The evolution of aging research: from theories to epigenetic reprogramming

Abstract

Over the past decades, numerous studies aimed to discover the fundamental cause of the aging process. Rather than a single root cause, multiple factors were identified, suggesting that aging manifests itself through a progressive degradation of different molecules, cells and in the end, entire systems, directly affecting an individual’s health. To address this rapidly growing challenge, various anti-aging strategies have been proposed, among which partial reprogramming has emerged as a promising approach capable of extending both lifespan and healthspan. In this review, we summarize the historical development of aging theories, the effects of established anti-aging strategies, and the evolution of partial reprogramming using Yamanaka factors. We also highlight recent advances in overcoming the efficacy and safety limitations of partial reprogramming, as well as the remaining challenges that must be addressed to fully realize its therapeutic potential.

Introduction

The diversity of living organisms across the phylogenetic tree highlights that aging cannot be reduced to a single, universal process shared by all organisms¹. Instead, aging varies widely in its demographic and biological manifestations, ranging from clear increases in mortality in some species^2,3 to negligible or even absent senescence in others^4,5. From this comparative perspective, aging is better understood as a heterogeneous and multifactorial phenomenon, with patterns that differ notably across species and even within species, and across evolutionary contexts¹. In humans, this complex biological process is commonly characterized by a progressive decline in physiological functions^6,7, mediated by a combination of internal and external factors^8,9. This gradual deterioration makes older individuals more susceptible to chronic diseases, such as cancer¹⁰, neurodegenerative disorders¹¹, cardiovascular diseases¹², type 2 diabetes¹³, and glaucoma^14,15. In this context, the interest in developing anti-aging therapies has grown significantly in recent decades, with the intention of reducing or even reversing the negative impact of the aging process. To provide insight into the existing knowledge and limitations, this review aims to present the evolution and the present state of anti-aging research, focusing on partial reprogramming as a promising emerging approach, highlighting its rejuvenative and regenerative potential.

Theories of aging

To better understand the biological mechanisms, numerous theories have been proposed, each capturing different aspects of this complex process. Rather than treating them as isolated or purely historical concepts, we organize them into an integrative classification based on their mechanisms. Specifically, aging theories can be broadly grouped into damage-based, programmed, evolutionary, epigenetic and cross-category mechanisms that span multiple frameworks. Chronologically (Fig. 1), the damage-based theories of aging, also known as the molecular and cellular damage accumulation theories, can be divided into different subgroups, as follows: “wear and tear” in 1882¹⁶, cross-linking in 1942^17,18, free radicals in 1956¹⁹, error catastrophe in 1963²⁰, somatic mutations in 1973²¹ and proteostasis in 2008²². The programmed theories are divided in three subgroups: the neuroendocrine theory in 1954²³, immunological theory in 1969²⁴, programmed longevity theory in 1988²⁵ and 1997²⁶. The epigenetic-based theories of aging include the epigenetic clock theory in 2018²⁷, the information theory in 2023²⁸ and the bioelectricity and the morphostatic information theory in 2024²⁹. The evolutionary theories include mutation accumulation³⁰ in 1952, antagonistic pleiotropy in 1957³¹ and disposable soma in 1977³². The telomere shortening in 1961³³, the gene regulation in 1975³⁴, inflammaging in 2000³⁵, the hyperfunction theory in 2006³⁶, and stem cell in 2007³⁷ although partially influenced by both damage-based and programmed processes, do not fully fall under those categories, since they primarily describe systemic, regulatory, or functional declines, rather than just accumulation of molecular damage.

Fig. 1. The old and the new theories of aging in chronological order.

The figure outlines the evolution of aging theories that have shaped our understanding of the aging process. Yellow boxes highlight the damaged-based theories of aging, pink boxes highlight the programmed theories, green boxes present the evolutionary theories, blue boxes represent the epigenetic theories, and white boxes contain the cross-category theories (Created in BioRender.com).

Damage-based theories

Damage-based theories propose that aging results from the progressive accumulation of molecular errors, which eventually overwhelms cellular repair and maintenance mechanisms³⁸. Early formulations, such as the “Wear and Tear” theory of aging proposed by August Weismann in 1882¹⁶, attributes aging to life-long exposure to external factors that leads to injuries and illnesses. This represents one of the first cumulative theories of aging, proposing that organisms function similarly to machines that, over time, accumulate damage through repeated used, which leads to functional decline and ultimately to failure³⁹. However, this theory was later considered insufficient on its own, as studies demonstrated that even animals that are raised in perfect conditions, separated from any aggressions caused by the environment or poor nutrition, still age⁴⁰. While this observation challenged this initial hypothesis, it is important to note that it led to an important refinement that paved the way for the next cumulative damage theories, this time also focusing on the accumulation of molecular damage caused by intrinsic biological processes.

Building on this framework, the cross-linking theory of aging, also known as the glycosylation theory of aging, was proposed in 1942 by Johan Bjorksten^17,41. He stated that, during aging, the glucose concentration in the blood tends to get higher. This leads to cross-linking between glucose and proteins resulting in the glycosylation of proteins⁴². Consequently, this buildup of cross-linked proteins tends to harm cells and tissues leading to a negative impact on biological functions^18,43.

Similarly, the initial free radical theory of aging (FRTA), also known as the oxidative stress theory, was developed by Dr. Denham Harman in 1956¹⁹. He claimed that excessive production of free radicals in cells generated from exposure to UV radiation, pesticides and herbicides can have a negative effect on tissues and organs, ultimately leading to oxidative damage and aging. The most prevalent free radicals in cells are reactive oxygen species (ROS) and even though they are seen as physiological cell signaling molecules that are crucial for cell differentiation and proliferation, they still can deteriorate the cellular membranes and leave the cells vulnerable to DNA damage, protein oxidation and lipid peroxidation^40,44,45. A few years later, Harman proposed an extension of his initial theory, called the mitochondrial theory of aging (MDTA)⁴⁶, based on the observation that the respiratory capacity of mitochondria declines with age in different organisms. This was linked to increased accumulation of ROS and decline of cytochrome c oxidase (CcO), a part of the mitochondrial electron transport chain (ETC). In essence, the two theories are interconnected since it was postulated that improving antioxidant defense will lower the free radical levels while slowing the decline in mitochondrial respiratory function^40,47–49. Also, knowing that mitochondrial DNA (mtDNA) is more susceptible to mutations compared to the nuclear DNA, researchers observed that mice induced to have higher mutational rates in mtDNA suffer accelerated aging phenotypes^50,51. However, antioxidants supplementation trials failed to demonstrate clear lifespan extension⁵².

In parallel, the error catastrophe theory of aging proposed by Leslie Orgel in 1963²⁰ attributes aging to the incorrect functioning and division of cells, due to malfunctions accumulated in cellular molecules. Over time, this error accumulation can reach a catastrophic point leading to cell death^20,46.

In 1973, Burnet replaced Weismann’s theory with the somatic mutation theory suggesting that aging is produced by the deterioration and malfunction of cells after accumulation of genetic mutations throughout the lifespan of an organism and the inability to repair them^21,53,54. DNA deterioration is an inevitable process, but most of these errors are repaired. However, when the rate of mutations that occur in DNA is too high, the DNA repair system cannot correct all the errors in time, so they tend to accumulate^21,43,55.

More recently, proteostasis theory of aging or loss of protein homeostasis proposed by Balch and colleagues in 2008 suggests that aging is due to dysfunctions in the proteostasis network of cells, leaving the proteome unable to correctly synthesize, fold, maintain and degrade. A disequilibrium in protein homeostasis leads to aging and age-associated diseases²². Together, these theories converge on the concept that aging arises from the gradual accumulation of molecular and cellular damage, originating from both exposure to environmental factors and intrinsic biological processes.

Programmed and system-level dysregulation theories

Programmed or regulatory theories propose that aging is driven by genetically programmed process that dictate the timing in our “internal clock”⁵⁶. Rather than being driven primarily by random damage, these models emphasize alterations in coordinated biological systems over time. Early formulations include the neuroendocrine theory of aging, also called the “grand biological clock” proposed by Vladimir Dilman in 1954²³, who links aging to progressive dysregulation of interactions between the nervous and endocrine systems, as two biological systems working together to adapt different tissues to the ever-changing environment^57,58. Toward the goal of maintaining homeostasis in the body, the hypothalamus can be left in a state of “hyperadaptosis”, leading to aging, due to dysregulations in hormone levels, such as the somatotropic axis, formed by the growth hormone (GH) and insulin-like growth factor 1 (IGF-1)^23,59.

Similarly, the immunological theory of aging introduced by Roy Walford in 1969²⁴, which links aging to a gradual decline in immune function, leading to increased susceptibility to infectious diseases that cause aging and mortality. When the efficacy of the immune system is compromised, the antibodies cannot fight new illnesses, ending up with cellular stress and apoptosis^24,43.

The programmed longevity theory is a subset of the programmed theories of aging, proposed by Libertini²⁵ and Skulachev²⁶. Their hypothesis revolves around the idea that death and lifespan are set in stone from the first moments of development. Instead of explaining aging as a progressive decline of body functions, this hypothesis defines aging as a requirement for evolution, to prevent overpopulation and resource consumption. This theory can be observed in nature, for example some plants die after they bloom, or some animals die after they reproduce. Likewise, species-specific lifespan ranges may be viewed as being consistent with this theory, although this pattern can also be interpreted within evolutionary theories. For example, the current median life expectancy in humans is approximately 72 years^{25,26,60–62}.

