README.md

metadata

tags:
  - synthetic-data
  - healthcare
  - cardiology
  - emr
  - survival-analysis
license: other
pretty_name: Synthetic UKB South Asian ASCVD (500K Patients)
size_categories:
  - 500K+
language:
  - en
task_categories:
  - text-generation
  - feature-extraction
  - summarization
  - tabular-to-text
  - table-to-text
  - text-retrieval

Synthetic UKBB South Asian ASCVD (500K Patients)

This dataset contains 0.5 million synthetic patients with ancestry (European and South Asian), demographics, anthropometrics, labs, lifestyle/SES, comorbidities, medications, and time-to-event outcomes for ASCVD, heart failure, and atrial fibrillation. The dataset inspired by the published Circulation study.

All data are artificially generated and contain no identifiable patient records.

---

Dataset Structure

Demographics

• id — unique patient identifier
• ancestry — European or SouthAsian
• age — years
• sex_male — binary (1 = male, 0 = female)

Anthropometrics

• height_cm, weight_kg, bmi
• waist_cm, hip_cm, whr (waist ÷ hip)
• waist_height_ratio, body_fat_pct, trunk_fat_pct

Blood Pressure

• sbp_mmHg, dbp_mmHg
• hypertension (binary)
• on_bp_meds (binary)

Lipids

• ldl_mgdl, hdl_mgdl, tg_mgdl
• apoA_mgdl, apoB_mgdl, lpa_nmolL
• on_statin (binary)

Metabolic / Inflammatory

• glucose_mgdl, hba1c_pct, crp_mgL
• diabetes (binary)
• on_dm_meds (binary)

Renal

• creatinine_mgdl, cystatin_c_mgL, egfr_ckd_epi
• ckd (binary)

Lifestyle / SES

• smoking_status (never / former / current)
• pack_years
• sedentary (binary)
• pa_met_h_wk (physical activity in MET-hours/week)
• diet_score
• education_level
• income_bracket
• townsend (deprivation index)
• psychosocial_stress
• pct_life_in_uk

Family / History

• family_history (binary)

Female-Specific (if sex_male = 0)

• early_menopause, gest_htn, gest_dm, preeclampsia, iugr

Outcomes

• time_ascvd_yrs, ascvd_event
• time_hf_yrs, hf_event
• time_af_yrs, af_event

(For outcomes: time_* = follow-up time in years (capped at 15), *_event = binary indicator)

---

Example Usage

import pandas as pd

df = pd.read_parquet("ukb_sa_ascvd_synthetic_500k.parquet")
print(df.shape)   # (500000, ~70 columns)
print(df.head())

---

Intended Use

• Educational & personal learning
• Benchmarking methods for EMR preprocessing, feature extraction, and survival analysis
• Synthetic data methodology testing

Not for clinical decision-making.