How Newer Antipsychotics Differ From Older Drugs

Newer antipsychotics, often called second-generation or atypical antipsychotics, were developed to treat psychosis with fewer of the stiffness and involuntary movement problems that plagued older drugs like haloperidol and chlorpromazine. They mostly succeed at that, but they introduced their own set of tradeoffs, particularly weight gain and metabolic disruption. The landscape has gotten more interesting in recent years as genuinely novel mechanisms have entered clinical trials, and long-acting injectable formulations have reshaped how psychiatrists think about keeping people well between crises.

How They Work Differently From Older Drugs

The older, first-generation antipsychotics all did roughly the same thing: they blocked dopamine D2 receptors, hard. That knocked down hallucinations and delusions effectively, but it also caused a suite of movement problems collectively called extrapyramidal symptoms, from muscle rigidity and tremor to a sometimes-irreversible condition called tardive dyskinesia.

The key pharmacological feature that most second-generation drugs share is a stronger blockade of serotonin 5-HT2A receptors relative to their weaker antagonism of dopamine D2 receptors. That serotonin-dopamine interaction is thought to be what gives them a different side-effect profile, reducing the risk of movement-related problems while still controlling psychotic symptoms.¹ Some drugs go further. Aripiprazole, for instance, does not simply block D2 receptors; it acts as a functional modulator at the receptor, behaving differently depending on the dopamine activity already present in a given brain region.² This more nuanced mechanism is one reason aripiprazole tends to produce less sedation and fewer metabolic side effects than many of its class.

Efficacy for Psychosis

When researchers pool data across dozens of trials, second-generation antipsychotics consistently reduce positive symptoms of schizophrenia, such as hallucinations and paranoia. A large network meta-analysis covering 32 antipsychotic drugs found that amisulpride, risperidone, olanzapine, and paliperidone were among the most effective at reducing positive symptoms and were significantly better than many other drugs in the class.³ In younger patients with early-onset schizophrenia, olanzapine, risperidone, and haloperidol showed statistically significant reductions in positive symptoms compared to placebo, though no drug achieved significant improvement in negative symptoms (the flat affect, social withdrawal, and motivational problems that are often harder to treat).⁴

There is preliminary evidence that some newer antipsychotics may modestly improve cognitive function in schizophrenia, with early data pointing to olanzapine and risperidone, though the research has described these findings as speculative until more rigorous trials are completed.⁵ Real-world studies that follow patients over longer periods generally confirm that second-generation drugs outperform older ones in keeping people out of the hospital and staying on treatment, though the evidence for cognitive gains remains weak.⁶

The Side-Effect Tradeoff

Newer antipsychotics did not eliminate side effects; they traded one set for another. The most clinically significant tradeoff is metabolic. A large comparison of 18 antipsychotic drugs found major differences in metabolic impact, with olanzapine and clozapine producing the worst profiles for weight gain, cholesterol, and blood sugar changes, while aripiprazole, brexpiprazole, cariprazine, lurasidone, and ziprasidone had the most benign metabolic profiles.⁷ This is not a minor issue: antipsychotic-related weight gain contributes to diabetes and cardiovascular disease, which are leading causes of early death in people with schizophrenia.

Movement problems are reduced but not eliminated. Second-generation drugs carry a significantly lower risk of acute extrapyramidal symptoms at recommended doses and may also reduce the risk of tardive dyskinesia, with some drugs even capable of suppressing pre-existing involuntary movements.⁸ But the anticipated reduction in tardive dyskinesia risk is less well documented than the reduction in acute stiffness and tremor.⁹ Among the atypicals, clozapine and quetiapine carry the lowest movement risk, while risperidone’s risk climbs with dose.¹⁰

Prolactin elevation is another concern that varies widely between drugs. Amisulpride, risperidone, and paliperidone are the worst offenders for raising prolactin levels, which can cause menstrual irregularities, breast tenderness, and sexual dysfunction. Aripiprazole and quetiapine have the most favorable profiles, and aripiprazole is sometimes added to other antipsychotics specifically to counteract their prolactin-raising effects.¹¹¹²

Use in Mood Disorders

Several newer antipsychotics have found a second life treating mood disorders, an expansion that reflects how these drugs act on brain circuits beyond the dopamine pathways that drive psychosis.

