USC, funded by NIA

Longitudinal MRI/PET imaging, biomarkers (CSF, blood, genetics), and clinical/cognitive data from ~2,500+ participants across the AD spectrum.

• Geography: Site-level, US and Canada (~60 sites), not population-representative
• Years: 2004-present (ADNI 1/GO/2/3/4 phases)

Why it would help: Gold-standard imaging/biomarker cohort for validating disease-stage definitions used to interpret national prevalence estimates.

adni.loni.usc.edu

University of Washington, funded by NIA, aggregating NIA-funded Alzheimer's Disease Research Centers (ADRCs)

Centralized clinical, neuropathology, imaging, and biomarker data pooled from 37+ active ADRCs (56,000+ participants, 217,000+ clinical visits).

• Geography: Multi-site (25 states with active ADRCs), not population-representative
• Years: ~1984-present (UDS harmonized data primarily 2005-present)

Why it would help: Clinically-adjudicated dementia diagnoses across research centers -- useful to sanity-check self-report/administrative prevalence estimates.

naccdata.org

CMS, technical assistance via University of Minnesota ResDAC

Access point/documentation for Medicare (Parts A/B/C/D) and Medicaid claims, enrollment, and assessment (e.g., MDS) research-identifiable files.

• Geography: Beneficiary/claim level, aggregable to county/state/national
• Years: Varies by file type; Medicare claims generally 1999-present

Why it would help: The actual source needed to split Medicare-only vs. Medicaid-only Alzheimer's spending, which the published Facts & Figures aggregates do not cleanly separate (see caveats on the *_spend_total_usd columns).

resdac.org

Institute for Health Metrics and Evaluation, University of Washington

Modeled estimates of deaths, DALYs, prevalence, and risk-factor attribution for Alzheimer's/dementia by cause.

• Geography: National and subnational (204 countries, 660 subnational locations)
• Years: 1990-2023 (GBD 2023 round)

Why it would help: Enables international/cross-national benchmarking of the US Alzheimer's burden trend against comparable modeled global estimates.

vizhub.healthdata.org

US Census Bureau

Annual demographic, economic, and housing survey (ACS) plus intercensal population estimates by age; source for the 65+ denominator and most of the demographic/socioeconomic panel columns.

• Geography: National, state, county, tract (ACS 1-yr/5-yr)
• Years: ACS ongoing since 2005; Population Estimates Program annual since 2000s

Why it would help: Provides the population-65+ denominator required to compute per-capita AD prevalence/mortality rates by state/year, plus income, education, race, and poverty covariates.

census.gov

University of Wisconsin Population Health Institute / Robert Wood Johnson Foundation

County- and state-level composite health outcome and health-factor measures (mortality, morbidity, social determinants).

• Geography: County and state
• Years: 2010-present annual releases

Why it would help: County-level social determinants (income, education, access to care) to correlate against local Alzheimer's mortality/prevalence at finer geography than state.

countyhealthrankings.org

HRSA, Bureau of Health Workforce / National Center for Health Workforce Analysis

6,000+ variables on health care professions, facilities, population characteristics, and health economics.

• Geography: County, state, national
• Years: Multi-decade historical files; dashboard covers 2019-2024

Why it would help: County-level physician/geriatrician and long-term-care workforce supply data to correlate against the dementia_care_workforce_requirement policy variable and healthcare-access columns.

data.hrsa.gov

Alzheimer's Association

Annual state-by-state Alzheimer's prevalence, projected growth, deaths, caregiver, geriatrician-supply, Medicare/Medicaid spending, and workforce estimates.

• Geography: State (all 50 + DC)
• Years: Annual editions, at least 2010-2026 confirmed

Why it would help: The primary published state-level dataset underlying most of panel_state.csv's Alzheimer's-specific columns and timeseries_national.csv.

alz.org

NIH Office of Extramural Research

Searchable database of NIH-funded grants/contracts, publications, and patents, including by RCDC disease category (Alzheimer's Disease, and Alzheimer's Disease and Related Dementias combined).

