The role of diet in vulnerability to noise-induced cochlear injury and hearing loss

There are roughly 60 physiologically essential nutrients (Baker, 2008). Work with experimental animal models has been extensively used to characterize the bioavailability of nutrients (i.e., the proportion of the substance entering the circulation and ability to have a biological effect). Interactions between nutrients that influence bioavailability of other nutrients are also of significant interest. Nutrients often contribute to a cascade of mechanisms that influence cell, organ, and bodily function, and it is likely that much of our current understanding of nutrient physiology is incomplete, with some effects of nutrients, and nutrient interactions, still undiscovered. The complex physiological interactions, which result in biostatistical collinearities and interactions within physiological and functional datasets, result in a need for complex statistical modeling within nutrition research. To complicate matters further, there is intrinsic variation across animal species in regards to how they metabolize, utilize, and excrete nutrients. In other words, the effects of nutrients may be different from one animal model to another, based on differences in the bioavailability of the nutrient in each model. Diet also has profound effects on gut microbiota structure, function, and secretion that modulate multiple inflammatory and metabolic pathways. Taken together, our understanding of the influence of diet on development, disease, genetics, and inter-generational implications is still evolving.

Despite the complexities associated with the identification of relationships between nutrients, overall diet, and auditory function (including the development and progression of hearing loss), the literature does contain a number of studies that provide evidence suggestive of a protective effect of nutrients with antioxidant (Ohinata , 2000), anti-inflammatory (Seidman , 2003), and anti-ischemic properties (Le Prell , 2007) in reducing or preventing acquired hearing loss. This article therefore discusses (1) animal models in nutritional research, (2) evidence of dietary nutrient-based otoprotection for age-related hearing loss (ARHL) and noise-induced hearing loss (NIHL), and (3) confounds and limitations to the study of nutrients and overall diet as otoprotective strategies.

Thirty years ago, the molecular and cellular biology of the inner ear was considered the “next frontier” of hearing science, with significant discoveries regarding membrane permeability, ionic gradients, neurotransmitters, and pathological implications emerging (Hughes, 1989). Basic knowledge of auditory molecular physiology, biochemistry, genetics, and immune regulation continue to accelerate and drive advances in the understanding of development and the potential for regeneration in the inner ear. Despite these gains, the specific influence of dietary nutrients on auditory physiology and function is not well understood. The vast majority of research on nutrients and hearing has focused on either the effect of deficiency or supplementation on auditory function and changes related to age, noise trauma, and/or ototoxic drug exposure with less emphasis on normal physiology and function. Nutrients and their metabolites not only serve as the building blocks of cellular structures and as fuel sources, but also serve as modifiers of protein function, signaling molecules, and gene expression, thus potentially influencing normal physiology and function as well as risk of damage related to aging and noise exposure.

Nutritional biochemistry is complicated, with nutrients serving numerous physiological roles, which can be dependent on a specific cell type and activity. Moreover, normal human physiology is a function of intrinsic and extrinsic factors that ultimately determine susceptibility to pathology and longevity. The ability to utilize nutrients delivered via dietary sources varies with species; Baker (2008) provides a summary of numerous species-specific nutritional characteristics worth reviewing. For example, most mammals can synthesize vitamin C endogenously; however, primates cannot (Drouin , 2011). Guinea pigs, one of the most common laboratory animal models for studies on prevention of acquired hearing loss, similarly do not synthesize vitamin C endogenously (Nandi , 1973; Chatterjee , 1975). β-carotene, a carotenoid with high vitamin A activity, has significant differences in bioavailability across species; gerbils and preruminant calves have been identified as the most suitable model for human vitamin A translation (Lee , 1999). Also, while beta-carotene and other carotenes such as beta-cryptoxanthin may be converted to vitamin A, these nutrients also have independent actions, directly effecting antioxidant and anti-inflammatory activity, playing a role in the modification of LDL cholesterol, inhibiting lipid peroxidation in arterial walls, and influencing plaque stability, vasomotor function, platelet aggregation, and thrombosis (for review see Zhang , 2014). As a third example, virtually all animal species use D-methionine almost effectively as L-methionine, but apes and humans cannot invert D-methionine to L-methionine efficiently (Baker, 2006; McIsaac , 2016). It is beyond the scope of this article to discuss the metabolism of each nutrient across species, or provide comprehensive physiological effects of each nutrient across species; however, these initial examples demonstrate the importance of understanding the effects of specific nutrients of interest within the animal model to be used in a given investigation.

