# Cycle 2 Data

## Overview

This component contains all data collected during Cycle 2 of a longitudinal case study of a 28-day cyclic hormone protocol in a post-vaginoplasty trans woman. Cycle 2 spans April 29, 2026 (CD1) through May 26, 2026 (CD28).

Unlike Cycle 1, there is no cross-component overlap with the baseline component. All data collected during this cycle belongs in this component.

## Protocol Summary

The cyclic protocol uses four EV injections totaling 11 mg per cycle, with vaginal progesterone luteal support and twice-weekly vaginal estradiol cream. A vaginal microbiome intervention (VagiBiom suppositories) runs in parallel. The protocol design is unchanged from Cycle 1.

| Component | Schedule |
|-----------|----------|
| EV injections (SubQ) | CD5: 2mg / CD10: 4mg / CD15: 3mg / CD22: 2mg |
| Vaginal P4 | 200mg nightly CD15-CD18; 400mg AM+PM CD19-CD26 |
| Vaginal estradiol cream | 0.01% cream, 1g, twice weekly (Wed/Sat) |
| VagiBiom suppositories | 5 days on, 2 days off; midday timing |
| Withdrawal window | CD27-CD28 (no EV, no P4) |

Detailed daily planned and actual administration are in `cycle2_protocol_log.csv`.

## Files

| File | Description | Source |
|------|-------------|--------|
| cycle2_protocol_log.csv | Day-by-day planned and actual protocol administration; deviation flags | Investigator-logged via Cycle Planner MCP |
| cycle2_inito.csv | Inito fertility monitor FMU readings (E3G, LH, PdG, FSH); 21 days with readings | Inito API via Cycle Planner MCP |
| cycle2_clue_daily.csv | Daily symptom and cycle tracking; 14 logged dates | Clue app JSON data export |
| cycle2_wearables.csv | Apple Watch biometrics (HRV, RHR, skin temp, sleep); CD1-CD28 | Apple Health via Clue integration |
| cycle2_serum_labs.csv | Serum hormone panels from 3 draws (CD6, CD13, CD20) | Lab report PDFs (CPL, Labcorp/Esoterix) |

Schemas for `cycle2_inito.csv`, `cycle2_clue_daily.csv`, `cycle2_wearables.csv`, and `cycle2_serum_labs.csv` are identical to the corresponding baseline files. See the baseline data dictionary for column definitions. The protocol log schema is documented in `cycle1_data_dictionary_additions.csv`.

## Files Not Included

- **cycle2_evvy_microbiome.csv / cycle2_evvy_summary.csv**: No microbiome retest was performed during Cycle 2. The baseline Evvy swab was collected April 2, 2026 (approximately 12 weeks before the end of this cycle). The next retest is planned for early July 2026.
- **cycle2_data_dictionary_additions.csv**: No new variables were introduced in Cycle 2. All stream file schemas match baseline. The Clue app recorded two "partying" entries (CD20, CD22) for a tracking category not present in the baseline schema; these are excluded from `cycle2_clue_daily.csv` per the schema continuity rule in `data_conventions.md`.

## Data Quality Notes

### Sources of truth

When the same value appears in more than one source, the following hierarchy applies:

| Data type | Authoritative source |
|-----------|----------------------|
| Inito readings (E3G, LH, PdG, FSH) | Inito API, synced into the Cycle Planner MCP (cycle tracker database) |
| Apple Watch biometrics | Apple Health JSON export (via Clue) |
| Self-reported symptoms and tags | Clue app JSON export |
| Serum lab values | Original lab report PDF |
| Protocol administration | Cycle Planner MCP, reconciled against memory and supporting evidence |

See `data_conventions.md` (project root) for the full data source documentation, including Inito API field mappings.

### Inito precision note

Cycle 2 Inito values are preserved at native API precision per the current `data_conventions.md` convention. Precision varies by reading (some 1 decimal, some 2 decimal, some integer). No LOD floor is applied; values below 0.1 are preserved as recorded. This differs from Cycle 1, which used 1-decimal rounding with a 0.1 LOD floor. The Cycle 1 convention predated the formalized precision rule.

### Logging gaps

- **Inito readings**: 21 of 28 days have readings. Missing days: CD1, CD2, CD4, CD5 (early follicular, pre-first-injection), CD7, CD26, CD27 (late luteal and withdrawal). Coverage is substantially better than Cycle 1 (10 of 23 eligible days).
- **Clue self-reporting**: 14 of 28 dates have at least one self-reported entry. Two logging clusters: CD3-CD7 (early/mid follicular) and CD19-CD25 plus CD28 (luteal through withdrawal). CD8-CD18 has no self-reported entries (11-day gap spanning ovulatory and early luteal phases). Wearable data is intact throughout.
- **Protocol administration logging**: CD2 has no entry in the cycle tracker. All other days have entries with action flags logged. Late-cycle VagiBiom status on CD28 was not explicitly confirmed in the tracker but is recorded as missed per investigator recall.