• Geography: Project/institution level, aggregable to state/national
• Years: FY1985-present (ExPORTER bulk files); categorical spending back to FY2008, FY2019-2025 fetched directly for this project

How this is linked: data/timeseries_national.csv column research_funding_nih_adrd_usd. data/datasets.json's own entry for this source names research_funding_nih_adrd_usd as its direct output, but neither “RePORTER” nor “RCDC” appears inside timeseries_national.csv itself, so the name scan above misses it.

Why it would help: Grant-level Alzheimer's research funding data; the direct source for the research_funding_nih_adrd_usd column in timeseries_national.csv.

report.nih.gov

NIH National Library of Medicine

Registered/results data for all US and international clinical trials, including condition, phase, status, enrollment, sponsor, location, and outcomes.

• Geography: Trial-site level, global (includes US state/city)
• Years: 1999-present (registry inception); ongoing

How this is linked: data/trials.csv column nct_id. data/trials.csv is this site's own catalogue entry, labelled “ClinicalTrials.gov API v2 pull”, and every one of its rows is keyed by an NCT trial ID from that API -- but the word “ClinicalTrials” never appears inside the file itself, so the name scan above misses it too.

Why it would help: Tracks the AD drug-development pipeline (trial counts/phases by year) as a leading indicator alongside the fda_approved_alz_drugs_cumulative column.

clinicaltrials.gov

US Environmental Protection Agency

Ambient air quality monitoring data including PM2.5 annual averages by monitor, county, and state (CBSA).

• Geography: Monitor, county, state, national
• Years: 1980-present (PM2.5 monitoring network established ~1999)

Why it would help: Raw monitor-level and county-level PM2.5 data would allow a finer-grained air-pollution correlate than the single state-average column included in panel_state.csv.

epa.gov

KFF (Kaiser Family Foundation)

Aggregated state-by-state health policy indicators (insurance coverage, chronic disease prevalence, provider supply, hospital/nursing-home capacity), sourced from CDC BRFSS, ACS, AHA Annual Survey, and CMS.

• Geography: State and national
• Years: Varies by indicator, generally 2019-2024

Why it would help: Single consistent access point for many of the risk-factor and healthcare-capacity columns already included in the panel; useful for re-pulling fresher vintages later.

kff.org

CMS (Centers for Medicare & Medicaid Services)

Medicare-certified hospice facility directory with CAHPS Hospice Survey and clinical quality measures, one row per facility x measure (453,493 rows nationally as fetched).

• Geography: Facility-level (CCN), aggregated here to county via facility address
• Years: 2026 (August 2026 snapshot)

Why it would help: Direct source for panel_county.csv's cms_hospice_facilities_count and cms_hospice_facilities_per_1000pop columns.

data.cms.gov

CMS (Centers for Medicare & Medicaid Services)

Medicare-certified home health agency directory (~12,460 agencies), including services offered, star ratings, and outcome measures.

• Geography: Agency-level (ZIP code), mapped here to county via a Census ZCTA-to-county crosswalk
• Years: 2026 (July 2026 snapshot)

Why it would help: Direct source for panel_county.csv's cms_home_health_agencies_count and cms_home_health_agencies_per_1000pop columns.

data.cms.gov

Longitudinal MRI/PET imaging, CSF/blood biomarkers, genetics, and clinical/cognitive data on ADNI participants (also gateway for AIBL, ADNIDOD, and other LONI-hosted studies).

Steps a human would take:

1. 1. Create an IDA account at https://ida.loni.usc.edu/explore/jsp/register/register.jsp (requires email verification through a Google reCAPTCHA challenge).
2. 2. Log in and open https://ida.loni.usc.edu/collaboration/access/appApply.jsp?project=ADNI
3. 3. Read and click 'I Agree' to the ADNI Data Use Agreement (no personal data entered at this step - it is a click-through acknowledgment of ~9 terms plus publication rules).
4. 4. Complete the online application: verify institutional affiliation with a scientific/educational institution, and describe the proposed research use of ADNI data.
5. 5. Wait for review by the ADNI Data Sharing and Publications Committee (DPC) - generally within two weeks.
6. 6. On approval, receive IDA login credentials by email and download data from the LONI Image and Data Archive.