The vast majority of non-human auditory research performed in mammals is done so in rodent models. This is consistent with medical research in general. Rodents, from Latin rodere, “to gnaw” are mammals of the order Rodentia. Mice (Mus musculus) and rats (Rattus norvegicus), in particular, have had a tremendous role in the broad areas of nutritional research and hearing research (see review by Ohlemiller, 2019). Mice have become prominent models due to their genetic manipulability and similarity to the human genome (85% of human genes have orthologs in mice; Makalowski , 1996). Rats, similar to mice, have several inbred strains (see review by Trevino , 2019). Rats are also very well suited for behavioral experiments. Further, rats show more complex feeding behavior and eat a larger variety of foods compared to mice. Another rodent, the guinea pig (Cavia porcellus) was the first to have inbred strains developed, they have been critical in understanding dietary cholesterol and lipoprotein metabolism (Rubio-Aliaga, 2012) and are popular in auditory research (see review by Naert , 2019). A final rodent model is that of the chinchilla (Chinchilla lanigera), though a highly utilized animal model in hearing research (see review by Trevino , 2019), very little nutritional research has been performed in chinchilla models. Guinea pigs and chinchillas, like the rat, are well suited to use in behavioral experiments. However, unlike the rat, these species are “covered” by the United States Department of Agriculture (USDA) which increases the cost to acquire and house these animals as laboratory subjects. Each of these animal models has contributed significantly to our understanding of auditory physiology and pathology, yet there is considerable variance in regards to nutritional factors such as nutrient metabolism across these (and other) species (Rubio-Aliaga, 2012).

The translation of animal model-based findings to humans is an overarching goal for many nutritional research studies; i.e., better understanding of factors that are associated with, and may perhaps directly influence, longevity and fitness of mice (or other rodent models) will hopefully lead to identification of interventions that improve longevity and fitness of humans. Of course, humans are much more complex in regards to both intrinsic (gender, age, weight, genetics, etc.) and extrinsic (diet, environment, etc.) factors, which creates challenges identifying and appropriately adjusting for confounding variables. Most research using animal models does not adjust for intrinsic differences beyond sex and perhaps age, and extrinsic factors such as diet are rarely considered in statistical modeling of laboratory animal data. Specifically, rarely do the methods sections in research papers include complete descriptions of the dietary intake of the animals, such as the nutrient content of their chow, logging of additional food items, or a record of daily caloric intake.

Understanding species differences in nutrient metabolism is an important consideration, in addition to understanding species differences in auditory anatomy and physiology. Studies examining otoprotective agents derived from dietary nutrients must therefore consider variance in animal models as part of the study design. As introduced previously, vitamin C (ascorbic acid) physiologically serves as a cofactor in the hydroxylation and consequent maturation of collagen, is a critical water-soluble antioxidant, and helps keep vitamin E in a reduced state (Traber and Stevens, 2011). L-Gluconolactone oxidase (GULO) catalyzes the rate-limiting step of vitamin C biosynthesis. Most rodents, including mice and rats are able to endogenously synthesize vitamin C. Primates, however, are a rare exception due to a mutation in the gene encoding GULO approximately 30 × 10⁶ years ago. Other animals lacking endogenous production of vitamin C include guinea pigs, capybara, and some birds and fish (Padayatty and Levine, 2016). All these species lacking the ability to synthesize vitamin C have a vitamin-C rich diet, even in the wild, reducing their reliance on endogenously generated vitamin C. From an evolutionary perspective, loss of the capacity to synthesize vitamin C may be advantageous, as vitamin C synthesis leads to formation of hydrogen peroxide (H₂O₂) and depletion of glutathione (Drouin , 2011).

When examining vitamin C as a potential otoprotective agent with the ultimate goal of translation to human populations, rats and mice presumably would not be the preferred model due to the potential that their endogenous capacity to produce vitamin C will confound the influence of dietary supplementation or deficiency. Guinea pigs, which do not produce vitamin C endogenously, or, as an alternative, mouse models with GULO mutations, would likely be preferable. Although some investigations assessing vitamin C as an otoprotective agent have used the guinea pig as a model (i.e., McFadden , 2005; Le Prell , 2007), other studies have used animals that produce endogenous vitamin C in studies of otoprotection (Seidman, 2000; Derekoy , 2004; Tamir , 2010; Le Prell , 2011b; Loukzadeh , 2015). Selection of animal model could underlie differences in the effectiveness of vitamin C supplementation across studies.

