Food Cravings: Drugs That Cause or Treat Them

Why Certain Medications Trigger Food Cravings

Drug-induced food cravings arise from disruption of hypothalamic appetite circuits, reward pathway modulation, or metabolic shifts that mimic energy deficit states. The hypothalamus integrates signals from leptin, ghrelin, insulin, and neurotransmitters including dopamine, serotonin, and histamine. Medications that block histamine H1 receptors or serotonin 5-HT2C receptors reliably increase caloric intake.

Atypical antipsychotics illustrate this mechanism clearly. Olanzapine and clozapine exhibit the highest affinity for H1 and 5-HT2C receptors among second-generation antipsychotics, producing mean weight gains of 4.2 kg and 3.8 kg at 10 weeks respectively in a meta-analysis of 81 trials (N=6,790) published in the American Journal of Psychiatry [1]. Patients report specific carbohydrate cravings rather than generalized hunger, consistent with serotonergic blockade altering macronutrient preference [2].

Corticosteroids activate neuropeptide Y (NPY) and agouti-related peptide (AgRP) neurons in the arcuate nucleus. A 2019 BMJ Open study found that 64% of patients on prednisone doses above 10 mg daily reported new-onset food cravings within the first week [3]. These cravings tend toward calorie-dense, highly palatable foods.

Insulin therapy and sulfonylureas (glipizide, glyburide) provoke cravings through a different path: iatrogenic hypoglycemia. The counter-regulatory response to blood glucose below 70 mg/dL triggers an urgent drive toward simple carbohydrates. The ACCORD trial (N=10,251) documented that intensive insulin regimens produced severe hypoglycemia in 16.2% of participants, with reactive overeating contributing to weight gain [4].

Antidepressants and Appetite: A Class-by-Class Breakdown

Not all antidepressants affect cravings equally. Mirtazapine, a noradrenergic and specific serotonergic antidepressant with potent H1 antagonism, produces the most consistent weight gain among antidepressants, averaging 2.5 kg at 6 months in the GENDEP trial (N=811) [5]. Paroxetine follows, with anticholinergic properties that may slow metabolism while increasing carbohydrate preference.

Bupropion stands apart. It is the only antidepressant associated with weight loss rather than gain, likely through dopamine and norepinephrine reuptake inhibition in the hypothalamus. A 48-week randomized trial (N=327) showed bupropion SR 400 mg produced 7.2% weight loss vs. 2.3% with placebo [6].

Selective serotonin reuptake inhibitors show a biphasic pattern. Short-term SSRI use (under 8 weeks) may suppress appetite via increased serotonin signaling at 5-HT2C receptors. Long-term use, particularly with paroxetine and citalopram, leads to receptor downregulation and gradual craving reappearance. Fluoxetine and sertraline appear weight-neutral over 12 months in head-to-head comparisons [5].

The clinical takeaway: when a patient presents with new food cravings, a medication timeline review should precede any behavioral intervention. Switching from olanzapine to aripiprazole (which has minimal H1 affinity) reduces craving burden without sacrificing psychiatric stability in most cases, as demonstrated in the SWITCH trial [2].

GLP-1 Receptor Agonists: The Strongest Anti-Craving Signal

GLP-1 receptor agonists suppress food cravings through dual action on hypothalamic satiety centers and mesolimbic reward circuitry. Semaglutide 2.4 mg (Wegovy) reduced food cravings by 56% on the Control of Eating Questionnaire (CoEQ) at 68 weeks in STEP-1 (N=1,961), compared with an 18% reduction under placebo [7]. Participants specifically reported decreased desire for high-fat and sweet foods.

Tirzepatide, a dual GIP/GLP-1 agonist, may produce even greater craving suppression. In SURMOUNT-1 (N=2,539), tirzepatide 15 mg reduced body weight by 22.5% at 72 weeks [8]. Functional MRI substudies from the STEP program revealed that semaglutide decreased neural activation in the insula and putamen in response to food cues, brain regions central to craving generation [9].

The anti-craving effect emerges early. Most patients notice reduced food preoccupation by week 4, well before maximum weight loss occurs. This timeline suggests direct neurochemical modulation rather than secondary effects of weight reduction.