Together, the programmed theories highlight that aging may emerge from progressive dysregulation of biological control systems, rather than solely from the accumulation of molecular damage.

Evolutionary theories of aging

Evolutionary theories attribute aging as a consequence of declining selective pressure with age⁶³, rather than as a programmed or damaged driven process. In this framework, traits that are beneficial early in life can be favored by natural selection even if they have detrimental effects later.

The mutation accumulation theory proposed by Medawar in 1952. He stated that accumulation and expression late in life of deleterious germline mutation leads to a “genetic drift” that promotes aging and its consequences³⁰. Similarly, the antagonistic pleiotropy theory of aging, initially introduced by George C. Williams, in 1957³¹, explaining aging as an evolutionary balance between early life benefits and late-life costs. The theory proposes that certain wild-type genes can have beneficial effect on fitness during early life, such as increased fertility, improved growth, or survival, but later produce harmful consequences that contribute to aging and age-related diseases. Based on his hypothesis, senescence is not a programmed process, but rather an unavoidable byproduct of natural selection that favors traits that enhance reproductive success early in life^64,65.

Disposable soma theory of aging was proposed in 1977 by Thomas Kirkwood³². His hypothesis is based on the compromise between survival and reproduction. Basically, an organism must choose between longevity or reproduction because the body or “soma” requires enough energy and resources only for one or the other, but not for both. For example, if an organism chooses to reproduce, then the soma is “disposable”, in a way the organism does not need to live any longer because it has already fulfilled its purpose. This phenomenon can be seen in different species, for instance the ones with a shorter lifespan have a higher reproduction rate, while the ones with an increased lifespan have a lower reproduction rate^32,66–68.

More recent perspectives, such as those proposed by Mitteldorf (2019), suggest an alternative interpretation based on evidence showing that some mutations can extend lifespan, without compromising reproductive capacity. From this perspective, pleiotropy should not be viewed solely as an unavoidable physiological constraint, but rather as a feature shaped by evolutionary dynamics. In his view, the inverse relationship between longevity and fertility is potentially contributing to limit excessive population growth and reduce the risk of resource depletion and ecological collapse⁶⁴. Together, these theories frame aging as an evolutionary trade-off, arising from the balance between survival, reproduction, and resource allocation across the lifespan.

Epigenetic and information-based theories

Epigenetic based theories attribute aging to changes in epigenetic patterns or the loss of youthful epigenetic information. Rather than focusing on damage alone, these models emphasize changes in how genetic information is interpreted over time. One of the main determinants of cell type and function are the patterns of DNA methylation (DNAm) established throughout embryogenesis¹⁴. The relationship of this pattern and aging was first discovered by Vanyushin et al. in 1973⁶⁹, and their observations were later confirmed and supported by numerous studies^70–73. The finding that methylation patterns have the tendency to alter with age, for presently unknown reasons¹⁴, led to the development of age predictors, generally referred as “epigenetic clocks”^73–78. By combining the methylation levels of distinct subsets of methylated cytosines (5mC) throughout the genome⁷⁹ and the chronological age of an individual, machine learning algorithms can determine the epigenetic age⁸⁰. To put it simply, people whose epigenetic age exceeds their chronological age are said to exhibit positive epigenetic age acceleration, whereas those who have an epigenetic age that is lower than predicted show negative age acceleration, indicating accelerated or decelerated biological aging in the analyzed tissue²⁷.

Age prediction based on these DNAm clocks is highly accurate across a wide range of tissues throughout the life course, linking developmental and maintenance processes to epigenetic aging and contributing to the emerging epigenetic clock theory of aging²⁷ proposed by Horvath and Raj in 2018.

Recently, in 2023, Lu and his team proposed the information theory of aging (ITOA) where they state that aging is caused by the loss of youthful epigenetic information during the lifetime²⁸. This speculation was based on the information theory postulated by mathematician and engineer Claude Shannon⁸¹. By applying this knowledge in biology, Lu et al. separated cells “information” into two storage systems²⁸. One system stores “digital information” under the form of DNA nucleotides, and the other system stores “analogue information” under the form of the epigenome.

Gradual change in epigenetic information, particularly in DNAm levels⁸², correlates with perturbations in gene expression profiles during the aging process^83,84. While the methylation pattern changes, somatic cells seem to gradually lose their differentiated identity. In mammals, DNAm predominantly occurs at specific sites called cytosine-guanine dinucleotides (CpGs) and carries out distinct functions depending on the genomic regions involved⁸⁵. The DNAm patterns (referred to as the ‘methylome’) are established and subsequently preserved through cell divisions to maintain cellular identity⁸⁴. In the context of aging, it was noted that the level of DNA methylation in the human methylome changes by only 2–5%⁸⁶.

The problem with the analogue storage system, in both computers and the epigenome, is that information is easily altered and lost. The advantage of analogue information is that it can be restored. In computers this can be fixed with the help of a “backup copy”, but in living organisms this can be restored with partial reprogramming. In previous work, it was demonstrated that cellular identity loss, mainly caused by the alteration of epigenetic information during aging can be restored to a more youthful pattern with the help of Yamanaka factors²⁸. How, why and to which extent this works without “a backup copy”, are unanswered questions currently under investigation. This subject is important and is extensively covered later in this review.

In 2024, Pio-Lopez and Levin proposed the bioelectricity and the morphostatic information theory. They state that endogenous bioelectric signals function as epigenetic regulators that encode instructive patterns for cellular organization and homeostasis. Aging is therefore understood as gradual degradation of bioelectric signaling, leading to a loss of the instructions that tell the body how to maintain its structure and function. As this regulatory information deteriorates, cells increasingly fail to maintain their physiological homeostasis and decline in regenerative capacity and tissue integrity²⁹. Together, these theories present aging as a loss of biological information and regulatory fidelity, providing a conceptual bridge between molecular mechanisms and rejuvenation-based interventions.

Cross-category or hybrid mechanisms

Several aging theories do not fit perfectly within a single mechanistic category, as they integrate elements of damage accumulation, regulatory dysfunction, and systemic decline. These hybrid models highlight the interconnected nature of aging processes at the molecular and cellular levels.

The telomere shortening theory of aging started in 1961, after Hayflick and Moorhead proposed the so-called Hayflick’s limit. They discovered that human somatic cells undergo a limited number of divisions before the onset of replicative senescence³³. The molecular mechanisms behind this loss of proliferation weren’t explained until years later, after the discovery of deoxyribonucleoprotein complexes, at the end of chromosomes^87–90. These complexes are called telomeres, and their role is to protect the coding sequences in DNA, against stressors accumulated through time. In the majority of human somatic cells, the DNA polymerase implicated in the maintenance of telomeres, known as telomerase, is inactive following cellular differentiation, leading to progressive telomere shortening with each replication cycle^7,91–94. However, in highly proliferative cell types, such as most cancer cells^95,96, stem cells^97,98 and activated lymphocytes⁹⁹, telomerase remains active. While it has been proposed that the rate of telomere shortening correlates with longevity across mammalian species¹⁰⁰, telomere dynamics are in fact highly species-specific. In contrast to humans, mice (Mus musculus) exhibit constant telomerase activity in many somatic cells^101–103. Although earlier studies reported telomere shortening in mice^104,105, more recent experiments using high-resolution sequencing techniques suggest that telomere length may remain stable with age¹⁰¹, highlighting differences in telomere regulation between mammalian species.

Similarly, the gene regulation theory of aging proposed by Kanungo in 1975³⁴, states that aging is caused by dysregulations in the genes that are activated and silenced during development. This accumulation of alterations in gene expression is responsible for flaws in cell functions, leading to loss of homeostasis and ultimately to aging^68,106.

“Inflamm-aging” or “inflammaging” theory of aging was first introduced by Franceschi et al., in 2000³⁵. Their theory states that lifelong exposure to stressors, such as infections and cellular damage, leads to a progressive increase in proinflammatory status, which contributes to aging and age-related diseases. In a healthy, young individual the immune response to infection is tightly regulated, with an initial production of proinflammatory molecules (e.g., cytokines) to eliminate pathogens, followed by anti-inflammatory signals that restore immune homeostasis. In older individuals, the immune system is dysregulated, and the initial proinflammatory response is not balanced by anti-inflammatory pathways, leading to persistent inflammation that develops into chronic low-grade inflammation, referred to as inflammaging^35,107–111.

The hyperfunction theory of aging was proposed by Blagosklonny in 2006¹¹². In contrast, to many other proposed theories that attribute aging to gradual accumulation of molecular damage, Blagosklonny proposed the opposite. In his view, aging results from the continued activity of normal, wild-type genes that promote growth and development. A central example from his studies is the mTOR signaling pathway, which is essential for growth and anabolic processes in early life, but becomes deleterious when insufficiently downregulated in later life³⁶. Based on his work on rapamycin, an mTOR inhibitor, he suggested that aging is a form of biological hyperfunction, that drives age-related dysfunctions and diseases^65,113.

The stem cell theory of aging, introduced in 2005–2006 by Campisi¹¹⁴ and Rando¹¹⁵ and further developed by Sharpless and DePinho in 2007³⁷, proposes that aging is caused by problems interfering with the self-renewal function of stem cells. Particularly, the progressive decline in their ability to differentiate, due to DNA damage and changes in the stem cell microenvironment. Since stem cells have an important role in the maintenance of tissue homeostasis, a decline is responsible for a reduced regenerative capacity. Beside the external and internal factors that lead to dysfunctions in the self-renewal process of stem cells, other mechanisms known to protect the organism against cancer, such as senescence, apoptosis, telomere shortening, also play a role in the reduced capacity of stem cells with age^37,116–119.