One early experimental evaluation of the effects of dietary fat manipulation on vulnerability to NIHL was provided in a study by Harold Pillsbury (Pillsbury, 1986). Pillsbury was influenced by the work of Samuel Rosen in the 1960s. Rosen studied hearing in the Mabaan tribe, a tribe located in a remote region of the Sudan with healthy diets, active lifestyles, minimal noise exposure, and essentially normal hearing sensitivity through the eighth decade of life (Rosen , 1962). As described in Pillsbury (1986), rats were assigned to a normal diet or atherogenic diet (in this case a high fat diet) to induce hypertension. The two groups of rats were then exposed to noise, with auditory thresholds measured via auditory brainstem responses (ABR) before and after noise exposure. The rats maintained on the atherogenic diet sustained greater noise-induced permanent threshold shift (PTS) and showed greater loss of outer hair cells. Chinchillas fed a high cholesterol diet similarly showed increased vulnerability to noise-induced damage (Sikora , 1986). In parallel to these studies in animal models, Goycoolea (1986) replicated the results from the Mabaan tribe, showing an association between environment and hearing, specifically including better hearing in natives to Easter Island who had only lived on the island and poorer hearing in natives who had lived in modern civilization, with both different diets and more exposure to noise associated with modern civilization experiences.

Even prior to the Pillsbury (1986) study, mineral deficiencies were being identified as factors associated with vulnerability to NIHL. Ising (1982) demonstrated that magnesium (Mg) deficient guinea pigs had increased susceptibility to noise exposure. Rats fed Mg-deficient diets similarly showed an increased susceptibility to noise injury, and NIHL was negatively correlated with concertation of Mg in the perilymph (Joachims , 1983). Lower Mg concentrations were associated with greater NIHL. Other dietary nutrient deficiencies have also been reliably associated with increased noise-induced pathology. For example, Biesalski (1990) demonstrated that vitamin A deficiency increased sensitivity to temporary threshold shift (TTS) induced by a 90 dB sound pressure level (SPL) broad band noise exposure for 15 min using guinea pigs as a model.

Given evidence of relationships between dietary deficiencies and NIHL, several more recent studies have directly assessed the potential role of dietary supplements in perhaps conferring protection from NIHL by manipulating the levels of individual nutrients, or combinations of nutrients. This work was stimulated by (1) the demonstration that noise induces reactive oxygen species (ROS) formation in the cochlea and dietary glutathione attenuates both ROS formation and subsequent NIHL (Yamane , 1995; Ohlemiller , 1999; Ohinata , 2000); and (2) both dietary antioxidants and caloric restriction mediate age-related hearing loss (ARHL) (Seidman, 2000). These key findings in both NIHL and aging rodent models influenced a cascade of otoprotection studies examining single nutrients and nutrient compounds including: vitamin E (Scholik , 2004), vitamin C (Derekoy , 2004; McFadden , 2005; Loukzadeh , 2015), creatine and tempol (Minami , 2007), L-N-acetyl-cysteine (Kopke , 2000), D-methionine (Kopke , 2002), the synthetic selenium-like glutathione peroxidase-mimic ebselen (Pourbakht and Yamasoba, 2003; Lynch , 2004), acetyl-L-carnitine (ALCAR) (Kopke , 2002), alpha-lipoic acid (Quaranta , 2012), vitamin B-12 (Quaranta , 2004), polyphenols like resveratrol (Seidman , 2003), coenzyme Q10 (Cascella , 2012), and cocktails of nutrients (e.g., vitamin A, C, E, and Mg) (Le Prell , 2007). Reviews on these agents for noise protection and proposed mechanisms are available to the interested reader; see for example Le Prell and Spankovich (2013) and Hammill and Campbell (2018).

Although there are many successes in animal models, where a single nutrient or combination of specific nutrients can be selectively manipulated, translation to human benefit has not been simple or broadly successful. N-acetyl-cysteine (NAC), one of the first nutrient-based antioxidants brought to clinical trials, has demonstrated efficacy in animal models of noise (Kopke , 2000; Duan , 2004; Bielefeld , 2007; Wu , 2010; Motalebi Kashani , 2013; Ada , 2017) despite some notable negative findings (Hamernik , 2008; Davis , 2010). Kramer (2006) performed a randomized, double blind, place controlled study examining temporary threshold shift (TTS) in young adults attending discotheques. No significant difference was observed between experimental and placebo groups. Though some positive findings were suggested in two other human studies with noise exposed workers for TTS (Lin , 2010; Doosti , 2014), variation in noise exposure across participants and small TTS changes were cited as study limitations. In a single center, prospective, randomized, double-blind placebo-controlled study performed with the United States Marine Corps, NAC was examined for efficacy in reducing evidence of NIHL after weapons training. Although some of the findings were positive, statistically significant differences were restricted to subgroups compared in secondary analyses (Kopke , 2015).