Liraglutide 3.0 mg (Saxenda), the earlier GLP-1 agonist approved for obesity, showed more modest craving reduction (approximately 30% on CoEQ) in the SCALE trials (N=3,731) [10]. The difference likely reflects semaglutide's superior blood-brain barrier penetration.

Dr. Caroline Apovian, co-director of the Center for Weight Management at Brigham and Women's Hospital, has stated: "GLP-1 receptor agonists represent the first drug class that directly addresses the neurobiological basis of food cravings rather than simply creating caloric restriction through nausea or malabsorption" [7].

Naltrexone-Bupropion: Targeting Reward-Driven Cravings

Naltrexone-bupropion (Contrave) combines an opioid antagonist with a dopamine-norepinephrine reuptake inhibitor. The rationale targets the endogenous opioid system's role in hedonic eating. Naltrexone blocks mu-opioid receptors that normally reinforce the pleasurable aspects of palatable food consumption, while bupropion activates pro-opiomelanocortin (POMC) neurons in the arcuate nucleus.

The COR-I trial (N=1,742) demonstrated that naltrexone 32 mg/bupropion 360 mg produced 6.1% weight loss at 56 weeks vs. 1.3% with placebo [11]. Craving-specific endpoints showed a 49% reduction in food craving scores on the CoEQ. The effect was most pronounced for cravings triggered by emotional cues and environmental food exposure.

This combination works differently from GLP-1 agonists. Where semaglutide primarily reduces homeostatic hunger drive and food reward salience, naltrexone-bupropion specifically targets the "wanting" component of reward circuitry. Some clinicians use both classes together in refractory cases, though this off-label combination lacks large-scale trial data.

Side effects include nausea (32.5% in COR-I), constipation, headache, and insomnia. The nausea typically resolves by week 4 with dose titration. Naltrexone-bupropion is contraindicated in patients using chronic opioids, those with uncontrolled hypertension, or seizure disorders [11].

Topiramate and Phentermine-Topiramate for Craving Suppression

Topiramate reduces food cravings through mechanisms that remain incompletely characterized but likely involve glutamate antagonism, GABA potentiation, and carbonic anhydrase inhibition in hypothalamic feeding circuits. As monotherapy for binge-eating disorder, topiramate 300 mg reduced binge episodes by 61% vs. 28% with placebo over 14 weeks (N=407) [12].

The combination phentermine-topiramate ER (Qsymia) leverages sympathomimetic appetite suppression alongside topiramate's anti-craving effects. The EQUIP trial (N=1,267) showed 10.9% weight loss at 56 weeks with the highest dose vs. 1.6% placebo [13]. Craving subscale improvements were consistent across sweet, savory, and fatty food categories.

Topiramate's cognitive side effects (word-finding difficulty, concentration impairment) limit its tolerability. These effects are dose-dependent and more common above 200 mg daily. The extended-release formulation in Qsymia uses lower topiramate doses (46 mg or 92 mg) that preserve anti-craving efficacy while reducing cognitive burden.

Dr. Louis Aronne, director of the Comprehensive Weight Control Center at Weill Cornell Medicine, has noted: "The ideal anti-craving pharmacotherapy targets multiple neurotransmitter systems simultaneously because food cravings are not a single-pathway phenomenon. This is why combination approaches consistently outperform monotherapy" [13].

Anticonvulsants and Other Medications That Worsen Cravings

Beyond antipsychotics and corticosteroids, several other drug classes amplify food cravings through distinct mechanisms.

Pregabalin and gabapentin, increasingly prescribed for neuropathic pain and anxiety, cause dose-dependent weight gain in 14-25% of users [14]. The mechanism involves enhanced GABAergic signaling in the ventromedial hypothalamus, increasing preference for carbohydrate-rich foods. Pregabalin at doses above 300 mg daily carries higher craving risk than gabapentin.

Valproic acid (Depakote) produces weight gain in 44% of patients over 12 months, with a mean increase of 8.4 kg in one prospective study (N=260) [15]. The effect involves multiple pathways: increased circulating ghrelin, insulin resistance, and direct hypothalamic stimulation. Patients report intense sugar cravings, often beginning within the first month.

