The biological responses to low-dose radiation (LDR) are frequently distinct from those to high-dose radiation (HDR). For example, LDR is able to promote growth and development, enhance immune functions, prevent or delay carcinogenesis and cancer progression, and suppress the aging process^[1-5]. The phenomenon that LDR may be beneficial is often referred to as radiation hormesis or adaptive response^[6-7]. Although the mechanisms underlying LDR-triggered hormesis or adaptive response are not fully understood at the molecular, cellular, tissue, or organism levels, significant increases in immune function, antioxidants, DNA repair capacity, and cell proliferation have been extensively discussed^[8-9].

The responses of the central nervous system (CNS) to stressors and injuries such as ionizing radiation have been a subject of attention for several decades. It is known that exposure of brain tissue to HDR leads to the expression and release of biochemical mediators of neuro-inflammation such as pro-inflammatory cytokines and reactive oxygen species (ROS), thereby leading to tissue destruction. In contrast, LDR may reduce the vulnerability of exposed tissues (such as those of the CNS) to subsequent HDR-induced damage, which may be mediated largely through LDR-induced adaptive responses such as anti-inflammation and antioxidant defense mechanisms. These disparate responses may be reflective of non-linear differential microglial activation, which manifests as an anti-inflammatory or pro-inflammatory functional state. It should be noted that the LDR dose that induces hormesis and/or an adaptive response is mostly referred to as being less than 100 mGy (0.1 Gy)^[10-12].

One specific question that has been asked is whether LDR has any detrimental or preventive effect on the development of Alzheimer′s disease (AD). AD is the leading cause of dementia among the elderly, and the USA Centers for Disease Control and Prevention (CDC) stated that (https://www.cdc.gov/aging/aginginfo/alzheimers.htm) "In 2014, as many as 5 million Americans were living with AD". The symptoms of the disease usually first appear after the age of 60, with the risk increasing with age. Globally, the prevalence of AD is predicted to reach 80 million by 2040^[2]. Therefore, as part of an AD prevention strategy, it is very important to better understand the various factors contributing to it. Considering the increasing clinical use of medical instruments that emit radiation, the question of whether the development of AD is related to the effect of radiation, particularly at LDR levels, is an important topic. Therefore, this review summarizes the available literature to examine whether there is any effect of LDR on AD, either good or bad. On the basis of the updated information, it seems that exposure to LDR at a dose of less than 100 mGy does not increase the risk of AD, and may even provide some beneficial effects under certain conditions. The therapeutic effect of LDR on a patient with AD has been reported, and even though there is only one such case reported, this possible effect should be further explored and evaluated by further clinical evidence.

There is considerable uncertainty concerning the impact of LDR on the CNS. An examination of publication trends reveals relatively less literature regarding LDR and the brain than there is on the effects of higher radiation doses, which emphasizes the lack of uncertainty surrounding the neurobiological effects of low doses. In one of the earliest research articles into the acute effects of LDR on the adult brain, Yamaoka et al^[13] demonstrated evidence of the activation of protective mechanisms in the brain via the induction of the antioxidant activity that mitigates lipid peroxides. It is possible that LDR may not induce deleterious alterations in cognition, cell functioning, DNA, and gene expression, nor induce apoptosis or pathological signs in vivo^[14], and that it may in fact stimulate molecular and cellular protective mechanisms^{[7, 15]}. Mitochondrial redox balance and microglial responses are also critical for modulating responses to LDR, largely through the stimulation of antioxidant defense mechanisms. Lower doses can confer protection to cell functioning, molecular structures, synapses, and key brain mechanisms such as neurogenesis, and induce reparative mechanisms in the face of CNS pathology^{[7, 15]}.

Nowadays, LDR from medical diagnostic techniques constitutes the largest proportion of the average yearly radiation exposure worldwide. According to the literature, one of the more prominent manifestations of radiation-induced injury is seen in the hippocampus, a radiosensitive region housing populations of proliferating progenitor cells^[16-17].

