Based on the introduction above, what is commonly referred to as “homeostasis” involves several closely related aspects. The observable stability of the IE reflects the surface-level phenomenon, the regulatory processes that produce this stability represent the underlying mechanisms, and the outcome of these processes corresponds to what Bernard described as a free and relatively independent life. These are not competing definitions, but different dimensions of the same complex concept.

Importantly, homeostasis cannot be adequately understood by considering these aspects in isolation. As a holistic property of the organism, homeostasis necessarily depends on the internal functional organization of the body, in which bodily functions are interconnected and arranged in a hierarchy. This internal, interconnected, hierarchical organization is therefore fundamental to homeostasis, constituting an essential dimension of it.

Together, these considerations support a four-dimensional understanding of homeostasis (Kuang, 2023): internal functional organization (first dimension), functional manifestation (second dimension), regulatory mechanisms (third dimension), and outcome (effect or consequence, fourth dimension). Built upon the core insights of Bernard and Cannon, this framework integrates different perspectives on homeostasis and helps clarify the ambiguity in its definition.

Through logical and conceptual analysis, Kuang (2023) proposed that homeostasis may be understood in terms of homeostatic tendency, rather than as a homeostatic phenomenon or the regulatory mechanisms underlying it. As a tendency, homeostasis does not denote a fixed state, nor does it reduce to specific control processes; instead, it refers to a dynamic property of the IE that is oriented to achieve relative internal stability through two types of complementary regulations as follows.

On the one hand, homeostatic tendency may be expressed through stabilizing regulation, by maintaining or restoring a previous steady state of the IE when perturbations are transient or limited. When stabilizing mechanisms are insufficient to preserve the steady state in the face of sustained internal or external challenges, homeostatic tendency is expressed through adaptive regulation, leading to the establishment of a new, variable steady state. Once the new steady state is achieved, stabilizing mechanisms again predominate to maintain this new condition. If environmental demands continue to change and the new steady state can no longer be sustained, further adaptive transitions may occur to establish a newer, variable steady state. In this sense, homeostasis is neither static nor episodic, but a continuous, layered process that integrates stability and adaptability over time (Kuang, 2023). Here, “homeostatic tendency” is not proposed as a replacement term, but as an alternative term that foregrounds its dynamic, layered, and adaptive character and facilitates its understanding.

Historically, however, this dual nature of homeostasis has been obscured in physiology education. As Carpenter (2004) pointed out, the concept has often been reduced in textbooks to the point of being treated as nearly equivalent to negative feedback regulation around a fixed set point. In contrast, Cannon’s original writings on homeostasis emphasized the capacity of organisms to maintain relative internal stability under changing external conditions (Cannon, 1926; 1932). Such adaptability is not optional. It is an inherent requirement for internal stability in a changing environment.

Moreover, Cannon’s analyses of biological regulation provided not a single regulatory template, but a variety of biological exemplars that later inspired the development of control theory (Carpenter, 2004). Regulatory strategies now formalized in control theory, including internal feedback, feedforward control, predictive regulation, and hierarchical organization, can all be traced back to the biological examples Cannon described carefully. Control theory is taken as a framework that emphasizes regulation under changing conditions, this historical continuity supports the view that adaptability was already implicit in Cannon’s original conception of homeostasis. Nevertheless, in recent textbook presentations of homeostasis, adaptation has often been overlooked.

In response to these limitations, several concepts have been introduced to supplement homeostasis, including allostasis (Sterling and Eyer, 1988; McEwen and Stellar, 1993; McEwen and Wingfield, 2003), predictive homeostasis or predictive regulation (Sterling, 2012; Schulkin and Sterling, 2019), and adaptive homeostasis (Davies, 2016). These developments reflect both the longstanding ambiguity surrounding the concept of homeostasis and the growing recognition of the complexity of physiological regulation. A more effective approach is to clarify the original concept of homeostasis by explicitly articulating its dynamic and multidimensional character.

