Salt-Taste Polymorphism TRPV1-rs8065080 Is Associated with Increased Likelihood of Depression in an Elderly Cohort Open Access

Depression and anxiety are common mental health conditions, impacting more than 280 million people worldwide [1]. Both conditions contribute to significant reductions in disability-adjusted life years [1] and cost the global economy an estimated USD 1 trillion each year [2]. Concerns over worldwide prevalence in older adults (depression 7%; anxiety 3.8%) are heightened by the impact of rapidly ageing populations [3]. Furthermore, depression severity increases, and prognosis worsens with age [4, 5]. The aetiologies of depression and anxiety are unclear, and diagnoses can be difficult [6, 7]. Treatment effectiveness varies for both conditions [8‒15] and medication side effects lead to non-adherence [14‒17]. Therefore, additional information that assists with detection, prevention, and treatment is needed.

There is evidence indicating a healthy diet that includes a variety of nutrient-rich foods can have positive effects on mental health [18‒20]. Nutrient adequacy and diet quality are inversely associated with depression [21‒27] and anxiety [25, 28, 29]. Poor quality diets are typically characterized by higher intakes of discretionary and processed foods, which are often high in sodium [30‒32]. Data suggesting a potential link between sodium intake and mental health have begun to emerge with a large cross-sectional study (n = ∼10,000) finding people with depression were more likely to add salt to their food [33]. In contrast, women with low salt intakes were more likely to be depressed [33]. In this study by Goldstein et al. [33], sodium intake was higher in men, which was hypothesized to explain their lower rates of depression. However, other studies have shown that high sodium intake is associated with increased inflammation and oxidative stress [34‒37], which are thought to play a role in the development of depression and anxiety [38, 39].

Variations in taste can influence dietary preferences and intakes of foods [40‒44] important to mental health [18, 45, 46]. Changes in the way we perceive taste have been found in mental health conditions [47‒49] such as increased salt-taste thresholds in chronic anxiety [48] and depression [49]. Taste sensitivity, dietary preferences, and food intakes are linked to variance in genes coding for taste receptors [50‒54]. The TRPV1 (transient receptor potential cation subfamily V member 1) gene codes for a cation channel involved in the detection and modulation of salt taste [51, 55, 56]. Single-nucleotide polymorphisms (SNPs) in this gene have been associated with changes to salt-taste perception [57‒61]. The most frequently implicated SNP is TRPV1-rs8065080 for which T allele carriage is associated with higher taste sensitivity to salt [57‒59], reduced liking of salty foods [59], and higher consumption of sodium [57]. Other TRPV1 SNPs have been studied with varying outcomes. For example, the rs4790522 and rs150908 SNPs have been associated with alterations in salt taste sensitivity [60, 61] while an earlier study found no associations [58]. Whether the relationships between salt-taste genetics, sodium intake, and diet quality are involved in depression and anxiety remains to be explored.

Taste receptors are also expressed extra-orally and have additional non-gustatory functions [62]. The presence of taste receptors in brain regions [63, 64] that demonstrate functional and structural abnormalities in anxiety [65, 66] and depression [67, 68] further highlights possible involvement of these receptors in the pathology of these conditions. Anxiolytic [65, 69] and antidepressant behaviours have been observed in TRPV1-knockout mice [69]. Changes in serotonin (5-HT) and gamma-amino-butyric-acid (GABA) receptors of the brain in the absence of TRPV1 receptors have also been observed [69]. Other neurotransmitters are modulated by TRPV1 channel activation such as the release of glutamate from synapses [70, 71] and the reduction of GABA, dopamine, and other catecholamines [69, 71]. Neurotransmitters released after salt stimulation of taste receptors, such as 5-HT, GABA, and norepinephrine (NE) [72, 73], are altered in anxiety and depression [74, 75] and have also been found to modulate taste thresholds [76]. Together, these studies indicate possible involvement of TRPV1 receptors in anxiety and depression; however, the role of SNPs in genes coding for salt-taste receptors is yet to be investigated.

