How To Improve Facial Dimorphism: The Full Protocol

Facial sexual dimorphism is the set of sex-typical differences in the face after puberty. This page is how to improve facial dimorphism in men by changing cortical geometry, not by chasing one hormone. The levers are androgen signalling, thyroid T3, organized mechanical load, and substrate. Testosterone can masculinize facial cortical geometry through conversion to DHT, androgen-receptor activation on osteoblasts, and periosteal apposition that widens bone externally. Mandibular osteoblasts express functional AR and respond to DHT (PMID 9383273). DHT and testosterone increase periosteal bone formation even without GH-receptor signaling (PMID 17014385). Cyclic masticatory strain then tells those cells where to deposit matrix through integrin/PIEZO → YAP/TAZ mechanotransduction. Serum T is substrate; DHT, AR, load, and GH/IGF-1 are the machinery. For the trait itself, start at what facial dimorphism is.

Facial sexual dimorphism is how male and female faces differ after puberty: jaw width, brow position, lip height, eye aperture, facial hair, skin. It is a developmental and endocrine trait, not a single lab number. Male-typical faces run wider in the mandible and heavier in the brow. Female-typical faces run higher in the brow, fuller in the lips, narrower in the jaw. The longer definition, including whether dimorphism equals attractiveness, is what facial dimorphism is. This page is the male protocol.

Most dimorphic facial bone is laid down in adolescence. Adult periosteal apposition continues, slower (PMID 5917324, PMID 2399958). Week to week, a photo moves more from fat, water, beard, masseter tone, and lens than from new cortex. Bone is months. Fat, hair, and muscle are faster. An honest plan uses both.

To improve facial dimorphism you work four levers at once: androgen signalling (DHT reaching androgen receptors on osteoblasts), thyroid T3, organized mechanical load on the sutures and palate, and enough energy and mineral substrate to build matrix. No single one of them moves facial sexual dimorphism on its own.

Most facial skeletal development occurs during adolescence and early adulthood. Bone changes in adults are slower and limited; visible differences are more often caused by muscle, body fat and posture.

Practical steps:

• Chewing. Use plain Chios mastic resin only. Start with five to ten minutes and increase only if there is no joint pain, clicking or locking.
• Breathing. Breathe through the nose when possible and keep the tongue resting comfortably on the palate. In adults, this may support breathing and posture but has not been shown to reshape facial bones. See oral posture in facial analysis.
• Posture and training. Maintain a neutral head position and use regular resistance training. These affect posture and muscle, not mandibular width.
• Nutrition. Meet energy, protein and mineral requirements. Raw milk supplies lactoferrin and myo-inositol, compounds with osteogenic and mandibular-growth evidence discussed below.

Bone remodeling takes months. Stack androgen/T3 signaling with Suture Traction Therapy, static thumbpulling, and cyclic thumbpulling with Buteyko. Hormonal drugs and T3 dosing still need medical supervision.

Male facial skeletal development depends on androgen-receptor signaling, GH/IGF-1, thyroid hormone, periosteal apposition, mechanical loading and osteoblast activity. DHT binds the androgen receptor without aromatization. Human osteoblasts, including cells derived from the mandible, express functional androgen receptors and respond to DHT. PMID 9383273. Androgen-receptor signaling drives cortical expansion and periosteal bone formation. DHT and testosterone increased periosteal bone formation even in GH-receptor-deficient males, while androgen-receptor disruption cut normal cortical expansion during growth. PMID 17014385, PMID 15312246, PMID 25407961. Periosteal apposition deposits new cortical matrix on the external surface of bone and increases cross-sectional diameter. At the pubertal dimorphic peak, males showed about 40% greater radial expansion and roughly 70% higher periosteal bone formation than females, with GH/IGF-1 also contributing. PMID 19888832.

Best practical results come from stacking endocrine signaling with a mechanical routine. Run Suture Traction Therapy first, then a static-force thumbpulling foundation, then cyclic thumbpulling locked to Buteyko nasal breathing, then myofunctional tongue work while the tissues are still adapted. Static holds create a sustained directional field across the hard palate and circummaxillary sutures. Cyclic reload amplifies oscillatory mechanotransduction on top of that field. Matching peak force with a cyclic waveform grows sutures harder than a static hold alone (PMID 12619937). Brief high-frequency cyclic tension or compression widens sutures, raises cell density, and increases osteoblast-occupied surface (PMID 18032124). Small oscillatory strain doses accelerate sutural growth (PMID 12454093). Cranial suture traction therapy is a structured facial-asymmetry protocol (PMID 35888588). Expanded maps and timing live in the Craniofacial Mechanobiology Atlas.

