Peptides are involved in the transmission of biological signals throughout the body, and some of them are studied in the context of sleep, memory, cognitive functions, anxiety, neuroendocrine regulation, and neurological diseases. This has led to a growing interest in terms such as „sleep peptides,” „brain peptides,” or „nootropic peptides.” However, such names can suggest that sleep, memory, mood, attention, and overall brain function are one biological outcome. That is not the case.
Changing a molecular marker associated with the brain does not indicate an improvement in memory. Reducing anxiety does not mean better learning. The onset of drowsiness does not necessarily indicate an improvement in the architecture of sleep or its regenerative properties. Similarly, the results obtained in people with cognitive disorders cannot be automatically considered evidence of improved cognitive function in healthy adults.
Results from human studies are ambiguous and strongly depend on the specific intervention, the population being studied, the outcome being evaluated, the formulation, and the study design. In a small controlled study of the delta sleep-inducing peptide known as DSIP, some beneficial changes were observed in objective sleep parameters, but the authors described the overall effect as weak and concluded that the short-term intervention probably does not provide a significant therapeutic benefit. [1] A significantly larger study of insulin in people with mild cognitive impairment or dementia in the course of Alzheimer’s disease, however, showed no improvement in cognitive or functional performance in the main analysis. [2]
Other studies provide different signals. In a randomized study of growth hormone-releasing hormone analogs in older adults, beneficial results were reported regarding cognitive functions, especially executive functions, whereas the Selank study focused primarily on anxiety symptoms rather than memory or sleep.[3][4] Preclinical studies of DSIP, Semax, and related compounds provide additional mechanistic information, but in themselves they do not confirm clinical improvements in sleep or cognitive functions.
In this article, we analyze what actually results from these different types of studies. We explain which endpoints are used in sleep and cognitive function studies, why the delivery of peptide compounds to the central nervous system presents a particular challenge, how mechanistic evidence differs from clinical evidence, and what gaps still exist in terms of efficacy and safety.
The article is educational and scientific in nature. It is not a guide on the selection, combination, dosage, or administration of peptides or other experimental substances.
Sleep studies and neurological functions answer various questions
The term „functioning of the brain” is extremely broad.
The brain is involved in processes such as attention, learning, immediate and delayed memory, executive functions, emotion regulation, vigilance, language, movement, decision-making, and perception.
A substance that affects one of these areas cannot automatically be presented as improving all the others.
The same problem applies to research on sleep.
Researchers can measure the time it takes to fall asleep, the number of awakenings, the time spent awake at night, the total time spent sleeping, the effectiveness of sleep, the breakdown of its individual phases, subjective sleep quality, morning alertness, or functioning during the day.
These results are related to each other, but they are not equivalent.
Someone may fall asleep faster without extending the total sleep time. The objective sleep efficiency may improve, although the participant does not experience an improvement in its quality. There may also be a change in the specific sleep phase without a measurable improvement in memory or the next day's performance.
The study can therefore show a positive result for one sleep parameter and a slight change or absence of change in the case of other parameters. This does not necessarily mean that there is a contradiction.
The problem arises when one chooses a single favorable endpoint and on the basis of it formulates a much broader statement that the intervention „improves sleep.”.
The same principle applies to cognitive functions.
Improvements in one test of executive functions do not indicate better memory, and a favorable change in a complex cognitive outcome does not automatically mean an improvement in every cognitive area.
Insomnia does not mean any type of sleep problem.
Insomnia is a specific clinical problem and should not be used as a general term for all sleep disorders.
A clinical study conducted in people with chronic insomnia does not automatically address questions about circadian rhythm disorders, sleep breathing disorders, excessive daytime sleepiness, problems associated with shift work, or sleep disorders resulting from another medical condition.
It is the studied population and the diagnostic criteria that determine which question the data actually answer.
The same principle applies to neurological research.
Research on people with mild cognitive impairment or dementia in the course of Alzheimer's disease answers a completely different question than research testing whether a healthy student remembers more information after taking a specific substance.
Recovery after stroke, age-related anxiety disorders, cognitive impairment, common forgetfulness, and attempts to improve cognitive function in healthy individuals constitute separate research contexts.
The evidence should therefore correspond to a specific claim.
When discussing a potential sleep-related peptide, it is necessary to determine whether the researchers were actually measuring sleep, analyzing an interference with sleep, or merely observing a biological process that could theoretically be associated with sleep.
What exactly does the name „delta sleep-inducing peptide” mean?
Delta sleep-inducing peptide is a well-established research name, but the term „sleep-inducing,” i.e., „inducing sleep,” should not be taken as clinical evidence.
The names of substances may result from the historical circumstances of their discovery or the initial directions of research. They do not guarantee that subsequent controlled clinical trials will confirm the effect suggested by the name.
In a double-blind study from 1992, DSIP was evaluated in 16 people with chronic insomnia over five laboratory nights. The researchers noted some favorable results in objective sleep parameters. However, the effects were considered weak, and the authors indicated that some of the observed differences may have resulted from a random change in the placebo group. Subjective sleep quality did not improve, and overall interpretation indicated no significant short-term therapeutic benefit. [1]
The study is important because it provides direct data on sleep in humans.
It would therefore be incorrect to state that DSIP has never been studied in people with insomnia.
At the same time, a study of 16 people over several nights does not allow us to determine lasting clinical efficacy, effects that persist for months or years, long-term neurological safety, or repeatable benefits in various sleep disorders.
