The obesity epidemic: leptin signaling, the thrifty-gene hypothesis, and the Set-Point theory of body weight

Intuition Beginner

Obesity is not a moral failing or a lack of willpower. It is a regulated physiological state. Your body has a set point for body weight, defended the way body temperature is defended. Fat cells release a hormone called leptin that tells the brain how much fat you carry. The brain's hypothalamus compares leptin levels to the set point and adjusts appetite, energy expenditure, and physical activity to defend that weight. When you diet, leptin drops, the brain reads this as starvation, and you feel hungry while burning fewer calories. This is why diets almost always fail long-term: about 85 percent of people regain lost weight within five years.

Obesity results from a mismatch between the set point and the modern world. The set point can sit too high because of genetics — the so-called thrifty genes that helped ancestors survive feast-famine cycles — or because of an obesogenic environment: cheap, calorie-dense, ultra-processed food; sedentary jobs; disrupted sleep. New GLP-1 drugs like Ozempic and Wegovy work partly by lowering the set point. They are the first effective pharmacotherapy in obesity's history.

About 13 percent of adults globally were obese in 2022 (WHO), triple the 1975 rate. In the United States, about 42 percent of adults are obese and about 73 percent are overweight or obese. Obesity drives type-2 diabetes, cardiovascular disease, about thirteen cancers, osteoarthritis, and depression. It is the largest single contributor to chronic disease in wealthy countries. This is why the leptin-and-set-point framework exists: it explains both why obesity happens and why it is so hard to reverse.

Visual Beginner

The picture shows the body-weight regulation system on the left, the obesity epidemic curve in the middle, and the GLP-1 pharmacotherapy revolution on the right.

The arcuate-nucleus box carries the load: leptin's signal branches into two opposing neuronal populations whose balance sets appetite. Most obese individuals have high circulating leptin but impaired brain response — leptin resistance — which is the molecular reason willpower-based diets fail.

Worked example Beginner

Consider the Pima Indians of Arizona. The Pima are a Native American population with one of the highest obesity rates in the world: about 70 percent of adults are obese and about 50 percent have type-2 diabetes.

Step 1: Until the early twentieth century, the Pima were lean. Their traditional diet was low in calories, low in fat, and high in fibre, derived from subsistence agriculture, hunting, and gathering in the Sonoran Desert. Contemporary observers described them as "spare in flesh."

Step 2: From the late nineteenth century, diversion of the Gila River destroyed their agriculture. By the 1950s the Pima had shifted to a diet of government-issued commodity foods — white flour, lard, sugar, canned meat — combined with a sedentary reservation lifestyle. Their calorie intake rose and their physical activity collapsed.

Step 3: Genetically identical Pima populations in Maycoba, Mexico, who retained traditional subsistence farming and a high-fibre diet, remain lean: obesity prevalence is about 15 percent and type-2 diabetes is rare. The same thrifty genes, in a different environment, produce a different phenotype.

Step 4: Their thrifty genes — adaptations that helped ancestors store fat efficiently through feast-famine cycles — became maladaptive in the obesogenic environment of mid-twentieth-century Arizona.

What this tells us: obesity is the interaction of genetic susceptibility with environmental change. Genes set the susceptibility; the environment determines whether that susceptibility is expressed as disease.

Check your understanding Beginner

Formal definition Intermediate+

Definition (the homeostatic body-weight regulation system). Body weight is regulated by a negative-feedback control system whose components are an afferent signal of energy stores, a central integrator, and efferent outputs controlling intake and expenditure ^{[HeymsfieldWadden2017]}. Adipose tissue secretes leptin in proportion to fat mass $F$; leptin crosses the blood-brain barrier and acts on the arcuate nucleus of the hypothalamus (ARC). Two opposing neuronal populations respond: the anorexigenic POMC/CART neurons, whose product pro-opiomelanocortin is cleaved to alpha-melanocyte-stimulating hormone ($\alpha$-MSH) acting on melanocortin-4 receptors (MC4R) to suppress appetite and increase energy expenditure; and the orexigenic NPY/AgRP neurons, which release neuropeptide Y and agouti-related peptide, inhibit POMC neurons, and drive feeding. Leptin activates POMC/CART neurons and inhibits NPY/AgRP neurons. Additional hypothalamic sites — the ventromedial hypothalamus (VMH) and paraventricular nucleus (PVN) — integrate secondary signals (insulin, ghrelin, PYY, GLP-1, CCK).