A second nutrient-based otoprotection strategy that moved to human clinical testing is a combination of beta-carotene, vitamins C and E, and Mg (ACEMg). Comparable to NAC, ACEMg has shown benefit in several animal studies in prevention of NIHL, both temporary and permanent (Le Prell , 2007; Le Prell , 2011a; Le Prell , 2011b). A human trial using a double-blind placebo-controlled design was therefore conducted with a controlled music player exposure to elicit a small but reliable TTS, addressing a limitation reported in the Kramer (2006) study. However, consistent with Kramer (2006), no reliable effect of ACEMg was observed for prevention of TTS (Le Prell , 2016).

With dose optimization or changes in combination of agents delivered, it remains possible that either NAC or ACEMg might prevent either PTS or TTS. Increased doses of the active agents may be feasible, or longer pre-noise delivery of the active agents may improve outcomes, as many nutrients are slow to accumulate in tissue, reaching stable plateau levels only after multiple days or weeks of dosing (Levine , 1996). However, the null hypothesis must be considered as well (for discussion see Le Prell , 2016); it is possible that dietary supplements will not confer protection under any dosing scheme when the control group is composed of healthy adult volunteers with no dietary deficiencies.

Not all clinical trial findings using TTS models have been negative. The synthetic organselenium drug ebselen has also advanced to human clinical trials. Comparable to other antioxidant molecules, ebselen has been shown to be effective in prevention of TTS and PTS in animal models [for recent review see Lynch (2016)]. Using the same study design as Le Prell (2016), Kil (2017) demonstrated that ebselen significantly reduced TTS in humans.

A fourth molecule to mention is D-methionine (D-met). D-met has been effective in prevention of TTS and PTS, as well as rescue of hearing when provided after noise exposure, in rodent models (Campbell , 2011). D-met is concluding a phase III clinical trial to determine efficacy for prevention of NIHL in humans (NCT02903355). A potential limitation in translation of D-met to humans, is unlike other animals, humans do not convert D-met to L-met efficiently (Baker, 2006; McIsaac , 2016); suggesting protection may not translate well from rodents to humans based on differences in the metabolism of the molecule. However, D-met may nonetheless be biologically active in humans, based on biological effects achieved through other mechanisms of action. For example, given the potential negative implication of elevated methionine on longevity, it is plausible that D-met in humans decreases the utilization of L-met and in that way is protective or has some role in downstream mediation of the methionine/transsulfuration cycle (McIsaac , 2016).

Other nutrient-based protection strategies that have been examined in small human TTS studies include alpha-lipoic acid (Quaranta , 2012), Mg (Attias , 2004) and vitamin B-12 (Quaranta , 2004), with each shown to have some level of protection against TTS. Although Attias (1994) showed Mg supplements protected against PTS in a military population, only weak correlations were observed between serum Mg levels and PTS (Walden , 2000) and TTS (Attias , 2004).

All of the reports listed in this section thus far discuss the controlled manipulation of diet to either increase or decrease the bioavailability of specific macro or micro nutrients. In addition to these experiments, that directly assess the role of specific nutrients or nutrient compounds for otoprotection, there is a small but growing literature on associations between dietary quality and hearing loss in humans (Spankovich and Le Prell, 2013, 2014; Spankovich , 2017; Curhan , 2018). Particularly relevant to the issue of NIHL, Spankovich and Le Prell (2014) provided a seminal analysis of the relationship between dietary quality, noise history, and hearing status. Using National Health and Nutrition Examination Survey (NHANES) data, the relationships between the Healthy Eating Index (HEI), a 100-point scale of dietary quality per USDA recommendations (higher score infers high dietary quality), low and high frequency pure tone average (LFPTA and HFPTA), and noise exposure as indicated by self-reported exposure to four noise sources (occupational noise, military service, recreational firearm use, other non-occupational noise), were examined. The results demonstrated a significant negative association between HEI and HFPTA, but not LFPTA, consistent with Spankovich and Le Prell (2013). In addition, the number of reported noise sources was positively related to HFPTA, i.e., increased noise exposure was associated with higher (poorer) HFPTA, with no relationships observed for noise exposure and LFPTA. During further analysis, the potential for interactions in the relationships between noise, HEI, and HFPTA were examined. In brief, there was a statistically significant interaction effect in which poorer diet was related to poorer hearing and poorer hearing was further exacerbated by noise history in the poorer diet grouping; in contrast, better diet was associated with better hearing and noise exposure did not exacerbate hearing deficits suggesting better dietary quality may have attenuated the effects of noise that were observed in the poorer dietary quality cohort.