A balanced proposal lies somewhere in between these extremes.
There are studies of DSIP in people with insomnia, but this specific controlled study provided weak, rather than strong, evidence of a clinically significant benefit. [1]
Does the name DSIP itself prove that the peptide improves sleep?
No. The research name does not replace clinical evidence. In a small controlled study of DSIP in people with chronic insomnia, some favorable objective results were noted, but the effects were considered weak and no corresponding improvement in subjective sleep quality was observed. [1] The name may reflect the history of research on the molecule, but a reliable effect on sleep must be demonstrated experimentally.
Research on diseases is not the same as research on improving cognitive function
One of the most common sources of misunderstandings is the transfer of the results of studies conducted on diseases to claims regarding the improvement of cognitive abilities in healthy individuals.
The substance under study in people with cognitive disorders cannot be automatically considered a „memory enhancer” for healthy adults, students, working people, or athletes.
The initial level of functioning matters.
In a person with progressive neurological disease, even a slow decline in function can be a significant result, although the results of cognitive tests will never exceed the baseline level.
In the case of healthy participants, the study of cognitive improvement answers another question: is it possible to increase performance beyond the already optimal level?
The natural course of the changes also differs.
A healthy young adult, an older person with mild cognitive impairment, and a patient with dementia in the course of Alzheimer's disease begin the study with a different level of cognitive function and a different biological situation.
Evidence should therefore be presented with respect to the population being studied, the comparator used, and the expected clinical course.
Bringing all these situations together to define a „nootropic” removes precisely the information that is needed to properly understand the results.
Cognitive functions, mood and sleep are separate outcomes
Studies of cognitive functions become much easier to interpret when the assessed domain is precisely defined.
Executive functions typically involve more complex processes such as planning, organizing behavior, switching between tasks, inhibiting inappropriate responses, maintaining a goal, and adapting to changing information.
Memory itself also covers many areas.
Researchers can analyze immediate recall, deferred recall, verbal learning, visual memory, working memory, recognition, and other specific processes.
Attention is another group of functions that can also be divided into several types.
The statement that a given substance „improved cognitive functions” can therefore conceal significant differences regarding what actually changed.
In research, complex results are sometimes used, combining the results of several tests according to a previously established method of analysis.
A complex positive result may be significant, but it does not mean that every element of it has improved.
Similarly, a statistically favorable result from one test does not confirm overall improvement in brain function.
This distinction is particularly important when a clearer effect is observed in one domain than in another.
The GHRH study shows why individual cognitive functions should be analyzed separately
A good example is a randomized study from 2012 involving 152 people aged 55 to 87 years. Among the participants were both healthy older people and people with mild cognitive impairment. For about 20 weeks, an analog of the growth hormone-releasing hormone was studied. [3]
The study noted a favorable overall result regarding cognitive functions and a clearer signal related to executive functions.
The result regarding verbal memory was less certain and constituted a trend, rather than a statistically significant difference at the conventional threshold.[3]
This distinction matters.
It can be said that the study provided data that justified further analyses of the impact on cognitive functions in older adults.
It would be less accurate to simplify this result to say that „GHRH improves memory”.
The study also does not confirm the prevention of dementia, the improvement of cognitive functions in healthy young adults, or the treatment of insomnia.
In this same interpretation, adverse effects should be taken into account. In the active intervention group, they occurred more frequently than in the placebo group, although they were described as mild. [3]
Effectiveness and tolerance should be considered together, rather than separating beneficial cognitive outcomes from observed adverse effects.
Does a positive result of cognitive function tests mean an improvement in any type of memory?
No. Cognitive function studies often involve several domains, and the results may vary between them. In a 2012 study of GHRH analogs, the stronger signal was related to executive functions, while the verbal memory score was less certain. [3] Responsible interpretation should acknowledge these differences. A favorable overall cognitive performance does not automatically mean an improvement in all types of memory, attention, and executive functions.
The result regarding mood cannot automatically become a result regarding memory.
Selank is another good example showing why conclusions should remain tied to the area that was actually measured.
In a small comparative study, 62 patients with generalized anxiety disorder or neurasthenia participated. Selank was compared with medazepam, using, among other things, psychometric assessments; the results obtained mainly concerned anxiety symptoms.[4]
The study therefore provides information about anxiety-related outcomes in the analyzed population.
It does not constitute evidence of a controlled placebo effect that improves memory.
It was also not designed as a controlled study of insomnia treatment.
Reducing anxiety can indirectly affect the way a person assesses their concentration or sleep. A person experiencing less anxiety may find it easier to focus their attention or fall asleep.
However, these possibilities do not change the research on fear into direct evidence of improvements in learning, memory consolidation, sleep efficiency, or its architecture.
The proposal should therefore remain linked to the result that is actually being sought.
The term „calming” should not automatically be associated with improved memory or more restorative sleep.
Does reducing anxiety lead to improved memory?
No. Anxiety and memory are separate psychological and cognitive domains. Anxiety can affect concentration and subjective assessment of one's own cognitive abilities, but improving anxiety symptoms does not indicate better learning or memory. Selank's study focused primarily on anxiety-related outcomes in a specific clinical population, rather than a controlled placebo study of memory improvement. [4]
Objective and subjective sleep measurements answer different questions
Sleep can be studied using objective physiological measurements and subjective assessments by the participants.
Polysomnography records several physiological signals that allow the characterization of sleep and the differentiation of its individual phases.