Exploring associations between taste genetics, depression, and anxiety may identify new markers to support risk assessment and diagnosis and develop new dietary approaches useful to their management and prevention. Therefore, this cross-sectional study explored the associations between the TRPV1-rs8065080 salt-taste genotype, sodium intake, diet quality, and indicative levels of depression and anxiety in an elderly cohort. This study used the Hospital Anxiety and Depression Scale (HADS) to obtain data on these mental health states and a food frequency questionnaire (FFQ) to generate three diet quality indices and estimated sodium intake data.

Data for this secondary cross-sectional analysis were obtained from the Retirement Health and Lifestyle Study (RHLS). Participants were eligible to take part in the RHLS if they were aged 65 years or older, were residents of the Wyong and Gosford local government areas of the NSW Central Coast of Australia, and were either living independently in private dwellings or retirement villages [77]. Participants were excluded if they were unable to provide written informed consent. They were not excluded based on the presence of pre-existing health conditions, including depression or anxiety [77]. This sub-study (n = 536) included those who had provided blood samples, were successfully genotyped for TRPV1-rs865080, and completed a valid FFQ. Ethics approval for the RHLS research was granted by the Human Research Ethics Committee of the University of Newcastle (Reference No. H-2008-0431) [78, 79].

Prior studies have found age, sex, and diet quality are associated with taste genetics and/or mental health [53, 80‒84]. Smoking is known to modify genetic expression [85] and has been linked to the presence of bitter TAS2R38 haplotypes [86]. Depression and anxiety have been correlated to income, education [87], and body mass index (BMI) [88]. Therefore, these variables were included as possible confounders to the associations between TRPV1-rs8065080 and HADS. Data on age, sex, income, education, and smoking history were obtained through interviewer-administered questionnaires. The measurement of height and weight followed the International Society for the Advancement of Kinanthropometry (ISAK) guidelines [89]. Digital scales (Wedderburn© UWPM150 Platform) were used to measure weight, which was recorded to the nearest 0.01 kg [77, 89]. The stretch stature method was used to measure height, which was recorded to the nearest 0.01 cm [77, 89]. BMI (BMI = weight [kg]/height [m²]) calculations were derived from the height and weight measures.

Fasted whole blood samples were collected from participants into EDTA-lined tubes, then stored at −20°C. The QIAGEN QIAmp DNA mini kit was used to isolate DNA from peripheral blood cells following the manufacturer’s instructions. TaqMan^™ assay code C___11679656_10 (Applied Biosystems^™, ThermoFisher Scientific, CA, USA) was used for allelic discrimination of TRPV1-rs8065080. Further detail on the genotyping protocol can be found in a prior publication [90].

Indicative depression and anxiety scores were obtained for participants through the self-administered HADS assessment tool. The HADS has been widely used in hospital (non-psychiatric) and general population research [91] and is recommended as a reliable measure of anxiety and depression [91, 92]. The HADS is composed of 14 questions equally divided to produce scores measuring each subscale of depression (HADS-D) or anxiety (HADS-A) [92]. The subscales are each scored between 0 and 21 with a score of ≥8 within a subscale indicating the presence of either depression or anxiety [92].

Three diet quality indices and daily sodium intake estimates were obtained via data from an FFQ containing 225 items [93]. Each of the indices focuses on different aspects of diet quality. The Dietary Guideline Index (DGI) provides an assessment of key nutrient intakes and the proportions obtained from healthy food types, food diversity, and unhealthy foods [94]. The Australian Recommended Food Score (ARFS) assesses dietary variety as the indicator of diet quality [95], and the Australian Healthy Eating Index (Aust-HEI) includes measures of dietary variety, healthy food choices, and consumption of fruits, vegetables, fats, and discretionary foods [96]. Inverse associations have been found in previous studies between depression and DGI [27], ARFS [22, 24], and Aust-HEI [26] diet quality scores and between anxiety and ARFS [22, 97] and Aust-HEI [26] diet quality scores. Further information about each of the indices and the scoring methods can be found in a prior publication [90].