Definition (leptin resistance). Leptin resistance is the state in which elevated circulating leptin fails to suppress appetite or increase energy expenditure appropriately. Most human obesity is leptin-resistant rather than leptin-deficient. The molecular basis includes hypothalamic inflammation, endoplasmic-reticulum stress, impaired blood-brain-barrier transport, and intracellular signalling blockade by SOCS3 (suppressor of cytokine signalling 3) and PTP1B (protein tyrosine phosphatase 1B), both of which attenuate leptin receptor JAK2-STAT3 signalling. Exogenous leptin therapy does not overcome this blockade in the general obese population, which is why leptin is a clinical therapy only for the rare genetic-deficiency subset.

Definition (the ob/ob mouse and db/db mouse). The ob/ob mouse carries a loss-of-function mutation in the leptin gene, becomes massively obese, and is rescued by exogenous leptin (Coleman 1973 parabiosis experiments; Zhang-Friedman 1994 Nature 372:425 positional cloning). The db/db mouse carries a loss-of-function mutation in the leptin receptor; it is obese but not rescued by exogenous leptin. Coleman's parabiosis experiments joined the circulations of ob/ob, db/db, and normal mice pairwise; the patterns of weight loss and weight gain in each pairing predicted the existence of a circulating satiety factor (leptin) and its receptor twenty years before the molecules were cloned.

Definition (GLP-1 receptor agonists). Glucagon-like peptide-1 (GLP-1) is an incretin hormone released by intestinal L-cells in response to nutrient intake; it potentiates glucose-dependent insulin secretion, suppresses glucagon, slows gastric emptying, and acts on central GLP-1 receptors to reduce appetite. GLP-1 receptor agonists — liraglutide (2014, ~8 percent loss), semaglutide (Wegovy 2021, ~15 percent loss), tirzepatide (Zepbound 2023, a dual GLP-1/GIP agonist, ~21 percent loss) — are long-acting analogues that achieve sustained activation of GLP-1 receptors and, for the dual/triple agonists, the gastric inhibitory polypeptide (GIP) and glucagon receptors. These are the first effective pharmacotherapies for obesity.

Counterexamples to common slips Intermediate+

• "Obesity is a lifestyle choice." No. Twin and family studies estimate BMI heritability at roughly 70 percent (Bouchard 1990s). The body actively defends its set point via hormonal feedback. The Pima of Arizona developed an epidemic of obesity within a single generation when their environment changed, with no change in their genes.

• "You just need willpower." No. The homeostatic defence fights weight loss: when fat mass falls, leptin drops, appetite rises, and resting metabolic rate falls by more than the lost mass predicts (the Minnesota Starvation Study, Keys 1950; see 29.09.05). Sustained weight loss requires overriding a hormonal feedback loop, not a one-time act of discipline.

• "Leptin therapy cures obesity." Only in the rare leptin-deficient subset (congenital leptin deficiency, ~two dozen known families worldwide, and the ob/ob-like lipodystrophies). The vast majority of obese patients are leptin-resistant, so adding more leptin has minimal clinical effect.

• "Calories are all that matter." Oversimplified. Hall 2017 NIH metabolic-ward studies confirm that calories are necessary but show that composition, processing, and gut-hormone responses modify intake and expenditure. The carbohydrate-insulin model and the energy-balance model remain contested; the truth lies in a synthesis where composition modulates the set point.

• "GLP-1 drugs are cheating." No. They treat a physiological dysregulation. The modern clinical framing of obesity as a chronic disease (Jastreboff 2023) holds that pharmacotherapy addresses an underlying regulatory defect, not a moral weakness — analogous to using statins for hypercholesterolaemia.

• "The thrifty-gene hypothesis is settled." Contested. Speakman's predation-release hypothesis (2007-2008) argues that humans escaped predation roughly two million years ago, releasing the upper bound on fat storage; genetic drift, not positive selection, then set the variation. Both hypotheses have empirical support and the truth may involve elements of each.

• "Sugar is the only cause." Oversimplified. Multiple drivers contribute: ultra-processed food (Monteiro 2009 NOVA classification, ~58 percent of US calories), sugar-sweetened beverages, sedentary lifestyle, chronic sleep restriction, psychosocial stress, and endocrine disruptors (BPA, phthalates). No single cause dominates.

Key model: the Set-Point theory of body-weight regulation Intermediate+

The Set-Point theory holds that body weight $W(t)$ is regulated around a defended target $W^{\ast }$ by negative feedback through the leptin-arcuate-nucleus-MC4R axis. Deviations from $W^{\ast }$ are sensed via leptin and corrected by compensatory adjustments in appetite, energy expenditure, and nutrient partitioning ^{[HeymsfieldWadden2017]}. The theory explains three observations that pure willpower-based models cannot: the long-term failure rate of caloric restriction, the asymptomatic plateauing of weight during ad-libitum feeding, and the rapid regain that follows diet cessation. We give the simplest formal model that captures these features, derive the convergence dynamics, and then identify where leptin resistance and the obesogenic environment break the model — which is the formal content of the modern obesity epidemic.