Sleep diaries and questionnaires, on the other hand, describe the experiences of the participant themselves.
Both types of measurements can be valuable.
They simply answer other questions.
Sleep efficiency describes what portion of the time spent in bed was actually spent sleeping.
Sleep latency, on the other hand, determines the time it takes to fall asleep.
None of these parameters in themselves do not inform whether a person feels rested the next morning or functions better during the day.
Similarly, the subjective perception of deep or regenerative sleep does not confirm that a specific sleep phase change has occurred in physiological measurements.
The DSIP study in people with insomnia is interesting precisely because it evaluated both objective and subjective results, and the improvement was not uniform. [1]
An accurate review of the evidence should preserve this discrepancy rather than choosing only the measure that creates the most favorable image.
Does better sleep efficiency mean that a person feels more rested?
No. Sleep efficiency is determined by the proportion of time spent in bed during which a person actually sleeps. The feeling of being rested is a subjective experience and can also be assessed through functioning during the day. These parameters can vary independently. In the DSIP study, no uniform improvement in objective and subjective outcomes was shown.[1] Therefore, changes in the physiological parameters of sleep should not be automatically presented as evidence of more restorative sleep.
Individual sleep parameters and brain functions require separate interpretation
Sleep latency indicates how much time the participant needed to fall asleep under the study conditions. Reducing it may be meaningful, but it does not confirm better sleep architecture, longer total sleep time, or improved daytime functioning.
Sleep efficiency determines the proportion of time spent in bed that actually corresponds to sleep. A beneficial change may mean fewer periods of wakefulness, but it does not indicate an improvement in all phases of sleep or a more restorative nature of sleep.
Sleep phase measurements describe its physiological properties and its distribution. They help determine whether an intervention changes the architecture of sleep, but the change in the individual phases alone does not indicate an improvement in memory, mood, or daytime performance.
Subjective measures of sleep quality describe the way in which participants perceive their sleep. They are clinically important because the experience of sleep is important, but they do not confirm a specific physiological mechanism.
Cognitive tests measure the performance of specific tasks or sets of tasks. Their interpretation should remain tied to those tasks. A better score on one test does not necessarily indicate a universal improvement in „brain functioning.”.
The anxiety and mood scales measure symptoms in the emotional or psychiatric domains. They may be related to concentration or the perceived quality of sleep, but an improvement on these scales is not evidence of an increase in cognitive abilities.
Brain biomarkers and imaging studies provide information about biological processes, brain structure, activity, metabolism, connectivity, or other physiological characteristics. These results can be of great scientific importance, but without corresponding clinical results they do not confirm a significant improvement in daily cognitive functions.
The effect of learning the test may make it difficult to assess improvement in cognitive function
Repeating cognitive tests introduces an additional methodological challenge.
People often achieve better results simply because they have already taken a given test before.
They may remember instructions, recognize the structure of a task, develop a better strategy, or simply feel more comfortable in the testing environment.
This phenomenon is referred to as the practice effect.
The test result can therefore improve, even if the basic cognitive ability has not changed to the same extent.
Researchers can mitigate this problem through placebo groups, alternative versions of tests, blinding, appropriate time intervals, and statistical methods.
It is also important to define the primary and secondary endpoints in advance and to take into account the problem of multiple comparisons.
A similar principle applies to sleep studies.
During the first night in the laboratory, participants may sleep differently because they are in an unfamiliar environment. Adaptation during subsequent nights may alter the measurements regardless of the intervention being studied.
Expectations can also influence subjective evaluations.
That is why baseline measurements, placebo, blinding, and repeated assessments are so important.
The improvement itself compared to the participant's first visit does not prove that the substance under investigation caused it.
The presence in the blood does not mean the presence in the brain
Issues related to the delivery of peptides require special caution, since systemic circulation and central nervous system exposure are separate issues.
The peptide hormone secreted by the anterior pituitary gland can enter the bloodstream and be transported throughout the body.
It does not require a separate neural connection leading to each target tissue.
Blood carries the hormone to many tissues, while the biological response depends, among other things, on whether certain cells possess the appropriate receptors and intracellular signaling mechanisms.
Many hydrophilic peptide hormones act through receptors located on the surface of target cells.
The hormone binds to a receptor and triggers intracellular signaling without the need for free passage through the cell membrane.
This mechanism explains how peptide hormones can effectively transmit signals to many peripheral tissues.
However, this does not mean that the same molecule easily penetrates into the brain tissue.
The central nervous system is protected by specialized barriers, primarily the blood-brain barrier and the barrier between blood and the cerebrospinal fluid.
Therefore, the concentration measured in the blood, the peripheral hormonal response, the concentration in the cerebrospinal fluid, and the active concentration in a specific area of the brain are different observations.
They should not be treated as equivalent evidence.
The most important question concerns whether the active molecule actually reaches its target.
The statement that a peptide „penetrates the brain” can be misleading if it is not known exactly what has been measured.
Researchers can detect the intact peptide, its metabolite, fragment, radioactive marker, fluorescent material, or a signal recognized by a test using antibodies.
Such measurements do not necessarily mean the same thing.
The detection of a marker associated with a molecule does not automatically prove that the original biologically active peptide has reached its target in unchanged form.
The measurement location also matters.
The detection of the substance in the blood does not prove its presence in the cerebrospinal fluid.
The presence in the cerebrospinal fluid does not mean its uniform distribution in the brain tissue.
The detection in one area of the brain does not confirm the appropriate exposure of all target areas.