Viewed from this perspective, homeostasis, or homeostatic tendency, can be understood as the unity of two inseparable yet complementary aspects: stability and adaptability (Kuang, 2023). Such a unified framework has the potential to serve as the true central organizing concept or framework of physiology, extending well beyond the oversimplified, single-dimensional definitions that dominate current teaching resources. Within this framework, newer terms are naturally encompassed rather than treated as competing alternatives, thereby preserving conceptual coherence.

Here, we further abstract the two types of complementary regulatory mechanisms as stabilizing and adapting modes. This terminological shift reflects a conceptual refinement rather than a change in meaning: each physiological parameter can operate in either a stabilizing or an adapting mode, whereas the concrete mechanisms through which these modes are implemented are parameter-specific and may differ substantially across systems.

Both Bernard and Cannon noted the importance of IE stability, but given the limitations of the methods, they could not actually study the stability of the IE as a whole. What they studied instead was the regulation of individual parameters, such as blood glucose, electrolytes, pH, or body temperature. However, a collection of parameters is not the same as the IE itself.

As mentioned in Kuang (2023), this also shaped the way physiology education developed. In textbooks and classrooms, the concept of homeostasis is often introduced as it refers to the stability of the IE. However, because the complexity of the IE is not always emphasized, the discussion quickly deviates into discussing the homeostasis of individual parameters without a clear distinction or transition. As a result, students are left confused regarding whether homeostasis is about the stability of the IE as a whole or about the regulation of individual parameters and their fundamental differences and interdependence. Physiology education needs to make this distinction clear by properly addressing the similarities and differences between the two. This logical step was missing in the past, but it is essential for both conceptual clarity and effective teaching.

The four dimensions of homeostasis and the concept of homeostatic tendency mentioned above apply to both the homeostasis of the IE and the homeostasis of a parameter.

Building on Kuang (2023), the present article focuses on how to strengthen physiology education in the area of homeostasis of a parameter, since this is where most of the teaching actually happens and also where major misconceptions exist. We then turn to the future and discuss how new technologies, such as multi-omics, and especially metabolomics, may help us to study and teach the homeostasis of the IE itself.

Readers are encouraged to refer to Kuang (2023), in which a dedicated subsection (Section 3.3) systematically examined the relativity of homeostasis. In brief, homeostasis does not imply an absolute or context-free internal stability. Rather, any steady state—whether of an individual parameter or of the IE—is meaningful only with respect to a specific temporal and spatial reference, such as the time scale of adaptation or the environmental and physiological context. Transitions among modifiable steady states therefore reflect reference-dependent adaptations, not departures from homeostasis.

As an inherent characteristic and a defining feature of homeostasis, this relativity should be explicitly emphasized in physiology education. However, among the 21 definitions of homeostasis surveyed in the supplemental table of Kuang (2023), only three explicitly acknowledged the relative nature of stability. This omission highlights a conceptual insufficiency in current teaching resources and underscores the need to strengthen how the relativity of homeostasis is presented and taught in the future.

Framing relativity as a new development may overstate its novelty, since the relative nature of homeostasis has long been implicit in physiological practice. The contribution here lies in making this implicit feature explicit and treating relativity as an indispensable component of how homeostasis should be understood and taught. In this sense, teaching homeostasis without relativity risks reducing a fundamentally dynamic process to an oversimplified and static account.

The relativity of homeostasis is straightforward to illustrate at the level of individual parameters (see Section 3), whereas the relativity of the IE’s homeostasis is more complex, yet more revealing of why homeostasis is inherently relative (see Section 4).

When Cannon coined the word homeostasis, he was very deliberate. Homeo means “similar” or “unchanging,” and stasis means “standing” or “condition.” Put together, the word conveys the idea of a steady state. We might also call this a dynamic equilibrium, although Cannon preferred to avoid the term “equilibrium” ¹ .

In most teaching resources, the description of steady state leads to a common misunderstanding of a constant, unchanging value of a parameter, such as body temperature fixed at 37 °C. However, the parameter value fluctuates within limits around a set point, and continuous regulatory processes supported by energy expenditure maintain this condition. It is important to educate learners that a steady state or a dynamic equilibrium does not imply that the parameter value itself is steady; rather, the pattern by which that value fluctuates over time is steady. Only by making the meaning of steady state explicit can students understand why the homeostasis of a parameter is inherently dynamic rather than static.

