Does love nourish or consume? Let’s imagine an organism immersed in water. If the water is clear, the body floats effortlessly, the muscles relax, and breathing finds its natural rhythm. If the water is murky or agitated, the same body begins to stiffen: every movement becomes a defense, every second consumes energy.
Love, on a clinical level, works in the same way. It is not what we feel that determines its healthiness, but the physiological and emotional environment it creates around us. A bond can be as nourishing as a favorable current or as corrosive as water that slowly poisons.
Healthy love is not just a “beautiful” feeling: it is a stable neuropsychological balance, a choreography of mutual regulation that, repeated over time, becomes biology. In clinical practice, it can be recognized by a seemingly simple fact: with that person, the body stops living in alert mode. The body does not have to “earn” closeness, it does not have to earn peace through performance, nor pay for security with self-deprecation. Intimacy, in a mature bond, does not coincide with fusion: rather, it resembles a pact of proximity that preserves two identities. It is precisely this balance—closeness without annihilation, desire without possession, care without control—that distinguishes nourishing love from consuming love.
Contemporary neuroscience describes falling in love and attachment as highly salient motivational states, in which the partner becomes a privileged stimulus for the reward circuits and for the systems that guide attention, memory, and decision-making processes. Numerous neuroimaging studies show how the experience of love recruits a repertoire of brain networks involved in reward, social evaluation, and interoceptive integration, confirming its deeply embodied and motivational nature.
In recent years, research has begun to move beyond an undifferentiated view of love, distinguishing more finely between its different forms—romantic, parental, friendly—and their respective neural imprints. This approach helps to redefine the idea, still widespread in popular culture, of a single “brain signature” of love. A 2024 study showed how different objects of love selectively modulate the circuits of reward and social cognition, proposing a more articulated neuroaffective taxonomy that is more in line with the complexity of human experience.
In the same year, a meta-analysis of fMRI activations in the different stages of romantic love highlighted patterns that, in part, resemble those observed in addiction disorders. This parallelism invites a non-moralistic but neurofunctional interpretation of the craving-like quality that characterizes some intense romantic experiences. However, the clinical point remains decisive: the activation of reward circuits, in itself, does not allow us to distinguish a healthy bond from an addictive dynamic. What discriminates is not the intensity of the activation, but the quality of the regulation—affective, cognitive, and relational—that supports it.
In a healthy bond, regulation is bidirectional and reparative. We influence each other without using the other as an emotional prosthesis. The difference is subtle yet measurable: functional love produces predictability, and predictability is one of the most powerful nutrients for the nervous system. When the relationship is reliable, the brain reduces vigilance and frees up resources for exploration, creativity, non-compulsive sexuality, and planning.
This is where neuroimmunomodulation comes into play: intimate relationships, for better or worse, are not just ‘psychology’, they become inflammatory tone, endocrine balance, cellular immunity. This is a well-known idea in general terms; less well known is how subtle some recent evidence is. A 2023 study based on daily life highlighted how time spent in physical presence with a partner is associated with lower levels of C-reactive protein (CRP), controlling for variables such as perceived relationship quality, hostility, and loneliness.
To put it bluntly: it is not just a matter of ‘getting along’; it is also a matter of being physically present in the other person’s sensory perimeter.
Conversely, chronic conflict and, even more so, emotional incommunicability seem to leave their mark on the body through immune and genomic pathways. In recent years, relational psychoneuroimmunology has begun to measure not only cytokines or plasma markers, but even changes in gene expression in response to marital stress.
One particularly telling strand shows transient increases in pro-inflammatory expression during laboratory tasks that evoke couple distress or tension; Surprisingly, in some paradigms, listening to a partner’s distress can evoke an inflammatory response even more pronounced than open conflict, suggesting that it is not only confrontation that ‘does damage’, but also empathic impotence when a framework of security and containment is lacking.
A healthy love, then, is also a favorable “immunological environment”: not because it eliminates stress (it would be childish to expect that), but because it metabolizes it. A well-functioning couple does not avoid friction; it transforms it. Repair—knowing how to return, recognizing the impact, renegotiating boundaries—is more important than the absence of friction.
From a clinical point of view, I often put it this way: compatibility is not about never hurting each other, it is about knowing how to heal together without creating identity scars.
Emotional dependence, on the other hand, is love without ‘breathing space’. It is not too much love: it is too much fear masked as love. Its core is asymmetry: one of the two (sometimes both, alternately) does not use the relationship to grow, but to quell anxiety. Here, neurobiology takes on an almost didactic form. The other becomes an intermittent stimulus: sometimes available, sometimes opaque, sometimes seductive, sometimes punitive. This intermittent unpredictability is the perfect fuel for compulsive consolidation: the brain learns that the reward is not guaranteed, and for this very reason increases its attentional investment, rumination, and search for signals.