Taken together, there are multiple nutrients, and nutrient combinations, that have provided protection in animal models. Human investigations have yielded more mixed results, and there is certain to be continued interest in the translation of positive outcomes in animal models to humans. The importance of understanding the pharmacokinetics and pharmacodynamics of drugs of interest for otoprotection have recently been discussed by Lynch (2016), and the adoption of these more rigorous development strategies may be useful in improving on the design of studies employing dietary supplements for potential otoprotection purposes.

One of the most comprehensive efforts to identify associations between diet and hearing status across the lifespan is the work by Gopinath and colleagues as part of the Blue Mountains Hearing Study. This long-term longitudinal cohort investigation has revealed associations between carbohydrate and sugar intake (estimated using glycemic index) and hearing (poorer hearing associated with increased glycemic index, see Gopinath , 2010a). This body of work has similarly revealed associations between dietary cholesterol intake and hearing (poorer hearing associated with increased cholesterol), although statin use mediated this relationship, decreasing calculated odds ratios for hearing loss (see Gopinath , 2011a). In contrast to the negative associations with carbohydrates and cholesterol, increased antioxidant intake has been associated with reduced odds of hearing loss, with vitamins A and E having the most robust relationships (see Gopinath , 2011b). Increased consumption of fish that are high in omega-3 polyunsaturated fatty acid has also been associated with reduced odds of hearing loss in this population (see Gopinath , 2010b).

Curhan (2015) followed some 65 000 women enrolled in the Nurses Health Study II and detected statistically significant relationships between sources of vitamin A (β-carotene and β-cryptoxanthin) and hearing status, with better hearing associated with higher intake of these nutrients. Similar results with other carotenoids (lutein and zeathanthin) were recently reported by Wong (2017), although it should be noted that no other nutrients were considered in this latter investigation. Interestingly, Curhan (2015) also reported higher intake of vitamin C was associated with poorer hearing, when women taking ≥1000 mg/day were compared to women taking ≤75 mg/day. The US recommended daily allowance (RDA) is 75 mg/day, and the upper limit is 2000 mg/day. In contrast, Kang (2014) had reported that intake of vitamin C was significantly associated with hearing with better hearing at 2000 and 3000 Hz (the mid-frequencies) in those with higher vitamin C intake. This study is also one of the few studies to report that dietary supplement use was positively associated with better hearing. Results such as these highlight the possibility for interactions among nutrients, and illustrate the possibility that supplements may not be beneficial, or could even be harmful, when supra-physiological doses are taken on a regular basis.

Rosenhall (2015) completed a cross-sectional investigation of 524 70–75 year olds and found that higher intake of fish was associated with better hearing in males, but not females. Peneau (2013) similarly detected relationships in which higher intake of fish was associated with better hearing in males, but not females, although the results were borderline with respect to statistical significance (p-values less than 0.10). The above results in which benefits were observed only in males contrast with Curhan (2014) who reported lower risk of hearing loss in women who consumed two or more servings of fish per week compared to women who rarely consumed fish (i.e., less than once/month); men were not included as participants in this latter investigation. Results such as these highlight the need for study not only of food groups, but also the potential for sex differences.

In contrast to healthy foods that are associated with improved hearing, the consumption of “junk food” (juice and soft drinks, cakes and biscuits, refined sugar, honey, and sweets) was negatively associated with hearing in both men and women (Rosenhall , 2015). No correlation between fruit and vegetable intake and hearing status was revealed in this cohort, in contrast to other studies described above, that generally suggest benefits of antioxidant nutrients as part of a healthy diet. New data from Croll (2019) are notable in that no statistically significant relationships were detected for overall diet quality and hearing thresholds. However, they did detect relationships between higher body mass index (BMI) and higher fat mass index and hearing thresholds, with poorer hearing as BMI and fat index increased at study onset; they also observed that hearing was poorer with sugary beverage intake and hearing was better with higher unsaturated fat intake.

Statistically significant relationships listed above should be interpreted with caution, as the effect size is often modest in magnitude. For more detailed information on nutrition and hearing loss we refer the reader to other recent reviews (Spankovich, 2015; Puga , 2018).

