Abstract
Objectives
To assess the cross-sectional and longitudinal associations between nut consumption and depression in two cohorts of older adults.
Design, Setting, Participants and Measurements
The first cohort (Seniors-ENRICA-I or SE-I) included a representative sample of Spanish noninstitutionalized adults aged ≥65 years interviewed in 2010 and 2013. The second cohort (SE-II) included individuals from the Madrid region, Spain, aged ≥65 years interviewed in 2017 and in 2019. Nut consumption was estimated with a validated computer-based diet history. Depression was defined as self-reported physician-diagnosed depression or the use of antidepressants. Logistic regression models were adjusted for the main confounders. The DerSimonian and Laird random-effect method was used to meta-analyze the results from both studies. A participant-level pooled analysis was conducted to examine the robustness of our analyses.
Results
The SE-I included 2278 individuals (233 prevalent cases) in the cross-sectional analysis and 1534 (108 incident cases) in the longitudinal analysis; the corresponding figures for SE-II were 2726 (407 prevalent cases) and 1566 (74 incident cases). In the meta-analysis of cross-sectional results from the two studies, compared to consuming <1 serving (30 g) of nuts/week, the odds ratio (95% confidence interval) for depression was 0.90 (0.64, 1.16) for consuming 1 to <3 servings/week and 0.92 (0.70, 1.13) for consuming ≥3 servings/week; the corresponding figures for the longitudinal results were 0.90 (0.41, 1.38) and 0.66 (0.35, 0.97).
Conclusion
Nut consumption was associated with a lower risk of depression in a pooled longitudinal analysis using data from two cohorts of older adults. Nuts should be recommended as part of a healthy diet in older adults.
Key words: Depression, mental health, Mediterranean diet, nuts, older adults
Introduction
Depression is the most prevalent mental disease and was ranked as the seventh leading cause of disability-adjusted life years in 2019 (1). The global burden of mental disorders, specifically depression, has increased due to the lockdown and the health and economic challenges imposed by the COVID 19 pandemic (2). In addition to the increase in the incidence of depression, those who suffered from depression were more likely to report more severe depressive symptoms after the lockdown (3). Although depression can occur throughout life, its prevalence is particularly high among older adults, who are commonly more exposed to risk factors for depressive symptoms, such as social isolation, reduced physical activity, the presence of chronic diseases, and associated drug treatments (4, 5). Therefore, public health systems should focus on preventive and management strategies to reduce symptoms and health-related consequences of depression (6).
Among lifestyle behaviors, nutrition has gained attention in the last decade as a modifiable factor potentially relevant to prevent and manage mental disorders (7). In addition to the benefits of healthy dietary patterns (i.e., Mediterranean diet, Dietary Approaches to Stop Hypertension [DASH] diet, MedDiet-DASH Intervention for Neurodegenerative Delay [MIND], etc.) in cardiovascular health (8), some authors have suggested that these diets are associated with lower cognitive decline (9, 10) and depression risk (11). These dietary patterns are characterized by moderate-to-high consumption of plant-based food such as vegetables, fruits and legumes. Among these foods, nuts may play a key role because they are rich in bioactive compounds and non-sodium minerals (i.e., unsaturated fatty acids, B vitamins, folate, polyphenols, phytosterols, tocopherols, calcium, magnesium, potassium, etc.) (12).
Nut consumption, particularly walnut consumption, has been shown to reduce cardiovascular risk factors (8), and some studies have explored its potential effects on cognitive performance and mental disorders (13, 14). Regarding depression, a recent systematic review by our group found some indication that nut consumption may lower the risk of depression and improve mood state in the general population, but the evidence was inconclusive and particularly scarce in older adults (14). In fact, only a small 3-month (n = 45) randomized controlled trial of a low-carbohydrate diet supplemented with 2 servings/day of almonds compared with a low-fat diet in older adults (mean age: 72±6, and 55.6% women) with type-2 diabetes reported a protective effect of almond intake on depression 15. Specifically, compared to the control diet, the almond-based diet led to significantly lower depression scores measured by the Adult PROMIS short-form Depression 8b, which measures self-reported negative mood, self-criticism, self-worthlessness, loneliness, interpersonal alienation, and decreased positive affect and engagement (15).
Therefore, this study aimed to assess the cross-sectional and longitudinal associations between nut consumption and depression in two cohorts of older adults in Spain.
Methods
Study design and participants
The data were taken from two cohorts of older adults with a very similar design, data collection and measurement procedures; their methods have been reported elsewhere (16, 17).
The first cohort is the Seniors-ENRICA-I (SE-I), established in 2008–2010 with 3289 individuals selected through stratified random sampling of the noninstitutionalized population aged ≥ 60 years in Spain. Baseline data were collected in three sequential stages: 1) a computer-assisted telephone interview to obtain information on lifestyle, subjective health, morbidity and health services use; 2) a first home visit to perform the physical examination and collect blood and urine samples; and 3) a second home visit to measure anthropometric variables and to administer a computerized dietary history (18). To enable comparability between the two cohorts in the present study, only individuals aged 65 years or older were selected. Of the 2533 older adults included in the sample, 139 lacked data on nut consumption, 5 on depression, 44 on body weight and height, 9 on sleep duration, 7 on the MEDAS score and 51 on any of the comorbidities considered. Therefore, cross-sectional analyses of the SE-I were performed with 2278 (89.9%) individuals. Of these, between 2010 and 2013, after a mean follow-up time of 3.2 years, 233 participants were excluded because they had depression at baseline, 70 participants had died, and another 441 were lost to follow-up, resulting in 1534 individuals for the longitudinal analyses (Figure 1).
Figure 1.
Flow diagram of the study participants in each Seniors-ENRICA study
The second cohort is the Seniors-ENRICA-II (SE-II) cohort, which was established in 2015–2017 with noninstitutionalized individuals aged ≥65 years from the region of Madrid, Spain 19. Baseline data were collected through telephone interviews and household visits as in the SE-I study (19). Of the 3271 participants included in the sample, 478 lacked data on nut consumption, 4 on depression, 11 on body weight and height, 4 on alcohol drinking, 2 on sleep duration, 41 on physical activity and 3 on the MEDAS score. Thus, 2726 (83.3%) participants were included in the cross-sectional analysis. For the longitudinal analysis, we excluded 407 participants with depression at baseline, 38 who died and 715 who were lost to follow-up for different reasons (i.e., were not located, refused to participate, etc.). Thus, between 2017 and 2019, after a mean follow-up time of 2.3 years, 1566 participants were included in the longitudinal analysis (Figure 1).