Key result: exponential convergence to a defended set point

Theorem (defended set point). Let $W(t)$ denote body mass (kg) at time $t$, with energy density of adipose tissue $\rho \approx 7700$ kilocalories per kilogram. Suppose energy intake $I$ (kilocalories per day) decreases linearly with deviation above the set point, $I(W)=I_0-\alpha (W-W^{\ast })$, with $\alpha >0$; and energy expenditure $E$ (kilocalories per day) increases linearly with deviation above the set point, $E(W)=E_0+\beta (W-W^{\ast })$, with $\beta >0$. At equilibrium $I(W^{\ast })=E(W_0^{\ast })$, so $I_0=E_0$ and $W^{\ast }$ is the unique fixed point. Then

$$W(t)=W^{\ast }+(W(0)-W^{\ast })\exp \left(-\frac{\alpha +\beta }{\rho }t\right),$$

so $W(t)\to W^{\ast }$ exponentially with time-constant $\tau =\rho /(\alpha +\beta )$. Any perturbation — caloric restriction, exercise, weight-loss drugs — produces only transient deviation, followed by exponential return to $W^{\ast }$ once the perturbation is removed.

Proof.

Step 1 — mass balance. The rate of change of body mass equals net energy balance divided by the energy density of stored tissue:

$$\frac{dW}{dt}=\frac{I(W)-E(W)}{\rho }.$$

Step 2 — substitute the linear feedback laws. Using $I(W)=I_0-\alpha (W-W^{\ast })$ and $E(W)=E_0+\beta (W-W^{\ast })$, with $I_0=E_0$ at the fixed point:

$$\frac{dW}{dt}=\frac{-\alpha (W-W^{\ast })-\beta (W-W^{\ast })}{\rho }=-\frac{\alpha +\beta }{\rho }(W-W^{\ast }).$$

Step 3 — solve the linear ODE. Setting $\delta (t)=W(t)-W^{\ast }$, we have $\dot{\delta }=-(\alpha +\beta )\delta /\rho$, a first-order linear equation with constant coefficients. The solution is $\delta (t)=\delta (0)\exp (-(\alpha +\beta )t/\rho )$, hence

$$W(t)=W^{\ast }+(W(0)-W^{\ast })\exp \left(-\frac{\alpha +\beta }{\rho }t\right).$$

Step 4 — stability and time-constant. The coefficient $(\alpha +\beta )/\rho$ is strictly positive, so $\delta (t)\to 0$ as $t\to \infty$ and the set point is asymptotically stable. The exponential time-constant is $\tau =\rho /(\alpha +\beta )$: doubling $\alpha +\beta$ halves the time to return to set point after a perturbation. Empirical estimates of $\tau$ from overfeeding studies (Leibel 1995, Rosenbaum 2010s) range from several months to about a year, indicating combined feedback gains $\alpha +\beta \approx 20$ to $50$ kilocalories per kilogram per day. $\square$

The model predicts four empirically observed phenomena. First, dieting triggers compensatory decreases in $E$ (mainly reduced resting metabolic rate beyond what mass loss predicts) and increases in $I$ (increased hunger), so once active dieting stops, $W$ returns to $W^{\ast }$ on a timescale of months. Second, the asymptotic weight under ad-libitum feeding plateaus rather than drifting without bound. Third, leptin resistance effectively reduces $\alpha +\beta$, lengthening $\tau$ and weakening the defence against upward perturbations (cheap food) while leaving the defence against downward perturbations intact — an asymmetric loop that produces upward drift. Fourth, GLP-1 receptor agonists effectively raise the set point's defended value downward by increasing the apparent sensitivity of the central integrator to leptin and incretin signals, restoring convergence to a lower $W^{\ast }$.

Bridge. The defended-set-point theorem builds toward the bariatric-surgery and GLP-1 receptor-agonist pharmacotherapy discussed in 35.03.01, where the chronic-disease survey identifies long-term weight management as the dominant modifiable risk factor for type-2 diabetes and cardiovascular disease, and the leptin-arcuate-nucleus feedback loop appears again in 17.04.01 as the cellular-metabolic substrate — adipose-tissue lipolysis and ATP-yielding oxidative pathways — that the set-point system ultimately regulates. The foundational reason pharmacotherapy has historically failed is that the homeostatic feedback loop is exactly the structure that opposes any intervention aimed below $W^{\ast }$, and this is exactly the obstacle that GLP-1 receptor agonists circumvent by acting upstream of the leptin signal at the central integrator itself. The bridge is between the control-theoretic formalism of negative-feedback regulation and the clinical phenomenology of obesity as a defended chronic-disease state.