Research in animal models. Research into nutrient-based protection against age-related hearing loss (ARHL) has also been conducted in rodent models. The results across studies have not been consistent even though the majority of research has evaluated nutrients with antioxidant properties, as in noise models described above (Tavanai and Mohammadkhani, 2017). Although numerous studies have demonstrated an imbalance of redox status in the cochlea during aging (McFadden , 2001; Staecker , 2001; Jiang , 2007), there have been contradictory findings with antioxidant supplementation. Some studies show otoprotection (Le and Keithley, 2007; Heman-Ackah , 2010) and others show limited or no otoprotective effect (Davis , 2007; Bielefeld , 2008; Sha , 2012). Two considerations may influence the interpretation of these mixed results; (1) intrinsic factors unrelated to ROS may have ultimately mediated the cochlear pathology and resulting hearing loss and thus no change in nutrient intake would matter, or (2) the diet of the animals is already meeting all the recommended daily requirements and thus there is limited protection that can be mediated by additional supplementation.

It is interesting and not yet well understood why otoprotection against NIHL has been relatively well demonstrated in animal models, with mixed results in translation to humans. It is possible that in challenge experiments (e.g., noise exposure) in rodent models, supplementation in animal models has proven efficacious due to enhanced efficacy to deal with acute changes that task endogenous antioxidant defense systems such as superoxide dismutase, catalase, and glutathione peroxidase systems, which are depleted by acute challenge (for review see Le Prell and Bao, 2012). In contrast to the results in noise models, the epidemiological data show statistically significant interactions between diet and hearing in humans, whereas long-term otoprotection results have been mixed in animal studies used to more precisely probe the associations detected in human datasets. Cumulative long-term effects of nutrient dense diets are possible however. It is worth note that Alvarado (2018) recently reported that the ACEMg combination of active agents prevented age-related hearing loss in a rat model, findings that are consistent with the epidemiological analyses suggesting benefits of these and other dietary derived nutrients in aging humans even after adjusting for noise exposure (see Spankovich and Le Prell 2013, 2014). One possibility that must be considered is the potential that controlled feeding to prevent obesity in laboratory animals creates control conditions that make it difficult to detect benefits of antioxidants in ARHL models in the laboratory.

Discrepancies in the outcomes of studies using dietary/caloric restriction illustrate the importance of dietary considerations as a study factor. Caloric restriction (CR) has been a successful intervention across studies assessing potential life span extension. A significant portion of the CR lifespan studies have been completed in rodent models. The benefits are not universal in rodents, as there is significant variation in outcomes across strains (Mitchell , 2016; Vaughan , 2017). Dietary restriction (DR) is often used in the same context, but DR reflects more than just reduced calorie intake and includes other restrictive manipulations; to reduce confusion we will use CR to refer to both. It is worth noting that slight variations in study design, including age of CR onset, diet composition, and feeding regimens, can have significant implications (Vaughan , 2017). Until recently, it was believed that reduced caloric intake drove the effects of CR, as opposed to reductions in specific macronutrient intake that were driven by the CR. However, accumulating evidence now points to restriction of protein, which restricts intake of specific amino acids, as underlying CR benefits. Reduced methionine intake, in particular, induces specific protective molecular mechanisms (Fontana and Partridge, 2015). It should be noted that in general these CR studies show tent-shaped response, with peak life at moderate levels of intake and declined lifespan with either starvation or elevated intake (Partridge , 2005). Another dietary restriction method, intermittent fasting (IF), can also be associated with protective effects and the effects of CR may be partially mediated by the effects of IF when food is delivered at fixed intervals with fasting periods between feedings (Cummings and Lamming, 2017). CR has been associated with numerous metabolic changes and signaling pathways (Finkel, 2015).

Two non-human primate studies assessing the effects of CR were initiated nearly in parallel by the National Institute of Aging (started in 1987) and the University of Wisconsin (started in 1989) in rhesus monkeys to determine effects of CR on morbidity, mortality, and lifespan. Though much of the data have been positive for such relationships, one critical difference has led to significant debate. At UW, both morbidity and survival have been extended, however at the NIA, only morbidity has been improved. Several study design factors as related to diet have been implicated including the control group diet, feeding regimen, diet composition, age of treatment onset, and genetics. For example, in both the UW and NIA studies, neither group of control animals were fed in a true ad libitum manner; rather, a limited daily feeding was initiated. The lack of increased survival in the NIA study may be in part attributable to an additional 10% restriction applied to their control animals, which may have been sufficient to bring them close to peak effect of CR. In addition, monkeys in the Wisconsin study were fed once a day and in the NIA study twice per day, which may have created an IF phenomenon and contributed to differences in mortality. Also, diet composition was considerably different between the two studies, UW using purified ingredients and NIA natural ingredient. Natural diets provide a more complete source of nutrition, but may have less consistency batch to batch (Vaughan , 2017).