The data analyses were undertaken using JMP (Pro v.14.2.0; SAS Institute Inc., Cary, NC, USA 27513). The cohort characteristics are described for continuous variable distributions (means, 95% confidence intervals, and standard deviations) and categorical variable distributions (number and percentage of cohort). The TRPV1-rs8065080 polymorphism was analysed by presence or absence of the variant allele (C) and genotype occurrence. Both genotype and HADS data were reported as the number and percentage of the study cohort. Due to sex differences in salt taste [52, 59], in TRPV1 findings [58], and in anxiety and depression [1, 98], results were further stratified by sex.

Continuous variables were compared using standard least-squares regression analyses (least-squares mean, 95% confidence intervals, p values) and categorical variables through nominal logistic regression analyses (odds ratios, χ², p values). p values are presented to one significant number and a threshold of <0.05 was considered statistically significant. Where appropriate, results were adjusted for potential confounding factors such as age, sex, education, income, smoking status, and BMI. In these adjustment models, values missing at random conditional on variables were omitted from the analyses. As there are relationships between diet quality, sodium intake, and depression [21‒27, 33] and anxiety [26, 28, 29], further adjustments for these dietary measures were made.

The average age of the 536 participants was 77.4 (SD ± 6.8) years, the mean BMI was 28.5 kg/m² (SD ± 4.8, Table 1), and there were no differences by sex (55% female, Table 2) for either variable (Table 1). The mean diet-quality scores were 97.0/150 points (DGI), 29.2/74 points (ARFS), and 30.5/60 points (Aust-HEI) (Table 1). Females had higher mean scores than males across all indices (DGI: p = 0.0004, ARFS: p = 0.002, and Aust-HEI: p < 0.0001, Table 1). Estimated mean sodium intake in the total population was 2,053.0 (SD ± 842.8) mg/day (Table 1) and was higher in males than females (2,226.9 mg/day vs. 1,910.8 mg/day, respectively, p < 0.0001, Table 1).

The household income of the largest proportion of participants was between AUD 20,000–AUD 40,000/year (49.2%) with males reporting higher incomes than females (p <0.0001, Table 2). Males were more likely to be educated at trade level or higher (75.9% vs. 59.6%, p < 0.0001) and to have a history of smoking (67.2% vs. 36.9%, p < 0.0001) (Table 2). Where groups had insufficient numbers for statistical analyses as adjustment variables, they were consolidated. Income was collapsed to three categories (“<AUD 20,000”; “AUD 20,000–40,000”; “>AUD 40,000”), education to three categories (“≤school”; “trade/TAFE/certificate”; “≥diploma/bachelor”), and smoking to two categories (“history of smoking”; “never smoked”).

The TRPV1-rs8065080 variant allele (C) occurred at a frequency of 0.36 and the ancestral allele (T) at 0.64. 41.0% of participants were homozygous for TT, 46.7% were heterozygous for CT, and 12.3% were homozygous for CC (Table 3). The genotype frequencies in the sample (CC 12.3%, CT 46.7%, TT 41.0%) did not deviate from expected Hardy-Weinberg equilibrium frequencies (CC 16.8%, CT 46.7%, TT 34.7%) (p = 0.16). The TRPV1-C allele (CC or CT genotypes) was carried by 58.9% of participants, with no difference between female and male subjects (Table 3). Due to low numbers in the CC homozygous group, analyses were undertaken by presence versus absence of the variant allele for more robust statistical power.

HADS-D scores indicative of depression occurred in 8% of total participants, 8.1% of females, and 8.3% of males (Table 4). HADS-A scores indicative of anxiety occurred in 13.4% of total participants, 15.9% of females, and 10.4% of males (Table 4). There were no statistically significant differences between the sexes for either HADS-D or HADS-A scores.

There were no statistically significant associations between TRPV1-C allele presence and any of the dietary measures in the total population, in females, or males (Table 5). Participants with HADS-D scores indicative of depression had lower diet quality scores on both the DGI (89.7.3 vs. 97.7, p = 0.001) and ARFS (26.3 vs. 29.5, p = 0.02) than those with normal scores, but there was no association with the Aust-HEI index (Table 6). These significant associations were also present in females (DGI 90.0 vs. 100.0, p = 0.003; ARFS 26.8 vs. 30.5, p = 0.04) but not in males (Table 6). In the total population and by sex, there were no associations between HADS-A scores and the diet quality indices (Table 7) or between HADS-D or HADS-A scores and sodium intake (Tables 6, 7).