For bipolar depression, a network meta-analysis of FDA-approved atypical antipsychotics found that cariprazine, lurasidone, olanzapine, and quetiapine all had higher remission rates than placebo.¹³ A separate meta-analysis found that lurasidone and quetiapine had the largest effects on depressive symptoms and that lurasidone produced significantly better response odds than cariprazine, aripiprazole, or ziprasidone.¹⁴

In major depressive disorder that hasn’t responded well to standard antidepressants alone, several atypical antipsychotics are now used as add-on treatments. A recent network meta-analysis of over 10,000 participants found that lumateperone had the highest effect size for boosting antidepressant response, followed by aripiprazole and brexpiprazole.¹⁵ Another meta-analysis covering quetiapine, olanzapine, aripiprazole, and brexpiprazole found all four significantly outperformed placebo as augmentation agents for treatment-resistant depression.¹⁶ The tradeoff, as always, is tolerability: the metabolic and movement risks still apply, which is why these augmentation strategies are typically reserved for people who haven’t responded to safer first-line options.

The Adherence Problem and Long-Acting Injectables

One of the biggest obstacles in treating schizophrenia isn’t finding a drug that works. It’s keeping people on it. A study of over 46,000 Medicare beneficiaries with schizophrenia found that only about 35% were adherent to their newly started oral antipsychotic over 12 months, and nearly 80% discontinued treatment altogether, with a median time to stopping of roughly four months.¹⁷ That’s a staggering drop-off rate, and every lapse in treatment raises the risk of relapse, hospitalization, and progressive worsening.

Long-acting injectable (LAI) antipsychotics address this by delivering the drug through an injection given every two weeks, every month, or even every three months, removing the daily pill-taking burden. The results are striking. In a trial of people after their first psychotic episode, the relapse rate with long-acting injectable risperidone was about 5%, compared to 33% with oral treatment.¹⁸ A large adjusted analysis found that LAIs were associated with a 48% reduction in psychiatric hospitalizations and a 44% reduction in suicide attempts compared to oral antipsychotics.¹⁹

A three-year mirror-image study, which compared each patient’s outcomes before and after switching from oral to injectable formulations, found that average hospitalization days dropped from about 10 days to less than one day and total relapses fell significantly.²⁰ Among the injectable options currently available, real-world Medicare data suggest that paliperidone given every three months has the lowest risk of treatment failure, outperforming other LAIs and all oral antipsychotics in head-to-head comparisons.²¹ Despite this evidence, LAIs remain underused in many settings, partly due to patient reluctance about injections and partly due to logistical barriers in clinic scheduling.

Treatment-Resistant Schizophrenia and the Clozapine Question

Roughly one in three people with schizophrenia does not respond adequately to standard antipsychotic trials. For these individuals, clozapine remains the only antipsychotic with robust evidence for treatment-resistant disease, a position it has held since the 1980s. The uncomfortable reality is that no newer atypical antipsychotic has convincingly matched it in head-to-head studies for this population. A Cochrane review comparing newer atypicals to clozapine found that response rates were broadly similar on global symptom scales, with clozapine showing a trend toward greater effectiveness for positive symptoms.²² A separate systematic review noted the same pattern: newer atypicals appeared roughly equivalent overall, but the available studies were small and few, leaving the question incompletely answered.²³

Clozapine is underused for the same reasons it has always been underused: it requires regular blood monitoring because of a rare but serious risk of a life-threatening drop in white blood cells. It also causes significant weight gain, sedation, and excessive salivation. But for the sizable minority of patients who do not respond to anything else, it remains the best-supported option available, and the failure of newer drugs to clearly surpass it is one of the more frustrating gaps in psychiatric pharmacology.

Younger Patients

Children and adolescents metabolize drugs differently and are more vulnerable to certain side effects. In young people with schizophrenia-spectrum disorders, olanzapine and risperidone have the largest effect sizes for reducing symptoms, but olanzapine’s metabolic burden makes it a poor first choice. Lurasidone and aripiprazole show more modest efficacy but considerably cleaner side-effect profiles.²⁴

A large meta-review looking at the safety of psychiatric medications in children and adolescents across 78 different adverse effects found that lurasidone had the safest overall profile among the antipsychotics studied, while olanzapine raised the most concerns.²⁵ Another network meta-analysis focused on physiological effects confirmed that olanzapine, quetiapine, risperidone, paliperidone, and clozapine had the worst metabolic side-effect profiles in young people, while aripiprazole, ziprasidone, lurasidone, and a few less commonly used agents were relatively benign. For prolactin elevation, risperidone and paliperidone were again the worst offenders, and aripiprazole had the best profile.²⁶ For clinicians treating young patients, the evidence increasingly points toward lurasidone or aripiprazole as reasonable starting points, with more potent options held in reserve.