Hearing outcomes have also been examined within these two studies. In the UW study, ABR were of higher amplitude in the CR monkeys compared to controls. In brief, there was a statistically significant sex by diet interaction in which latencies were shorter and the amplitudes larger for the responses of females relative to males, and the latencies were shorter and the amplitudes were larger for the subjects under the CR condition. Age was also significant, with shorter latencies and larger amplitudes in the younger monkeys than the older monkeys (Fowler , 2002). Although there was a trend in which thresholds were better in the younger animals, the differences were not statistically significant (Fowler , 2002). The mechanism speculated to underlie the protective effects of CR on the inner ear was prevention of the oxidative damage to macromolecules such as DNA, proteins, and lipids that otherwise normally occurs with aging and age-related diseases (for detailed discussion, see Someya , 2010). On the contrary, the NIA study did not reveal any differences in hearing (Torre , 2004). As indicated, diet composition and dietary regimen (once/day versus twice/day feeding) may underlie these differences in outcomes for auditory function metrics, but genetics may also contribute as the NIA study included Indian and Chinese monkeys, whereas the UW study only included monkeys of Indian origin. Strain differences in CR and hearing have been observed in mice (Willott , 1995).

The concern for diet composition as a source of variation in experimental animal models is not limited to longevity studies (Jucker, 2010; Mirzaei , 2016). There is also concern for how these diets reflect dietary intake patterns typical of human consumption (Phelan and Rose, 2005; Giles , 2016). In other words, the increases in longevity associated with CR in rodents may prove either irrelevant for humans (Phelan and Rose, 2005) or potentially unsustainable (Cummings and Lamming, 2017) given that human eating habits and patterns differ from both control and experimental animal populations.

Limitations for translation of nutrient-based hearing protection strategies are multiplicative. Here we highlight two notable obstacles.

There are significant species-specific differences in how nutrients are acquired, metabolized, and excreted. Researchers should be cognizant of these limitations when choosing an animal model, in particular in research applying nutrient-based otoprotection strategies. Work in animal models allows for much greater control of study design including diet, and controlling the sample size is necessary to determine efficacy of an otoprotective strategy. Controlling diet in humans is more complicated, but diet quality should be a consideration. These factors can be compounded by influence of dosage difference, delivery vehicle, scheduling/regimen, and unknown long-term consequences. For example, an animal study may show long term benefit of an antioxidant supplement in rodent model. However, data in human epidemiological studies suggest that long-term use of antioxidant supplements at minimal is of no benefit and at worst may potentially be harmful (Mursu , 2011; Bjelakovic , 2014). A number of trials have targeted military populations for potential prevention of hearing loss during military basic combat training. Interestingly, assessment of dietary intake during military basic combat training has demonstrated significant improvement in HEI during training compared to pre-training (Lutz , 2013). In regards to trial design this may have positive and negative implications. First, the better diet may offer additional protection. Second, the participants may have more uniformity and consistency in their diet. Yet, if diets are already healthy and meeting USDA recommendation, this may diminish or alter additional otoprotection mediated by a supplemental agent. Stressing the point, any study in human examining agents for otoprotection should consider diet quality of the participants. A final critical factor in nutrient-based research, in particular human data, is biostatistical considerations. As addressed, nutrients significantly interact and are not acquired in the diet as single compounds, this can create collinearities and confounds that are often not addressed. Summary tables listing outcomes in animal studies (Le Prell and Bao, 2012) and human trials (Le Prell and Lobarinas, 2015; Le Prell, 2019b) are available.