Together, these hybrid theories highlight that aging is a multifactorial process arising from the interplay between molecular damage, regulatory dysregulation, and tissue-level dysfunction. This complexity makes it difficult to describe aging using a single defining parameter, which in turn motivated the search for measurable biomarkers that could integrate and reflect these heterogeneous biological processes.

Biomarkers of aging

In the early 1980s, as researchers struggled to find explanations for the discrepancy between chronological age and longevity, the idea of biomarkers of aging began to emerge in gerontological and geriatric literature¹²⁰. In 1988, Baker and Sprott defined these biomarkers as measurable characteristics that can predict physiological decline or functional capacity better than chronological age¹²¹. Further, American Federation for Aging Research (AFAR) created a list of criteria that biomarkers of aging must fulfill^122–124. In the year 1988, the National Institute of Aging (NIA) started the search for biomarkers of aging, with a 10-year-old in vivo study on mice^120,125. Over the years many groups of scientists attempted to develop potential biomarkers of aging, but no universally reliable biomarkers have yet been establised^120,125,126. In 2021, Hartmann et al. stated that there still aren’t any universally accepted biomarkers and given the multi-causal nature and complexity of the aging process, it seems doubtful that one will ever be created^127,128.

Despite this, to study aging and create anti-aging approaches researchers need biomarkers. So, the observed changes proposed in the aging theories and the markers used in age-associated diseases became the so-called biomarkers of aging^122,129. As anticipated, their effectiveness to quantify biological aging is contested and many scientists agree that only one biomarker cannot measure the pace of aging. However, each currently known indicator offers insight into aging, but only at certain levels, rather than systemic aging¹³⁰. The current list of biomarkers is abundant, and it reflects the many modifications that take place in an organism during aging.

These so-called biomarkers can be classified based on the observed changes at different levels of biological organization from changes we do not see in molecules (Tables 1 and 3) or cells, to the ones that are noticeable, like loss of functions, both physically and mentally (Table 2).

Table 1.

Molecular biomarkers of aging

Type	Biomarker	Change	Additional classification	Reference
Genetic	TL	↓	Research	322–324
Epigenetic	Aging-positive CpG sites	↑	Research	145,325–328
	Aging-negative CpG sites	↓	Research	145,325–328
	H4K20me3, H3K4me1, H3K4me2	↑	Research	329–331
	H4K20me1, H3K9me3, H3K27me3	↓	Research	332–334
	DNMT1	↓	Research	335
	DNMT3A, DNMT3B	↑	Research	336
Transcriptomic	IL1B	↑&↓	Research	136,137
	IL6	↑	Research	337,338
	CXCL8	↑	Research	339
	IL15	↓	Research	142,143
	TNFR1, TNFR2	↑	Research	340
	SIRT1	↓	Research	341
	SIRT6	↓	Research	342
	GDF15	↑	Research	343
	CXCL1	↑	Research	344
	TERT	↓	Research	342
	BRCA1	↓	Research	345
	KL	↓	Research	346
	IGF1	↓	Research	347
	CDKN1A (p21)	↑&↓	Research	135,348
	CDKN2A (p16INK4a)	↑&↓	Research	139
Proteomic	IL-1β	↑	Clinical and immunological	349
	IL-6	↑	Clinical and immunological	349,350
	IL-8	↑	Clinical and immunological	351
	IL-15	↑	Clinical and immunological	141
	TNFα	↑	Clinical and immunological	349,350
	TGF-β1	↑	Clinical and immunological	352
	SIRT1	↓	Research	353
	SIRT6	↓	Research	342
	GDF15	↑	Clinical and immunological	354,355
	CXCL1	↓	Clinical and immunological	356
	CRP	↑	Clinical and immunological	357
	TERT	↓	Research	342
	α-Klotho	↓	Research	358
	IGF-1	↓	Clinical and metabolomic	359
	PAI-1	↑	Clinical	360
Metabolomic	AGEs	↑	Clinical	361
	Carbamylated proteins	↑	Clinical	362,363
	TG	↑	Clinical	364,365
	Total cholesterol	↑	Clinical	365
	Glucose	↑	Clinical	364

TL telomere length, DNMT DNA methyltransferase, IL interleukin, CXCL C-X-C motif chemokine ligand, TNFR tumor necrosis factor receptor, SIRT sirtuin, GDF growth differentiation factor, TERT telomerase reverse transcriptase, KL klotho, IGF insulin-like growth factor, CDKN cyclin-dependent kinase inhibitor, TNF tumor necrosis factor, CRP C-reactive protein, PAI plasminogen activator inhibitor, AGEs advanced glycation end products, TG triglycerides, ↑increase, ↓ decrease, ↑&↓ context-dependent changes, additional classification column indicated the primary evidence context (e.g., clinical, experimental, or mixed), reflecting differences in study design, tissue source, and translational relevance.

Table 3.

Clinical and behavioural biomarkers of aging

Category	Type	Biomarker	Change	Additional classification	Reference
Clinical and imaging	Clinical	Hb	↓		398,399
		HbA1C	↑		400
		Lymphocytes	↓	Immunological	401,402
		Monocytes	↑	Immunological	403,404
		HCT	↓		405
		ALB	↓	Molecular	406,407
		Creatinine-cystatin C	↑	Molecular	408,409
		Urea	↑	Molecular	410
		BIL	↑	Molecular	411
		ALP	↑	Molecular	412
		Skin microbiome	↑	Molecular or omics-based	413,414
		Gut microbiome diversity	↓	Molecular or omics-based	415,416
	Imaging	BMD	↓		127,417
		Muscle mass	↓		127,418
		Brain volume	↓		127,419
		Atherosclerotic lesions	↑		420,421
Behavioral	Lifestyle	Smoking	↓		40,187,188
		Excessive alcohol intake	↓		40,187,188
		Nutritious diet	↑		422–424
		Caloric restriction or intermittent fasting	↑		425–427
		Physical exercise	↑		40,187,188

Hb hemoglobin, HbA1C glycosylated hemoglobin, HCT hematocrit, ALB albumin, BIL bilirubin, ALP alkaline phosphatase, BMD bone mineral density, ↑increase, ↓ decrease, additional classification column indicated the primary evidence context (e.g., clinical, molecular, or mixed), reflecting differences in study design, tissue source, and translational relevance.

Table 2.

Biological and functional biomarkers of aging

Category	Type	Biomarker	Change	Additional classification	Reference
Omics-based	Epigenomics	Genome-wide DNA methylation pattern	Altered	Research	69–75,325,366
	Transcriptomics	Gene expression pattern	Altered	Research	367–370
	Proteomics	The level of synthesized proteins	Altered	Research	160,371,372
	Metabolomics	Global profile of metabolites	Altered	Research	161,373–375
	Enzymomics	The expression level of enzymes	Altered	Research	376
Cellular	Senescence		↑	Research	111,377–382
	Extracellular vesicles		↓	Research	383
	Oxidative stress and mitochondrial dysfunction		↑	Research	50,111,384,385
	Stem cells self-renewal		↓	Research	37,116,118,386
Functional	Physiological	Mouth and nose width	↑	Imaging	387
		Sitting and standing balance	↓	Clinical (CGA)	388
		BMI	↑	Clinical (CGA)	389
		WC	↑	Clinical (CGA)	390
		Walking speed	↓	Clinical (CGA)	391
		Grip strength	↓	Clinical (CGA)	392
		Lung function	↓	Clinical (CGA)	393
		Blood pressure variability	↑	Clinical (CGA)	394
	Cognitive	Short-term memory	↓	Clinical (CGA)	395
		Processing speed	↓	Clinical (CGA)	395
		Verbal function	↓	Clinical (CGA)	395
	Sensorial	Vision	↓	Clinical (CGA)	396
		Smell	↓	Clinical (CGA)	396
		Hearing	↓	Clinical (CGA)	396
		Taste	↓	Clinical (CGA)	396
		Chronic pain	↑	Clinical (CGA)	397

CGA comprehensive geriatric assessment, BMI body mass index, WC waist circumference, ↑increase, ↓ decrease, additional classification column indicated the primary evidence context (e.g., clinical, experimental), reflecting differences in study design, tissue source, and translational relevance.

It’s important to note that the rate of biological aging is not the only factor that affects these biomarkers. Although they all are expressed in various human tissues, their expression levels differ according to tissue type (e.g., the expression of p21 and p16 increases in skin, and decreases in muscle, IL1B gene expression is lower in aged skin and higher in blood)^131–137, others show gender-related differences (e.g., females present a lower epigenetic age acceleration compared to males)¹³⁸, and certain biomarkers are drastically influenced by the presence of an age-associated pathological condition (e.g., p16 is increased in healthy older individuals and decreased in older individuals with Alzheimer’s disease)¹³⁹. Moreover, some biomarkers present at both transcriptomic and proteomic levels exhibit differences across regulatory layers, including post-transcriptional regulation and protein turnover (e.g., IL15 gene expression is decreased, while IL-15 protein levels are increased)^140–143.