Why no form was submitted: Step 1 (account creation) is gated by a reCAPTCHA challenge, which the binding rules forbid defeating. Independent of that, step 4 requires a true institutional affiliation and a real research description - facts this session does not have and is not authorized to invent.

ida.loni.usc.edu · checked 2026-09-28

Longitudinal UDS clinical/cognitive data, SCAN MRI/PET imaging and derived summaries, and neuropathology data pooled from 29+ NIA-funded Alzheimer's Disease Research Centers.

Steps a human would take:

1. 1. Search existing NACC-based publications to confirm the proposed research hypothesis doesn't overlap with an existing project (researcher's own responsibility, not NACC's).
2. 2. Open the request survey at https://nacc.redcap.rit.uw.edu/surveys/?s=KHNPKLJW8TKAD4DA
3. 3. Check the consent box and click 'Begin survey' to proceed into a multi-page REDCap questionnaire about the project (exact downstream questions are dynamically generated by REDCap after the consent step and were not reachable by a plain page fetch).
4. 4. Accept the NACC Data Use Agreement.
5. 5. NACC responds within 3 business days to acknowledge the request and ask clarifying questions if needed.
6. 6. On approval, receive a username, password, and link to download custom data files.

Why no form was submitted: The gate page itself (a single consent checkbox) has no personal-data fields and could technically be clicked through, but doing so commits to a multi-page REDCap survey asking for 'a unique research hypothesis' and project description - content that has to come from Alex, not invented. Submitting a hypothesis fabricated by this session would violate the no-fabrication rule and waste NACC staff time reviewing a project that doesn't exist.

naccdata.org · checked 2026-09-28

GWAS summary statistics, brain eQTL/pQTL, and network-analysis products derived from ROSMAP, Mayo, MSBB, Mount Sinai and other AD cohorts.

Steps a human would take:

1. 1. Create a free Synapse account.
2. 2. Accept the single 'ADKP Clickwrap' Terms-of-Use covering the ~190-entity ROSMAP/AMP-AD tree (no institutional review, no DUA application, no fee).
3. 3. Authenticate and download named files, e.g. ROSMAP_GWAS_cerebral_atherosclerosis_summary_statistics.txt (syn25998845).

Why no form was submitted: Verified in the 2026-09-27 session (see data/sources/ampad_synapse/README.md): the folder structure and access-requirement type were confirmed via the live Synapse API (HTTP 200 on entity/accessRequirement calls), but actually downloading a file still requires an authenticated Synapse session (HTTP 403 without one) - account creation was out of scope for that session and remains so here, since it is a registration action, not an information-only fetch.

adknowledgeportal.synapse.org · checked 2026-09-27 (prior session; re-cited, not re-fetched, this session)

Full individual-level clinical, cognitive, pathology, and biospecimen data from the Religious Orders Study and Memory and Aging Project - the underlying raw cohort behind the AMP-AD-hosted GWAS/QTL summary statistics above.

Steps a human would take:

1. 1. Create an account on the RADC Research Resource Sharing Hub at https://www.radc.rush.edu/.
2. 2. Review the RADC Data and Resource Sharing Policy.
3. 3. Submit a data/specimen request describing the proposed research.
4. 4. Review by RADC's Biospecimen Distribution Committee on scientific merit, feasibility, investigator capability, and funding source.
5. 5. On approval, execute a Biologic Material Transfer Agreement / Letter of Agreement before data or specimens are released.

Why no form was submitted: Requires a real funded research proposal and institutional details reviewed by a committee - not something this session can honestly provide.

radc.rush.edu · checked 2026-09-28

Access: MANAGED (participant-level data); OPEN (data dictionary)

Deep genetic, imaging, and health-record phenotyping (incl. precomputed Alzheimer's polygenic risk scores, fields 26206/26207) on ~500,000 UK adults.