To clarify that steady state does not apply to the absolute value of a parameter, but to its regulated behavior over time, we routinely use the following simple multiple-choice question in teaching (Figure 1).

Question 1

What is in steady state? Select all that apply.

(A) The value of body temperature (No).

(B) The set point of body temperature (Yes).

(C) The upper and lower limits of body temperature (Yes).

(D) The oscillation of body temperature (Yes).

This question allows students to see that steady state does not refer to the value of the parameter itself, but to the oscillation of the value around a set point. In this context, the stability of the set point, the persistence of upper and lower limits, and the continuous fluctuation of the parameter value converge on the concept of parameter homeostasis. Accordingly, it is not the parameter value itself but its oscillation that is stable; the homeostasis of a parameter is therefore most accurately described as a relatively stable oscillation.

Despite its precision, the concept of stable oscillation is rarely used in teaching, leading to persistent conceptual ambiguity. When explicitly identified as the defining feature, dynamic equilibrium becomes clear as an alternative term for stable oscillation.

From Bernard to Cannon, terms such as fixity, constancy, and stability played a central role in the early framing of the stable, IE and physiological regulation. In their historical context, Bernard’s emphasis on the “constancy of the milieu intérieur” (Bernard, 1865; Bernard, 1879) and Cannon’s focus on the “stability of the internal environment” (Cannon, 1926; Cannon, 1932) conveyed a crucial insight: life depends on maintaining internal order despite continuous external perturbations. From a historical perspective, this language was both necessary and illuminating.

However, when these terms are adopted in contemporary physiology teaching without clarification, they invite misinterpretation. Students may interpret constancy as an unchanging value and stability with immobility—interpretations that diverge from Bernard’s and Cannon’s original intent but arise naturally when historical context is no longer explicit.

For this reason, while the historical significance of classical terminology should be acknowledged, its pedagogical limitations must also be recognized. Rather than relying on fixity, constancy, or stability alone, explicitly framing parameter homeostasis in terms of relatively stable oscillation preserves Bernard’s and Cannon’s original insights while avoiding static interpretations.

Understanding that stable oscillation is the accurate phrase for what is in a steady state is an important first step. To reinforce this understanding, students should also be guided to ask a deeper question: What if all the physiological parameters did not oscillate at all, but instead remained completely static?

This kind of question seems not to be commonly asked in current physiology education, yet it is crucial. The answer is straightforward but profound: if parameters were truly fixed and unchanging, the organism would lose its adaptability. Without oscillation, the system would have no flexibility, no capacity to adjust to perturbations, and ultimately no ability to survive.

Oscillation of a parameter value, therefore, is not a minor detail but the very feature that connects stability to adaptability. It shows students why homeostasis of a parameter cannot be equated with simple constancy. By asking and answering this counterfactual question, teaching can reinforce the central role of stable oscillation of a parameter, ensuring that students grasp both its precision as a term and its essential biological meaning. Figure 2 illustrates how the oscillation of a parameter adapts to the environmental changes to achieve a new steady state, or in other words, a new stable oscillation by changing the set point of the oscillation.

For clarity, in this article the notation SS1, SS2, and SS3 (and higher-order extensions) is used exclusively to describe layered steady states of individual physiological parameters (Kuang, 2023). Briefly, this relatively stable oscillation can be preserved in two complementary ways: either by stabilizing the existing oscillation around an unchanged set point, or by adapting to a new set point and establishing a new relatively stable oscillation. In this sense, the homeostasis of a parameter reflects a homeostatic tendency expressed through the regulation of oscillatory behavior.

The sequence of layered steady states (Figure 2) provides a concrete example of the relativity of parameter homeostasis. What constitutes a relatively stable oscillation (SS1, SS2, SS3, etc.) depends on the prevailing regulatory context and cannot be reduced to a single, fixed set point. In this sense, parameter homeostasis is maintained through regulated transitions among modifiable steady states and thus reflects the expression of homeostatic tendency at the level of individual physiological parameters.