At the molecular level (Table 1), changes in the methylation frequencies of specific CpG sites or histones caused by disruption of DNA methyltransferases (DNMTs) activity can be the trigger of a “domino effect”, a cascade of interdependent molecular events^144,145. If located in intragenic regions of the genome, these modifications can further alter the gene expression, causing gradual disruptions to the transcriptome and, ultimately, to the proteome^146–148. This is why many biomarkers of aging are well conserved and have an interconnected pattern of disruption that can be observed at several omics’ layers. One of the most representative examples is the up-regulated expression of pro-inflammatory factors (IL-6, IL-1β, TNFα, monocyte chemoattractant protein-1 (MCP-1), C-reactive protein (CRP)) in aged tissues^111,149,150. The high levels of pro-inflammatory factors lead to the activation of pro-inflammatory signaling pathways, such as Nf-kB, and together with a dysregulated immune system, cause the onset of inflammaging^151,152. Another molecular marker that is directly proportional with lifespan and age-associated diseases, such as cardiovascular diseases^153,154 certain types of cancer^87,155 and neurological problems^156,157 is shortening of telomeres, specifically in leukocytes.

Instead of analyzing individual molecular changes, omics-based biomarkers (Table 2) rely on multiple CpG sites, genes or proteins to view the changes as a global pattern. They are multi-biomarkers that can predict a biological age by computing the level of DNA methylation^74,75,77,158, number of messenger RNA (mRNA) transcripts¹⁵⁹, protein or metabolites abundance^160–162 with statistical instruments, called biological aging clocks. With the current advancements in biohorology, the fields that studies biological aging, we now have multiple clocks that can measure and estimate an age from each omics layer¹⁶³. Epigenetic clocks are the most widely used and they are based on the methylation levels from only three CpGs¹⁵⁸ or from thousands¹⁶⁴. These clocks are versatile, some of them analyze the entire methylome landscape (CpG islands, CpG shelves, CpG shores and open sea), such as DNAmAge⁷⁴ and others are targeted toward specific regions, like histones¹⁶⁵ or retrotransposons, known as RetroAge¹⁶⁶. Although these instruments use multiple biomarkers, they still predict only one aspect of aging. As Johnson and Shokhirev (2024) suggested, a specific biological clock can only predict the age of what they analyzed, for example epigenetic clocks predict only epigenetic age (eAge), not biological age¹³⁰. To better portray biological aging, second generation clocks were developed, that include additional indicators during model training to identify CpG sites correlated with aging-related phenotypes. For example, GrimAge^164,167 includes plasma proteins (myoglobin, beta-2-microglobulin, adrenomedullin) and lifestyle habits like smoking and PhenoAge¹⁶⁸ includes clinical biomarkers (glucose, CRP, creatinine).

At the cellular level, biomarkers focus on the entire aspect of the cells, including morphological changes (senescent cells), altered cellular communication (the activation of proinflammatory pathways) or loss of function (stem cells self-renewal). They are closely correlated with molecular markers, as the cellular alterations are directly reflected in changes at the molecular level¹⁶⁹. For example, numerous researchers point out that the constant accumulation of senescent cells is one of the causes of aging^170,171. These cells release a by-product called senescence-associated secretory phenotype (SASP), predominately characterized by the previously mentioned pro-inflammatory factors^111,150. Increasing evidence suggests that aging arises early in the vascular system, particularly in the endothelial cells lining the blood vessels. As these cells become senescent, they secrete SASP factors in the blood stream, which may facilitate the spread of aging-related alterations throughout the entire organism¹⁷².

Functional changes can be included among the categories of biomarkers (Table 2), as some phenotypical changes can give insight into biological aging. Facial features have been used in phenotypical biological aging clocks, such as VisAgeX¹⁷³. The physiological, cognitive and sensoria functions can also be classified under clinical and medical imaging biomarkers (Table 3), because they are part of the comprehensive geriatric assessment (CGA)¹⁷⁴. Although, these biomarkers predict age-associated frailty, rather than the rhythm of aging, they are still a useful assessment, knowing that frailty increases with age and is frequently linked to a predisposition for chronic illnesses¹²⁷. Also, many changes in the levels of metabolic biomarkers (glucose, cholesterol, AGEs) that are used in clinic for cardiovascular disease or diabetes, have been proven to also be associated with accelerated aging^175,176.

Even our day-to-day habits reveal information about our biological aging and directly influence other molecular and cellular changes. As previously mentioned, behavioral biomarkers can be used as additional indicators in second generation clocks¹⁶⁴. Moreover, these biomarkers are normally the ones targeted by anti-aging strategies, because lifestyle changes, like a healthy diet or quitting smoking, represent accessible and efficient changes that can slow down the pace of aging^177–179. These strategies are discussed in the next section.

Anti-aging strategies

Advancements in science, medicine, technology, and improved living conditions have increased human life expectancy¹⁸⁰. The steady increase in the aging population observed in recent years does not necessarily represent an advantage, even if we can live a longer life, it does not mean that we also live a disease-free life¹⁸¹. These, together with progressive decline and age-related frailty, directly affect an individual’s quality of life. Moreover, the need for costly medical care, support and treatment will inevitably lead to a socio-economic burden on families and health systems worldwide^182–185.

While there is no way to stop aging with the current technologies, strategies and therapies aimed to restore cellular systems to optimal function may be able to slow down or reduce the effects of aging in older individuals¹⁸⁶. Throughout the years many anti-aging strategies have been proposed. So far, they are placed in two categories: behavioral and pharmacological approaches (Fig. 2)¹⁸⁷.

Fig. 2. Classification of anti-aging strategies.

This figure highlights the current anti-aging strategies, grouped into two main categories: behavioral strategies (lifestyle changes, such as diet and exercise), and pharmacological strategies (drugs approved for other diseases, that have shown potential anti-aging benefits, such as NMN, metformin, resveratrol, rapamycin and D&Q). A new potential anti-aging strategy is partial reprogramming (induction of OSKM factors). Although, a success in vitro and in vivo, this approach still requires the development of safe methods of administration, as well as a deeper understanding of the mechanisms by which it can reverse the epigenetic clock and promote regeneration and rejuvenation (Created in BioRender.com).

Behavioral strategies

Behavioral strategies consist of changes every individual can make in their day-to-day life. For example, avoiding or quitting smoking, avoiding excessive alcohol intake, exercise and eating a clean diet. According to previous studies, the best-known methods for extending lifespans in several model organisms are caloric restriction and fasting^187,188. Fitzgerald et al.^189,190 successfully demonstrated that only 8-weeks of consistent lifestyle changes can reduce the average biological age by up to 4.6 years for woman and 2.04 years for men. The 8-week program consisted of a diet rich in epinutrients, such as wild berries, turmeric, green tea, pumpkin seeds and more, adequate water intake, minimum of 30 minutes of daily physical activity, as well as a minimum of 7 hours of sleep and breathing exercises for stress management. One of the rules in the program was to practice a 12:12 intermittent fasting routine, meaning the participants were allowed to eat for 12 hours and fast for the remaining 12 hours^189,190.

Another potential therapy aimed to enhance health and slow down aging is the consumption of nutraceuticals. These are compounds found in food or dietary supplements, such as polyphenols, theaflavins, catechins and anthocyanins. Researchers have found that nutraceuticals can improve health, treat, prevent diseases, and slow down aging⁴⁰. For example, Menicacci et al. (2017) demonstrated, in an in vitro experiment on neonatal human dermal fibroblasts, that oleuropein aglycone, a polyphenol, can reduce the expression of the senescence-associated beta-galactosidase (SA-β-Gal) marker and downregulate the expression of p16, thereby contributing to a decrease in the number of senescent cells¹⁹¹. In addition, flavonoids, a class of polyphenols, have been shown to protect against age-related conditions such as Parkinson’s disease by protecting neurons against oxidative stress and neuroinflammation^192,193.

Even though behavioral strategies are accessible and represent a promising method to decelerate the speed of aging, many find it hard to follow a strict routine for an extended period¹⁸⁷.

Pharmacological strategies

In terms of pharmacological strategies, the approach is based on modulating signaling pathways that are dysregulated by aging in many species. Certain medications can influence signaling pathways the same way behavioral strategies do. For example, caloric restriction was seen to interact with the adenosine monophosphate-activated protein kinase (AMPK), mechanistic target of rapamycin (mTOR) and sirtuins^6,187. These medications that can slow down the aging process, by targeting different aging pathways, are called geroprotectors^194,195. Some examples are nicotinamide mononucleotide (NMN)^196–198, rapamycin¹⁹⁹, metformin²⁰⁰, resveratrol²⁰¹, dasatinib and quercetin (D&Q)²⁰². In 2014, a study by Bannister et al. highlighted the beneficial effects of metformin on slowing the aging process²⁰³. It was observed that patients with type 2 diabetes that were treated with this drug had a 15% longer survival time compared to non-diabetic matched individuals^203,204. Results from ongoing clinical trials evaluating metformin as anti-aging intervention in a healthy population were not published yet^205,206. Subsequently, after rapamycin was shown to prolong lifespan in mice²⁰⁷, Harinath et al. (2024) continued the research, this time on humans. Results revealed alleviated frailty biomarkers, such as an increase in muscle mass and reduction in joint pain, contributing to an overall self-declared well-being²⁰⁸. Also, in a preliminary study, senolytics such as D&Q were shown to be effective in reducing the expression of senescence and proinflammatory markers, in patients with chronic diabetic kidney disease (DKD)²⁰⁹.

Many geroprotectors have been tested on model organisms and even though they successfully increased lifespan, they also caused harmful effects on health and inflammation, due to their on-target and off-target toxicity^{199,210–213}. Because the long-term side effects of pharmaceutical treatments may outweigh the advantages, their usage is still controversial for clinical use¹⁸⁷.