Both the SE-I and SE-II studies were approved by the Clinical Research Ethics Committee of the La Paz University Hospital in Madrid, and participants provided informed written consent.
Study variables
Exposure variable: nut consumption
A validated computer-based diet history, representing the average intake during a typical week of the previous year, was conducted to assess the daily intake in grams of nut consumption among the participants (18). The diet history included 20 types of nuts, which were later grouped as follows: almond, hazelnut, peanut, chestnut, walnut, pine nut, pistachios, cashew, macadamia nut, and other nuts. Nut consumption was expressed as a continuous variable (g per week) and recoded into three categories: 1) 0 to < 1 serving per week; 2) 1 to < 3 servings per week; and 3) ≥ 3 servings per week, assuming that 1 serving = 30 grams, as previously reported (8).
Outcome variable: depression
In both cohorts, depression was defined based on two criteria. The first was when the participant answered positively to the question ‘Did a physician tell you if you currently have, or have had in the past year, depression (in need of treatment)?'. The second criterion was the use of antidepressant drugs, which was checked by a nurse at the home visit. The study participant was considered to suffer from depression when one or both criteria were met.
Covariates
Self-reported information was obtained for age (years), sex, education level (higher: secondary/university; lower: no studies/primary) and cohabitation (living with others or living alone). Tobacco consumption (never/former or current), alcohol drinking (never/former or current), and night's sleep duration (normal or long: ≥ 6.5; short: <6.5 h/d) were obtained through a computer-assisted telephone interview. Information on leisure-time physical activity was also obtained with the EPIC-cohort questionnaire (20); individuals were classified as less active (inactive or moderately inactive) or more active (moderately active or active). Body mass index (BMI) was calculated as weight divided by the square of the height (kg/m2), both objectively measured under standardized conditions. Adherence to the Mediterranean diet was assessed with the Mediterranean Diet Adherence Screener (MEDAS) (21). Finally, participants were asked about physician-based diagnosis of the following chronic conditions: diabetes mellitus, cardiovascular disease (ischemic heart disease, stroke, or heart failure), hypertension, obstructive sleep apnea (OSA) syndrome, cancer at any site, and neurodegenerative disease (Parkinson's disease, dementia, or Alzheimer's disease).
Statistical analysis
We first described the main baseline characteristics of participants in the cross-sectional and longitudinal analyses of each cohort. Additionally, the baseline characteristics of each cohort were stratified by nut consumption category. The differences between both cohorts and between categories of nut consumption were compared through the chi-squared test for categorical variables and the t-test for continuous variables. Next, logistic binary regression models were performed to estimate the odds ratio (OR) and the 95% confidence interval (95% CI) of the association between three categories of nut consumption (0 to < 1, 1 to < 3, and ≥ 3 servings per week) and depression (dichotomic dependent variable). First, we fitted a crude model (unadjusted) and then another two models with adjustments for potential confounders suggested by the available evidence (14, 22, 23, 24). The first adjusted model (Model 1) was adjusted for sex, age, education, and cohabitation. Model 2 additionally adjusted for smoking status, alcohol consumption, sleep duration, physical activity, BMI, MEDAS score, diabetes mellitus, cardiovascular disease, hypertension, OSA syndrome, cancer at any site, and neurodegenerative disease.
Finally, the DerSimonian and Laird random-effect method (25) was used to meta-analyze the results from both SE-I and SE-II studies and estimate the pooled ORs and 95% CIs of depression according to the category of nut consumption in cross-sectional and longitudinal analyses.
To examine whether the results were robust, we conducted complementary analyses by pooling the data from both cohorts into a single database (participant-level pooled analysis). For this purpose, multilevel models with participants nested in the study (SE-I or SE-II) were conducted for cross-sectional and longitudinal analyses. In the case of the longitudinal analyses, the corresponding follow-up time (y) was included as an adjustment variable. Finally, considering the use of antidepressants to be a more conservative criterion for defining current depression, these sensitivity pooled analyses were conducted including the dependent variable of using or not using antidepressant medications. Subgroup analyses considering sex and age (65 to < 72 y vs. ≥ 72 y) were conducted to verify the robustness of the main findings.
Statistical analyses were performed using STATA SE software, version 15 (StataCorp, College Station, TX, USA).
Results
The baseline population included 2278 individuals (mean age 71.7 ± 5.7 years, 56.5% females) in SE-I and 2726 (mean age 71.7 ± 4.4 years, 52.3% females) in SE-II. As presented in Table 1, some characteristics differed between the cohorts. Specifically, some variables presented higher values in the SE-I than in the SE-II, such as the mean BMI (p < 0.001) and the frequency of female sex (p = 0.021), cardiovascular disease (p < 0.001) and hypertension (p < 0.001). In contrast, other characteristics were more frequent in participants of the SE-II than in those from the SE-I, such as alcohol drinking (p < 0.001), less physical activity (p < 0.001), OSA syndrome (p < 0.001), and nut intake (p < 0.001) (Table 1).
Table 1.