The full assessment of delivery therefore includes the stability of the molecule, transport across biological barriers, distribution, concentration, duration of retention, and ultimately the interaction with the appropriate biological target.
This is one of the reasons why promising compounds can exhibit interesting laboratory mechanisms but still not lead to significant clinical results.
A biologically active molecule cannot produce the proposed effect in the central nervous system if the formulation used does not provide adequate active exposure at the target site.
Does the detection of a peptide in the cerebrospinal fluid mean that it reaches every area of the brain?
No. Detection in the cerebrospinal fluid does not confirm uniform distribution in brain tissue, adequate concentration at a specific neuronal target, or the maintenance of an intact active molecule. Older experiments regarding DSIP transport used animal models and specific analytical methods. [5] [6] A reliable claim regarding delivery to the brain should specify what was detected, where the measurement was performed, whether the molecule remained intact, and whether there was a biologically relevant interaction with the target.
What do the more recent DSIP transport studies actually contribute?
Older experiments involving DSIP show how transport across biological barriers can be studied.
In one study, perfused sheep vascular bundles and experiments with mouse cerebrospinal fluid were used to analyze the movement of DSIP across the blood-cerebrospinal fluid boundary. [5]
The results provided experimental information regarding the transport of DSIP in these specific animal models.
In another study, material resembling DSIP was measured in the cerebrospinal fluid of dogs after administration of DSIP or its related analogues. The results indicated dependencies related, among other things, to the time of retention in the plasma and to lipophilicity. [6]
These studies are useful because they show that exposure within the brain is a question dependent on the properties of a specific molecule and requires experimental verification.
However, they do not prove that the modern commercial preparation DSIP provides therapeutically relevant exposure of the brain in humans.
The species studied were not humans.
The formulations and analytical methods were characteristic of specific experiments.
A specific biological range was measured.
The term „material resembling DSIP” also requires caution, because detecting a signal in a test does not necessarily mean the presence of an intact and biologically active precursor peptide.
These limitations do not make the research insignificant.
They simply define the limits of what can be inferred from them.
A whistleblower report does not automatically solve the problem of delivering the substance to the brain.
The route of administration is often discussed in studies of the central nervous system because it can provide potential routes of exposure that differ from the classical systemic administration.
However, the mere term „confidential reporting” is not evidence that the substance reaches its intended target in the brain in a therapeutically relevant amount.
The properties of the formulations are important.
The device used for administration is important.
The significance lies in the deposition in the nasal cavity.
The stability of the molecule, clearance, absorption, transport, and biological activity are also important.
The multicenter study of reported insulin provides an important example from practice.
289 participants were randomized to the study; however, problems with the reliability of the original device led to the use of another device in the remaining 240 people, which formed the main population for the analysis.[2]
In this main cohort, no benefits were shown regarding cognitive or functional performance during the blinded treatment period.[2]
This result should not be interpreted as proof that any substance administered by injection will be ineffective.
At the same time, the problem with the device cannot be used to turn the negative result of the main analysis into evidence of hidden effectiveness.
The scientifically valid conclusion is that the main analysis did not confirm the benefits, while the problem with the device constitutes an important methodological aspect related to the delivery of the substance.
Does reporting a crime guarantee action in the brain or an improvement in cognitive functions?
No. The route of administration is a method of administration, not a means of demonstrating exposure to the intended target or clinical efficacy. The formulation, the device’s action, the site of deposition in the nasal cavity, the stability of the particle, and its transport and biological activity still require confirmation. A multicenter study of the route of administration of insulin has encountered problems with the reliability of the device, and in the main analyzed population no improvement in cognitive or functional function was demonstrated. [2]
The modified peptide cannot automatically take over the evidence regarding the parent molecule
Chemical modification can significantly change the properties of a peptide.
The fusion construct, analog, conjugate, or modified formulation may differ from the original molecule in terms of stability, susceptibility to enzymatic degradation, half-life, receptor activity, transport, tissue distribution, and biological activity.
The evidence should therefore relate precisely to that construct that was actually tested.
In the 2024 experiment, a fusion preparation containing DSIP was evaluated in a chemically induced mouse model of insomnia and the results regarding neurotransmitters and sleep-related parameters were described. [7]
These results concern a specific experimental fusion construct and an applied mouse model.
They do not automatically confirm the operation of the ordinary, unmodified DSIP.
They also do not prove that a commercial product intended for humans behaves the same as the experimental construct.
This distinction is particularly important when the modification was introduced specifically to change the transport properties or stability.
If a modification changes the pharmacological behavior, the modified and the original molecule cannot be considered interchangeable.
Cellular studies cannot show improvement in human memory
Cell culture studies can provide very valuable mechanistic information.
Researchers can analyze receptor signaling, gene expression, cell viability, inflammatory responses, neurotransmitter-related pathways, and cellular responses to experimental stimuli.
However, cells cultured in the laboratory cannot remember conversations, learn lists of words, function in everyday environments, maintain attention during work, or demonstrate improvements in executive functions.
The cellular marker associated with neuronal plasticity is therefore not synonymous with improved memory.
Studies of animal behavior are closer to functional assessment, but they also require careful interpretation.
The performance of the task by an animal can be influenced by factors such as motivation, physical activity, stress, sensory functions, or appetite, and not only learning and memory.
The evidence becomes significantly stronger when mechanistic observations are subsequently confirmed in repeated studies involving humans using appropriate clinical cognitive tests.