Partial reprogramming experiments

One of the first aims of aging research was and still is the development of an anti-aging therapy that not only prolongs lifespan, but also preserves physiological function, in other words prolongs healthspan. Recent years have seen the emergence of cellular reprogramming using transcriptional factors, as an effective new method to reverse aging-related cellular phenotypes by revitalizing aged cells, restoring epigenomic, transcriptomic, proteomic and metabolomic patterns^214–216. This new method was developed from Takahashi’s and Yamanaka’s creation of induced pluripotent stem cells (iPSCs)^217,218. To generate these cells, they created a hypothesis based on the known fact that animals have previously been successfully cloned by somatic cell nuclear transfer into oocytes²¹⁹, for example the Dolly sheep²²⁰. Another successful discovery was that by fusing with embryonic stem cells (ESCs), somatic cells can also be converted to an embryonic-like state. Based on these results, Takahashi and Yamanaka postulated that specific transcriptional factors are responsible for the pluripotency of differentiated cells and ESC identity^{217,218,221,222}.

To experimentally test this hypothesis, they tested 24 transcriptional factors on an adult mouse fibroblast cell line, using retroviral transduction and they observed that four of these factors are implicated in generating and maintaining iPS cells²¹⁷. A year later, the same team of scientists successfully generated iPS cells, this time from adult human fibroblasts, using the same combination of four transcriptional factors (OCT4, SOX2, KLF4 and MYC). Both iPSC from mouse and human resembles embryonic stem cells in terms of their morphology, proliferation, expression of genes and epigenetic status²¹⁸.

In the following years, anti-aging scientists approached this unique method in the hopes of finding a cure for aging and age-associated diseases. But they soon realized that once a cell is fully reprogrammed, the epigenetic landscape (and consequently the epigenetic clock) is completely reset to zero⁷⁴, and the cells lose their identity and dedifferentiate^14,223–226.

Since dedifferentiation in vivo carries the risk of cancer²²⁷, many scientists looked for ways to prevent it. They soon discovered that a potential anti-aging treatment could be achieved by modifying the original protocol in ways to make it safer. Therefore, Ocampo et al. (2016) demonstrated that a short-term induction of those four reprogramming factors achieves a youthful state without the complete removal of the epigenetic signature that determines the cell’s identity^223,228. The concept of reversing the aging process without the dedifferentiation of cells was named “partial reprogramming”^227,229,230.

From the initial protocol to the most recent, partial reprogramming using Yamanaka factors has proved itself to be a viable anti-aging treatment that can both ameliorate aging and regenerate tissues that are typically affected by age-related diseases (Fig. 3)^14,227,228.

Fig. 3. Schematic representation of partial reprogramming protocols used in anti-aging studies, in chronological order.

The figure illustrates nine panels showing the research team and the year of publication, the induction and delivery method used, the number of repetition cycles, the type of experimental model (mice/rat for in vivo and cells for in vitro studies) and the age at the beginning of the experiment, and lastly, a note highlighting important experimental details. The last panel, in blue, contains the legend, where: pink boxes indicate the induction days, the blue boxes represent the break days, the pink-blue boxes indicate an on/off induction protocol, in which OSK factors are activated for one week and stopped the following week, and the circle shows the total duration of the protocol, expressed in weeks (Created in BioRender.com).

Ocampo et al., 2016—the groundbreaking protocol

Despite the importance of previously done in vitro investigations²³¹, a more in-depth understanding of the potential impact of cellular reprogramming using Yamanaka factors in the field of aging requires an in vivo approach²²⁸. Following those investigations, they speculated that lifespan can be extended, and the biological process of aging can be delayed or even reversed with the right technique of OSKM induction. From this assumption, Ocampo and colleagues created the first successful protocol for in vivo partial reprogramming.

For their experiment they used LAMIN-A knock-in (LAKI) mice, a model for progeria syndrome, crossed with four-factor (4 F) transgenic mice, carrying one copy of OSKM polycistronic cassette, used to activate the expression of the four transcriptional factors with 1 mg/mL doxycycline (DOX) in drinking water.

Firstly, before they created the current cyclic induction protocol, they tested a long-term continuous induction experiment, even though it was known from previous full reprogramming in vivo studies that a continuous expression of these factors leads to tumorigenesis^{217,218,232,233}. And just like the previous results, they observed severe weight loss, teratoma formation and death after only 4 days of continuous induction²²⁸.

Secondly, after trials and errors, they created the cyclic induction protocol, which consisted of 2 days of DOX administration to activate the OSKM factors and 5 days break. For the longevity experiment the OSKM treatment was induced in 8-week-old LAKI 4 F mice and continued until death, for a period of approximately 16 to 22 weeks. With this protocol, the high rate of premature death and weight loss previously observed in the continuous expression of OSKM was prevented. Also, to test the safety of this treatment they cyclically induced OSKM for 35 weeks in wild-type (WT) 4 F mice and no signs of teratomas, dysplasia or iPSCs were observed.

This protocol of partial reprogramming was enough to increase lifespan and improve age-associated symptoms in progeroid mice models. Compared to control mice, the treated mice not only had an enhanced exterior appearance, but also the age-associated histological alterations were enhanced. For example, the organs that tend to be altered by physiological aging, such as skin, kidneys, stomach, spleen, had a noticeably improved appearance^{228,234–237}. These findings show that age-associated histological alterations can be ameliorated and ultimately longevity increased in LAKI 4 F mice, since the most obvious feature of aging is the decline in regenerative capacity and subsequently loss of tissue homeostasis.

Furthermore, with this protocol, the presence of biomarkers of aging was reduced in some examined tissues. For example, IL-1α and stress response genes in the p53 pathway were downregulated in liver, kidneys and stomach. Also, H3K9me3 and H4K20me3, two histone markers with role in heterochromatin preservation, showed restored normal levels, in kidneys and spleen. This proves that the cyclic treatment of OSKM induction could inhibit age-related epigenetic modifications^228,238,239.

Additionally, they demonstrated that in naturally aged mice, short-term exposure to Yamanaka factors reduces damage in tissues affected by age, like muscles and pancreas. Furthermore, they used the short-term protocol to induce OSKM in mouse and human cells, for an in vitro experiment. They revealed that aging cells can be restored to a youthful state.

Ocampo’s et al. research opened the doors to partial reprogramming, a potential anti-aging treatment that can delay the aging process by inhibiting the molecular changes linked to aging, such as cellular senescence pathways being activated and the unfolding of epigenetic modifications. Ultimately, these changes can extend lifespan and improve tissue homeostasis²²⁸ by partially attenuating aging phenotypes¹⁸⁶.

Sarkar et al., 2020—first mRNA protocol

Even though Ocampo and his team showed that partial reprogramming can be achieved in progeroid mice, no one attempted to test this theory on naturally aged human cells. Moreover, in vitro studies done before only used retroviral or adeno-associated viruses to induce OSKM in cells²⁴⁰. Sarkar et al. acknowledged these limitations and managed to achieve partial reprogramming by non-integrative expression of mRNA containing the OSKMLN genes²²³.

For their assay, they cultivated fibroblasts and endothelial cells, from human donors aged 25–70 years and 45–60 years, respectively. They induced partial reprogramming by using a non-integrative protocol, where they transfected LNP-mRNAs expressing six transcriptional factors known to induce iPSCs, also known as the OSKMLN cocktail, beside OSKM, they used two additional transcriptional factors Lin28 (L) and Nanog (N), previously described by Takahashi, Yamanaka and Thomson^217,241.

To begin with, they applied their protocol continuously, to find the window of opportunity or the previously described point of no return (PNR), before the cells are fully reprogrammed²⁴². By measuring the expression of long non-coding RNAs (lncRNAs) associated with the initiation of pluripotency, they observed that on day 5, the PNR takes place. After this finding, they transfected the mRNA cocktail for 4 days and analyzed the gene expression of the cells 2 days post OSKMLN transfection, in comparison to control untreated cells from the same donors²⁴³. Additionally, they analyzed the transcriptome and found that OSKMLN transfected cells had a gene expression profile similar to the young cells. Using this transient induction protocol, the expression of pluripotency genes was not observed, meaning that the cells did not lose their identity following this transient induction protocol of reprogramming factors. Without altering the expression of genes associated with cell identity, the transcriptome signature analysis showed that OSKMLN expression stimulates a very quick activation of a more youthful gene expression profile that is cell-type specific. Also, using Horvath’s pan-tissue clock⁷⁴, the authors observed that DNA methylation age was more markedly reverted in endothelial cells than in fibroblasts. Their findings demonstrate that, in human somatic cells, similar to observations in progeroid mouse models, cellular age can be rapidly reversed without loss of cellular identity²²³.

Lu et al., 2020—OSK-only epigenetic rejuvenation part 1

Lu’s et al. experiment started from the idea that tissues lose their homeostasis because of epigenetic errors that accumulate over time, leading to a decline in the regenerative potential, phenomenon observed in many species¹⁴. To test this theory, they used a mouse model (C57BL6/J) for glaucoma, since it is one of the most common causes of blindness in older individuals. This disease manifests with a high level of intraocular pressure, a gradual atrophy of the optic-nerve and loss of retinal ganglion cells (RGCs). An interesting fact noticed in embryos and neonatal mice is that the optic-nerve, if damaged, can fully be regenerated by RGCs. However, this regenerative power is lost after birth^14,244,245.