Baseline characteristics of the participants in Seniors-ENRICA-I and Seniors-ENRICA-II
|
Characteristic |
Cross-sectional analysis |
Longitudinal analysis |
||||
|---|---|---|---|---|---|---|
| Seniors-ENRICA-I | Seniors-ENRICA-II | p-valuea | Seniors-ENRICA-I | Seniors-ENRICA-II | p-valuea | |
| Total, n | 2278 | 2726 | 1534 | 1566 | ||
| Age (y), mean ± SD | 71.7 ± 5.7 | 71.7 ± 4.4 | 0.937 | 71.4 ± 5.5 | 71.4 ± 4.2 | 0.924 |
| Female sex, n (%) | 1288 (56.5) | 1425 (52.3) | 0.021 | 819 (53.4) | 724 (46.2) | 0.001 |
| Lower educational level, n (%) | 1435 (63.0) | 1733 (63.6) | 0.148 | 891 (58.1) | 918 (58.6) | 0.351 |
| Living alone, n (%) | 453 (19.9) | 598 (21.9) | 0.070 | 300 (19.5) | 317 (20.2) | 0.652 |
| Current smoking, n (%) | 210 (9.2) | 255 (9.4) | 0.716 | 151 (9.8) | 144 (9.2) | 0.546 |
| Current alcohol drinker, n (%) | 1342 (58.9) | 1999 (73.3) | <0.001 | 936 (61.0) | 1227 (78.4) | <0.001 |
| Short (< 6.5 h/d) sleep duration, n (%) | 917 (40.3) | 1083 (39.7) | 0.326 | 605 (39.4) | 620 (39.6) | 0.432 |
| Less physically active,b n (%) | 1131 (49.7) | 1532 (56.2) | <0.001 | 783 (51.1) | 826 (52.8) | 0.215 |
| Body mass index (kg/m2), mean ± SD | 28.7 ± 4.4 | 27.9 ± 4.5 | <0.001 | 28.6 ± 4.2 | 27.7 ± 4.4 | <0.001 |
| MEDAS score, mean ± SD | 7.1 ± 1.7 | 7.1 ± 1.7 | 0.998 | 7.3 ± 1.7 | 7.2 ± 1.7 | 0.243 |
| Diabetes mellitus, n (%) | 426 (18.7) | 503 (18.5) | 0.979 | 268 (17.5) | 271 (17.3) | 0.855 |
| Cardiovascular disease, n (%) | 150 (6.6) | 128 (4.7) | 0.016 | 90 (5.9) | 53 (3.4) | 0.001 |
| Hypertension, n (%) | 1557 (68.4) | 1521 (55.8) | <0.001 | 1064 (69.4) | 848 (54.2) | <0.001 |
| Obstructive sleep apnea syndrome, n (%) | 82 (3.6) | 201 (7.4) | <0.001 | 52 (3.4) | 117 (7.5) | <0.001 |
| Cancer at any site, n (%) | 59 (2.6) | 78 (2.9) | 0.328 | 31 (2.0) | 41 (2.6) | 0.216 |
| Neurodegenerative disease, n (%) | 28 (1.2) | 34 (1.3) | 0.842 | 15 (1.0) | 16 (1.0) | 0.479 |
| Depression, n (%) | 233 (10.2) | 407 (14.9) | <0.001 | - | - | - |
| Nut consumption (servingsc/week), mean ± SD | 1.6 ± 3.2 | 2.6 ± 3.5 | <0.001 | 1.6 ± 3.2 | 2.9 ± 3.6 | <0.001 |
| Nut consumption (servingsc/week), n (%) | ||||||
| 0 to < 1 | 1568 (68.8) | 1331 (48.8) | <0.001 | 1040 (67.8) | 729 (46.6) | <0.001 |
| 1 to < 3 | 247 (10.9) | 429 (15.7) | 171 (11.2) | 243 (15.5) | ||
| ≥ 3 | 463 (20.3) | 966 (35.5) | 322 (21.0) | 594 (37.9) | ||
MEDAS: Mediterranean Diet Adherence screener; SD: standard deviation. a P-values were obtained through the chi-squared test for categorical variables and the t-test for continuous variables. b Less active (inactive or moderately inactive) during leisure time according to the EPIC-physical activity questionnaire. c 1 serving = 30 g.
The most consumed type of nut was walnuts both in SE-I (13.0% consumed walnuts) and in SE-II (34.9% consumed walnuts), followed by almonds (SE-I: 5.1% consumed almonds; SE-II: 18.9% consumed almonds). The other types of nuts were consumed by less than 3% of the study population.
Table 2 presents the baseline characteristics of participants in the two cohorts by category of nut consumption. In both databases, compared with adults who consumed ≥ 3 servings per week of nuts, those who consumed fewer nuts (0 to <1 serving per week) had higher frequencies of low educational level (p = 0.024 for SE-I and < 0.001 for SE-II), higher BMI (p <0.001 for both SE-I and SE-II), lower diet quality MEDAS scores (p < 0.001 for both SE-I and SE-II), and higher frequency of diabetes mellitus (p = 0.016 for SE-I and 0.007 for SE-II).
Table 2.
Baseline characteristics of the participants in Seniors-ENRICA-I and Seniors-ENRICA-II by category of nut consumption
|
Characteristic |
Seniors-ENRICA-I Nut consumption (servingsa/week) |
Seniors-ENRICA-II Nut consumption (servingsa/week) |
||||||
|---|---|---|---|---|---|---|---|---|
| 0 to < 1 | 1 to < 3 | ≥ 3 | p-valueb | 0 to < 1 | 1 to < 3 | ≥ 3 | p-valueb | |
| Total, n | 1568 | 247 | 463 | 1331 | 429 | 966 | ||
| Age (y), mean ± SD | 71.9 ± 5.8 | 71.4 ± 5.3 | 71.1 ± 5.6 | 0.041 | 71.9 ± 4.6 | 71.5 ± 4.3 | 71.5 ± 4.2 | 0.061 |
| Female sex, n (%) | 865 (55.2) | 148 (60.1) | 274 (59.2) | 0.140 | 715 (53.7) | 233 (54.3) | 477 (49.4) | 0.079 |
| Lower educational level, n (%) | 1028 (65.6) | 147 (59.3) | 260 (56.2) | 0.024 | 888 (66.7) | 280 (65.3) | 565 (58.5) | <0.001 |
| Living alone, n (%) | 343 (21.9) | 38 (15.4) | 72 (15.6) | 0.001 | 282 (21.2) | 102 (23.8) | 214 (22.2) | 0.519 |
| Current smoking, n (%) | 151 (9.7) | 27 (10.8) | 31 (6.8) | 0.122 | 143 (10.7) | 33 (7.7) | 79 (8.2) | 0.050 |
| Current alcohol drinker, n (%) | 889 (56.7) | 162 (65.7) | 290 (62.7) | 0.052 | 941 (70.7) | 313 (72.9) | 745 (77.1) | 0.003 |
| Short (< 6.5 h/d) sleep duration, n (%) | 623 (39.7) | 97 (39.4) | 197 (42.5) | 0.423 | 529 (39.7) | 158 (36.8) | 396 (41.0) | 0.341 |