Again, work in animal models allows for much greater control of study design including calibrated acoustic exposure to induce a reliable trauma. Human workplaces have limitations on noise exposure, and laboratory paradigms are similarly limited to prevent injury to participants. The days of blasting ears to loud sound to elicit a 50 dB TTS are long gone (Ward and Glorig, 1960) given that it is now well accepted that TTS exposures can be hazardous (Kujawa and Liberman, 2009). Smaller TTS changes with lower level exposures do not appear to have long term effects (Fernandez , 2015) and contemporary study designs have therefore attempted to reduce risk by either using lower risk TTS based noise exposures or identifying populations with higher risk for TTS or PTS (e.g., military). Nonetheless, TTS and PTS are not equivalent and the lack of protection from TTS does not exclude potential protection from PTS or the reciprocal. Concerns for pathology (e.g., synaptopathy) with TTS designs are worth noting, though the exposures that have been used in more recent human TTS studies are not likely substantial enough to generate a synaptopathy-like event; noise-criterion for synaptopathy vary across animal models and are currently unknown for humans (Dobie and Humes, 2017). This highlights the need for further understanding of inter-species differences in susceptibility to noise-induced pathology. In general, higher noise exposures for longer durations are required to elicit TTS and PTS at comparable magnitudes in higher species and recovery may differ. For example, a common noise exposure method for eliciting synaptopathy in mice uses a narrowband of noise (NBN) at 100 dB SPL for 2 h (Kujawa and Liberman, 2009). In comparison, guinea pigs (Lin , 2011a; Lin , 2011b; Furman , 2013), rats (Lobarinas , 2016), and non-human primates (Valero , 2017) required higher levels, longer durations, or both, to elicit a similar synaptopathic event [see Le Prell (2019a) for recent review]. In addition, much work in noise, animals, and otoprotective strategies is limited to a single noise exposure in a naive animal, whereas many humans are exposed to noise on a repeated basis either as a consequence of daily work tasks or during common weekend recreation. Further, humans have greater genetic heterogeneity and higher variance for the effects of noise (Lindgren and Axelsson, 1986). This in turn increases sample sizes necessary and requires greater consideration of potential confounding variables including diet. Recent reviews with data listed in Table form are available (Le Prell, 2019a; Bramhall , 2019).

There is data from mouse models showing that long-term exercise preserves threshold sensitivity in aged mice (Han , 2016), and some data from humans suggest similar relationships (Haas , 2016), although these relationships have not been consistently observed across populations (Loprinzi and Joyner, 2017). Many adults that have healthy eating habits also have healthy exercise habits, and thus exercise may serve as a covariate in mediating what the relationships between diet and hearing. It is also true that some exercise classes are conducted at high noise levels (Sinha , 2017) and thus there may be some cases in which exercise classes need to be included in estimates of noise exposure.

The influence of dietary intake on successful hearing with age and in mediating otoprotection from noise exposure and other ototoxic factors is an important relationship yet to be fully appreciated. Dietary intake creates a stream of effects, from providing essential nutrients for cellular processes to influencing stress response, immune response, cardidometabolic status, neural status, and psychological well-being. There are differences between rodent physiology and human physiology and these differences have likely limited the overall success of various efforts to translate dietary supplements from rodent models directly to human clinical trials. In moving from rodent models to human testing, comprehensive consideration of the specific differences between the rodent model of interest (mouse, rat, guinea pig, chinchilla) and the human must be considered with respect to nutrient metabolism, utilization, and excretion.

Dietary quality has been shown to alter risk for essentially all chronic disease including hearing loss and tinnitus (Spankovich , 2017). A key issue regarding the relation of dietary intake and risk of hearing loss that should be considered in future research is whether “adequate intake” is sufficient for maintenance of auditory health, or if there is a need to obtain higher plasma threshold levels through greater dietary or supplemental intakes. Separate but related, researchers must be cognizant of the differences in recommendations coming not only from national governmental bodies (i.e., the USDA) but also health related organizations advocating specific diets (for review and discussion, see Shao , 2017).

Only when “adequate intake” has been more precisely defined can the potential for adverse effects of specific nutritional deficiencies be identified, including how these relations may be modified by age and whether the effects of these deficiencies can be ameliorated by other nutrients consumed at higher levels. Factors that influence serum biochemical status, including gut absorption, metabolism, and intrinsic and extrinsic factors that influence micronutrient bioavailability, may confer inter-individual differences that are not routinely captured using the current clinical trial designs which are typically limited to auditory threshold assessment, and, perhaps, some basic plasma or serum analyses in a subset of studies. Unraveling the intricate biochemistry and complete multitude of interactions may not be realistic, but being informed on species-specific physiological differences and other potential confounding variables is critical to further progression in this understudied area of hearing research. At a minimum, clinical trials evaluating otoprotective benefits of nutrients should account for dietary quality, noise exposure history, and exercise habits as potential covariates that may influence the efficacy and effectiveness of test agents, and pharmacokinetic measures are encouraged as well.