Their aim was to regenerate the optic nerve, and ultimately to restore vision, using only three of the Yamanaka factors (OSK). First, they created a mouse model of optic nerve crush injury, after that an adeno-associated virus (AAV) polycistronic system containing the OSK factors was injected into the eye. The transcription factors were activated by DOX (2 mg/mL) in drinking water. They chose to express only OSK, since c-Myc (M) was observed to be associated with cancer and death in previously done studies on mice^14,246. Before the regeneration experiment, they tested the lifetime induction on wild-type C57BL6/J mice, to see if the AAV-OSK method is safe. As a result, no tumors or structural alterations were observed, even after a continuous period of induction (78 weeks). For the glaucoma mice, 4 weeks of OSK activation resulted in partial regeneration of the optic nerve, but not full restoration of vision.

In terms of analyzing the transcriptome, after 5 days of OSK induction they observed that gene expression was like a young mRNA profile, without any significant signs of cellular identity loss. The gene expression of Lmnb1, Chaf1b, two genes known to change with age, was restored back to a youthful level. They also monitored and found no expression of Nanog, and found no sign of differentiation loss.

Therefore, Lu and his team demonstrated that OSK-expressing cells retained their cellular identity over a period of 10-18 months of continuous OSK induction, no tumors or other modifications that could have a negative effect on health were detected¹⁴.

Chondronasiou et al., 2022—one-week of treatment can reverse aging

After it was demonstrated that partial reprogramming can be achieved by different protocols, a few questions remain unanswered about the exact mechanisms that take place during reprogramming and what happens after you stop activating the transcriptional factors. Taking that into consideration, Chondronasiou et al. decided to do a multi-omics analysis on different tissues that are known to be altered with age²¹⁶. For the experiment they used naturally aged transgenic i4F-B mice (55 weeks old; 100 weeks old) and the OSKM expression was activated by administering a low dose of DOX (0.2 mg/ml) in drinking water.

The transient expression of OSKM for 1 week did not lead to teratoma formation or cell identity loss, but histological modifications were noticed in the pancreas. However, these modifications were reverted to the original state 2 weeks after OSKM expression was stopped.

They also observed that the stomach and intestine suffer histological enhancements, while the liver and the spleen remained unrestored. and interestingly, the epigenome and transcriptome changes happened after OSKM treatment was stopped.

Additionally, they compared DNA methylation sites between young (13 weeks old), OSKM mice and 55 weeks old mice. They noticed that the group of mice treated with OSKM had a DNA methylation age located between young and old groups, suggesting that the epigenetic age was reset by partial reprogramming. They also noticed that some sites that are hypomethylated with age, were hypermethylated after OSKM induction and vice versa. Also, to find if these modifications take place during partial reprogramming or after, they compared the same methylation regions at day 7 and day 21. The results showed that demethylation takes place after the OSKM expression is stopped and methylation takes place during the one-week period of induction.

Analyzing the transcriptome, they detected no expression associated with genes that are upregulated during pluripotency, like Nanog, endogenous Oct4 or Tfe3. They also tested this protocol for a period of 2 weeks, and the expression of the genes mentioned above was upregulated in the pancreas, indicating that full reprogramming was achieved²¹⁶.

Browder et al., 2022—long-term protocol

Browder et al. continued the initial partial reprogramming experiment from 2016²²⁸, this time for a longer period, demonstrating that a long-term cyclic expression of OSKM is safe to use on naturally aged mice and it has rejuvenating effects. Even though long-term continuous expression of OSKM factors reprogramming has been attempted before and all the experiments resulted in teratomas, this time no teratomas or any other irregularities were noticed due to cyclic, discontinuous expression^214,232,233.

Browder’s team was previously involved in creating the first successful in vivo partial reprogramming protocol²²⁸. That study used progeria mouse models, so this time they wanted to test a long-term and short-term protocol at different time-points during physiological aging in 4 F reprogrammable mice.

For the long-term partial reprogramming protocol, they had 2 groups: in the first group the OSKM induction treatment was administered via DOX (1 mg/mL) in drinking water for approximately 30 weeks (started at 15 months of age and ended at 22 months), in the second group the OSKM induction was performed for around 43 weeks (started at 12 months of age and ended at 22 months of age).

After the long-term partial reprogramming treatment, they used DNA methylation clocks to analyze the epigenetic age and RNA-seq to analyze the gene expression. They noticed that the kidneys and the skin were reverted to a more rejuvenated molecular phenotype, compared to other tissues, or untreated mice.

In addition to the long-protocol, they also attempted a late-onset short-term protocol (4 weeks) in 25 months old mice. The short expression of OSKM in older mice did not result in rejuvenation and no significant changes were observed by analyzing the epigenome and transcriptome. They speculated that old cells are in a state that prevents reprogramming, other researchers have also noticed that reprogramming usually fails to affect senescent cells^216,247.

Compared to the protocol created by Ocampo et al., they managed to obtain better results, proving that a long-term protocol is more effective than a short-term induction of OSKM. Ultimately, this experiment showed that the beneficial effects of long-term expression of OSKM are directly proportional to the duration of the treatment²¹⁴.

Alle et al., 2022—early-life prevention

Alle et al. aimed to prove that OSKM induced earlier in life, at a lower dose, for only a short period of time can induce results like a continuous or cyclic induction protocol. So, they created a transgenic modified version of LAKI progeria mice that produces lower levels of progerin, similar to the level expressed during normal aging²⁴⁸.

As a first experiment Alle et al. modified Ocampo’s et al. first protocol for cyclic OSKM induction²²⁸. Instead of inducing reprogramming with 1 mg/mL DOX administered for 2 days in drinking water with a 5-day break, for 35 cycles, they induced OSKM by administering a lower dose of DOX (0.2 mg/mL) in drinking water, continuously throughout life. Even with a lower concentration of DOX the life span was improved significantly, like Ocampo’s results²⁴⁸.

Alongside the cyclic induction protocol, they attempted a short-term induction protocol, where they activated OSKM with 0.2 mg/mL DOX, respectively 0.5 mg/mL DOX for a short period of time (2.5 weeks). They use 2 months old progeria-like mice. The group that received only 0.2 mg/mL of DOX in drinking water did not obtain any significant results, but from the 0.5 mg/mL concentration of DOX they managed to increase the lifespan.

To assess the changes, the mice were analyzed after the short-term protocol when they reach an older age. They noticed that the histology of organs affected by old age was enhanced and by analyzing the epigenome they noticed that the DNA methylation pattern was rejuvenated. The short protocol did not affect the body weight of the mice, but it prevented the mice from gaining musculoskeletal deteriorations as they aged. Altogether, they proved that a short induction of OSKM at an early onset is enough to give the anti-aging and regenerative benefits of partial reprogramming.

Karg et al., 2023—OSK-only epigenetic rejuvenation part 2

Karg’s et al. experiment is a more in-depth version of their experiment in 2020^14,15, when they observed that AAV-OSK injected directly into the eye can restore vision and reverse aging phenotypes in mice models with induced glaucoma. In this follow up study, they monitored the changes made by OSK partial reprogramming for 1 year, to better understand when the exact window of opportunity takes place and see the extent of the benefits after OSK induction is stopped¹⁵.

As previously done, they used mice models (C57BL6/J) for glaucoma (3-4 months old). Using a dual AAV system injected into the eye and DOX (1 mg/mL) in drinking water to activate the OSK expression. They noticed that vision was restored after 8 weeks of continuous administration. To see how long the OSK genes are expressed after treatment is stopped, they analyzed the expression in RGCs and noticed that after only 2 weeks of DOX administration the level of OSK returned to the beginning level only 1 month after administration was stopped.

They noticed that vision was fully restored after a period of 8 weeks of DOX administration, and the vision remained restored for up to 11 months after OSK induction was stopped. To prove the safety and effectiveness of the OSK treatment, the continuously activated the three factors for a period of 91 weeks. No teratoma formation, weight loss or changes in the anatomy of the eyes were observed.

In conclusion, their results prove that partial reprogramming can be a useful treatment for glaucoma, an age-associated disease.

Macip et al., 2024—even old age can be reversed

The AAV-OSK-inducible system is a reprogramming technique that has been used before, but only on transgenic mouse models¹⁴. Macip and his team stated the idea that future treatments for age reversal in humans cannot be effectively applied to transgenic mouse models¹⁸⁶. So, instead they used naturally aged WT mice. They managed to partially reprogram 2-year-old/124 weeks old mice using retro-orbital injection of an AAV-OSK system induced by DOX (2 mg /mL) in drinking water in one week on and one week off manner. Compared to control mice there was a significant increase in remaining lifespan and overall health. The purpose of their experiment was to prove that even very old mice can receive the benefits of partial reprogramming.

Using DNA methylation clocks, they observed that the life span was increased compared to control. Also, the liver and heart, the two organs where OSK was highly expressed, presented a lower epigenetic age. Also, by using quantitative PCR (qPCR) they observed that OSK was not as strongly expressed in the brain. Another key point to consider is that by analyzing the tissues at the end of the experiment they notice no teratomas.

Overall, the results show that partial reprogramming using OSK can extend lifespan and reverse aging phenotypes even in very old mice.

Horvath et al., 2025—age rejuvenation and cognitive function restauration

Just like vision, the power to learn and memorize tends to decline with age. Thankfully, it was proven that Yamanaka factors can regenerate vision in glaucoma mice models, while also exhibiting epigenetic rejuvenation^14,15. As far as learning and memory loss, there were no studies to attest the regenerative potential of the hippocampus using OSKM factors until recently. In 2024, Horvath and his team directly injected adenoviral vector carrying OSKM genes into the hippocampus of old rats (25.3 months) following a continuous protocol for a period of 5.6 weeks. After the treatment has ended, the rats were analyzed and no morphological changes were observed in the brain, meaning that the treatment did not induce pluripotency in the hippocampal cells²⁴⁹.