Participants carrying the TRPV1-C allele were more likely to have HADS-D scores indicative of depression in the unadjusted model (p = 0.003) and in the models adjusting for age and sex (p = 0.003) and age, sex, income, education, smoking, and BMI (p = 0.01) than those not carrying the TRPV1-C allele (Table 8). With lower certainty, males carrying the TRPV1-C allele were more likely to have HADS-D scores indicative of depression in the unadjusted model (p = 0.003) and in the models adjusting for age (p = 0.003) and age, income, education, smoking, and BMI (p = 0.002) (Table 8). No associations between presence of the TRPV1-C allele and depression (HADS-D) scores were found in females (Table 8).

The association between presence of the TRPV1-C allele and HADS-D scores indicative of depression remained statistically significant after adjusting for the dietary measures in the total population (DGI, p = 0.004; ARFS, p = 0.001; Aust-HEI, p = 0.003; sodium, p = 0.003) and in males (DGI, p = 0.004; ARFS, p = 0.003; Aust-HEI, p = 0.003; sodium, p = 0.003) but not in females (Table 9).

There were no statistically significant associations between presence of the TRPV1-C allele and HADS-A scores indicative of anxiety in the total cohort or by sex, in unadjusted and adjusted models (Table 10). Therefore, no further analyses were undertaken.

The key finding in this cross-sectional analysis is that presence of the TRPV1-rs8065080 variant allele (C) was found to increase the likelihood of having a HADS-D score indicative of depression in an elderly cohort. The association remained significant after adjusting for the demographic variables of age, sex, income, education, BMI, and smoking and after adjusting for the three diet quality indices and sodium intake. This suggests that the association may be related to the function of the salt-taste receptor, rather than the receptor being linked to depression through changes in diet quality or sodium intake.

The association between TRPV1-C allele presence and HADS-D scores indicative of depression (with and without adjustments) was sex specific, with significant findings in males but not females. This is not due to differences in sex-specific distributions in TRPV1-C allele carriage or HADS-D scores as these did not vary between groups. However, females had higher diet quality scores on all three indices than males, and females with lower diet quality scores (DGI and ARFS) were more likely to have indicative depression while there were no associations between diet quality and depression in males. This indicates there may be a threshold effect where the impacts of genotypic variance manifest in the context of poor diet quality in males, or a sex hormone-specific interaction may be occurring.

As a missense mutation [99], TRPV1-rs8065080 may be modifying protein structure and function [57, 58, 100‒103]. Neuropathic pain irregularities such as higher tolerance to cold pain [101, 102], lower risk of wheezing in childhood asthma [100], and protective effects against hypertension (TRPV1-C allele) [103] are associated with the SNP. TRPV1-rs8065080 has been shown to decrease TRPV1 channel activity by approximately 20–30% after heat and capsaicin stimulation [100]. Decreased salt sensitivity has been found in TRPV1-C allele carriers [57] and they have been found to rate salt solutions as less intense than non-carriers [58]. Therefore, as TRPV1-rs8065080 likely causes a reduction in taste function [57, 58], and the current study suggests there is a greater likelihood of having HADS-D scores indicative of depression in the presence of the SNP, the results indicate that people with genes that reduce salt-taste function are more likely to experience depression.

The findings in this study indicate that the hypothesized functional changes in the TRPV1-rs8065080 receptor may have extra-oral implications aside from changing taste. Sodium intake was not a significant determinant of HADS-D threshold scores, TRPV1-C allele presence did not influence sodium intake, and sodium intake did not impact on the association between presence of the TRPV1-C allele and HADS-D scores. Therefore, the analyses suggest that TRPV1-rs8065080 genotype is driving the association with depression independent of sodium intake and that the functional changes in the SNP may be occurring extra-orally. TRPV1 mRNA is expressed in the human brain including in the insula, amygdala, orbitofrontal cortex, thalamus, and hippocampus [63, 64]. These same brain regions demonstrate structural and functional abnormalities in human depression [67, 68]. The missense mutation may be creating functional changes in TRPV1 and the presence of these receptors in regions of the brain affected by depression could be an avenue for further exploration.