In terms of inflammation, decreases in pro-inflammatory markers in the mouse hippocampus and cortex were present in animals primed with a 100 mGy dose prior to subsequent exposure to 2 Gy, suggesting that early exposure to LDR can prevent the upregulation of inflammation by HDR^[18]. A study by Yin et al^[19] demonstrated that 0.1 Gy gamma radiation induced alterations in the expression of genes involved in neuroprotective functions, notably DNA repair, cell-cycle control, lipid metabolism, and stress responses. Interestingly, later-stage changes concerned genes involved in metabolic functions, myelin and protein synthesis, and increases in transcripts for antioxidative enzymes.

Analysis of transcriptome profiles after LDR of 100 mGy also demonstrated acute alterations in the expression levels of genes involved in damage responses, signaling pathways associated with cognition, and downregulation of ERK/MAPK, with these alterations being different to the pathways induced by HDR^[20].

Antioxidative mechanisms are implicated in the neuroprotective and reparative responses that LDR can confer^[21]. Corroborating evidence from Veeraraghavan et al^[22] demonstrated that mice subjected to single dose exposures of 100 mGy gamma radiation showed increased expression of the SOD2 gene and SOD activity facilitated through the NF-κB and SOD signaling network, which persisted for 8 days after exposure. Notably, this antioxidant upregulation was not present after exposure to 20 mGy, suggesting that the mitochondrial machinery behind this response requires sufficient exogenous stimulation from LDR.

AD is the leading cause of dementia among the elderly and the fourth leading cause of death, and affects about 24 million people worldwide^[23-24]. Researchers are working to uncover as many aspects of AD and other dementias as possible, and some remarkable progress has shed light on how AD affects the brain. The hope is that this better understanding will lead to new treatments, and many potential approaches are currently under investigation worldwide^[25-27].

Although radiation therapy is an important tool in the treatment of primary and metastatic brain tumors^[28], it is also responsible for various adverse neurological effects such as cognitive dysfunction and dementia, which may occur in brain tumor patients aged 50 years or over who are treated with radiotherapy^[29]. Evidence suggests that exposure to HDR can lead to the development of AD^[28-29]; however, whether LDR has such an effect will be the focus of discussion in this section.

Early brain responses to LDR exposure involve molecular networks and pathways associated with cognitive functions, advanced aging, and AD^[20]. For example, LDR of 100 mGy induced genes that were not affected by HDR, and these LDR-induced genes were associated with unique pathways and functions similar to those seen in the aging brain and brain tissue from patients with AD^[20]. The molecular responses of the mouse brain a few hours after LDR involve the downregulation of neural pathways associated with the cognitive dysfunctions that are also downregulated in normal human aging and AD^[20]. A study examining the level of total background ionizing radiation from natural radon in U.S. States and its association with AD^[30] demonstrated a positive correlation between radon background ionizing radiation and AD death rate in 50 states and the District of Columbia. Further detailed analysis revealed significant correlations of AD death with radon background, age, hypertension, and diabetes. In summary, these studies show that several molecular processes induced by chronic LDR overlap with those of AD, suggesting their potential contribution to the pathogenesis of AD.

However, the above studies focused on early transcriptional responses of the murine brain to LDR at 100 mGy, but whether these early responses really influenced cognitive functions, advanced aging, and AD is unclear. To this end, Wang et al^[31] investigated the acute and late transcriptional, pathological, and cognitive consequences of LDR. Using X-rays, they applied an acute dose of 100 mGy total body irradiation to C57BL/6 J mice, and then collected hippocampi and analyzed the expression of 84 AD-related genes. The learning ability and memory of the mice were assessed with the Morris water maze test, and the fibrillary amyloid beta peptide (Aβ) accumulation was analyzed by in-vivo PET scan. Immunohistochemical staining for amyloid precursor protein (APP), Aβ, tau, and phosphorylated tau was also performed. The authors found that the LDR only significantly downregulated the mRNA expression of two of the 84 AD-related genes, Apbb1 and Lrp1, at 4 h after irradiation, and of only one gene, Il1α, at 1 year after irradiation. Spatial learning ability and memory were not significantly affected at 1 or 2 years after irradiation. No induction of amyloid fibrillogenesis or changes in APP, Aβ, tau, or phosphorylated-tau expression were detected at 4 months or 2 years after irradiation. LDR induced early or late transcriptional alterations in only a few AD-related genes, but did not significantly affect spatial learning, memory, or AD-like pathological change in the mice. Similar outcomes were obtained when they repeated the study used high linear energy transfer (LET) carbon ion (50 or 100 mGy) radiation. On the basis of the results from pathological and behavioral studies, it is reasonable to speculate that although these genes have a close association with AD, the acute early transcriptional alterations were transient, and were probably mainly due to factors such as apoptosis-related events in some cells, rather than induction of AD-related effects. These acute transcriptional alterations are probably insufficient to induce any late biological consequences such as significant AD-like pathogenesis. These findings indicate that low-dose carbon ion irradiation did not cause behavioral impairment or AD-like pathological change in mice.