Without such a connection, cellular outcome remains a mechanistic observation or hypothesis regarding possible action at a higher biological level.
Can cell research confirm memory improvement?
No. Cell experiments can reveal changes in receptor signaling, gene expression, cell viability, or neurotrophin-related pathways, but the cells do not perform the behavioral tasks necessary to demonstrate improvement in human memory. Such results may provide a biological rationale for further research, but improving memory requires appropriately designed cognitive and functional assessments in humans.
Studies of Semax show a difference between the mechanism and the clinical result
Semax is a good example of this distinction.
In a study in rats, changes in the expression of the genes Bdnf and Ngf were analyzed after exposure to Semax. The reactions differed depending on the brain region and time, instead of leading to a uniform, steady growth throughout the brain.[8]
This is an important mechanistic information.
BDNF and NGF play important roles in the biology of the nervous system, therefore changes in their expression can provide clues regarding potential molecular pathways.
However, the experiment did not show an improvement in memory in healthy people.
It did not confirm the prevention of dementia.
He did not show improvement in his academic performance.
He also did not confirm the treatment for insomnia.
The difference between these statements is fundamental.
The conclusion that „Semax changed the expression of genes related to neurotrophins in specific areas of the brain of rats and at specific time points” describes a mechanistic result.
The statement that „Semax improves memory in humans” is a claim regarding clinical efficacy and requires direct studies of cognitive functions in humans.
It is also necessary to distinguish the information contained in the introduction of the publication from the results of the experiment itself.
Authors of scientific papers often summarize previous hypotheses and claims in the introductory part. This does not mean that they were confirmed by the experiment currently described.
The review of the evidence should always identify what the researchers actually measured.
Does the increase in the marker associated with BDNF indicate an improvement in cognitive function?
No. BDNF and related neurotrophic pathways are important for the functioning of the nervous system, but changes in their expression are mechanistic observations rather than direct measurements of human learning or memory. In the Semax study in rats, changes in the expression of genes related to neurotrophins were shown to be region- and time-dependent. [8] It did not confirm an improvement in memory in humans.
Studies of cognitive function in humans include both positive and inconclusive results.
Studies involving humans do not lead to a single, uniform conclusion.
A 2012 study of a GHRH analog yielded a favorable signal regarding cognitive functions in older adults, particularly executive functions.[3]
A much smaller, controlled placebo pilot study of tesamorelin from 2026, involving 22 participants, did not show significant immediate changes in the main evaluated parameters after the shorter intervention. Exploratory analyses using machine learning indicated possible imaging-related patterns, but this does not change the immediate clinical results in evidence of cognitive function improvement. [9]
The studies differed in terms of sample size, duration, characteristics of the participants, and analysis methods.
A smaller subsequent study does not automatically invalidate the earlier result.
On the other hand, a previous positive result does not mean that later data that do not confirm the effect become irrelevant.
Overall, they show that the available evidence does not constitute a universal and completely conclusive result.
A larger study of reported insulin provides another important example of the failure to demonstrate benefits in the main analysis, although it concerns a different intervention and a different population.[2]
It would be incorrect to use the results of an insulin test to conclude that all peptide approaches to cognitive function are ineffective.
It would also be equally inappropriate to omit this study just because its main result was negative.
A balanced analysis of the evidence includes positive, uncertain, and inconclusive results.
Pilot studies reduce uncertainty, but do not eliminate it
Pilot studies play an important role in science.
They help assess the feasibility of recruiting participants, the feasibility of the project, tolerability, measurement procedures, imaging methods, adherence to study hypotheses, and preliminary biological signals before starting a larger experiment.
However, the small number of participants is associated with significant statistical limitations.
A small study may not detect a real, but moderate effect.
It can also show an apparently large effect, which in a larger sample turns out to be significantly smaller or completely disappears.
It is particularly difficult to assess rare adverse reactions.
Advanced statistical or computational methods do not eliminate these limitations.
Analyses using machine learning can, for example, detect complex patterns in image data or biomarkers.
Such patterns can be valuable in creating new hypotheses.
However, they require independent confirmation, and researchers still need to demonstrate that they are equivalent to results that have real significance for patients.
Therefore, in a good analysis of evidence, it is necessary to distinguish between previously defined clinical endpoints and exploratory analyses.
Hypotheses derived from the analysis of data should not be presented with the same certainty as the repeatable main endpoint.
Results regarding different peptides cannot be combined into a single category of „brain peptides”
DSIP, insulin, GHRH analogs, Semax and Selank are separate interventions.
They differ in molecular structure, biological targets, pharmacology, delivery issues, populations studied, and the history of research.
The fact that all of them can be discussed in the broader context of peptides and brain function does not mean that they have a common mechanism or clinical effect.
Combining the most favorable results for each of these compounds could create a misleading image that „brain peptides” simultaneously improve sleep, memory, anxiety, neurotrophic signaling, and cognitive functions.
No single study or single peptide needs to show all of this set of effects.
A reverse error is also possible.
A negative result of a reportable insulin test does not prove that DSIP, Semax, Selank, GHRH analogs, or all other peptides have no research potential in neurology.
The appropriate unit of evidence is a specific intervention combined with a specific claim in a specific population.
The question should therefore be whether DSIP improved certain sleep parameters in people with insomnia, whether the GHRH analog affected specific cognitive domains in older adults, whether Selank affected anxiety symptoms in the studied population, or whether Semax altered specific molecular pathways in an animal model.