There is an additional biological pathway to consider as a possible mechanism involved in the findings of this study and for further research. Salt triggers the release of 5-HT and NE from type III taste cells [104, 105]. These neurotransmitters are found to be low in depression [106] and to downregulate depression [74, 75]. Salt also stimulates the release of anandamide which activates TRPV1 [107, 108] and restores 5-HT and NE levels in the brain [109, 110]. Absence of TRPV1 in knockout mice shows changes in 5-HT receptors and decreased depressive behaviours [69]. Therefore, the association between TRPV1-C allele carriage and higher likelihood of scores indicative of depression may be a result of SNP-related functional changes to levels of neurotransmitters linked to depression.

As the nature of this study is observational and mental health conditions and genetic expression are multifactorial, the findings are limited to association. Additional limitations include those around the collection of subjective data such as the HADS test and FFQ. For example, rates of depression and anxiety in the study sample differed to those of similar age in the general population at the time of data collection (2010–2012). HADS-A threshold scores occurred in 13.4% and HADS-D threshold scores in 8.0% of the study population. In comparison, the Australian Health Survey data from 2011 to 2012 reported that approximately 3.3% of people over 65 experienced anxiety and 10.6% experienced depression [111]. Therefore, significant associations to depression may be under-stated.

The study’s strengths include the large cohort size and availability of broad demographic and dietary data that enabled adjustments to be made for sex, age, income, education, smoking, BMI, sodium intake, and three diet quality indices. As age was not a confounding factor in the significant associations, the results can be more widely considered in other age groups. The sex distribution is reflective of the Australian statistics for people aged 65 years and older [112] and HADS is a reliable and valid measure of levels of depression and anxiety in research settings [91]. While cross-sectional studies have limitations, they are a necessary first step towards experimental studies to test the results under controlled conditions. Future studies should investigate how salt-taste receptors, TRPV1 channels, and SNPs in the brain are involved in the biology of depression, which may generate new approaches to prevention and treatment. Studies that elucidate the role neurotransmitters have in salt taste and mental health and that determine if salt-taste SNPs alter neurotransmitter levels may also contribute to improved management of depression.

Overall, the novel findings of this exploration of associations between genetic variance of a salt-taste receptor (TRPV1) and levels of anxiety and depression in an elderly population suggest that the TRPV1-rs8065080 SNP is more likely to be present when depression is indicated. This association appears to be independent of sodium intake and diet quality in the total sample. However, poor diet quality in males may provide the context of a threshold effect for the impacts of the gene. This knowledge may aid in the development of improved risk assessment or diagnosis of depression and the development of preventative strategies and new therapies. Given the significant and increasing disease burden of depression, further research is needed to explore these associations to improve outcomes and the quality of life of sufferers.

Authors acknowledge the role of Paul Roach in the design of the original study and the roles of Charlotte Martin, Zoe Yates, Katrina King, and Suzanne Niblett in sample collection and data management.

Ethics approval for the Retirement Health and Lifestyle Study was granted by the Human Research Ethics Committee of the University of Newcastle (Reference No. H-2008-0431). Written informed consent was provided by all subjects included in this study.

There are no conflicts of interest to declare.

This research was conducted as part of the Retirement Health and Lifestyle Study, with initial and ongoing funding provided by the Australian Research Council (G0188386), Central Coast Local Health District Public Health Unit (G0190658/G1700259), UnitingCare Ageing NSW/ACT (G0189230), Urbis Pty Ltd. (G0189232), Valhalla Village Pty Ltd. (G1000936), and Hunter Valley Research Foundation.

Conceptualization, formal analysis, methodology, and writing – original draft: C.F. and E.L.B.; data curation: C.F., M.L., and E.L.B.; funding acquisition, project administration, and resources: M.V., M.L., and E.L.B.; and investigation and writing – review and editing: C.F., C.S., M.V., M.L., T.B., and E.L.B. All authors have read and agreed to the published version of the manuscript.

All data generated or analysed during this study are included in this article. Further enquiries can be directed to the corresponding author.