These findings indicate that radiation-induced changes in the expression of genes associated with AD are not necessarily predictors of the emergence of AD, as measured by either amyloid plaque formation or cognitive impairment. The lack of AD-like pathogenesis and symptoms may be due to different signal transduction cascades being activated at the same time in the hippocampi of mice in response to low and high LET irradiations^[14]. As illustrated in Figure 1, LDR may induce acute alterations in the expression of genes as in the A-pathway; however, this A-pathway alone would not be able to cause an AD-like syndrome at later age-points because LDR also activates other anti-inflammatory and antioxidant functions as in the B-pathway, and increases DNA repair mechanisms and stimulates cell proliferation in a tissue repair capacity as in the C-pathway. In both the B-and C-pathways LDR can both prevent the A-pathway from potentially-inducing AD-like damage, and may even prevent other AD pathogenic damage. Therefore, although LDR generates a small amount of ROS that in turn generate a small amount of DNA damage or gene expression with apparent detrimental effects, at the same time the LDR also activates body tissue or cells in some comprehensive defense mechanisms (Figure 1).

Figure 1 Possible effects of LDR on the development of AD. Under normal conditions (top panel), LDR can generate a small amount of oxidative stress and inflammation, which may cause acute brain alterations in the expression of several genes, as in the A-pathway (blue arrow), and as often seen in the brains of aging and AD patients. However, whether these acute responses initiate the development of AD at the late-age stage is uncertain, as LDR also activates other pathways such as the B-(dark gray line) and C-pathways (green line), which help to prevent AD development. In patients with AD (lower panel), LDR may still stimulate the A-pathway, but its activation of the B-and C-pathways would play a more important role in preventing and improving AD symptoms. Therefore, for patients with AD, the late potency of LDR-induced AD, or even cancers, would be less of a concern, as either of these would take many years to develop, while there is an urgent need to improve the AD symptoms over the potentially shorter time period of the patient′s remaining lifespan

Recent work suggests that even relatively low-dose radiation (such as that resulting from computed tomography [CT] scans) could trigger mechanisms associated with the cognitive dysfunctions seen in normal aging and AD. Cuttler et al^[32] reported on the first case of a patient with AD who was given CT exposure for therapeutic purposes. In the January-March issue of Dose-Response 2017 and the January-March issue of Dose-Response 2018, the authors gave updates on the status of this patient^[33-34]. On the basis of the three letters, the key features and progression of the AD patient treated with CT are summarized in Table 1.

表 1(Table 1)