| Less physically active,c n (%) | 745 (47.5) | 139 (56.3) | 246 (53.2) | <0.001 | 760 (57.1) | 236 (55.0) | 536 (55.5) | 0.643 |
| Body mass index (kg/m2), mean ± SD | 28.9 ± 4.4 | 28.5 ± 4.5 | 28.2 ± 4.2 | <0.001 | 28.3 ± 4.5 | 27.8 ± 4.7 | 27.4 ± 4.4 | <0.001 |
| MEDAS score, mean ± SD | 6.9 ± 1.5 | 7.1 ± 2.1 | 7.9 ± 1.9 | <0.001 | 6.6 ± 1.6 | 7.1 ± 1.7 | 7.8 ± 1.6 | <0.001 |
| Diabetes mellitus, n (%) | 321 (20.5) | 40 (16.1) | 66 (14.2) | 0.016 | 267 (20.1) | 88 (20.5) | 148 (15.3) | 0.007 |
| Cardiovascular disease, n (%) | 108 (6.9) | 18 (7.5) | 23 (5.0) | 0.451 | 76 (5.7) | 15 (3.5) | 37 (3.8) | 0.048 |
| Hypertension, n (%) | 1101 (70.2) | 146 (59.3) | 309 (66.8) | 0.002 | 764 (57.4) | 238 (55.5) | 519 (53.7) | 0.214 |
| Obstructive sleep apnea syndrome, n (%) | 56 (3.6) | 12 (4.7) | 14 (3.1) | 0.302 | 99 (7.4) | 36 (8.4) | 66 (6.8) | 0.585 |
| Cancer at any site, n (%) | 36 (2.3) | 9 (3.5) | 15 (3.2) | 0.364 | 34 (2.6) | 11 (2.6) | 33 (3.4) | 0.437 |
| Neurodegenerative disease, n (%) | 23 (1.5) | 2 (0.7) | 4 (0.8) | 0.646 | 20 (1.5) | 7 (1.6) | 7 (0.7) | 0.186 |
MEDAS: Mediterranean Diet Adherence screener; SD: standard deviation. a 1 serving = 30 g. b P-values were obtained through the chi-squared test for categorical variables and the t-test for continuous variables. c Less active (inactive or moderately inactive) during leisure time according to the EPIC-physical activity questionnaire.
Cross-sectional analyses
As presented in Table 3, the total prevalent cases of depression were 233 (10.2%) in SE-I and 407 (14.9%) in SE-II.
Table 3.
Cross-sectional associations between nut consumption and depression in the Seniors-ENRICA-I and Seniors-ENRICA-II studies
| Cross-sectional analysis | Seniors-ENRICA-I | Seniors-ENRICA-II | ||
|---|---|---|---|---|
| Total n (prevalent cases of depression) | Odds ratio (95% CI) | Total n (prevalent cases of depression) | Odds ratio (95% CI) | |
| Total | 2278 (233) | 2726 (407) | ||
| Crude model | ||||
| Nut consumption (servingsa per week) | ||||
| 0 to < 1 | 1568 (156) | 1.00 | 1331 (226) | 1.00 |
| 1 to < 3 | 247 (30) | 1.27 (0.80, 2.03) | 429 (60) | 0.80 (0.58, 1.08) |
| ≥ 3 | 463 (47) | 1.03 (0.68, 1.55) | 966 (121) | 0.70 (0.55, 0.89)* |
| p-for-trend | 0.832 | 0.014 | ||
| Adjusted model 1 | ||||
| Nut consumption (servingsa per week) | ||||
| 0 to < 1 | 1568 (156) | 1.00 | 1331 (226) | 1.00 |
| 1 to < 3 | 247 (30) | 1.27 (0.79, 2.04) | 429 (60) | 0.78 (0.57, 1.07) |
| ≥ 3 | 463 (47) | 1.06 (0.69, 1.61) | 966 (121) | 0.74 (0.58, 0.94)* |
| p-for-trend | 0.951 | 0.062 | ||
| Adjusted model 2 | ||||
| Nut consumption (servingsa per week) | ||||
| 0 to < 1 | 1568 (156) | 1.00 | 1331 (226) | 1.00 |
| 1 to < 3 | 247 (30) | 1.33 (0.80, 2.20) | 429 (60) | 0.83 (0.60, 1.16) |
| ≥ 3 | 463 (47) | 1.33 (0.85, 2.08) | 966 (121) | 0.86 (0.66, 1.12) |
| p-for-trend | 0.278 | 0.545 | ||
Model 1 adjusted for sex, age, educational level and living alone. Model 2 included the variables of Model 1 and smoking status, alcohol consumption, sleep duration, physical inactivity, body mass index, MEDAS score, diabetes mellitus, cardiovascular disease, hypertension, obstructive sleep apnea syndrome, cancer at any site, and neurodegenerative disease. a 1 serving = 30 g. * p <0.05.
The consumption of nuts was not associated with the frequency of depression in the SE-I in the crude and adjusted analyses. On the other hand, in the SE-II, consuming ≥3 versus <1 serving per week of nuts was associated with a lower likelihood of depression in the crude model (OR=0.70; 95% CI: 0.55, 0.89) and in Model 1, which was adjusted for sociodemographic covariates (OR=0.74; 95% CI: 0.58, 0.94). This association lost statistical significance when lifestyle and health condition covariates were adjusted in Model 2 (OR=0.86; 95% CI: 0.66, 1.12).
In the meta-analysis of the cross-sectional data from both studies, although we found a lower frequency of depression associated with nut intake, statistical significance was not achieved (Figure 2); compared to consuming <1 serving of nuts/week, the odds ratio (95% CI) for depression was 0.90 (0.64, 1.16) for consuming 1 to <3 servings/week and 0.92 (0.70, 1.13) for consuming ≥3 servings/week. No cross-sectional association was observed in the results of the participant-level pooled analysis after adjusting for sex, age, study level and living alone (Table S1, Supplementary Material). Likewise, no association with nut consumption was observed in the complete adjusted model (Model 2) in complementary analyses considering the use of antidepressants as the dependent variable (Table S2, Supplementary Material) and when the analyses were stratified by sex and age group (Table S3, Supplementary Material).
Figure 2.