Separating these questions prevents a situation where a broad term like „peptide” begins to replace real evidence.
Short studies do not confirm lasting improvement in sleep or cognitive function
The duration of the study determines the period for which there is actually direct evidence.
Several nights of laboratory measurements cannot confirm that the effect will persist for six months.
A five-week study of cognitive function does not prove that dementia can be prevented for many years.
A separate question is also whether the effect persists after the intervention has ended.
The improvement observed during the exposure may persist, disappear, or change after the intervention ends.
Without further observation, it is not possible to determine what actually happened.
Longer studies also allow for the evaluation of results that may not be immediately apparent, such as compliance with recommendations, quality of life, daily functioning, tolerance, and delayed adverse effects.
Therefore, a beneficial short-term physiological result should be presented as evidence of this only as a short-term result.
It should not be automatically converted into a claim of lasting neurological benefit.
Safety depends on the specific substance, formulation, route of administration, and population.
There is no common safety profile for „peptides”.
The identity of the molecule, its sequence, purity, formulation, stability, route of administration, biological target, level of exposure, and patient characteristics can all influence the risk.
The study of GHRH analogs shows why information regarding tolerance should be presented along with the results regarding efficacy.
In the active treatment group, adverse reactions occurred more frequently than in the placebo group, although they were described as mild.[3]
This does not specify the safety profile of DSIP, Selank, Semax, insulin, or other peptide interventions.
It merely shows that the benefit-risk analysis must be conducted separately for each substance under study.
Current FDA information regarding the risk associated with the preparation of formulations also indicates a lack of certain data regarding the safety dependent on the route of administration and potential issues related to the characteristics or immunogenicity of emedyltydine, the other name used for DSIP.[10]
This introduces an important distinction between historical experimental studies and the safety of modern products.
A small sleep study from several decades ago cannot characterize the quality, composition, route of administration-related risk, immunogenicity, or long-term safety of any current product sold under the same or similar name.
Sleepiness does not necessarily mean better sleep
Sedation, drowsiness and restorative sleep are not synonymous.
The substance can cause drowsiness without improving the structure of the sleep.
It can shorten the subjectively perceived time needed to fall asleep without improving functioning the next day.
The reverse situation is also possible: intervention can change the objective parameters of sleep without noticeable subjective improvement.
Therefore, sleep studies should, as far as possible, include several mutually complementary endpoints.
The consequences that occur during the day also have significance.
If the intervention increases nighttime drowsiness but simultaneously causes an undesirable deterioration in functioning the next day, this information is important for the overall clinical interpretation.
A similar caution should be exercised regarding individual opinions.
A person who claims to have „slept very deeply” can correctly describe their own experience.
Such a relationship does not, however, provide polysomnographic data, standardized sleep assessment, blinding, or controlled comparison.
Subjective experiences can lead to valuable research questions.
They cannot independently confirm the physiological improvement in sleep or the clinical effectiveness.
What do the current evidence not confirm?
The analyzed data do not indicate a single universally best peptide for sleep.
They do not prove that peptides as a group improve memory.
They do not show that changes in BDNF, NGF, neurotransmitter-related pathways, imaging results, or other brain-related biomarkers reliably lead to improvements in daily cognitive functions.
They do not confirm that the substance studied in people with cognitive disorders improves cognitive abilities in healthy young adults.
They do not prove that reducing anxiety is equivalent to better memory.
They do not confirm that detecting the substance in the cerebrospinal fluid guarantees adequate active exposure throughout the brain.
They do not prove that the signaling pathway automatically ensures effective delivery to the brain.
They also do not justify the transfer of results between modified and unmodified versions of the same molecule.
These limitations should not be interpreted as evidence that all neurological hypotheses regarding peptides are incorrect.
They rather show the place where the current evidence ends.
Some interventions generated signals worth further investigation among people. In other studies, the main results did not confirm the benefits. Still other compounds are supported primarily by preclinical or mechanistic data.
Problems with the delivery of substances, small sample sizes, short observation times, differences between populations, inconsistent endpoints and incomplete safety data remain important sources of uncertainty.
Frequently asked questions about peptides, sleep, memory and brain function
Can sleep results be transferred between different peptides?
Not automatically. Individual peptides may have different molecular targets, pharmacological properties, exposure profiles, delivery constraints, and biological effects. The population studied and the type of sleep disorder also matter. The DSIP study in insomnia and the reported insulin study regarding cognitive function included different molecules, populations, endpoints, and research questions.[1][2] The results should therefore remain attributed to the actually studied compound.
Does the study in people with mild cognitive impairment confirm improved function in healthy young adults?
Not directly. Older individuals with mild cognitive impairment differ from healthy young adults in their initial level of function, biology, and expected course of change. GHRH and yield-enhancing insulin studies included specific older or cognitively impaired populations, not healthy individuals seeking improvement in cognitive abilities. [2] [3] The results may support research hypotheses, but direct claims of improvement in function require studies in the population they are intended to affect.
Do exploratory brain imaging results indicate clinical benefits?
No. Imaging studies can reveal important structural, metabolic, functional, or connectivity-related patterns, but they require confirmation and should preferably be linked to cognitive or clinical outcomes of relevance to the participants. In a 2026 GHRH pilot study, exploratory analyses using machine learning indicated possible differences in imaging data despite the absence of significant direct changes in the main study parameters.[9] Such results should remain defined as exploratory.