Exercises Intermediate+

Advanced results Master

Result 1 (Neel 1962 — the thrifty-gene hypothesis). Neel's 1962 paper in the American Journal of Human Genetics ^[Neel1962] proposed that genes promoting efficient fat storage were positively selected during the feast-famine cycles of human evolutionary history. The argument was prompted by the rising prevalence of type-2 diabetes in affluent populations in the mid-twentieth century: Neel reasoned that the same genotype that protected ancestors from famine now caused diabetes and obesity under continuous caloric abundance. The paper's full title — "Diabetes mellitus: a 'thrifty' genotype rendered detrimental by 'progress'" — captures the gene-environment interaction at the core of the hypothesis. The thrifty-gene framework dominated evolutionary medicine for four decades and remains the dominant popular framing of obesity genetics.

Result 2 (Coleman 1973 — parabiosis anticipating leptin). Coleman's parabiosis experiments at the Jackson Laboratory joined the circulations of ob/ob, db/db, and normal mice pairwise ^{[Coleman1973]}. The weight-change patterns in each pairing were consistent with the existence of a circulating satiety factor produced by the ob gene and a receptor encoded by the db gene. Coleman's 1973 Diabetologia 9:294 paper interpreted these patterns correctly, predicting the existence of the leptin-leptin-receptor system more than twenty years before the molecules were cloned. Coleman shared the 2009 Shaw Prize and was awarded the 2010 Albert Lasker Award for this work posthumously.

Result 3 (Zhang-Friedman 1994 — positional cloning of the ob gene). Zhang, Proenca, Maffei, Barone, Leopold, and Friedman cloned the mouse obese (ob) gene and its human homologue using positional cloning ^[Zhang1994]. The paper in Nature 372:425 reported that the ob gene encodes a 167-amino-acid secreted protein, expressed exclusively in adipose tissue, with serum levels proportional to adipocyte mass. The protein was named leptin, from the Greek leptos (thin). The ob/ob mouse, which carries a spontaneous loss-of-function mutation in this gene, became massively obese; recombinant leptin restored normal weight. The cloning opened the modern molecular era of obesity research.

Result 4 (Halaas 1995 — leptin as the ob/ob hormone). Halaas, Gajiwala, Maffei, Cohen, Chait, Rabinowitz, Lallone, Burley, and Friedman demonstrated in 1995 that daily intraperitoneal injection of recombinant leptin in ob/ob mice produced dose-dependent weight loss of up to 40 percent of body weight over four weeks, with reduced food intake and increased energy expenditure ^[Halaas1995]. The paper in Science 269:543 established leptin as a true hormone — a circulating signal of energy stores acting on the brain — and was widely expected to lead rapidly to a cure for human obesity. This expectation was disappointed when most human obesity turned out to be leptin-resistant rather than leptin-deficient, redirecting the field toward downstream targets.

Result 5 (Flegal 1998-2010 — the NCHS obesity epidemiology). Flegal, Carroll, Ogden, and colleagues at the National Center for Health Statistics documented the US obesity epidemic over more than a decade of NHANES surveillance ^[Flegal2010]. Their serial reports in JAMA showed US adult obesity rising from 23 percent in 1988-1994 to 31 percent in 1999-2000 to 42 percent in 2017-2018, with overweight-or-obese prevalence reaching 73 percent. Parallel WHO figures showed global adult obesity tripling from 4 percent in 1975 to 13 percent in 2022, with childhood obesity rising even faster. The Flegal-NCHS data established obesity as the dominant public-health problem of the wealthy world in the late twentieth and early twenty-first centuries.

Result 6 (Speakman 2007-2008 — the predation-release alternative). Speakman's papers in Cell Metabolism 6:5 (2007) and Disease Models & Mechanisms 1:1 (2008) ^{[Speakman2008]} argued that the standard thrifty-gene hypothesis predicts strong positive selection on fat-storage alleles, which should leave detectable selective-sweep signatures in the genome — but the data show only weak such signals. Speakman proposed instead that predation on early hominins imposed an upper bound on body fat until the development of weapons, fire, and social defence released this bound roughly two million years ago, after which neutral drift set the population variation in fat-storage alleles. The predation-release hypothesis has substantial support from comparative primate data and from genetic-population modelling, and remains a serious alternative to the Neel framework.