Table 1 Summary of the CT times and doses, and the key symptoms of the AD patient Date CT(mGy) Symptoms 4/1/2015 Diagnosed with AD with symptoms of dementia for 10 years. AD had progressed gradually to the final stages of advanced AD. Commenced hospice care, which is allowed only if life expectancy is less than 6 months. Patient was completely nonresponsive and frequently refused her medications. Patient was almost totally non-communicative. Patient would only rarely utter a single word and that word would be inappropriate. Patient was almost immobile; had not attempted to rise from her wheelchair in months. 7/23/2015 80 (2 CTs) 7/25/2015 Two days later, her caregiver, who was unaware of the CT acquisitions, reported a noticeable improvement. The caregiver was quoted as saying: "She is doing so well that it is amazing. I have never seen someone improve this much". She wanted to get up and walk, was talking a little with more sense, and was feeding herself again. 8/6/2015 40 8/12/2015 The patient′s spouse received an e-mail from the patient′s caregiver who was highly excited with the patient′s significant improvement. 8/20/2015 40 9/14/2015 During the week of September 14, the following behaviors were observed: A sign of an old memory was seen when she called her daughter′s old roommate byname and then said "roommate." The patient often talked in three-to five-word sentences, as well as shorter responses, such as "yes, no, and maybe". Whatever her verbal response, it almost always seemed appropriate. 10/1/2015 40 Almost immediately, a significant setback was observed with an estimated loss of about 80% of the gain. This was initially very discouraging, but a slow recovery of cognitive ability began again. 10/28/2015 A neuropsychological examination indicated that she was able to give some simple verbal responses. 11/20/2015 The patient′s slow progress continued until she was judged to be no longer eligible for hospital care after 6 months. 12/4/2015 The patient demonstrated restoration of appetite and responsiveness. Recognizing that any treatment efficacy of the CT scans would likely be transitory, further scans were planned. 2/24/2016 40 6/22/2016 40 10/27/2016 40 12/13/2016 40 12/29/2016 The patient fed herself with a spoon and opened hermouth to accept food. She drank glasses of juice by herself. She gave smiles and one-word answers to questions. 1/24/2017 80 (2 CTs) 2/4/2017 Patient was able to attempt to put words together, and gave many appropriate one-word responses, some two-word responses, and some three-word statements. However, this speaking ability gradually declined. She was still wheelchair bound, but moved her feet to help propel it. 3/6/2017 The patient was returned to hospice care. Over several months, the patient gradually lost her ability to swallow solid food and later could not swallow liquids such as nutritional drinks. The patient would hold the liquid in her mouth, wanting to swallow, but unable to because her brain could not control the appropriate muscles. During this period of reduced food intake, her weight declined from 185 to 160 lbs. The caregiver began to feed the patient ice chips and found that it would trigger the swallowing reflex. She was then given frozen drinks, such as cola, and blended mixtures of ice, fruit, and yogurt. 7/25/2017 40 8/3/2017 No immediate physical improvement was noted, but the patient began to repeatedly flash smiles. 10/28/2017 The patient had her 83rd birthday. The patient was able to chew and swallow chopped peaches at noon time. 11/-/2017 In the evening, she swallowed potato salad, watermelon, and pieces of cupcake. Through to November 2017, she continued to swallow solid food and appeared relatively happy, giving many smiles and much laughter. It was shown by mGy and CTDIvol across three reports (Cutter et al. 2016, 2017, 2018). CTDIvol was measured using a plastic patient phantom and used as a reference for the patient dose. The doses are the scanner output in mGy.

Table 1 Summary of the CT times and doses, and the key symptoms of the AD patient

The patient was 81 years of age in 2015, at which time she had suffered from AD with the symptoms of dementia for about 10 years, with a gradual progression to the final stage of advanced AD (Table 1 lists the detailed symptoms). The patient was placed in hospice care on April 8, 2015, with a life expectancy of less than 6 months. A neuropsychological examination on May 21 revealed her to be completely nonresponsive. The patient would frequently refuse her medications and was almost totally non-communicative, rarely uttering a single or appropriate word, and had been almost immobile in her wheelchair for several months. Considering that there was no efficacious medication available to improve her symptoms, and that the patient′s spouse was aware of potential protective effects of LDR on age-related deterioration, the patient′s husband requested a standard CT scan to determine any anatomical changes that had occurred, and also to stimulate neuroprotective systems. On July 23, 2015, two standard CT scans were performed on the patient, and the patient′s symptoms then showed noticeable improvement, as outlined in Table 1. During the following 6 months of hospice care, the patient received about five CT scans, each with a radiation dose of about 40 mGy, and her symptoms greatly improved (Table 1). On the basis of the authors′ description of the observed positive responses of the patient, we suggest that the cumulative dose of the first four radiation exposures, a total of 168 mGy, was within the range suitable for causing radiation-induced beneficial health effects.