It is the same principle by which variable reinforcements maintain stubborn behaviors: it is not constant gratification that creates addiction, it is inconsistent gratification.
It is not surprising, therefore, that some of the most recent literature is questioning “love addiction” as a clinical construct (albeit with diagnostic caution, because the official nosographic categories do not include it in an unambiguous way). A 2025 study published in Behavioural Brain Research linked symptoms of love addiction to subjective cognitive difficulties (daily memory, attentional lapses), mediated by anxiety and depression and with a significant role played by social media abuse as a behavioral amplifier (monitoring, control, comparison, algorithmic jealousy).
In parallel, longitudinal studies on digital dynamics in couples indicate that attachment anxiety can more easily translate into electronic surveillance and platform-mediated jealousy, with repercussions on relationship satisfaction: a circuit in which technology does not ‘create’ insecurity, but makes it more accessible, faster, and more obsessive.
The clinical criterion for distinguishing intense love from addiction is not how much one thinks about the other, but what happens when the other is not there (or does not respond). In healthy love, absence triggers longing and desire; in addiction, it triggers disorganization: panic, anger, protest behaviors, devaluation, control, self-humiliation. The other is not a ‘you’, they are an external regulator: without them, the system collapses.
Here, attachment patterns remain fundamental, but they must be read with contemporary eyes: anxious and avoidant attachment are not moral labels, they are learned survival strategies, often consistent with histories of emotional unpredictability. And today we know that these strategies are not just narrative: they modulate reactive stress, inflammation, and, in some cases, measurable immune responses during couple interactions, as shown by literature that has for years linked attachment styles and inflammatory indices in conflict contexts.
If we broaden our view, another uncomfortable truth emerges: the quality of our bonds interacts with the biology of loneliness. A recent study in Nature Human Behaviour (UK Biobank, over 42,000 participants, proteomics on thousands of plasma proteins) identified protein signatures associated with social isolation and loneliness, implicated in inflammation, antiviral responses, and complement systems; with prospective follow-up, many of these proteins were linked to morbidity and mortality. Even more interestingly, Mendelian randomization analyses suggested causal directions from loneliness to specific proteins (including ADM), outlining a plausible biological pathway through which relational deprivation can be “written” into the body.
This does not mean that “being in a couple” is automatically healthy; it means that the human system is designed for relational security, and when security is lacking—or when the bond becomes threatening—the organism pays a price.
That is why, in defining the prerequisites for healthy love, I use four criteria that are not romantic, but clinical: freedom, reliability, reciprocity, and mentalization. Freedom: being able to say “no” without fear of emotional retaliation. Reliability: consistency between words and behaviors, because consistency is the most natural sedative for the amygdala. Reciprocity: giving and receiving without obsessive accounting, but also without unilateral plundering. Mentalization: the ability to represent the other’s mind without confusing it with one’s own fears, that is, to distinguish what is happening from what is feared.
When these pillars exist, oxytocin (which popular discourse tends to mythologize) is not “the love molecule,” but part of a larger system: it facilitates trust and closeness in contexts perceived as safe, while in threatening contexts it can amplify social salience, even in a defensive sense.
Even epigenetics becomes relevant here, not as determinism, but as a trace: work on oxytocin receptor methylation (OXTR) continues to explore links between stress, relationship quality, and biological adaptations, especially in sensitive phases of life and attachment transitions (e.g., the transition to parenthood).
In a dependent relationship, however, these pillars are replaced by surrogates: freedom by concessions, reliability by intermittent promises, reciprocity by pursuit, mentalization by obsessive reading of signals (and often by paranoid interpretations).
Emotional language changes: we no longer talk about needs, we speak in allusions; we don’t ask, we test; we don’t expose vulnerability, we use vulnerability as a weapon or currency. And the body, predictably, adapts: hypervigilance, insomnia, nervous hunger or emotional anorexia, mood swings, cravings, controlling behaviors.
This is where the therapist must not limit themselves to ‘explaining’: they must offer a corrective experience of regulation, because emotional dependencies are not dissolved by logic, but by new neurophysiological trajectories made possible by repeated experiences of security.
The most interesting science on these issues is not that which “romanticizes” biology, but that which makes it ethical. If bonding can lower CRP through simple daily coexistence, then healing is not just a value, it is an intervention on the body. If marital distress can reverberate in pro-inflammatory signatures, then humiliation is not a character trait: it is a repeated microtrauma with potential somatic costs. If loneliness can be correlated with proteomic profiles associated with disease risk, then connection is not a luxury, but a dimension of prevention.
In the end, healthy love resembles fertile soil: it does not eliminate adverse seasons, but allows roots to deepen without rotting. Dependence, on the other hand, is a closed and overheated greenhouse: at first it accelerates growth, then it takes away oxygen.
The difference is not in the intensity of the light, but in the quality of the air we breathe.