Can a short study confirm the long-term neurological safety?
No. The small number of participants and the short observation period are poorly suited to detecting rare, delayed, cumulative, or exposure-dependent adverse reactions. The FDA information regarding DSIP/emediltydine also indicates gaps in the safety data related to the route of administration and potential issues related to characterization or immunogenicity. [10] The absence of serious problems in the limited trial does not confirm the long-term safety of the other formulation, route of administration, population, or exposure time.
Do user reviews allow us to identify the best sleep peptide?
No. Individual relationships typically do not include verified product identity, standardized measures of outcomes, blinding, randomization, or an appropriate comparator. Sleep is also particularly susceptible to expectations, natural differences between consecutive nights, changes in daily routine, and subjective interpretation. The controlled DSIP study itself showed that objective and subjective results can differ. [1] Personal experiences may suggest research questions, but they do not allow for reliably ranking sleep-related interventions or establishing a causal relationship.
What question is the most important when evaluating the claim regarding the functioning of the brain?
The most useful question is: has the precisely defined intervention improved exactly that outcome that the claim refers to, in the relevant population of people, using the appropriate comparator and validated measurement method? Sleep phase change, improvement in the anxiety scale score, cognitive test score, molecular biomarker, visual observation, and the brain-delivery experiment all answer different questions. The DSIP and reported insulin studies clearly show why these differences matter. [1] [2]
Limitations and other gaps in the evidence
In the literature on peptides, sleep and brain function, several important limitations are repeated.
The first is the wide variety of interventions. DSIP, insulin, GHRH analogs, Semax, Selank, and modified peptide constructs differ chemically and pharmacologically. Their results should not be lumped together simply because all of them can be considered within the broad category of peptides.
The second is the variability of endpoints. Sleep latency, its effectiveness, subjective sleep quality, anxiety, executive functions, memory, neurotrophin expression, presence in the cerebrospinal fluid, and imaging results are not equivalent measurements.
The third limitation concerns the transfer of pre-clinical research results. Cell and animal experiments can identify mechanisms and generate hypotheses, but they do not directly confirm the quality of sleep, memory improvement, or daily cognitive functions in humans.
The fourth problem is brain exposure. The presence of the substance in the systemic circulation does not automatically mean its availability in the central nervous system, and detection in the cerebrospinal fluid does not necessarily indicate adequate active exposure of each neuronal target.
The fifth limitation is the specific nature of the population. Studies conducted on older people, patients with mild cognitive impairment, dementia in the course of Alzheimer’s disease, generalized anxiety disorder, or chronic insomnia cannot be automatically considered evidence of improved function in healthy young people.
The sixth limitation is the size and duration of the studies. Several analyses are of a small, exploratory or short-term nature. Such projects can provide important information, but they do not provide the same level of precision or long-term safety data as large, repeated studies.
The seventh problem is the selective interpretation of endpoints. A beneficial outcome regarding executive functions should not be presented as a universal improvement in memory, just as a beneficial objective sleep parameter cannot be automatically considered evidence of better subjective sleep.
Finally, product safety and quality remain separate evidentiary questions. Historical experimental studies do not automatically describe the identity, purity, stability, immunogenicity, formulation, or long-term safety of contemporary products sold under similar names.
These limitations do not detract from the scientific significance of research on peptides in neurology. They merely determine the level of certainty that can be attributed to the current results.
It is worth keeping the same principle when reading studies concerning other uses of peptides. For example, in the article about Peptides and the regeneration and healing of injuries we discuss why biomarker change, symptom improvement, structural repair, and functional recovery also constitute different levels of evidence. A broader context of research on peptide applications can also be found in the discussion. peptide therapy.
Summary
Research on peptides in the context of sleep, memory, cognitive function, mood, and brain function is very diverse and cannot be properly summarized as evidence regarding a single common category of „peptides for the brain.”.
There are direct DSIP studies on sleep in humans, but a small controlled study of people with chronic insomnia showed weak effects and did not confirm a consistent improvement in objective and subjective sleep parameters.[1] Therefore, the results do not allow DSIP to be considered a well-established or durable method of treating insomnia.
Studies of cognitive functions in humans also yield mixed results.
In a 2012 GHRH analog study, older adults showed a favorable overall signal regarding cognitive function, with stronger results concerning executive functions than verbal memory. [3] In a smaller 2026 pilot study, no significant direct changes were observed in the main parameters, although exploratory imaging analyses provided hypotheses for further study. [9]
In a larger study of reported insulin, no improvement in cognitive or functional abilities was shown in the main analyzed population. This study also showed how the functioning of a device used to deliver substances can be an important component of research in the central nervous system. [2]
Selank studies provide data on anxiety in humans, rather than direct evidence of memory or sleep improvement.[4] Semax studies on neurotrophin-related gene expression provide mechanistic information, rather than direct evidence of memory improvement in humans.[8] Older experiments on DSIP transport and more recent studies of modified peptide constructs help to understand mechanisms and delivery problems, but do not themselves confirm routine clinical efficacy.[5][6][7]
In all these examples, several principles are repeated.
Molecular change is not automatically an improvement in cognitive function. Reducing anxiety is not automatically an improvement in memory. Being alert does not necessarily mean restorative sleep. Detecting a substance in the cerebrospinal fluid does not indicate proper exposure of the entire brain. Reporting results does not guarantee an effect on the target. Animal studies do not confirm equivalent effects in humans, and studies in neurological diseases do not demonstrate improvements in cognitive abilities in healthy individuals.