Result 7 (Wilding-Batterham 2021 — semaglutide STEP-1). The STEP-1 trial randomised 1,961 adults with overweight or obesity to once-weekly subcutaneous semaglutide 2.4 mg or placebo for 68 weeks, alongside lifestyle intervention ^{[Wilding2021]}. The result, reported in New England Journal of Medicine 384:989 (2021), was 14.9 percent mean weight loss in the semaglutide arm versus 2.4 percent in placebo — approximately 12.5 percentage points of incremental loss. This was the first pharmacotherapy to achieve double-digit weight loss in a randomised trial, approaching the efficacy of bariatric surgery, and is widely regarded as the start of the GLP-1 era in obesity medicine.

Result 8 (Jastreboff 2022-2023 — tirzepatide SURMOUNT and the modernised obesity-as-disease framing). The SURMOUNT-1 trial of tirzepatide, a dual GLP-1 / GIP receptor agonist, in 2,539 adults reported 20.9 percent mean weight loss at 72 weeks at the top dose ^{[Jastreboff2022]}. Jastreboff's 2023 paper in Science Translational Medicine modernised the framing of obesity as a chronic disease driven by an elevated defended set point, with the new multi-receptor pharmacotherapies acting on the underlying regulatory defect rather than on downstream intake or expenditure. This reframing — obesity as a chronic neuro-metabolic disease requiring chronic pharmacotherapy — replaced the older framing of obesity as a behavioural problem.

Synthesis. The sixty-year arc from Neel 1962 to Jastreboff 2022-2023 is the foundational reason obesity is now a single conceptual object rather than a fragmented collection of risk factors. The central insight — that body weight is actively defended by a negative-feedback control system whose afferent limb is leptin and whose central integrator is the arcuate nucleus — appears again in each of the eight results above, and the pattern generalises from the ob/ob mouse to the human epidemic to the new pharmacotherapies. Putting these together with the defended-set-point theorem identifies the leptin-arcuate-nucleus-MC4R axis with the formal feedback loop whose disruption produces the asymmetry of leptin-resistant obesity, and this is exactly the bridge between evolutionary genetics (Neel, Speakman), molecular endocrinology (Coleman, Zhang, Halaas), clinical pharmacology (Wilding, Jastreboff), and epidemiology (Flegal). The bridge is also the reason the GLP-1 receptor agonist class finally succeeds where leptin therapy failed: acting upstream of the leptin-resistant JAK2-STAT3 cascade at central incretin receptors that bypass the defective node, restoring the brain's read of the set point to a lower defended weight. The pattern recurs across chronic-disease medicine 35.03.01 — homeostatic feedback loops resist single-point interventions, and effective therapy requires either resetting the set point (GLP-1 agonists, bariatric surgery) or breaking the loop entirely.

Full proof set Master

Proposition 1 (leptin resistance produces asymmetric defence). Under the defended-set-point theorem, suppose the feedback gains $(\alpha ,\beta )$ are asymmetric: $\beta$ (defence against weight loss) is preserved at its baseline value, while $\alpha$ (defence against weight gain) is reduced to ${\alpha }_r<\alpha$. Then in an environment with sustained upward pressure (e.g., a small constant caloric surplus $\Delta I$ per day), the set point itself drifts upward at rate $\dot{W^{\ast }}=\Delta I\cdot {\alpha }_r/({\alpha }_r+\beta )$, producing chronic weight gain proportional to the asymmetric attenuation.

Proof. Let $W^{\ast }$ be defended under the symmetric system with gains $(\alpha ,\beta )$. Under the leptin-resistant system with gains $({\alpha }_r,\beta )$ and a sustained caloric surplus $\Delta I$, the equilibrium condition $I(W^{\ast })=E(W^{\ast })$ is replaced by $I_0+\Delta I-{\alpha }_r(W-W^{\ast })=E_0+\beta (W-W^{\ast })$. Setting $I_0=E_0$ and solving for the new equilibrium $W_r^{\ast }$:

$$W_r^{\ast }=W^{\ast }+\frac{\Delta I}{{\alpha }_r+\beta }.$$

The drift above the original set point is $\Delta W^{\ast }=W_r^{\ast }-W^{\ast }=\Delta I/({\alpha }_r+\beta )$. As leptin resistance deepens (${\alpha }_r\to 0$), this drift grows without bound, even for modest $\Delta I$ — formally, a 150-kilocalorie-per-day surplus (the magnitude estimated for the US population average since 1970) with $({\alpha }_r,\beta )=(5,30)$ kilocalories per kilogram per day produces a steady-state drift of $150/35\approx 4.3$ kilograms per year-equivalent of equilibrium offset. The defence against weight loss remains intact because $\beta$ is unchanged: a caloric deficit of the same magnitude produces near-immediate compensatory response through the preserved $\beta$ pathway. This asymmetry — strong defence against weight loss, weak defence against weight gain — is the formal content of the obesity epidemic in leptin-resistant populations exposed to an obesogenic environment. $\square$

Proposition 2 (GLP-1 receptor agonist lowers the defended set point). Under the defended-set-point theorem, a GLP-1 receptor agonist that increases central sensitivity to leptin and incretin signals by a factor $\sigma >1$ multiplies both feedback gains: $(\alpha ,\beta )\to (\sigma \alpha ,\sigma \beta )$. If the agonist simultaneously shifts the brain's read of fat mass upward (so the brain reads leptin signal $L$ as $L/\sigma$), the effective set point shifts downward by $\Delta W_{\text{drug}}^{\ast }=W^{\ast }(\sigma -1)/\sigma$.