Full session, in order. Warm first. Force stays on the hard palate, never on teeth, gingival margins, or soft palate. Build static control before you add rhythm.

How much force, and why. You are dosing strain into sutures and periosteum, not crushing bone. Equal bilateral contact and a clean directional vector beat maximum effort. Start at about 40–60% of a hard pull: firm enough that the palate feels a clear outward / diagonal stretch and the tissue creeps a little under a 20–30 second hold, light enough that you can breathe nasally, keep the jaw unclenched, and stop on command. If teeth hurt, the TMJ clicks/locks, one side dominates, or you need to brace the neck, you are too heavy or off the bone. Sutural growth responds to organized strain, including small oscillatory doses; the same peak force as a cyclic waveform outperforms a static slam (PMID 12619937, PMID 12454093). That is why the session is static field first, then cyclic reload, not one max yank. Progress by cleaner symmetry and longer clean holds, not by grinding harder each week.

• Warm-up. Hot shower or bath so face and neck are warm and adaptable.
• Masseter / temporalis release. External massage, then gentle intraoral release. Drop tonic clench before suture load.
• Suture Traction Therapy. Facial suture sequence first, especially zygomaticomaxillary and frontozygomatic for midface / asymmetry (PMID 35888588). Steady manual traction, not a sudden jerk.
• Static thumbpulling, middle palate. Thumbs on hard palate, outward and slightly diagonal. Hold ~30 seconds at the force described above, release, one calm nasal breath, repeat 4–6 holds.
• Static thumbpulling, anterior hard palate. Around the incisive papilla. Same force band. Controlled bilateral / diagonal traction. Zero load on the incisors.
• Cyclic thumbpulling. Only after static holds feel controlled and symmetrical. Progressive load → progressive unload on a steady rhythm at the same peak force, not higher. No bouncing.
• Buteyko lock-in. Mouth closed, light diaphragmatic nasal breathing. Match traction to the breath: load on one phase, unload on the other.
• Myofunction immediately after. The three tongue drills below, while tissues are still adapted. Lock the session into resting oral posture.
• Finish. Atlas massage, rib-cage and pelvis stacking, quiet nasal breathing.

Three tongue drills after the session. Traction is temporary. The tongue delivers thousands of low-level force cycles every day; if it stays low or asymmetric, the palate loses the resting load that keeps the work useful. Do these right after thumbpulling. Daytime default: nasal breathing, tongue on the palate, lips closed, jaw unclenched. More context: myofunctional analysis.

• Tongue–palate suction hold. Broad suction across the hard palate, posterior tongue elevated, teeth lightly apart, jaw relaxed. Hold 15–20 seconds. 5 reps. This is the resting load you want all day.
• Tongue against spoon. Press the tongue firmly into a spoon with lips gently closed and no jaw brace. Hold 30 seconds. 2 sets. Builds real upward force without teeth compensation.
• T-release + posterior elevation. Articulate a clear T while keeping the back of the tongue high on the palate for 10–15 seconds. 3 sets. Opens mobility, then parks the tongue where it belongs.

Buteyko breathing pairs with cyclic thumbpulling. Use nasal, reduced-volume breathing and coordinate palatal traction with the respiratory cycle during suture traction and thumbpulling. Mouth closed, light diaphragmatic nasal breathing. Three mechanisms matter for facial structure work: lip seal and tongue–palate contact prevent midface load from reverting to oral breathing; nasal inspiration carries paranasal nitric oxide into the lungs and improves oxygenation versus mouth breathing (PMID 8971255); control pauses and reduced breathing raise CO₂ tolerance and set a repeatable load–unload rhythm. Randomized asthma trials report fewer symptoms and less bronchodilator use with the Buteyko breathing technique (PMID 12885982, PMID 38212823). For facial dimorphism protocols, the target is nasal mechanics plus controlled cyclic force, not breathless maximum effort.