### Protocol deviations (14 days flagged)

| CD | Deviation |
|----|-----------|
| 1  | Estradiol cream missed |
| 3  | VagiBiom missed |
| 4  | Estradiol cream missed. VagiBiom missed |
| 5  | First injection delayed to CD6. VagiBiom missed |
| 8  | Unscheduled VagiBiom inserted (scheduling mix-up) |
| 18 | Estradiol cream missed; applied at bedtime CD19 |
| 20 | VagiBiom missed (travel) |
| 22 | Estradiol cream missed |
| 23 | PM P4 dose missed (fell asleep early) |
| 24 | VagiBiom missed |
| 25 | PM P4 dose missed. Estradiol cream missed |
| 26 | Both P4 doses missed. VagiBiom missed |
| 27 | VagiBiom missed |
| 28 | VagiBiom missed |

Summary of deviation categories:
- EV injection: 1 day delayed (CD5 to CD6); all 4 injections ultimately administered
- P4: 3 days with partial or missed doses (CD23 PM missed, CD25 PM missed, CD26 both missed). CD26 miss effectively advanced withdrawal onset by one day.
- Estradiol cream: 5 of 8 scheduled applications missed on their scheduled day (CD1, CD4, CD18, CD22, CD25). CD18 dose was applied late (at bedtime CD19).
- VagiBiom: 8 days missed, 1 unscheduled insertion. Loading protocol adherence was lower in the early follicular (CD1-CD5) and late luteal/withdrawal (CD24-CD28) windows.

### Period and withdrawal bleed

No period or spotting was logged during Cycle 2. The prior cycle's withdrawal bleed occurred during Cycle 1 CD25-CD27 (April 25-27, 2026), before Cycle 2 began. The protocol-designed withdrawal window for Cycle 2 was CD27-CD28 (May 25-26). P4 was missed entirely on CD26, one day before the designed withdrawal window.

### Serum lab notes

Three draws were collected during Cycle 2, the first serum monitoring since the CD2 washout-endpoint draw in the baseline component.

- **CD6 and CD13 draws** were performed at Clinical Pathology Laboratories (CPL, Austin TX) via Any Lab Test Now. Methodology: Roche Cobas ECLIA.
- **CD20 draw** was performed at Esoterix Inc (Calabasas Hills, CA) via Labcorp OnDemand. E2 methodology: LC-MS/MS. This creates a cross-lab, cross-methodology comparison for E2 between the CPL draws (immunoassay) and the Labcorp draw (mass spectrometry). Differences in reported values may reflect methodology as well as true hormonal change.
- **Progesterone units**: Labcorp reported P4 in ng/dL (434 ng/dL). This was converted to ng/mL (4.34 ng/mL) for unit consistency with the baseline serum labs file. The original lab-reported value and units are documented in the notes column.
- **P4 interpretation context**: The CD20 P4 value of 4.34 ng/mL reflects post-dose absorption kinetics (400mg vaginal P4 inserted same morning pre-draw), not steady-state concentration. This is an absorption datapoint for vaginal micronized progesterone in peritoneal neovaginal tissue.
- **FSH/LH suppression**: All FSH and LH values are below or at the low end of female reference ranges. This is expected on exogenous estrogen post-orchiectomy.

### Other notes

- **CD8 unscheduled VagiBiom**: The investigator inserted a VagiBiom suppository on an unscheduled off-day due to a scheduling mix-up. This is logged in the protocol log with a deviation flag.
- **CD16 and CD24 Inito quality**: CD16 FMU hold was approximately 4.5 hours (less than the recommended 6-hour minimum). CD24 Inito was explicitly noted as not with FMU and less than 6-hour hold. Both readings are retained as recorded.
- **Excluded Clue category**: Two "partying" entries (CD20 and CD22, both logging "alcohol") were recorded in the Clue app. This tracking category did not appear in the baseline export and has no corresponding column in the baseline schema. Per `data_conventions.md` schema continuity rules, these entries are excluded from `cycle2_clue_daily.csv`.

## File Conventions

- Encoding: UTF-8
- Date format: YYYY-MM-DD
- Time format: HH:MM (24-hour, local time)
- Multi-value cells: semicolon-separated (`;`)
- Empty cells: blank (no placeholder values)
- Phase naming: snake_case (e.g., `early_luteal`)
- Numeric precision: native Inito API precision (no rounding, no LOD floor)
- License: CC-BY 4.0