Forest plots of the meta-analyses of two cohort studies for the cross-sectional and longitudinal associations between nut consumption and depression in the Seniors-ENRICA-I (SE-I) and Seniors-ENRICA-II (SE-II)
The odds ratios (ORs) were adjusted for sex, age, educational level, living alone, smoking status, alcohol consumption, sleep duration, physical inactivity, body mass index, MEDAS score, diabetes mellitus, cardiovascular disease, hypertension, obstructive sleep apnea syndrome, cancer at any site, and neurodegenerative disease.
Longitudinal analyses
As shown in Table 4, the total number of incident cases of depression was 108 (7.0%) in SE-I and 74 (4.7%) in SE-II. The OR of depression was lower among those consuming ≥ 3 versus < 1 serving/week of nuts both in SE-I (OR=0.60; 95% CI: 0.30, 1.20) and in SE-II (OR=0.72; 95% CI: 0.41, 1.28) in adjusted Model 2.
Table 4.
Longitudinal associations between nut consumption and depression in the Seniors-ENRICA-I and Seniors-ENRICA-II studies
| Longitudinal analysis | Seniors-ENRICA-I | Seniors-ENRICA-II | ||
|---|---|---|---|---|
| Total n (incident cases of depression) | Odds ratio (95% CI) | Total n (incident cases of depression) | Odds ratio (95% CI) | |
| Total | 1534 (108) | 1566 (74) | ||
| Crude model | ||||
| Nut consumption (servingsa per week) | ||||
| 0 to < 1 | 1040 (81) | 1.00 | 729 (42) | 1.00 |
| 1 to < 3 | 171 (13) | 1.01 (0.51, 2.04) | 243 (11) | 0.78 (0.39, 1.53) |
| ≥ 3 | 323 (14) | 0.53 (0.27, 1.04) | 594 (21) | 0.60 (0.35, 1.02) |
| p-for-trend | 0.475 | 0.090 | ||
| Adjusted model 1 | ||||
| Nut consumption (servingsa per week) | ||||
| 0 to < 1 | 1040 (81) | 1.00 | 729 (42) | 1.00 |
| 1 to < 3 | 171 (13) | 0.99 (0.49, 2.00) | 243 (11) | 0.79 (0.40, 1.56) |
| ≥ 3 | 323 (14) | 0.51 (0.26, 1.02) | 594 (21) | 0.65 (0.38, 1.11) |
| p-for-trend | 0.477 | 0.161 | ||
| Adjusted model 2 | ||||
| Nut consumption (servingsa per week) | ||||
| 0 to < 1 | 1040 (81) | 1.00 | 729 (42) | 1.00 |
| 1 to < 3 | 171 (13) | 1.08 (0.54, 2.19) | 243 (11) | 0.80 (0.40, 1.60) |
| ≥ 3 | 323 (14) | 0.60 (0.30, 1.20) | 594 (21) | 0.72 (0.41, 1.28) |
| p-for-trend | 0.760 | 0.337 | ||
Model 1 adjusted for sex, age, educational level and living alone. Model 2 included the variables of Model 1 and smoking status, alcohol consumption, sleep duration, physical inactivity, body mass index, MEDAS score, diabetes mellitus, cardiovascular disease, hypertension, obstructive sleep apnea syndrome, cancer at any site, and neurodegenerative disease. a 1 serving = 30 g.
In the meta-analysis of longitudinal data, compared to consuming <1 serving of nuts/week, the odds ratio (95% CI) for depression was 0.90 (0.41,1.38) for consuming 1 to < 3 servings/week and 0.66 (0.35, 0.97) for consuming ≥ 3 servings/week (Figure 2). Consistently, similar results were obtained from the participant-level pooled analysis, where compared to consuming < 1 serving of nuts/week, the odds ratio (95% CI) for depression was 0.79 (0.50,1.25) for consuming 1 to < 3 servings/week and 0.62 (0.42, 0.93) for consuming ≥ 3 servings/week (Table S1, Supplementary Material). Last, a longitudinal association with higher nut consumption was observed in the complementary analysis considering the use of antidepressants as the dependent variable (Table S2, Supplementary Material). No association with nut consumption was observed in subgroup analyses by sex; in addition, the OR for depression for consuming ≥ 3 servings/week was 0.44 (0.22, 0.85) among participants aged 72 years and older but no association was observed in those aged less than 72 years (Table S3, Supplementary Material).
Discussion
Main results
This is the first study examining the association between nut consumption and the risk of depression in Spanish older adults. The meta-analysis of the results of two cohort studies supports that consuming ≥ 3 servings/week is prospectively associated with a lower risk of depression among Spanish older adults regardless of relevant confounders, including sociodemographic factors, lifestyle behaviors, BMI, adherence to the Mediterranean diet and several comorbidities. Additionally, the participant-level pooled analysis and sensitivity analyses were consistent with our main findings.
These findings are in line with those of a recent systematic review suggesting that nut-enhanced diets may reduce depression risk and depressive symptoms in the general population (14). For instance, our prospective results are consistent with those from a subanalysis of the SUN cohort (n=15,980), where moderate nut consumers showed a reduced risk of depression over a 10-y follow-up (22). Moreover, in other cross-sectional studies, consuming ≥ 30 g of nuts/week was associated with fewer depressive symptoms in Chinese adults (n=13,626) (23), and walnut consumers of 4 to 24 g/day had lower depression scores than nonnut consumers in the US (n=26,656) (24). The age of the population studied must be considered to compare our results and those from these studies. While their findings are based on individuals or with a mean age of 37.3 years 22, 44 years (23) or 46 years (24), our results correspond to older adults, with a mean age of approximately 70 years. Although our pooled longitudinal analyses showed a significant association between nut consumption and depression, this association was not observed in the separate analysis for each study (i.e., SE-I and SE-II). Since the prevalence of depression increases with age and the level of nut intake is more homogeneous in older individuals than in younger individuals, it might be more difficult to find differences in the frequency of depression according to nut consumption in older individuals. Additionally, the follow-up time of our prospective analyses (3.2 years for SE-I and 2.3 years for SE-II) might be too short to detect a large number of incident cases of depression, which ensures sufficient statistical power to confidently assess the study association. Furthermore, although it could be speculated that nuts could have a preventative effect on depression earlier in life, the current study cannot support this possibility because it is based solely on self-reported nut consumption from the past year. Therefore, trajectory-based studies with long follow-up are needed to draw solid conclusions on the association between nut and depression.