Breathwork, BDNF, and circulating progenitors. Brief intermittent hypoxia from controlled breath holds can increase BDNF and hippocampal neurogenesis through BDNF–TrkB signaling (PMID 20861371). In adult men, intermittent hypoxia mobilizes CD45⁺CD34⁺ hematopoietic progenitors during exposure (PMID 21962068). Osteoblasts express TrkB; BDNF increases VEGF secretion, osteoblast migration, and osteogenic programs (PMID 28098876, PMID 32803867). Protocol dose matters: mild nasal Buteyko during cyclic traction is the practical set point, distinct from pathologic sleep-apnea hypoxia.

Adult periosteal apposition continues after longitudinal growth. Tetracycline labeling in human ribs shows periosteal bone formation past age 20 and into the seventh decade, at a declining rate (PMID 5917324). Longitudinal skeletal studies document continuing periosteal apposition in adulthood (PMID 2399958). Male cortical widening depends on androgen and estrogen receptor biology (PMID 16303833); androgens stimulate periosteal formation even when GH-receptor signaling is absent (PMID 17014385). Adult facial bone remodeling is slower than puberty; stack DHT/AR, T3, mechanical load, and nasal breath rhythm and track change over months.

Figure 1 shows where four facial dimensions are read on a front portrait: brow position at the supraorbital rim, mandibular width at the lower-face contour, visible lip height, and palpebral aperture at the eyes. Masculine-typical averages tend toward a lower, heavier brow, a broader mandible, less lip height relative to midface width, and a narrower palpebral aperture. The portrait is illustrative; these callouts are not biometric measurements or a hormone assessment.

Brow positionJaw widthLip heightEye aperture

T3 is just as important and massively overlooked. Bone and cartilage are genuine T3 target tissues. Circulating T4 has to reach skeletal cells and be converted locally through DIO2 into active T3, while DIO3 opposes this by inactivating thyroid hormone. T3 enters cells through transporters such as MCT8/SLC16A2 and MCT10/SLC16A10, binds predominantly TRα1/THRA in the skeleton, alters recruitment of nuclear corepressors and coactivators at thyroid-response elements, and changes programs controlling chondrocyte maturation and osteoblast differentiation. This feeds into RUNX2, SP7/Osterix, SOX9, COL2A1, ACAN, IHH/PTHrP, COL10A1, MMP13, VEGFA, ALPL, osteocalcin and matrix mineralization. T3 also changes mitochondrial oxidative metabolism, ATP production and redox state, meaning thyroid signaling affects whether an osteoblast actually has the energetic capacity to synthesize collagen, secrete matrix and mineralize it. The skeleton is described in the literature as an exquisitely sensitive T3 target tissue for a reason. Start with PMID 26862888 and PMID 24783033, alongside the broader thyroid skeletal research library, including PMIDs 11014246, 31974498, 27959899, 28391432, 11524248, 20707983, 21886774, 3476664, 267106, 15126140, 11706949 and 16246940. T3 deficiency during development is not some cosmetic metabolic inconvenience. It directly changes skeletal maturation, growth-plate physiology and bone turnover. At the opposite extreme, chronically excessive thyroid hormone accelerates remodeling and can become catabolic, so the goal is strong physiological thyroid function, not randomly forcing T3 as high as possible.

Progesterone matters in males too. Human osteoblasts express progesterone receptor isoforms PR-A and PR-B, and progesterone directly increased human osteoblastic cell proliferation and increased IGF-2 production in vitro. PMID 9632845, PMID 1375800. The interesting part is that PR biology is highly dependent on sex and differentiation stage. A mouse conditional-knockout study found that progesterone receptor signaling in early mesenchymal/osteoprogenitor populations altered trabecular bone acquisition, with particularly strong changes in male osteogenic potential, while deleting PR later in mature osteoblasts produced very different results. PMID 28569405. Another experiment found progesterone stimulated osteoprogenitor proliferation and differentiation in adult female-derived cultures but not the male-derived cultures used in that particular model, which is exactly why progesterone should not be reduced to “female hormone good/bad.” PMID 8988343. Receptor abundance, developmental stage, androgen/estrogen environment and cell lineage determine the result. Progesterone is therefore part of male skeletal endocrinology, but it is not evidence that deliberately driving progesterone above the physiological male range will masculinize bone.