An emergent property refers to a system-level feature that arises from interactions among components and cannot be fully explained by examining those components in isolation (Emergent property, 2025; Kitano, 2002). From a teaching perspective, the idea can be introduced without formal systems theory, as educators are already familiar with physiological functions that depend on coordinated activity across multiple subsystems rather than any single mechanism.

Viewed in this way, the homeostasis of the IE represents a high-level emergent property at the organism level. It is not the sum of individual parameter regulation, such as temperature, glucose, or blood pressure, but the overall pattern formed by their coordinated behavior. Individual parameters provide partial views of this organization, yet none alone captures how the IE is maintained as an integrated whole.

The IE also shapes how subsystems operate. The global conditions of the IE influence subsystems’ responses to perturbations, the recruitment of regulatory mechanisms, and whether a system returns to a prior steady state or transitions to a new one. In this reciprocal sense, system-level order both arises from and constrains subsystems’ regulatory processes (Figure 3), a principle long recognized in systems physiology (Kitano, 2004; Noble, 2008; Noble, 2011; Woodward, 2020).

The complexity of the IE cannot be captured by enumerating individual variables or regulatory pathways. Instead, it requires a framework that organizes its conceptual richness in a coherent way. As developed in Kuang (2023) and introduced earlier in this article, the four dimensions of homeostasis provide such a framework, allowing, IE complexity to be articulated and examined rather than left as a vague intuition.

The relativity of IE homeostasis is more intricate than that of individual parameters and contributes directly to this complexity. IE homeostasis reflects the integrated behavior of multiple interacting parameters, each operating with distinct regulatory dynamics, adaptive capacities, and temporal profiles. As a result, stability and adaptation are not expressed uniformly, but emerge from heterogeneous yet coordinated responses across parameters.

Accordingly, IE homeostasis may be understood in terms of multiple IE overall steady-state patterns or configurations (denoted conceptually as H1, H2, etc.), each defined by its temporal and contextual conditions. Whether a given IE appears stable or adaptive depends on the time scale and observational window: processes that constitute adaptation over longer durations may appear as stability when examined within a short time window. For this reason, any discussion of IE homeostasis necessarily requires explicit temporal framing. A qualitative illustration of this relativity and heterogeneity is provided in the Supplementary Material (Supplementary Figure S1).

From a teaching perspective, the goal is not to introduce technical detail, but to highlight which technologies make organism-level homeostasis empirically accessible. Among these, metabolomics plays a central role by providing an integrative molecular snapshot of the IE, helping students conceptualize the IE as a coherent functional state rather than as a collection of independently regulated parameters. At a broader level, multi-omics approaches support this integrative view by conceptually linking system-wide patterns to underlying regulatory layers.

It is important to note that metabolomics and other multi-omics approaches do not directly represent the organizational or regulatory structure of the IE. Rather, they provide downstream, system-level readouts, reflecting the cumulative outcomes of underlying functional processes operating across multiple levels and timescales. As such, multi-omics data are best interpreted as context-dependent profiles of IE states, useful for identifying patterns, shifts, and comparisons, but not for inferring mechanisms or organizational principles in isolation. This distinction is particularly important in educational contexts, where omics data may otherwise be misconstrued as direct mappings of physiological organization rather than as outcome-level system summaries.

Finally, systems biology modeling and network analysis provide conceptual tools for understanding how interactions between subsystems give rise to global stability and how system-level configuration constrains the admissible behavior of individual subsystems (Angione, 2019).

By making connectivity and transitions among steady states explicit, these approaches support a shift in homeostasis education from parameter-centered descriptions toward a system-level understanding of IE homeostasis, consistent with the framework developed below.

Here, homeostasis or homeostatic tendency is expanded both in scope and in its constitutive relationship with the environment. This expansion does not impose a new conceptual structure, but reflects an interpretive realignment required to remain faithful to the biological reality that homeostasis was originally intended to capture.