Nut consumption and mental health
Although nuts have been suggested as natural pleiotropic nutraceuticals, mainly due to their effects on ameliorating cardiovascular risk factors and improving cardiovascular prognosis (12), there is a remarkable lack of literature on the effects of nuts on mental health conditions. This is even greater for older adults, in whom nutritional strategies are a feasible approach to maintain healthy aging, avoiding age-related issues such as frailty, sarcopenia, increased risk for chronic diseases, impairments in cognition or depressive disorders (26, 27). In addition to the increased risk of depression in older adults compared with younger adults, nut intake might counteract age-related diseases (28), assisting them in meeting potential micronutrient insufficiencies during aging (i.e., vitamin B6, vitamin E, iron, zinc, etc.) (27). Recently, we found that nut consumption (≥15 g/day) was associated with half the risk of impaired agility and mobility in males and with a lower risk of overall function impairment in females in comparison with nonconsumers (29). In addition, some studies have shown that high adherence to healthy dietary patterns characterized by moderate consumption of nuts (15~30 g/d) might help with these age-related issues by improving cognitive performance (9, 30, 31, 32). Moreover, antioxidants found in nuts and seeds, such as lignans and flavonoids, have been associated with lower depression or anxiety scores (33, 34).
Possible biological pathways
Some biological paths have been proposed for the potential benefit of nuts on mental health (12). They are related to diminished oxidative stress and inflammation (interleukin-1β and interferon-γ) 34 due to the nutrient profile of nuts (i.e., vitamins, antioxidants, polyunsaturated fatty acids, etc.) (35), which could result in neuroprotective effects. Nuts are rich in amino acids, including arginine, and lower levels of this amino acid have been associated with depression (26). Moreover, fiber and nutrients from nuts could optimize the gut microbiota, which may lead to lower depressive symptoms and a better mood state through the microbiome-gut-brain axis (36, 37). Additionally, increased production of essential neurotransmitters (gamma-aminobutyric acid and serotonin) could contribute to fewer depressive symptoms attributed to nut consumption (7, 38).
Study limitations
Our study has some limitations. First, nut consumption was self-reported, so some degree of measurement error due to recall and information biases may exist. Second, the results of the longitudinal analyses include few cases in the higher nut consumption categories (1 to <3 and ≥ 3 servings per day), so the lack of associations in separate analyses for each study may be due to limited statistical power. Third, due to the limited intake of most types of nuts, we could not elucidate the association between specific types of nuts (e.g., almond, walnuts, etc.) and depression. Fourth, although our analyses were adjusted for many confounders, residual confounding cannot be completely ruled out, as in all observational studies. Finally, our results were obtained exclusively from baseline data on nut consumption, so they assume that nut consumption was maintained during follow-up.
Conclusion
In conclusion, based on the meta-analyses of two cohorts of older adults in Spain, our results support an association between nut consumption and a reduced risk of depression. Given the established benefits of nuts on cardiovascular and metabolic health and body composition, our results on the lower depression risk associated with these foods add evidence to consume nuts as part of a healthy diet in older adults.
Acknowledgments
We thank all participants in the Seniors-ENRICA studies.
Supplementary Material
Supplementary material is available for this article at https://doi.org/10.1007/s12603-023-1927-8 and is accessible for authorized users.
Supplementary material, approximately 24.7 KB.
Author contributions: Rubén Fernández-Rodríguez contributed to the conception and design, analysis and interpretation of data, and drafting and revision of the paper for important intellectual content. Rosario Ortolá contributed to acquisition of data and drafting and revision of the paper for important intellectual content. Vicente Martínez-Vizcaíno contributed to the conception and design and the drafting and revision of the paper for important intellectual content. Bruno Bizzozero-Peroni contributed to drafting and revision of the paper for important intellectual content. Fernando Rodríguez-Artalejo contributed to the conception and design, acquisition of data, and drafting and revision of the paper for important intellectual content. Esther García-Esquinas contributed to acquisition of data and drafting and revision of the paper for important intellectual content. Esther López-García contributed to acquisition of data and drafting and revision of the paper for important intellectual content. Arthur Eumann Mesas contributed to the conception and design, analysis and interpretation of data, and drafting and revision of the paper for important intellectual content. All authors approved the final approval of the manuscript.
Funding: This work was supported by the Fondo de Investigación Sociosanitaria (FIS) grants 19/319 (State Secretary of R + D + I and FEDER/FSE), the REACT EU Program, Comunidad de Madrid and the European Regional Development Fund (ERDF) - European Union: FACINGLCOVID-CM project, and FEDER funds from the European Union (CB16/10/00477). A.E.M. is financially supported by a ‘Beatriz Galindo' contract (BEAGAL18/00093) from the Spanish Ministry of Education, Culture and Sport. B. B.-P. is supported by a grant from the Universidad de Castilla-La Mancha co-financed by the European Social Fund (2020-PREDUCLM-16746). R.F.-R. is supported by a grant from the Spanish Ministry of Education, Culture and Sport (FPU 19/00167). The funders played no role in the study design, data collection or analysis, decision to publish, or preparation of the manuscript.
Data availability: Data described in the manuscript, code book, and analytic code will be made available upon request pending application and approval.
Statement of Ethics: Both the SE-I and SE-II studies were approved by the Clinical Research Ethics Committee of the La Paz University Hospital in Madrid, and participants provided informed written consent.
Conflicts of interest: The authors declare no conflicts of interest.