Then comes the nutritional environment. I would build it around raw unfrozen, preferably never-refrigerated milk, raw butter, raw cheese, raw cream, raw honey, raw orange juice, raw sugar-cane juice, coconut cream, raw oysters, non-steamed dates and raw vanilla extract, with the point being to supply enough energy, protein, fats, minerals and micronutrients for endocrine and skeletal machinery instead of chronically running the body underfed.

Raw milk supplies lactoferrin, a protein with direct anabolic effects on bone cells. In primary human and rat osteoblast cultures, lactoferrin increased proliferation up to fivefold, increased differentiation, reduced apoptosis by 50–70%, and suppressed osteoclastogenesis. Applied locally over adult mouse calvariae, it increased new bone formation fourfold. PMID 15166119. The 2007 *Report in Favor of Natural Milk* also lists lactoferrin among the native proteins and enzymes in raw milk and cites Price's observation that processed-food diets paralleled poor facial-bone development.

Cow's milk also contains free and bound myo-inositol; mature cow's milk measured about 4.1 mg per 100 mL total myo-inositol. PMID 1151503. In growing mice, dietary myo-inositol increased mandibular length and condylar-cartilage thickness without increasing maxillary or femoral length. PMID 30682567. Growing rabbits reproduced the mandibular-selective effect. PMID 38534273. The evidence is direct for lactoferrin's osteogenic activity and for myo-inositol's effect on mandibular development in growing animals. What remains unquantified is how much ordinary raw-milk intake changes adult human facial dimensions.

Raw cream, butter and cheese then raise the energy, lipid, calcium, phosphorus and protein density around the same skeletal environment. Raw honey provides glucose and fructose together with enzymes and plant-derived compounds, but there is currently no human experiment showing that its endogenous enzymes specifically increase DHT, T3 or periosteal facial growth, so do not invent that claim. Orange juice, sugar-cane juice and dates provide rapidly available carbohydrate that can help prevent chronic low-energy availability, while coconut cream gives another dense fat source. Raw oysters are particularly interesting because of their zinc density. Human zinc-depletion experiments show that low zinc can reduce circulating testosterone, while correction of zinc deficiency can restore it. PMID 8875519, PMID 1609752. That does not mean zinc infinitely raises testosterone; it means zinc sufficiency is required for normal gonadal physiology.

Raw vanilla extract, around 2 teaspoons in a smoothie or juice, fits easily into the food framework and adds vanillin and related aromatic phenolics. It blends naturally with raw milk, cream, honey or orange juice and adds a concentrated aromatic layer to the stack.

T3 supplementation can help. Bone and cartilage are T3 target tissues. When local T3 signaling through TRα1/THRA is weak, chondrocyte maturation, osteoblast differentiation and matrix mineralization run underpowered. Restoring physiological T3 restores those programs. That can mean prescription T3, carefully dosed desiccated thyroid, or raw thyroid gland that actually delivers active hormone. The skeletal literature is blunt about this: T3 is not optional scenery for facial bone work (PMID 26862888, PMID 24783033, and the thyroid research library). Dose for strong physiological function. Chronic mega-dosing into a catabolic thyroid state is not the goal.

The actual objective is therefore not to obsess over one hormone. You want an endocrine and mechanical environment where androgen receptor signaling, physiological DHT, GH/IGF-1, T3, progesterone-receptor biology, adequate energy availability and mechanical loading converge on osteoblasts and periosteal progenitors. Those cells then integrate endocrine signals with integrin → FAK/Src → RhoA/ROCK, PIEZO-mediated Ca²⁺ signaling, ERK and YAP/TAZ, eventually feeding transcriptional programs involving RUNX2, SP7, COL1A1, ALPL and matrix mineralization. That is where masculine skeletal geometry actually comes from. You are not “boosting masculinity” with one supplement. You are building the biochemical conditions under which male skeletal dimorphism can actually be expressed. Trait definition: what facial dimorphism is. Broader sequence: the Healthmaxxing guides.

For the full mechanics, measurement maps, and ordered practice stack, continue with the Craniofacial Mechanobiology Atlas.