On November 20, 2015, the patient was judged to be no longer eligible for hospice care because her condition had sufficiently improved. After this date, she participated a stimulating dementia day-care program and demonstrated restoration of appetite and responsiveness. On February 24, 2016, as it was recognized that the efficacy of the CT scan treatments would likely be transitory, the patient′s spouse requested ongoing booster scans every 4 to 5 months, which were performed up until December 29, 2016 (Table 1). Over this period, the patient fed herself with a spoon and opened her mouth to accept food, she smiled, drank glasses of juice by herself, and gave one-word answers to questions. A physical therapist reported some improvement in her functioning shortly after this scanning. On February 4, 2017, a major improvement was noted in her attempts to put words together. She gave many appropriate one-word responses, some two-word responses, and some three-word statements. However, this speaking ability gradually declined.

On March 6, 2017, almost 2 years after her entry into hospice care, the physician and the patient′s spouse decided to return the patient to hospice care. Over several months, she gradually lost her ability to swallow solid food, and later could not swallow liquids such as nutritional drinks. The patient would hold the liquid in her mouth wanting to swallow, but unable to because her brain could not control the relevant muscles. During this period of reduced food intake, her weight declined from 185 to 160 lbs. The caregiver began to feed the patient ice chips and found that it would trigger the swallowing reflex. She was then given frozen drinks such as cola, and blended mixtures of ice, fruit, and yogurt.

On August 3, 2017, after a CT scan on July 25, no immediate physical improvement was noted, but the patient began to repeatedly flash many smiles. On October 28, 2017, on her 83rd birthday, the patient was able to chew and swallow chopped peaches at noon. In the evening, she swallowed potato salad, watermelon, and pieces of cupcake. The patient continued to swallow solid food and appeared relatively happy, giving many smiles and much laughter, through to November 2017, after which no further updates on the patient status are so far available.

Although there are not many similar reports with such important information, a couple of pilot clinical trials have been approved and registered in the NIH online system (ClinicalTrials.gov Identifier: NCT02359864; NCT03352258; NCT03597360), allowing small numbers of patients to be exposed to CT or radiotherapy to evaluate the safety and efficiency of LDR.

While the health effects of LDR (< 100 mSv) exposure are still a subject of controversy, human epidemiological and clinical studies indicate that LDR exposure may induce or prevent carcinogenesis, depending on age, sex, race, radiation components and sources, genetics, lifestyle, and other environmental exposure factors and diagnostic accuracy^[35]. Radiation exposure may increase CNS diseases such as AD, particularly when young individuals are exposed to radiation, especially HDR. Whether exposure to LDR induces the development of AD remains unclear, with there being no direct evidence to support a link between LDR of < 100 mGy and the development of AD, as illustrated in Figure 1. The lack of development of AD-like diseases in the late ages of mice exposed to LDR at an early age and showing alterations in the expression of several AD-related genes may be related to the fact that LDR also simultaneously upregulates several protective mechanisms to prevent the potential development of AD due to AD-like transient changes in gene expression (Figure 1).

In contrast to the above chronic effects, whether exposure to LDR of 100 mGy can provide a preventive or therapeutic effect on AD is also an important topic, as there is no efficacious medication available. In addition, any drug that could efficiently improve AD may have side effects on other organs, as these drugs would be metabolized in the liver and eliminated via the kidneys; it may be the case that elderly patients with AD have generally weak organs unable to tolerate the side effects of drugs suitable for other patients. However, if exposing AD patients to LDR could provide a therapeutic effect as shown by the reported case (Table 1), it would be a non-invasive approach, and the LDR could be delivered locally to the head without any impact on other organs in the body. Although we cannot eliminate the AD or cancer risk from exposure to LDR of < 100 mGy, even though it remains uncertain, it would not be a concern for most AD patients since they are mostly older than 65 years, and the latency for radiation-induced solid tumors is considered to be 15-30 years.

Conflict of Interest statement  All authors declare they have no conflicts associated with this study

Acknowledgment  We thank Bioedit for scientific editing of the manuscript

Contribution statement of authors  Feng Li, Liu Qiang, Wang Bing, and Cai Lu conceived and designed the paper. Wang Bing, Cai Lu coordinated the data acquisition for long-term follow-up. Feng Li, Liu Qiang, Wang Bing, and Cai Lu contributed to critical revision of the manuscript for important intellectual content and approved the final version of the manuscript. Feng Li, Wang Bing, and Cai Lu were responsible for the overall content of article and data analysis