References
- 1.Mental GBD. Collaborators D. Global, regional, and national burden of 12 mental disorders in 204 countries and territories, 1990–2019: a systematic analysis for the Global Burden of Disease Study 2019. The Lancet Psychiatry. 2022;9(2):137–150. doi: 10.1016/S2215-0366(21)00395-3. 10.1016/S2215-0366(21)00395-3 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Daly M, Sutin A, Robinson E. Longitudinal changes in mental health and the COVID-19 pandemic: Evidence from the UK Household Longitudinal Study. Psychol Med. 2020;(01). 10.1017/S0033291720004432 [DOI] [PMC free article] [PubMed]
- 3.Rodríguez-Gómez I, Sánchez-Martín C, García-García F, et al. The medium-term consequences of COVID-19 lockdown on lifestyle among Spanish older people with hypertension, pulmonary, cardiovascular, and musculoskeletal-diseases, depression, and cancer. Epidemiol Heal. 2022;Feb 21:e20. 10.4178/epih.e2022026 [DOI] [PMC free article] [PubMed]
- 4.Institute of Health Metrics and Evaluation. Global Health Data Exchange (GHDx). Published 2019. Accessed March 28, 2019. http://ghdx.healthdata.org/gbd-results-tool?params=gbd-api-2019-permalink/d780dffbe8a381b25e1416884959e88b
- 5.Zenebe Y, Akele B, W/Selassie M, Necho M. Prevalence and determinants of depression among old age: a systematic review and meta-analysis. Ann Gen Psychiatry. 2021;20(1):1–19. doi: 10.1186/s12991-021-00375-x. 10.1186/s12991-021-00375-x [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Patel V, Chisholm D, Parikh R, et al. Addressing the burden of mental, neurological, and substance use disorders: Key messages from Disease Control Priorities, 3rd edition. Lancet. 2016;387(10028):1672–1685. doi: 10.1016/S0140-6736(15)00390-6. 10.1016/S0140-6736(15)00390-6 PubMed PMID: 26454360. [DOI] [PubMed] [Google Scholar]
- 7.Sarris J, Logan AC, Akbaraly TN, et al. International Society for Nutritional Psychiatry Research consensus position statement: Nutritional medicine in modern psychiatry. World Psychiatry. 2015;14(3):370–371. doi: 10.1002/wps.20223. 10.1002/wps.20223 PubMed PMID: 26407799; PMCID 4592666. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Estruch R, Ros E, Salas-Salvadó J, et al. Primary Prevention of Cardiovascular Disease with a Mediterranean Diet Supplemented with Extra-Virgin Olive Oil or Nuts. N Engl J Med. 2018;378(25):e34. doi: 10.1056/NEJMoa1800389. 10.1056/NEJMoa1800389 PubMed PMID: 29897866. [DOI] [PubMed] [Google Scholar]
- 9.Nishi SK, Babio N, Gómez-Martínez C, et al. Mediterranean, DASH, and MIND Dietary Patterns and Cognitive Function: The 2-Year Longitudinal Changes in an Older Spanish Cohort. Front Aging Neurosci. 2021;13(December). 10.3389/fnagi.2021.782067 [DOI] [PMC free article] [PubMed]
- 10.Van Den Brink AC, Brouwer-Brolsma EM, Berendsen AAM, Van De Rest O. The Mediterranean, Dietary Approaches to Stop Hypertension (DASH), and Mediterranean-DASH Intervention for Neurodegenerative Delay (MIND) Diets Are Associated with Less Cognitive Decline and a Lower Risk of Alzheimer's Disease-A Review. Adv Nutr. 2019;10(6):1040–1065. doi: 10.1093/advances/nmz054. 10.1093/advances/nmz054 PubMed PMID: 31209456; PMCID 6855954. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Shafiei F, Salari-Moghaddam A, Larijani B, Esmaillzadeh A. Adherence to the mediterranean diet and risk of depression: A systematic review and updated metaanalysis of observational studies. Nutr Rev. 2019;77(4):230–239. doi: 10.1093/nutrit/nuy070. 10.1093/nutrit/nuy070 PubMed PMID: 30726966. [DOI] [PubMed] [Google Scholar]
- 12.Ros E, Singh A, O'keefe JH N. Natural pleiotropic nutraceuticals. Nutrients. 2021;13(9):1–28. doi: 10.3390/nu13093269. 10.3390/nu13093269 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Theodore LE, Kellow NJ, Mcneil EA, Close EO, Coad EG, Cardoso BR. Nut Consumption for Cognitive Performance: A Systematic Review. Adv Nutr. 2021;12(3):777–792. doi: 10.1093/advances/nmaa153. 10.1093/advances/nmaa153 PubMed PMID: 33330927. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Fernández-Rodríguez R, Jiménez-López E, Garrido-Miguel M, Martínez-Ortega IA, Martínez-Vizcaíno V, Mesas AE. Does the evidence support a relationship between higher levels of nut consumption, lower risk of depression, and better mood state in the general population? A systematic review. Nutr Rev. Published online April 7, 2022:nuac022. 10.1093/nutrit/nuac022 [DOI] [PubMed]
- 15.Ren M, Zhang H, Qi J, et al. An almond-based low carbohydrate diet improves depression and glycometabolism in patients with type 2 diabetes through modulating gut microbiota and glp-1: A randomized controlled trial. Nutrients. 2020;12(10):1–21. doi: 10.3390/nu12103036. 10.3390/nu12103036 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Rodríguez-Artalejo F, Graciani A, Guallar-Castillón P, et al. [Rationale and methods of the study on nutrition and cardiovascular risk in Spain (ENRICA)] Rev Esp Cardiol. 2011;64(10):876–882. doi: 10.1016/j.recesp.2011.05.019. 10.1016/j.recesp.2011.05.019 PubMed PMID: 21821340. [DOI] [PubMed] [Google Scholar]
- 17.Ortolá R, García-Esquinas E, Sotos-Prieto M, et al. Mediterranean Diet and Changes in Frequency, Severity, and Localization of Pain in Older Adults: The Seniors-ENRICA Cohorts. J Gerontol A Biol Sci Med Sci. 2022;77(1):122–130. doi: 10.1093/gerona/glab109. 10.1093/gerona/glab109 PubMed PMID: 33839765. [DOI] [PubMed] [Google Scholar]
- 18.Guallar-Castillón P, Sagardui-Villamor J, Balboa-Castillo T, et al. Validity and reproducibility of a Spanish dietary history. PLoS One. 2014;9(1):1–11. doi: 10.1371/journal.pone.0086074. 10.1371/journal.pone.0086074 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Cabanas-Sánchez V, Esteban-Cornejo I, Migueles JH, et al. Twenty four-hour activity cycle in older adults using wrist-worn accelerometers: The seniors-ENRICA-2 study. Scand J Med Sci Sport. 2020;30(4):700–708. doi: 10.1111/sms.13612. 10.1111/sms.13612 [DOI] [PubMed] [Google Scholar]
- 20.Pols MA, Peeters P, Ocke M, Slimani N, Bueno-de-Mesquita HB, Collette HJA. Estimation of reproducibility and relative validity of the questions included in the EPIC Physical Activity Questionnaire. Int J Epidemiol. 1997;26 Suppl 1:S181—9. 10.1093/ije/26.suppl_1.S181 [DOI] [PubMed]
- 21.Schröder H, Fitó M, Estruch R, et al. A short screener is valid for assessing Mediterranean diet adherence among older Spanish men and women. J Nutr. 2011;141(6):1140–1145. doi: 10.3945/jn.110.135566. 10.3945/jn.110.135566 PubMed PMID: 21508208. [DOI] [PubMed] [Google Scholar]
- 22.Fresán U, Bes-Rastrollo M, Segovia-Siapco G, et al. Does the MIND diet decrease depression risk? A comparison with Mediterranean diet in the SUN cohort. Eur J Nutr. 2019;58(3):1271–1282. doi: 10.1007/s00394-018-1653-x. 10.1007/s00394-018-1653-x PubMed PMID: 29516224. [DOI] [PubMed] [Google Scholar]
- 23.Su Q, Yu B, He H, et al. Nut Consumption Is Associated With Depressive Symptoms Among Chinese Adults. Depress Anxiety. 2016;33(11):1065–1072. doi: 10.1002/da.22516. 10.1002/da.22516 PubMed PMID: 27097046. [DOI] [PubMed] [Google Scholar]
- 24.Nutrients, 2019, 11, 2.