Older Adults With Dementia

At the other end of the age spectrum, antipsychotics are frequently prescribed for agitation and psychotic symptoms in people with dementia, despite carrying serious risks in this population. Atypical antipsychotics in elderly dementia patients have been associated with increased mortality, stroke, pneumonia, falls, and cognitive worsening.²⁷ These risks are not theoretical. A large population-based cohort study found that current antipsychotic use in dementia was associated with roughly double the risk of pneumonia, a 72% increase in acute kidney injury, and a 61% increase in stroke risk.²⁸

A review covering literature through 2022 reinforced that these medications may increase mortality and are associated with pneumonia, cerebrovascular events, parkinsonian symptoms, and higher rates of blood clots.²⁹ All antipsychotics, both first- and second-generation, carry an FDA black-box warning about increased mortality in elderly patients with dementia-related psychosis. In practice, these drugs are still prescribed when behavioral symptoms are severe enough to pose a safety risk, but the goal should always be the lowest effective dose for the shortest necessary time, with regular reassessment of whether the medication is still needed.

Drugs That Skip Dopamine Entirely

For decades, every approved antipsychotic acted primarily through dopamine receptors. That changed in 2024 with the FDA approval of xanomeline-trospium, the first antipsychotic that works through an entirely different system. Xanomeline targets muscarinic acetylcholine receptors (specifically M1 and M4 subtypes) in the brain, while trospium, which doesn’t cross the blood-brain barrier, blocks muscarinic receptors in the rest of the body to prevent gastrointestinal and other peripheral side effects.³⁰ Because the drug doesn’t block dopamine, it appears to avoid the movement problems, prolactin elevation, and some of the metabolic effects associated with conventional antipsychotics. Early clinical trial results were promising enough to win approval, though longer-term real-world data are still accumulating.

Another non-dopaminergic approach targets trace amine-associated receptor 1 (TAAR1). Ulotaront, a TAAR1 agonist with additional serotonin 5-HT1A activity, entered clinical development with considerable excitement.³¹ However, a systematic review and meta-analysis pooling data from four trials with nearly 1,300 participants found that TAAR1 agonists showed little difference from placebo in improving overall schizophrenia symptoms over four to six weeks. One trial directly comparing a TAAR1 agonist to risperidone found the TAAR1 drug was clearly less effective.³² The TAAR1 story is a useful reminder that a compelling mechanism doesn’t always translate into a compelling drug. Development continues, but the path forward for this target is uncertain.

Brain Volume Changes

One of the more unsettling lines of research involves the relationship between antipsychotic treatment and brain structure. A longitudinal study following people from their first episode of schizophrenia found that greater cumulative antipsychotic exposure was associated with reductions in both gray and white matter volume over time, even after accounting for illness severity, substance use, and duration of illness.³³ This finding has generated ongoing debate because schizophrenia itself causes progressive brain changes, and separating the effects of the disease from the effects of treatment is methodologically difficult.

The picture is not uniformly negative. A study of quetiapine treatment in first-episode schizophrenia found that after just three weeks of treatment, gray matter increased in the amygdala-hippocampal region, and the size of that increase correlated with quetiapine blood levels. Other brain regions showed decreases.³⁴ The takeaway is that antipsychotics reshape the brain in complex, region-specific ways. Whether these changes are harmful, adaptive, or a mix depends on the region, the drug, and the duration. It is one of the strongest arguments for using the lowest effective dose and periodically reassessing whether continued treatment at a given level is warranted.

Toward Personalized Prescribing

Right now, choosing an antipsychotic is largely trial and error. A psychiatrist picks a drug based on the symptom profile, the side effects the patient most wants to avoid, and clinical experience. If it doesn’t work or isn’t tolerated, they try another. This cycle can take months and leave patients struggling through multiple failed trials. Pharmacogenomics, using someone’s genetic information to predict which drug will work for them, is the field’s long-term answer to this problem, though it remains in its early stages.

A genome-wide analysis of patients in the CATIE trial, a large government-funded antipsychotic comparison study, identified genetic variants on chromosome 4p15 that were associated with treatment response, along with variants in genes called ANKS1B and CNTNAP5 that appeared to influence how well olanzapine and risperidone improved negative symptoms.³⁵ More recently, researchers have built prediction models combining clinical information, genetic risk scores, and epigenetic markers that achieved fairly strong accuracy in predicting which patients would respond to antipsychotic treatment, with an area under the curve above 0.85 in both discovery and external validation samples.³⁶ These tools are not yet ready for routine clinical use, but they point toward a future where the months-long trial-and-error process could be shortened considerably by a blood test or cheek swab done before the first prescription is written.