Steps a human would take:

1. 1. Register on the UK Biobank Access Management System.
2. 2. Submit an application describing the research, with institutional backing and an access fee.
3. 3. Await Access Sub-Committee review.

Why no form was submitted: The apply-for-access page returned HTTP 403 with a live Cloudflare bot challenge (cf-mitigated: challenge) to this session's automated request - per the binding rules, a challenge is not a puzzle to defeat. The data dictionary itself (OPEN, no login) was already fetched.

ukbiobank.ac.uk · checked 2026-09-27 (prior session; re-cited, not re-fetched, this session)

Cross-sectional and longitudinal structural MRI, PET amyloid/tau imaging, and clinical data across four release generations (416 to 1,378 subjects per release).

Steps a human would take:

1. 1. Register a NITRC account at nitrc.org/account/register.php.
2. 2. Apply for the specific OASIS project in NITRC-IR.
3. 3. Read and accept the OASIS Data Use Agreement (Data-Use-Agreement_July2025.pdf).
4. 4. Wait for OASIS staff to review and grant the NITRC-IR project application.

Why no form was submitted: Requires account registration plus a staff-reviewed application - a genuine gate, not a submit-and-download form. Notably, this corrects an initial assumption that OASIS-1 specifically was still openly downloadable; it is not.

sites.wustl.edu · checked 2026-09-27 (prior session; re-cited, not re-fetched, this session)

PET amyloid imaging, MRI, cognitive/psychometric data, and blood biomarkers on an Australian AD cohort partnered with ADNI.

Steps a human would take:

1. 1. Download the current EOI form: https://aibl.org.au/wp-content/uploads/2026/08/AIBL-EOI-v2026-07.docx
2. 2. Complete it with: legal name of the applying entity, business address, ABN (if an Australian organisation), principal contact name/phone/email, description and outline of the proposed project, which AIBL data domain is needed (Diagnostics & Biomarkers / Psychometrics / Neuroimaging / Lifestyle Interventions), how results will be used/published, other institutions involved, any cash/in-kind contributions sought, IP considerations, a detailed CV of the Principal Investigator, and which institution(s) will provide ethics approval.
3. 3. Sign as 'director/duly authorised representative'.
4. 4. Email or post to Christopher Fowler, AIBL Study Manager, christopher.fowler@florey.edu.au.
5. 5. Review and approval by the AIBL Scientific Management Committee and Business Development and Administration Committee.
6. 6. If approved, execute a separate legal contract (EOI Agreement / Project Agreement) before any data or samples are released.

Why no form was submitted: This is a legal/commercial EOI form (it asks for a legal entity name, ABN, signature of an authorised representative, and IP terms), not a researcher self-service form. It requires facts about Alex's institution/entity, a signed CV, and ethics-approval details that only Alex can honestly supply.

aibl.org.au · checked 2026-09-28

CSF/blood biomarkers, PET amyloid/tau imaging, MRI, and cognitive/neurologic/psychiatric assessments in a longitudinal AD/PD cohort.

Steps a human would take:

1. 1. There is no online application form. Access is 'through collaboration with PI only' per the study's own description (local/regional/national/international/industry/pilot-study tiers, all PI-mediated).
2. 2. Identify and contact a current member of the BioFINDER executive group (Oskar Hansson, Erik Stomrud, Niklas Mattsson-Carlgren, Sebastian Palmqvist, or Rik Ossenkoppele) directly, e.g. via their Lund University research-portal profile, to propose a collaboration.

Why no form was submitted: No self-service form exists to submit. No contact email is published directly on biofinder.se or the study-group page; finding a current, correct email would require an unverified guess, which the binding rules forbid. This needs Alex to make direct academic contact.

biofinder.se · checked 2026-09-28

Access: OPEN (study summary, variable dictionary, study documents); MANAGED (individual-level genotype/phenotype data)

Individual-level genotype and phenotype data for dozens of AD studies (e.g. ADGC, ADSP); also hosts open study summaries and variable dictionaries.

