Primary vs Secondary Stability: The Stability Dip

Implant stability is not a single quantity but the running sum of two reciprocal components that change in opposite directions over the weeks following placement. Primary stability is mechanical — the friction and interlock the surgeon achieves between the implant body and the prepared bone walls at the moment of insertion. Secondary stability is biological — the anchorage that develops as new bone forms along and bonds to the implant surface through osseointegration. The two do not coexist statically; they hand the load-bearing role from one to the other, and the geometry of that handover is the single most clinically consequential idea in early implant care.¹²

The handover is not seamless. Primary stability begins to decline almost immediately as the compressed and cut bone at the interface is remodeled, while secondary stability takes weeks to accumulate. For a brief interval the falling mechanical curve and the still-rising biological curve cross at a low point — the stability dip, classically reaching its nadir around weeks 1–4 and most often near week 3 to 4.¹² This trough is not incipient failure; it is the mechanical signature of normal remodeling. But it is also the implant's most vulnerable window, and much of contemporary loading-protocol thinking exists precisely to respect it. This chapter develops the biology of the two stabilities, the dip between them, the determinants the surgeon actually controls, and how serial measurement — insertion torque at placement and resonance-frequency analysis (RFA/ISQ) over time — lets the clinician watch the handover unfold.

The figure below is the signature schematic of this chapter. Primary (mechanical) stability is highest at placement and decays as early bone is remodeled; secondary (biological) stability begins near zero and rises sigmoidally as new bone forms; total stability is their sum. The two component curves cross around weeks 3–4, and at that crossing the total curve reaches its nadir — the stability dip. Read the figure vertically at any timepoint to see which component is carrying the implant, and follow the dashed total curve to see how shallow or deep the trough becomes.¹²

Why should stability fall at all before it rises? Because the very bone that provides the initial mechanical grip must be partly dismantled before durable bone can replace it. The compressed bone chips and cut osteotomy walls that generate placement-day friction are devitalized or under strain; osteoclasts resorb them, and for a time the interface loses mechanical contact faster than new lamellar bone is laid down. Serial RFA studies capture this directly: mean ISQ characteristically falls from a placement peak to a low point around week 4, then rebounds toward and often above baseline by weeks 11–12.² A falling reading in that window is biology, not failure — provided the absolute value stays within a safe band and the trajectory subsequently recovers.

Primary stability is the one part of the system the surgeon directly controls on the day of surgery, and it is governed by three families of factors: the bone the implant is placed into, the implant chosen, and the surgical technique used to prepare the site. Bone quality is largely a given — dense Type I–II bone offers far more cortical engagement and friction than soft Type IV posterior maxilla — but design and technique are levers. Macro-design features such as a tapered body, aggressive or self-tapping threads, and greater thread depth increase mechanical interlock. Technique can compensate for poor bone through under-preparation (drilling a final osteotomy narrower than the implant) and osteotome or osseodensification condensation, both of which raise insertion torque and primary stability — though excessive compression risks ischemia and pressure necrosis, which paradoxically deepens the subsequent dip.¹³ The table below summarizes the determinants and the direction of their effect.

Two complementary measures let the clinician quantify stability, and the distinction between them is conceptually central. Insertion torque (N·cm) is recorded during placement and reflects primary mechanical stability at that single moment; it is useful for gating accelerated-loading decisions but cannot be repeated to follow healing without disturbing the implant. Resonance-frequency analysis (RFA) reports the Implant Stability Quotient (ISQ, 0–100) — derived from the resonance frequency of a transducer attached to the implant, which scales with the stiffness of the implant–bone interface. Because RFA is non-invasive and repeatable, serial ISQ readings can be taken at placement and at intervals through healing, tracing the dip and the recovery directly.² Higher ISQ reflects greater stiffness; values in the 55–65 range maintained over time are conventionally read as a safe level of stability, with the trend mattering more than any single value.² The quick-reference table maps each timepoint to its dominant stability source and expected ISQ trend.

The reason this biology matters chairside is that loading protocol must be matched to where the implant sits on the curve. The danger in the dip is micromotion: excessive relative movement between implant and bone during the vulnerable window provokes fibrous encapsulation instead of bone, and the conventionally cited threshold above which integration is jeopardized is roughly 50–150 µm of micromotion.¹ Adequate primary stability exists precisely to keep micromotion below that threshold until secondary stability takes over. This is why a high day-0 reading is necessary but not sufficient for immediate loading in soft bone, and why conservative (conventional, delayed) loading is preferred where the dip is expected to be deep and long. The full loading-protocol decision tree — torque and ISQ thresholds against bone type — is developed in its own chapter (see Loading Protocol Selection →). The biology of bone quality and the drilling adjustments it demands are covered in Bone Quality & Drilling →, and the timing of placement relative to extraction in Placement Timing →.