An Early Isoboramine Dip Predicts Symbiont–Host Incompatibility After Joining: A Retrospective Cohort of 188 Joinings in the Trill Symbiosis Commission Register, 2340–2372

1. Introduction

A joined Trill is two beings in one body: a humanoid host and a long-lived symbiont that carries the memories of its previous hosts. The Trill Symbiosis Commission selects hosts from among trained candidates, and the joining itself is performed under its supervision. Most joinings settle without incident, including among joined Trill serving with Starfleet (Dalvane & Orrin, 2371). A minority do not. In the days or weeks after the procedure, some hosts show signs that host and symbiont are failing to integrate, and a physician must then decide whether to support the joining medically or to separate the pair before either is harmed (Kellun, 2359; Starfleet Medical, 2370).

The difficulty is timing. Clinical signs of incompatibility, such as disordered sleep, intrusive memory from earlier hosts and autonomic instability, tend to appear after the first few days, and by the time they are unmistakable the options have narrowed. Within roughly ninety-three hours of joining, host and symbiont become biologically interdependent, and from then on removing the symbiont would kill the host (Trill Symbiosis Commission, Medical Directorate, 2356). A physician who wishes to keep separation available must therefore decide within four days, usually before any clinical sign has appeared.

Host serum isoboramine is the obvious candidate for an early signal. Its level rises after joining and settles at a new plateau as symbiont and host integrate, and Commission physicians have measured it routinely for decades (Tavor, 2361). What the register has never been used for is the question of whether the *shape* of the early curve predicts what happens next. We addressed that question with thirty-three years of Commission records.

2. Methods

Every joining performed under Commission supervision is entered in the Joining Register, with the host's candidate file, the symbiont's history and scheduled clinical follow-up (Trill Symbiosis Commission, 2373). Between 2340 and 2372 it lists 203 joinings. Serum isoboramine is drawn before the procedure and on days 1, 3, 7, 14, 30 and 90. We included every joining with complete values through day 7, which left 188. The 15 exclusions were spread across the period and did not differ from the included joinings in host age or in the symbiont's number of prior hosts.

The outcome was early incompatibility, defined as it is in the Commission's own guidance: a clinical decision, recorded within 90 days of joining, either to separate host and symbiont or to begin sustained medical treatment for failing integration. Two of us, one from the Commission and one from Starfleet Medical, read each case file independently and agreed on every classification after discussion of three files. The reviewer from Starfleet Medical was blind to isoboramine values when coding outcomes.

Our exposure was the day-3 dip. In an uncomplicated joining, isoboramine rises on day 1 and continues upward. We defined a dip as a fall of 20% or more between day 1 and day 3, a threshold fixed before any outcome was examined and drawn from the assay's reported day-to-day variability of about 8% (Trill Symbiosis Commission, Medical Directorate, 2356). A switch to the revised assay in 2356 changed absolute values but not within-person ratios (Varma-Lindqvist, 2368), which is one reason we used a ratio.

Because early incompatibility is uncommon, we fitted a logistic regression with penalised likelihood (Achterberg, 2371), which reduces small-sample bias in odds ratios when events are few. Covariates were host age in decades, the symbiont's number of prior hosts and the host's years of candidate training. With 17 events we limited the model to these three covariates and the exposure. We also report sensitivity, specificity and predictive values for the dip as a screening rule, and a secondary model treating the day-1 to day-3 change as continuous.

3. Results

The 188 hosts were aged 23 to 71 at joining (median 34). Their symbionts had between none and eight prior hosts (median three). Seventeen joinings (9.0%) met the definition of early incompatibility. Four were separated and thirteen were treated medically. All four separations were performed on day 3 or early on day 4, before interdependence, in joinings that showed both a day-3 dip and early clinical signs; all four symbionts were returned to the pools of Mak'ala and all four hosts recovered. Of the thirteen medically treated joinings, eleven were intact and stable at day 90 and two remained under treatment. Clinical onset fell at a median of day 9 (range 3–41).

A day-3 dip occurred in 29 joinings (15.4%). Ten of these 29 developed incompatibility (34%), against 7 of the 159 joinings without a dip (4.4%). Table 1 sets out the cross-tabulation. In the adjusted model, the dip carried an odds ratio of 10.8 (95% CI 3.6–32.4; p < .001). None of the covariates was associated with the outcome: host age, odds ratio 1.2 per decade (95% CI 0.7–2.0; p = .49); number of prior hosts, 0.9 per host (95% CI 0.6–1.4; p = .66); and years of candidate training, 0.8 per year (95% CI 0.6–1.1; p = .17).

As a screening rule, the dip had a sensitivity of 59% (10 of 17) and a specificity of 89% (152 of 171). Its positive predictive value was 34% and its negative predictive value 96% (152 of 159). In the six true-positive cases that were not separated, the dip preceded clinical onset by a median of six days (range 2–19). In the secondary model, each further fall of 10 percentage points in isoboramine between day 1 and day 3 multiplied the odds of incompatibility by 2.3 (95% CI 1.6–3.3).

The seven incompatibilities without a day-3 dip were different in character. Five had onset after day 20, and in four of those the case notes attribute the crisis to an intrusive memory of a specific earlier host, which the Commission's physicians treated as a psychological failure of integration. These cases had no day-3 prodrome: isoboramine followed the ordinary rising curve and fell only once the crisis had begun.

4. Discussion

A fall in isoboramine on the third day after joining marks a joining at substantially higher risk of early incompatibility, and it does so before the clinical picture declares itself. The timing is what gives it value. Day 3 falls just inside the ninety-three-hour window after which separation would cost the host's life. In this register, every separation was performed inside that window, and every one followed a dip. For the other thirteen joinings the only option once incompatibility was recognised was medical support, which kept all thirteen joinings intact, although two were still under treatment at day 90.

The physiological reading we favour is that the dip reflects a failure of the second phase of integration. On day 1 the rise in isoboramine is probably driven by the symbiont's own output, which is large and immediate. The continued rise to day 3 and beyond requires the host's neural tissue to take up and sustain that signalling. A fall at day 3 would then indicate that the host is not taking over its share. This is our hypothesis, not established Trill physiology, and the register contains nothing that tests it directly.

No single marker substitutes for clinical judgment, and the dip is no exception. It missed seven of the seventeen events, and roughly two in three dips were false alarms. The misses had a recognisable pattern: late crises driven by the memory of a particular earlier host, which a metabolic marker would not be expected to detect. We would treat the dip as a trigger for immediate review while separation is still possible, and for daily review through the first fortnight, but not as grounds for separation on its own.

Several things that Commission lore holds to matter did not register (Kellun, 2359). Older hosts were not at higher risk, joinings to symbionts with many prior hosts were not either, and longer candidate training did not protect. The small number of events means these null results are imprecise, and the confidence intervals are wide enough to admit modest effects in either direction. Still, none of the three pointed where tradition expects.

5. Limitations

Every host in the register had passed Commission selection, so these results describe that population and no other. The register is maintained by the Commission itself and has not been audited independently; we used it as the Commission supplied it in its 2373 edition. Joinings performed outside Commission supervision, which are rare and which the register does not record, may behave differently. Incompatibility was defined by a clinical decision, and decisions may have been influenced by physicians' own reading of the isoboramine curve. Blinding one of the two outcome coders reduces but does not remove that risk, because the treating physician was not blind. Seventeen events support only a small model, and the wide intervals around every estimate reflect that. Finally, the fifteen excluded joinings, although similar on the characteristics we could check, may have differed in ways we could not.