Steps a human would take:

1. 1. Open study summary (no login): https://www.ncbi.nlm.nih.gov/projects/gap/cgi-bin/study.cgi?study_id=phs000372 - browse variable summaries and study documents.
2. 2. For controlled individual-level data: the requester's institution must first be registered in NIH's eRA Commons system with a designated Signing Official (SO).
3. 3. The individual researcher gets an eRA Commons account under that institution, then creates a dbGaP Data Access Request (DAR) selecting the desired dataset(s).
4. 4. The institutional Signing Official co-signs the request.
5. 5. NIH's Data Access Committee (DAC) reviews and approves or denies.
6. 6. For AD-specific data, both the investigator and the institutional Signing Official must also sign the NIA Genomic Data Sharing Plan / NIAGADS Data Distribution Agreement.

Why no form was submitted: There is no individual-researcher path into dbGaP controlled access - it structurally requires an institution already registered with NIH eRA Commons and a Signing Official, which is an institutional fact, not something an individual fills in on a form. This session has no information about whether Alex has such institutional standing.

ncbi.nlm.nih.gov · checked 2026-09-28

Access: OPEN (cohort discovery/summary tools); underlying raw data at each partner remains gated per that partner's own listing above

Federated cohort-level metadata and interactive cohort-building/comparison tools spanning ~72 named AD/dementia cohorts, including ADNI, AIBL, NACC, OASIS, and NIAGADS - not raw individual records, but a real open discovery layer over otherwise-gated sources.

Steps a human would take:

1. 1. Visit https://gaain.org/ (no account required to browse).
2. 2. Use Cohort Scout (https://www.gaaindata.org/partners/online.html) to search data attributes across all federated partners and build/save cohorts.
3. 3. Use the Interrogator tool to visually explore variable distributions and run preliminary cross-cohort comparisons.

Why no form was submitted: Not applicable - no form is required for the discovery/summary layer. Downloading the underlying individual-level data from any specific partner cohort still requires going through that partner's own gate (see ADNI, AIBL, NACC, OASIS entries above).

gaain.org · checked 2026-09-28

• 25 Survives
• 20 Collapses
• 21 Collapses extended
• 6 Collapses strict
• 18 Negligible
• 12 Null
• 1 Sign reverses
• 9 Consistency check
• 1 Artefact documented
• 3 Artefact collapses
• 2 Artefact test survives
• 8 Twin
• 2 Uninterpretable
• 1 Cannot compute

49–51 states (incl. DC).

• 4 Artefact documented
• 9 Consistency check
• 2 Twin
• 3 Uninterpretable
• 9 Survives
• 36 Null
• 4 Collapses strict
• 3 Attenuated
• 3 Collapses
• 3 Artefact collapses

51–192 countries.

• 5 Artefact documented
• 1 Twin
• 1 Survives
• 3 Collapses extended
• 12 Collapses
• 3 Survives sign flipped
• 3 Null
• 2 Sign reverses
• 2 Uninterpretable

2103–10073 studies.

• 2 Artefact documented
• 3 Survives
• 2 Raw only
• 4 Negligible
• 1 Null
• 1 Uninterpretable
• 1 Twin
• 1 Consistency check

73–107 curated variant records.

• 1 Null
• 2 Raw only
• 1 Consistency check

9–9 drug-program pairs.

• 1 Uninterpretable

45 of 161 coefficients that carry both a raw Pearson r and a strict-control partial r flip sign between the two: the raw and the confounder-adjusted relationship point in opposite directions. The open avenue is asking, for each one, which sign (if either) survives a study designed to settle it — not assuming the raw number is the real one.

7 of 53 coefficients that carry a stated literature expectation flip sign against it, and 3 of those 7 still carry a classification containing “survives” after confounder control — meaning the flip is not simply explained away by the named controls.