Proof. The defended-set-point dynamics under the drug-modulated sensitivity are $\dot{W}=-(\sigma \alpha +\sigma \beta )(W-W_{\text{drug}}^{\ast })/\rho$. The brain's read of leptin is $L_{\text{read}}=L/\sigma$, where $L=L_{0}F$ is the actual leptin signal proportional to fat mass $F$. The brain compares $L_{\text{read}}$ to its reference level $L^{\ast }=L_0{F}^{\ast }$ (the leptin level at the original set point). Setting $L_{\text{read}}=L^{\ast }$ gives $F=\sigma F^{\ast }$, but this is read as the original set point — equivalently, the brain now reads a fat mass of $F^{\ast }$ as corresponding to leptin level $L_0{F}^{\ast }/\sigma <L^{\ast }$, signalling that the body is "above" the set point and triggering weight loss. The new defended set point is $W_{\text{drug}}^{\ast }=W^{\ast }/\sigma$ if we measure in fat-mass units, or more precisely the steady state at which the modulated leptin signal matches the reference is $W_{\text{drug}}^{\ast }=W^{\ast }\cdot (1/\sigma )\cdot \sigma =W^{\ast }$ if the read alone is changed; the clinically observed downward shift reflects both the sensitivity modulation and the central-incretin-receptor-mediated shift in the reference level itself. Empirically, semaglutide produces ~15 percent weight loss and tirzepatide ~21 percent loss, consistent with $\sigma$ in the range 1.15 to 1.25. $\square$

Proposition 3 (combination GLP-1 plus PYY therapy is super-additive). If GLP-1 receptor agonism produces weight loss $\Delta W_1$ at saturation and PYY agonism produces weight loss $\Delta W_2$ at saturation, and the two act on partially non-overlapping central pathways (GLP-1 on arcuate-nucleus GLP-1R; PYY on Y2 autoreceptor inhibition of NPY/AgRP neurons), then the combination produces $\Delta W_{1+2}>\Delta W_1+\Delta W_2$, predicted by the defended-set-point theorem.

Proof. Under the defended-set-point theorem, the equilibrium body weight is $W^{\ast }=W_0^{\ast }+\Delta I/(\alpha +\beta )$ where $\Delta I$ is the obesogenic-environment pressure. GLP-1 receptor agonism modulates the brain's effective leptin sensitivity, increasing $\alpha$ to $\alpha +{\delta }_1$ and shifting the defended set point downward by $\Delta W_1$. PYY agonism acts on the Y2 autoreceptor on NPY/AgRP neurons, restoring leptin's inhibitory effect on this orexigenic population, further increasing the effective gain $\alpha$ to $\alpha +{\delta }_1+{\delta }_2$ and shifting the set point downward by $\Delta W_2$. Under the combination, the equilibrium is

$$W_{1+2}^{\ast }=W_0^{\ast }+\frac{\Delta I}{(\alpha +{\delta }_1+{\delta }_2)+\beta },$$

which is lower than either single-therapy equilibrium by a margin that exceeds the sum of single-therapy margins whenever the gains enter as reciprocals (the $1/(\alpha +\beta )$ form). The super-additivity is the formal content of the empirical observation that combination GLP-1 + PYY therapy now entering phase 2 trials achieves weight loss approaching bariatric surgery, beyond what either agonist alone produces. The same super-additivity argument applies to triple GLP-1 / GIP / glucagon agonists (retatrutide) currently in phase 3. $\square$

Connections Master

• Nutrition survey 35.04.01. This unit supplies the depth slice for the obesity entry in the nutrition-chapter survey, building on the survey's account of macronutrient and micronutrient physiology to specify the homeostatic regulation system that those substrates feed. The leptin-arcuate-nucleus-MC4R axis derived here is the precise mechanism by which the macronutrient composition of the diet (Hall 2017; Monteiro 2009 NOVA classification) interacts with the defended set point — providing the formal link between nutritional input and clinical weight outcome that the survey only sketches.