The most reliable summaries of the evidence should therefore be specific to a particular intervention, a particular outcome, and a particular population.
Research on peptides in neurology continues to yield new hypotheses, and in some cases also interesting results in humans. However, the available data do not confirm the existence of a single universal peptide solution for sleep, memory, or overall improvement in brain function. For many proposed applications, larger controlled studies, longer observation periods, more accurate characterization of brain exposure, direct functional measurements, and better safety data are still needed.
Disclaimer
The article is intended for educational and scientific informational purposes only. It does not constitute medical, neurological or psychiatric advice, recommendations regarding the treatment of sleep disorders, information on the prescription of medications, dosage instructions, instructions for administration, or recommendations for the use of DSIP/emeldetide, Semax, Selank, temsoromyl, commercial insulin, modified peptide constructs, or any other experimental substance.
Results from cell experiments, animal models, small observational studies, pilot studies, and controlled clinical trials should be interpreted in relation to the specific molecule, formulation, route of administration, population, clinical condition, endpoint, and duration of the study. Results should not be automatically transferred between modified and unmodified molecules, animal and human preparations, neurological diseases and attempts to improve function in healthy individuals, or between different sleep disorders and cognitive functions.
Individuals experiencing persistent insomnia, worsening cognitive function, significant memory changes, anxiety, neurological symptoms, or other health problems should consult with a suitably qualified health care professional.
References
[1] Bes, F., Hofman, W., Schuur, J., & Van Boxtel, C. (1992). Effects of delta sleep-inducing peptide on sleep of chronic insomniac patients. A double-blind study. Neuropsychobiology, 26(4), 193–197. 10.1159/000118919.
[2] Craft, S., Raman, R., Chow, T. W., Rafii, M. S., Sun, C. K., Rissman, R. A., Donohue, M. C., Brewer, J. B., Jenkins, C., Harless, K., Gessert, D., & Aisen, P. S. (2020). Safety, efficacy, and feasibility of intranasal insulin for the treatment of mild cognitive impairment and Alzheimer disease dementia: A randomized clinical trial. JAMA Neurology, 77(9), 1099–1109. 10.1001/jamaneurol.2020.1840.
[3] Baker, L. D., Barsness, S. M., Borson, S., Merriam, G. R., Friedman, S. D., Craft, S., & Vitiello, M. V. (2012). Effects of growth hormone-releasing hormone on cognitive function in adults with mild cognitive impairment and healthy older adults: Results of a controlled trial. Archives of Neurology, 69(11), 1420–1429. 10.1001/archneurol.2012.1970.
[4] Zozulia, A. A., Neznamov, G. G., Siuniakov, T. S., Kost, N. V., Gabaeva, M. V., Sokolov, O. I. u., Serebriakova, E. V., Siranchieva, O. A., Andriushenko, A. V., Telesheva, E. S., Siuniakov, S. A., Smulevich, A. B., Miasoedov, N. F., & Seredenin, S. B. (2008). [Eficacia y posibles mecanismos de acción de un nuevo péptido ansiolítico Selank en el tratamiento de trastornos de ansiedad generalizados y neurastenia]. Journal of Neurology and Psychiatry named S. S. Korsakov, 108(4), 38-48. PubMed record.
[5] Zlokovic, B. V., Segal, M. B., Davson, H., & Jankov, R. M. (1988). Passage of delta sleep-inducing peptide (DSIP) across the blood-cerebrospinal fluid barrier. Peptides, 9(3), 533–538. 10.1016/0196-9781(88)90160-x.
[6] Banks, W. A., Kastin, A. J., Coy, D. H., & Angulo, E. (1986). Entry of DSIP peptides into dog CSF: Role of physicochemical and pharmacokinetic parameters. Brain Research Bulletin, 17(2), 155–158. 10.1016/0361-9230(86)90111-5.
[7] Mu, X., Qu, L., Yin, L., Wang, L., Liu, X., & Liu, D. (2024). Pichia pastoris secreted peptides crossing the blood-brain barrier and DSIP fusion peptide efficacy in PCPA-induced insomnia mouse models. Frontiers in Pharmacology, 15, Article 1439536. 10.3389/fphar.2024.1439536.
[8] Agapova, T. I., Agniullin, I. A., Silachev, D. N., Shadrina, M. I., Slominskiĭ, P. A., Shram, S. I., Limborskaia, S. A., & Miasoedov, N. F. (2008). [Effect of Semax on the temporary dynamics of brain-derived neurotrophic factor and nerve growth factor gene expression in the rat hippocampus and frontal cortex]. Molecular Genetics, Microbiology and Virology, (3), 28–32. PubMed record.
[9] Stewart, C. E., French, K. P., Wright, T. J., Wilhoit, K., Randolph, K. M., Danesi, C. P., Gilkison, C. R., Karmonik, C., Lu, L., Dillon, E. L., Durham, W. J., Urban, R. J., Sheffield-Moore, M., & Masel, B. E. (2026). The effect of growth hormone-releasing hormone on cognition and brain connectivity in adults with cognition ranging from normal to mild cognitive impairment. eNeurologicalSci, 44, Article 100616. 10.1016/j.ensci.2026.100616.
[10] U.S. Food and Drug Administration. (n.d.). Certain bulk drug substances for use in compounding that may present significant safety risks. Retrieved September 20, 2026, from Certain bulk drug substances for use in compounding that may present significant safety risks.