- 25.DerSimonian R, Kacker R. Random-effects model for meta-analysis of clinical trials: an update. Contemp Clin Trials. 2007;28(2):105–114. doi: 10.1016/j.cct.2006.04.004. 10.1016/j.cct.2006.04.004 PubMed PMID: 16807131. [DOI] [PubMed] [Google Scholar]
- 26.Rusu ME, Mocan A, Ferreira ICFR, Popa DS. Health benefits of nut consumption in middle-aged and elderly population. Antioxidants. 2019;8(8):1–37. doi: 10.3390/antiox8080302. 10.3390/antiox8080302 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.Tan SY, Tey SL, Brown R. Can nuts mitigate malnutrition in older adults? A conceptual framework. Nutrients. 2018;10(10):1–15. doi: 10.3390/nu10101448. 10.3390/nu10101448 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.Tan SY, Tey SL, Brown R. Nuts and older adults' health: A narrative review. Int J Environ Res Public Health. 2021;18(4):1–26. doi: 10.3390/ijerph18041848. 10.3390/ijerph18041848 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29.Arias-Fernández L, Machado-Fragua MD, Graciani A, et al. Prospective Association Between Nut Consumption and Physical Function in Older Men and Women. J Gerontol A Biol Sci Med Sci. 2019;74(7):1091–1097. doi: 10.1093/gerona/gly171. 10.1093/gerona/gly171 PubMed PMID: 30052782. [DOI] [PubMed] [Google Scholar]
- 30.Bishop NJ, Zuniga KE. Investigating walnut consumption and cognitive trajectories in a representative sample of older US adults. Public Health Nutr. 2021;24(7):1741–1752. doi: 10.1017/S1368980020001287. 10.1017/S1368980020001287 PubMed PMID: 32618237. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31.Rakic JM, Tanprasertsuk J, Scott TM, et al. Effects of daily almond consumption for six months on cognitive measures in healthy middle-aged to older adults: a randomized control trial. Nutr Neurosci. Published online 2021. 10.1080/1028415X.2020.1868805 [DOI] [PubMed]
- 32.Tan SY, Georgousopoulou EN, Cardoso BR, Daly RM, George ES. Associations between nut intake, cognitive function and non-alcoholic fatty liver disease (NAFLD) in older adults in the United States: NHANES 2011-14. BMC Geriatr. 2021;21(1):1–12. doi: 10.1186/s12877-021-02239-1. 10.1186/s12877-021-02239-1 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 33.Jia S, Hou Y, Wang D, Zhao X. Flavonoids for depression and anxiety: a systematic review and meta-analysis. Crit Rev Food Sci Nutr. Published online April 9, 2022:1-11. 10.1080/10408398.2022.2057914 [DOI] [PubMed]
- 34.Cofán M, Rajaram S, Sala-Vila A, et al. Effects of 2-Year Walnut-Supplemented Diet on Inflammatory Biomarkers. J Am Coll Cardiol. 2020;76(19):2282–2284. doi: 10.1016/j.jacc.2020.07.071. 10.1016/j.jacc.2020.07.071 PubMed PMID: 33153588. [DOI] [PubMed] [Google Scholar]
- 35.Bolling BW, Chen CYO, McKay DL, Blumberg JB. Tree nut phytochemicals: Composition, antioxidant capacity, bioactivity, impact factors. A systematic review of almonds, Brazils, cashews, hazelnuts, macadamias, pecans, pine nuts, pistachios and walnuts. Nutr Res Rev. 2011;24(2):244–275. doi: 10.1017/S095442241100014X. 10.1017/S095442241100014X PubMed PMID: 22153059. [DOI] [PubMed] [Google Scholar]
- 36.Maslowski KM, Vieira AT, Ng A, et al. Regulation of inflammatory responses by gut microbiota and chemoattractant receptor GPR43. Nature. 2009;461(7268):1282–1286. doi: 10.1038/nature08530. 10.1038/nature08530 PubMed PMID: 19865172; PMCID 3256734. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 37.Simpson CA, Diaz-Arteche C, Eliby D, Schwartz OS, Simmons JG, Cowan CSM. The gut microbiota in anxiety and depression–A systematic review. Clin Psychol Rev. 2021;83:101943. doi: 10.1016/j.cpr.2020.101943. 10.1016/j.cpr.2020.101943 PubMed PMID: 33271426, July 2020. [DOI] [PubMed] [Google Scholar]
- 38.Swann OG, Kilpatrick M, Breslin M, Oddy WH. Dietary fiber and its associations with depression and inflammation. Nutr Rev. 2020;78(5):394–411. doi: 10.1093/nutrit/nuz072. 10.1093/nutrit/nuz072 PubMed PMID: 31750916. [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Supplementary material, approximately 24.7 KB.