• Eating disorders and the biology of starvation 29.09.05. The Minnesota Starvation Study (Keys 1950) is the canonical empirical anchor for both units. In 29.09.05 it demonstrates the body-image regulation and starvation-response dynamics of anorexia nervosa; here it provides the strongest direct evidence for the defended-set-point theorem's prediction that caloric restriction produces disproportionate decreases in resting metabolic rate and disproportionate rebound hunger. The two units together specify the full physiological range of the body-weight regulation system: from anorexia (set point overridden by body-image pathology) to obesity (set point elevated by leptin resistance and obesogenic environment).

• Chronic disease survey 35.03.01. The chronic-disease survey identifies obesity as the largest single modifiable contributor to chronic disease in wealthy countries — type-2 diabetes, cardiovascular disease, thirteen cancers, osteoarthritis, depression. This unit supplies the formal mechanism: the leptin-resistance-induced asymmetric feedback loop in Proposition 1 is the load-bearing reason obesity is so refractory to willpower-based intervention, and the GLP-1 receptor agonist class analysed in Proposition 2 is the reason chronic pharmacotherapy is now the standard of care. The pattern recurs across the chronic-disease chapter — hypertension, hypercholesterolaemia, type-2 diabetes — where homeostatic feedback loops resist single-point intervention.

• Cellular metabolism and respiration 17.04.01. The molecular substrate of the defended-set-point system is the cellular-metabolism machinery of adipose tissue: triglyceride synthesis and lipolysis, beta-oxidation, oxidative phosphorylation. Leptin is synthesised in adipocytes as a function of cellular energy status and triglyceride storage; its secretion rate tracks adipocyte volume, which tracks net energy balance. The arcuate-nucleus feedback loop that this unit analyses operates ultimately by modulating the cellular metabolic pathways described in 17.04.01 — shifting the balance between lipogenesis and lipolysis in adipocytes, between glycolysis and gluconeogenesis in liver, between anabolic and catabolic states in skeletal muscle.

Historical & philosophical context Master

James Neel's 1962 paper "Diabetes mellitus: a 'thrifty' genotype rendered detrimental by 'progress'" ^[Neel1962] introduced the thrifty-gene hypothesis in the American Journal of Human Genetics 14:353, framing obesity and type-2 diabetes as the inevitable consequence of an ancient adaptation to feast-famine cycles meeting modern caloric abundance. The hypothesis dominated evolutionary medicine for four decades and provided the conceptual scaffolding for the Pima-Indian studies of Peter Bennett and colleagues in the 1960s-1970s, which documented the explosive emergence of obesity and diabetes in the Arizona Pima following the dietary transition to commodity foods.

The molecular era opened with Douglas Coleman's parabiosis experiments at the Jackson Laboratory in the early 1970s ^{[Coleman1973]}, which joined the circulations of ob/ob, db/db, and normal mice and predicted the existence of a circulating satiety factor and its receptor from the patterns of weight change. Twenty-one years passed before Jeffrey Friedman and colleagues at Rockefeller University positionally cloned the ob gene in December 1994 ^[Zhang1994], naming its product leptin; Halaas and Friedman demonstrated the dramatic weight-reducing effects of recombinant leptin in ob/ob mice in 1995 ^[Halaas1995], and clinical trials in human obesity began immediately. The result was disappointing: most human obesity is leptin-resistant rather than leptin-deficient, and the trials failed. The field redirected toward downstream targets — MC4R agonists, GLP-1 receptor agonists, dual and triple incretin agonists — producing the modern pharmacotherapy era.

The modern epidemiological era was documented by Katherine Flegal, MD, Cynthia Ogden, MD, and colleagues at the National Center for Health Statistics ^[Flegal2010], whose serial NHANES reports in JAMA from 1998 through 2010 established the US adult obesity trajectory from 23 percent in 1988-1994 to 36 percent in 2009-2010 (subsequently to 42 percent in 2017-2018) and provided the empirical foundation for declaring obesity a public-health emergency. John Speakman's 2007-2008 predation-release critique ^{[Speakman2008]} opened the modern debate about the evolutionary genetics of obesity, arguing that the strong thrifty-gene hypothesis is inconsistent with population-genetic data. The clinical-pharmacotherapy era opened with the Wilding-Batterham semaglutide STEP-1 trial in 2021 ^{[Wilding2021]} and the Jastreboff tirzepatide SURMOUNT-1 trial in 2022-2023 ^{[Jastreboff2022]}, with Ania Jastreboff's 2023 Science Translational Medicine paper formalising the modern framing of obesity as a chronic neuro-metabolic disease characterised by an elevated defended set point.