Firstly, they analyzed the learning performance and spatial memory, using a behavioral test following a modified Barnes maze protocol and observed significant improvement compared to untreated old rats. Furthermore, to attest whether the cells in the hippocampus were epigenetically rejuvenated after the 39 days of treatment, they used Horvath’s rat pan tissue DNAmAge epigenetic clock. By comparing the results from the DNA extracted from old-OSKM rats and the control-old rats they discovered a number of 671 CpG sites that presented significant differences in methylation levels. Also, by comparing the methylome of OSKM-old rats, control-old rats and control-young rats they observed 174 CpGs that are hypomethylated in the treated rats and the young ones, and hypermethylated in old rats. These results point to the fact that the OSKM treatment caused a demethylation at these specific sites leading to epigenetic rejuvenation^250,251.

Ultimately, it is clear that their study proves that OSKM direct injection can rejuvenate the hippocampus causing an improved learning ability and memory in old rats, just like it was previously observed that OSKM injection directly into the eyes of old mice can rejuvenate vision. Moreover, their results show the epigenetic rejuvenation potential of the OSKM treatment, by discovering a subset of CpG that are hypermethylated with old age in rats but hypomethylated after the induction. This led the scientists to hypothesize that the hypomethylation of these specific CpGs could increase the expression of genes associated with learning and memory, but further research is required to prove this theory due to insufficient data²⁴⁹.

Anton-Fernandez et al., 2024; 2025—age rejuvenation and brain regeneration

Anton-Fernandez and his colleagues also observed anti-aging benefits while performing partial reprogramming protocols for neurodegenerative disorders^252,253. They followed a standard cyclic induction (3-4 days DOX + , 3-4 days DOX -) of OSKM via DOX (2 mg/mL) activation in 6 months old neuron-restricted reprogrammable mice²⁵² and in i4F-B mice crossed with P301S transgenic mice, a reprogrammable model for Alzheimer’s disease²⁵³. At the end of their experiments, after 16–20 weeks, they reported improved cognitive function, respectively a decrease in phospo-tau aggregates and neuroinflammation. The OSKM treated mice also showed anti-aging changes in neurons, such as the increase of H4K20me3 and H3K9me3, two biomarkers that get hypomethylated with age in the mouse brain²⁵⁴. Just like in previous cyclic protocols, no alterations linked to full cellular reprogramming were remarked, probably because they have targeted neurons, which are post-mitotic cells, and they do not divide^252,253.

Multi-omics profiling of partial reprogramming

Although the precise molecular and cellular mechanisms underlying partial cellular reprogramming with Yamanaka factors are not yet fully understood^216,255, anti-aging effects can be observed through molecular biomarkers across multiple omics layers.

Epigenomics

DNAm plays a central role in the acquisition of pluripotency during OSKM-mediated cellular reprogramming^{228,238,256–260}. Studies using epigenetic clocks have shown that continuous expression of Yamanaka factors can reset the epigenetic age of adult somatic cells to levels comparable to those of embryonic stem cells^27,74. By analyzing the methylome after OSKM factors expression we can get a clearer look at the changes that happen after OSKM induction. For example, it is known that DNA demethylation takes place in the early stages of reprogramming and appears to be a crucial requirement for pluripotency²⁶¹. Olova et al. investigated aging dynamics during iPSC transformation and using four epigenetic clocks^{75,158,168,262}, observed that partial reprogramming reduces the epigenetic age of HDFs²²⁷. Moreover, the authors proposed the existence of a “safe window” during which rejuvenation effects can be achieved without loss of cellular identity or increased oncogenic risk. In this cellular model, the safe window extended from day 3 to day 15 of reprogramming. Within this period, epigenetic age was reduced by approximately 10 years by day 7 and 20 years by day 11 (median clock error = 3.6 years)²²⁷. These differences in reported “safe windows” across studies reflect the strong dependence of partial reprogramming dynamics on experimental context. As a result, values such as PNR or optimal rejuvenation time cannot be directly compared across studies, but rather represent context-specific thresholds within each experimental system.

Transcriptomics

Transcriptomic analyses provide additional ways of determining a cell’s position along the reprogramming trajectory by assessing the expression patterns of age-associated and pluripotency-related genes^263,264. By comparing the transcriptional profiles to those of iPS cells, the progression and extent of partial reprogramming can be monitored. For example, upregulation of pluripotency-associated genes, such as OCT4 (POU5F1), SOX2, or NANOG indicates the acquisition of a pluripotent state^265,266. On the contrary, if the expression of the somatic genes is upregulated and the expression of the pluripotency genes mentioned above is downregulated²⁶¹, then the cell is still in the “safe window” of partial reprogramming²²⁷.

Using qPCR, Sarkar et al. identified strong similarities between the gene expression profiles of partially reprogrammed cells and young cells, as well as marked differences compared to aged cells, indicating transcriptional rejuvenation. To determine the extent of the reprogramming, they analyzed the transcriptome using RNA sequencing (RNA-seq) and they did not notice pluripotency-associated mRNAs (endogenous OSKMLN mRNAs)²²³.

Subsequently, Gill et al. (2022) used two transcriptomic aging clocks²⁶⁷ and determined that 13 days is the optimal duration of OSKM treatment. During this day, the eAge dropped by 30 years and at earlier or later time point, the results were lower. It is important to note that the median error rate of these clocks is higher than that of previously used epigenetic clocks (5.5–12.5 years)²⁶⁸.

Proteomics

The proteome can be analyzed during partial reprogramming to assess if the anti-aging effects are also reflected at the functional level. For example, Gill et al. (2022) used immunofluorescence and 3D imaging to quantify the signal intensity of functional proteins known to reduce with age in human dermal fibroblasts. They observed that after 13 days of OSKM factors induction, the declining signal of structural proteins such as collagen I, IV, cytokeratin 8, cytokeratin 18 and heterochromatin marker H3K9me3 was reversed to levels characteristic to young fibroblasts^268–272.

Since partial reprogramming is also intended as a regeneration strategy of age-related diseases, proteomic analysis can quantify proteins involved in cell stability and protection against these disruptions. For example, amyloid beta precursor protein binding family A member 2 (APBA2) is downregulated with aging, and this modification is associated with an increased risk of Alzheimer’s disease. During partial reprogramming, scientists observed that this protein is expressed at normal, youthful levels, indicating not only a reversal of the aging clock, but also recovery of cellular function. Additionally, monitoring the proteome can be used to identify the safety window before jumping to full reprogramming. In this case, researchers used surface protein aminopeptidase N (CD13) as a marker for fibroblast identity and stage-specific embryonic antigen-4 (SSEA-4) as a pluripotency marker. In successful partial reprogramming experiments, rejuvenated cells should express CD13^268,273.

Metabolomics

Partial reprogramming can be assessed not only by changes in the epigenome, transcriptome, proteome, but also by modifications in serum metabolites. With the help of mass spectrometry, the rejuvenating effects of partial reprogramming were seen in a short-term continuous experiment. After 1 week of induction, four blood metabolites that were affected by aging were also brought back to youthful levels. The metabolites in question are 4-hydroxyproline, thymine, trimethyl-lysine and indole-3-propionic acid, and the first two are known to play different roles in aging²¹⁶. During aging, it was observed that a decrease in 4-hydroxyproline levels leads to the decrease of collagen levels. Also, thymine was shown to increase lifespan in C. elegans^{216,274–276}.

Multi-omics profiling reveals that the molecular changes induced by this treatment influence chromatin organization, gene expression, protein abundance and metabolic pathways. Therefore, demonstrating the ability of Yamanaka factors to shift the molecules impacted by aging back to a rejuvenated and regenerated self. However, as in physiological aging, the kinetics of reprogramming vary across tissues²⁷⁷ and cell types²²³, with differences in response timing and extent of rejuvenation depending on experimental context.

Challenges and solutions in partial reprogramming

To date, partial reprogramming experiments have proven successful both in vitro and in vivo and are currently advancing to human clinical trials. This progress was enabled by overcoming several limitations, including challenges related to delivery methods, the use of c-Myc, which is a known proto-oncogene, uncertainties regarding administration protocols (long-term or cyclic), low reprogramming efficiency, and the lack of clarity concerning treatment withdrawal and the potential need for re-administration.

Viral delivery methods and safety concerns

Since 2016, numerous in vivo partial reprogramming protocols have been developed^{14,15,186,214,216,223,228,248,249,253}. One of these approaches has been translated into clinical trials for age-related diseases (glaucoma) however, none of these approaches are currently used as clinical treatments for aging²⁶⁵. One of the primary barriers to clinical translation is the method of induction. Notably, this was also one of the main limitations even before the partial reprogramming protocols were invented.

During the early era of full reprogramming, Takahashi and Yamanaka used retroviral methods to generate iPSCs. These retroviral vectors also integrated into the host genome, leading to aberrant gene expression and tumor formation^{217,218,261,278}. Similar genomic disruptions were later observed with other viral systems, including lentiviral vectors²⁷⁹.

Non-integrative gene delivery strategies

Due to safety concerns, integrative methods of induction are not considered reliable for clinical use. Consequently, non-integrative approaches have been developed, including AAVs, synthetic mRNAs, plasmids, recombinant proteins, and episomal vectors. Their use in partial reprogramming led to a safe induction, carrying a lower risk for cancer or other health problems seen in the integrative methods^{261,265,280–284}. Although considered a non-integrative measure, AAV vectors have been associated with potential immunogenicity, genotoxicity, and the risk of partial or complete integration into the host genome^285,286. Recently, following their successful in vivo studies and the FDA approval of AAV vectors for clinical applications²⁸⁷, Lu et al. and Karg et al. advanced their AAV-OSK glaucoma treatment to a Phase I clinical trial, making it the first human application of partial reprogramming with OSK²⁸⁸.