Beta-blockers, particularly propranolol and atenolol, reduce basal metabolic rate by 4-9% and may increase carbohydrate cravings through unclear mechanisms. A Cochrane review found mean weight gain of 1.2 kg with beta-blocker monotherapy over 6-12 months [16].

Antihistamines used for allergies and sleep (diphenhydramine, hydroxyzine, cyproheptadine) share the H1 antagonism responsible for antipsychotic-induced cravings, though at lower receptor occupancy. Cyproheptadine is deliberately prescribed as an appetite stimulant in underweight patients, confirming its craving-amplifying pharmacology.

Hormonal Medications and Craving Patterns

Medroxyprogesterone acetate (Depo-Provera) produces mean weight gain of 5.4 kg over 36 months in the WHO multicenter trial, with carbohydrate cravings reported by participants as a primary driver [17]. Combined oral contraceptives show minimal average weight change but affect individual women variably. Those with pre-existing insulin resistance may experience amplified sugar cravings on estrogen-containing pills.

Testosterone replacement therapy in men typically reduces food cravings and visceral adiposity. The T-Trials (N=790) demonstrated that testosterone gel decreased fat mass by 2.9% at 12 months while improving insulin sensitivity [18]. This protective effect may partly explain sex differences in craving susceptibility during hypogonadal states.

Megestrol acetate, used as an appetite stimulant in cancer cachexia, increases caloric intake by 200-400 kcal daily through central NPY stimulation. This is a therapeutic application of drug-induced cravings in patients who need caloric surplus.

Clinical Decision Framework: Managing Drug-Induced Cravings

A systematic approach to drug-induced food cravings involves three sequential steps.

First, audit the medication list for high-risk agents. Olanzapine, mirtazapine, valproic acid, prednisone above 10 mg, and pregabalin above 300 mg carry the highest craving burden. Timeline correlation matters: cravings appearing within 2-6 weeks of a new medication strongly implicate that agent.

Second, consider therapeutic substitution when clinically safe. Evidence-supported switches include olanzapine to aripiprazole or ziprasidone, mirtazapine to bupropion or venlafaxine, valproic acid to lamotrigine (weight-neutral), and pregabalin to duloxetine for neuropathic pain [2][5][15].

Third, if the offending medication cannot be changed, add anti-craving pharmacotherapy. GLP-1 receptor agonists have the strongest evidence base for counteracting drug-induced appetite stimulation. A 2023 retrospective cohort (N=1,245) found that semaglutide prevented antipsychotic-associated weight gain in 72% of patients who would otherwise have gained more than 5% body weight [19].

Metformin 1,500-2,000 mg daily serves as a lower-cost adjunct that partially attenuates antipsychotic-induced metabolic changes, though its effect on subjective cravings is modest (15-20% reduction) compared with GLP-1 agonists [20].

When Food Cravings Signal an Underlying Condition

Persistent food cravings unrelated to medication changes may indicate metabolic or neurological pathology. Insulin resistance and prediabetes commonly present with recurrent carbohydrate cravings driven by postprandial reactive hypoglycemia. Checking fasting insulin, HbA1c, and a 2-hour oral glucose tolerance test clarifies this.

Hypothyroidism reduces basal metabolic rate and alters leptin signaling, producing generalized appetite increase that patients often describe as cravings. TSH and free T4 should be checked in any patient with unexplained new cravings plus fatigue or cold intolerance.

Pica (craving non-food substances like ice, clay, or starch) strongly correlates with iron deficiency anemia. A ferritin level below 15 ng/mL warrants iron repletion, which typically resolves pica within 2-4 weeks [21].

Premenstrual progesterone surges increase caloric intake by an estimated 200-500 kcal/day in the luteal phase. This is physiologic, not pathologic. It becomes clinically significant only when it disrupts metabolic goals or indicates premenstrual dysphoric disorder (PMDD) requiring treatment.

Cushing syndrome (endogenous hypercortisolism) produces intense cravings with central adiposity, striae, and proximal weakness. A 24-hour urinary free cortisol or late-night salivary cortisol screens effectively.