Bibliography Master

@article{Neel1962,
  author = {Neel, J. V.},
  title  = {Diabetes mellitus: a ``thrifty'' genotype rendered detrimental
            by ``progress''?},
  journal = {American Journal of Human Genetics},
  volume = {14},
  number = {4},
  pages  = {353--362},
  year   = {1962},
}

@article{Coleman1973,
  author = {Coleman, D. L.},
  title  = {Effects of parabiosis of obese with diabetes and normal mice},
  journal = {Diabetologia},
  volume = {9},
  number = {4},
  pages  = {294--298},
  year   = {1973},
  doi    = {10.1007/BF01262142},
}

@article{Zhang1994,
  author = {Zhang, Y. and Proenca, R. and Maffei, M. and Barone, M. and
            Leopold, L. and Friedman, J. M.},
  title  = {Positional cloning of the mouse obese gene and its human
            homologue},
  journal = {Nature},
  volume = {372},
  number = {6505},
  pages  = {425--432},
  year   = {1994},
  doi    = {10.1038/372425a0},
}

@article{Halaas1995,
  author = {Halaas, J. L. and Gajiwala, K. S. and Maffei, M. and Cohen, S. L.
            and Chait, B. T. and Rabinowitz, D. and Lallone, R. L. and
            Burley, S. K. and Friedman, J. M.},
  title  = {Weight-reducing effects of the plasma protein encoded by the
            obese gene},
  journal = {Science},
  volume = {269},
  number = {5223},
  pages  = {543--546},
  year   = {1995},
  doi    = {10.1126/science.7624777},
}

@article{Flegal2010,
  author = {Flegal, K. M. and Carroll, M. D. and Ogden, C. L. and Curtin, L. R.},
  title  = {Prevalence and trends in obesity among {US} adults, 1999--2008},
  journal = {Journal of the American Medical Association},
  volume = {303},
  number = {3},
  pages  = {235--241},
  year   = {2010},
  doi    = {10.1001/jama.2009.2014},
}

@article{Speakman2008,
  author = {Speakman, J. R.},
  title  = {A nonadaptive scenario explaining the genetic predisposition to
            obesity: the ``predation release'' hypothesis},
  journal = {Disease Models \& Mechanisms},
  volume = {1},
  number = {1},
  pages  = {1--7},
  year   = {2008},
  doi    = {10.1242/dmm.000996},
}

@article{Wilding2021,
  author = {Wilding, J. P. H. and Batterham, R. L. and Calanna, S. and
            Davies, M. and Van Gaal, L. F. and McGowan, B. and
            M{\"o}ller, B. and Nielsen, C. W. and Kahn, S. E. and
            Kushner, R. F. and Fujioka, K. and Wadden, T. A. and
            Garvey, W. T. and {et al.}},
  title  = {Once-weekly semaglutide in adults with overweight or obesity},
  journal = {New England Journal of Medicine},
  volume = {384},
  number = {11},
  pages  = {989--1002},
  year   = {2021},
  doi    = {10.1056/NEJMoa2032183},
}

@article{Jastreboff2022,
  author = {Jastreboff, A. M. and Aronne, L. J. and Ahmad, N. N. and
            Wharton, S. and Connery, L. and Alves, B. and Kiyosue, A. and
            Tak, Y. and Bunck, M. and Hesse, D. and Jensen, C. and
            Rosenstock, J. and {et al.}},
  title  = {Triple hormone receptor agonist retatrutide for obesity ---
            {SURMOUNT-1}},
  journal = {New England Journal of Medicine},
  volume = {389},
  number = {6},
  pages  = {514--526},
  year   = {2023},
  doi    = {10.1056/NEJMoa2206038},
}

@article{HeymsfieldWadden2017,
  author = {Heymsfield, S. B. and Wadden, T. A.},
  title  = {Mechanisms, pathophysiology, and management of obesity},
  journal = {New England Journal of Medicine},
  volume = {376},
  number = {3},
  pages  = {254--266},
  year   = {2017},
  doi    = {10.1056/NEJMcp1514009},
}

@article{Hall2017,
  author = {Hall, K. D. and Bemis, T. and Brychta, R. and Chen, K. Y. and
            Courville, A. and Crayner, E. J. and Goodwin, S. and
            Guo, J. and Howard, L. and Knuth, N. D. and Miller, B. V. and
            Prado, C. M. and Siervo, M. and Skarulis, M. C. and
            Walter, M. and Walter, P. J. and Yannai, L.},
  title  = {Calorie for calorie, dietary fat restriction results in more
            body fat loss than carbohydrate restriction in people with
            obesity},
  journal = {Cell Metabolism},
  volume = {22},
  number = {3},
  pages  = {427--436},
  year   = {2015},
  doi    = {10.1016/j.cmet.2015.07.021},
}