Synthetic mRNA–LNPs as a clinically relevant delivery platform

A major limitation in early partial reprogramming was the use of c-Myc along with other transcriptional factors (OSK). Because c-Myc is an oncogene, its overexpression led to tumour formation in mice^186,261,278. This issue was later resolved when it was shown that c-Myc is not required for partial reprogramming^14,186. OSK represents a safer combination of Yamanaka factors, leaving the method of delivery as the primary barrier to clinical translation.

The most rational approach is to deliver these transcriptional factors using a platform that is already tested, proven safe, effective, and clinically approved. To date, the leading method is synthetic mRNA encapsulated in lipid nanoparticles (LNPs), a technology that has emerged after the approval of the Moderna and Pfizer-BioNTech mRNA vaccines during the COVID-19 pandemic²⁸⁹.

This approach has already been used in vivo. For example, Jo et al. (2025) delivered OSKM factors using mRNA in LNPs targeting the liver of mice with induced acute liver injury and observed regenerative effects comparable to those previously reported with OSKM partial reprogramming²⁹⁰. While an anti-aging mRNA-based partial reprogramming protocol was first developed by Sarkar et al. in 2020, the use of LNPs could enhance treatment stability and allow site-specific targeting if needed^223,291,292.

Treatment duration, safety windows, and persistence of rejuvenation

Another critical factor in partial reprogramming is the duration of treatment. Some protocols involve long-term induction^214,232,233, while others use short-term induction^216,248. In addition, transcription factors may be expressed continuously¹⁴ or transiently^223,228. Continuous expression of OSKM factors in vivo can lead to teratoma formation in multiple organs^232,293. However, continuous expression of OSK alone appear safe, even over long periods, and short-term continuous induction of OSKM has also been shown to safely achieve rejuvenating effects in mice^14,248. Both in vitro and in vivo studies indicate that rejuvenation is most effective when OSKM factors are expressed briefly, followed by a recovery period^216,223,228.

A further consideration for clinical translation is the persistence of the rejuvenation effects, as modifications induced by partial reprogramming are not permanent. Once treatment is stopped, benefits typically persist only for a limited period before returning to baseline¹⁵. For example, in an in vivo study by Chondronasiou et al., 1 week of OSKM induction rejuvenated the pancreas, but the histological improvements lasted only for 2 weeks after OSKM activation was stopped²¹⁶. On the other hand, old mice subjected to 1 month of OSKM induction during early life retained a rejuvenated musculoskeletal system²⁴⁸.

Strategies to enhance reprogramming efficiency and rejuvenation outcomes

It is known that only a small percentage of cells expressing the four Yamanaka factors undergo full reprogramming^294,295. By 2012, the mechanisms underlying cellular reprogramming towards pluripotency were poorly understood. To detect what facilitates or inhibits reprogramming in cells, researchers examined the DNA-binding patterns of reprogramming factors by inducing OSKM in human fibroblasts using a lentiviral transduction and analyzing subsequent molecular events^279,294.

The study revealed that a key barrier to pluripotency is the heterochromatin present at the transcription factor binding sites, referred to as Differentially Bound Regions (DBRs)^256,294,296. Heterochromatin is a tightly compacted region of chromatin, transcriptionally silent, and resistant to reprogramming^296–299. Histone H3 trimethylations, such as H3K9me3 and H3K27me3, are known as markers for heterochromatin²⁹⁶. Soufi et al. observed that reprogramming efficiency was initially very low (approximately 0.05%), a finding consistent with previous reports. They further demonstrated that knocking down histone H3 methyltransferases increased both the efficiency and speed of reprogramming^218,294,300.

Although targeting histone-modifying enzymes enhances reprogramming, its safety is uncertain, as epigenetic modifications such as histone methylation are essential for normal cellular function and survival^301–303. Recently, lysine-to-methionine (K-to-M) mutations in histone H3 where lysine (K) is replaced with methionine (M) was proposed as an improved alternative. The difference between the two approaches is that one completely stops the methylation of histones and the other, only reduces the methylation level by blocking the methylation from happening^{301,304–310}.

Hoetker et al. demonstrated that the depletion of dimethylation and trimethylation at H3K36 with the H3K36M mutation can lead to pluripotency in almost all human fibroblast cells after OSKM induction in vitro. Moreover, the speed of iPSC colony formation was significantly accelerated: colonies formed in 4 days in H3K36M cultures, compared to 6–8 days in normal cell cultures³⁰¹. Overall, combining OSKM induction with the K-to-M mutation at lysine residues such as K9M, K27M and K36M, can generate a major increase in the production of iPS cells in less time, compared to only OSKM induction. These findings indicate that the low efficiency of iPSC formation is due to the mechanisms that maintain differentiation, and that overcoming these barriers allows nearly all cells to be fully reprogrammed using Yamanaka factors^{261,295,301,311}.

More recently, combining OSK treatment with TERT gene therapy (OSKT), a strategy previously shown to increase lifespan in mice³¹², was found to enhanced youthful gene expression patterns in replicative senescent cells³¹³, a widely used in vitro model of cellular senescence and aging-related processes^314–316.

Chemical partial reprogramming as an alternative approach

Although, not induced by Yamanaka factors, partial reprogramming using chemically induced small molecules avoids the safety concerns associated with viral or integrative delivery methods. Compared to transcription factor-based approaches, this method can induce a more youthful state by directly inhibiting or activating components of cellular pathways known to change with aging, such as senescence-associated pathways^{14,186,317–321}.

In a recent in vitro study, Yang et al. developed a novel method called epigenetic programming of old cell health (EPOCH). This approach uses chemical cocktails containing different combinations of small molecules to target aging-associated pathways. Human fibroblasts from donors of various ages (22, 94, and 14 years old with HGPS) were treated for 4 days with six different chemical cocktails, while OSK-treated cells were used as a control. Transcriptomic analysis revealed that aging-associated genes, including SASP genes, were downregulated following chemical treatment. Biological age, assessed using transcriptomic clocks, was reversed by all six chemical cocktails, with the VC6TF cocktail (valproic acid, CHIR-99021, E-616452, tranylcypromine, forskolin) showing the greatest efficacy in inducing partial reprogramming and rejuvenation³²¹.

Overall, chemical partial reprogramming provides a cost effective and rapid method to reverse cellular aging, without the risks typical associated with full reprogramming. However, as noted by the authors, further in vivo, studies are required to determine the potential of this approach as a clinically viable anti-aging intervention.

Conclusions

It is becoming increasingly important to understand the mechanism of aging and create efficient interventions, as the world’s population lifespan increases and the health span decreases. For almost 150 years, scientists have emphasized the many faces of aging, by proposing numerous theories based on the so-called biomarkers of aging. These many measurable changes observed with advancing in age highlight how complex and multifactorial aging really is. A multitude of interconnected biological alterations, from one molecule to an entire pathway, ultimately leading to the visible functional deterioration and susceptibility to diseases.

The current behavioral and pharmacological anti-aging strategies may reduce the rate of biological aging, but the risk of diseases remains, and the potential side effects may compromise the outcome. So, over the course of the last decade, many teams of researchers perfected a novel technique that has the potential to be a true anti-aging therapy. It is called partial reprogramming, and it not only promotes rejuvenation, but it also induces regeneration. Moreover, this strategy is notable for its ability to tackle the complexity of aging. As scientists have proved, the affected cellular processes and molecules, extending from the epigenome to the proteome level, can be partially remodeled toward a more youthful state, particularly at the epigenetic and cellular function levels.

Numerous partial reprogramming protocols that target aging have been developed by researchers, each introducing modifications to the original 2016 protocol, either to enhance efficacy or to demonstrate the safety of this therapeutic approach. These methodologies have been tested on naturally aged or transgenic model organisms, as well as on different cell types at various chronological ages, using different combinations of transcriptional factors, from OSK to OSKMLN, delivered by viral vectors or mRNA molecules, with administration ranging from a few days to years, activated either in a continuous or cyclic manner, all with the aim to optimize this strategy. Partial reprogramming protocols using Yamanaka factors often produce convergent biological responses across different experimental conditions.

As shown above, this strategy has been proven effective in experimental models, and many challenges and limitations were addressed with each new protocol. With the current knowledge, scientists will continue to improve this approach in the upcoming clinical trials. However, partial reprogramming remains incompletely understood, and its effects across all hallmarks of aging, including genomic stability, mitochondrial integrity, extracellular matrix remodelling, and systemic regulation, are not yet fully established. Importantly, current evidence does not yet support a universal or complete reversal of aging, highlighting the need for cautious interpretation of existing findings. Moving forward, it is essential to map the path from the moment these factors are expressed to the anti-aging and regenerative effects. Understanding exactly how this works and what mechanisms are involved can lead to a successful clinical approach.

Author contributions

E.C.G., G.C.M. and R.G.P. contributed to the conceptualization, writing, figure design, and editing of the review. A.A.S. and S.E.G. contributed through substantive revision of the manuscript.

Data availability

No datasets were generated or analyzed during the current study.

Competing interests

Authors G.C.M. and R.G.P. were employed by the company Blue Screen SRL. The remaining authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest. E.C.G., A.A.S., S.E.G. declare no financial or non-financial competing interests.