Dietary Intake is Positively Associated with Cognitive Function of a Chinese Older Adults Sample

Abstract

Objective

The relationship between cognitive function and dietary intake in older adults was understudied in China. This study examined this relationship in a Chinese sample while controlling for the effects of sleep quality and socio demographic confounders.

Methods

The sample consisted of 340 Chinese older adults (age > 60) who were randomly selected from Wuhan city in central China. Cognitive function was assessed by the Mini-mental State Examination [MMSE], sleep quality by the Pittsburgh Sleep Quality Index [PSQI], and dietary intake by frequencies of intake of meat products, fruits, fish/seafood/aquatic products, nuts and mushroom/algae over the past year. First, confirmatory factor analysis (CFA) was conducted to evaluate the measurement properties of cognitive function, dietary intake, and sleep quality. Second, structural equation modeling (SEM) was conducted to examine the relations of cognitive function, dietary intake, and sleep quality.

Results

Dietary intake was found to be positively related to cognitive function. Older age, lower education status, monthly income, and living alone or without a spouse were significantly associated with poorer cognitive function. SES status had an indirect effect on cognitive function via dietary intake.

Conclusion

Dietary intake may be critical to maintain normal cognitive function of older adults in China.

Introduction

The relationship between dietary intake and cognitive function in older adults has been studied in two lines: the role of a single dietary nutrient and the role of dietary patterns. In the former line, fruits, meat, aquatic products, mushrooms, and nuts have been found to be important in improving cognitive function (1, 2, 3, 4, 5), but this finding is unclear in Chinese older adults. A population-based cross-sectional study reported that education and WHO recommendations for fruit and vegetable intake are associated with better cognitive function in the Brazilian elderly population (6). Several epidemiological studies found the association of seafood intake with lower risk of cognitive decline (7, 8, 9). A dose-response meta-analysis of 21 cohort studies indicated that fishery products are recommended as dietary sources and are associated with lower risks of cognitive impairment (10). A population-based prospective cohort study showed that higher long-term total nut intake was related to better average cognitive state for all cognitive (11). Mushrooms are a rare source of ergothioneine as well as fiber, selenium, and several other minerals and vitamins. Some preclinical and clinical studies showed that intake of mushrooms influenced cognition (12, 13). In spite of these findings, it was unknown to what extent such nutrients are related to cognitive function in Chinese older adults, as few studies have been done to examine this in China (14, 15, 16).

Effects of meat on cognitive function were inconsistent in previous studies. Some studies found a negative relationship. For example, Bajerska J et al. (17) observed that the consumption of red meat and meat products was negatively related to executive and global cognitive function. A casecontrol study reported that high consumption in red meat might contribute to the cognitive impairment in older Chinese adults (18). According to research published in the Journal of the American College of Nutrition, rearchers used dietary data from 10 countries and assessed Alzheimer's disease risk for several dietary factors and found that consumption of meat increased disease risk the most (19). There is also emerging evidence that protein meat diet and its constituent amino acids are related to cognition in older adults (20, 21, 22). There are some possible explanations for conflicting results, such as dietary assessments, the outcome measures, residual confounding, possible over-adjustment and different covariates, as well as the type of meat consumed, method of cooking, and portion size. In sum, it was unknown how these important nutrients were related to the cognitive function in older adults in China, as relevant studies were still scant.

The other line of research focused on diet pattern as a preventive measure of age-related cognitive decline. Dietary pattern refers to quantities, proportions, variety, and the frequency of habitually consumed diet. Priori patterns are defined by the adherence to a pre-defined healthy diet using a score such as the Healthy Diet Indicator (HDI) (23), etc. Posteriori dietary patterns are defined independently of the existing nutrition-disease knowledge. Posteriori patterns were inconsistently found to be related to dementia and cognitive decline. One review indicated a relation between dietary pattern and cognitive function (24). Of the cross-sectional studies, one study observed that a «healthy» cluster characterized by higher consumption of fish in men and fruits and vegetables in women had significantly lower mean number of errors in MMSE (25). The other study indicated that higher «vegetables-fruits» and «snacks-drinks-milk products» pattern scores were associated with reduced risk of cognitive impairment in Chinese older women but not in men (26). More studies are therefore needed to corroborate the linkage of dietary intake to the cognitive function of older adults in China.

Sleep duration and quality, in addition to dietary intake, have also been inconsistently found related to cognitive function. A cross-sectional study of sleep quantity and quality and Amnestic (ACF) and Non-Amnestic (nACF) cognitive function in an aging population reported that the associations between sleep disturbances and cognitive function vary between younger (50–64 years) and older adults (age > 65). The relationship was significant between subjective sleep quality and both ACF and nACF in the older age group, whereas not in the younger age group. At any level of sleep duration, older adults had a steeper correlation between sleep quality and amnestic (memory, ACF) than the younger group (27). A longitudinal study used samples of older adults (≥ 50 years old) in 6 middle-income countries (China, Ghana, India, Russian Federation, South Africa, and Mexico) from non-Western societies (28), and found that selfreported sleep quality was positively related to a cognitive test z-score (p < .05). Thus, sleep quality seemed to be associated with cognitive function in older adults.

As far as we were aware, there were few studies about the relationship between sleep and cognition in China. One epidemiological study from Chinese older adults reported that both short (3–4 hours) and long (>10 hours) self-reported sleep duration were independently correlated to memory impairment measured by a delayed word recall test (29). Another longitudinal study of older adults in a community over a 1-year period indicated that self-reported poor quality measured using the Pittsburgh Sleep Quality Index (PSQI) predicted the decline of cognitive function measured with the Chinese version of Mini-Mental State Examination (CMMSE) (30). In addition, poor sleep efficiency and short sleep duration were related to greater CMMSE decline and a higher risk of developing substantial CMMSE decline. Thus, sleep quality should be controlled when examining the relations between dietary intake and cognitive function.

We hypothesize that sufficient nutrition intake is critical to maintain normal cognitive functions in the older adults. This study was thus aimed to examine the associations of dietary intake of a few nutrients and cognitive function of an older adults sample in China, while controlling for sleep quality and certain demographic variables related to cognitive function. Our hypothesized model was depicted in Figure 1. We also explored if the effect of dietary intake on CMMSE depended on the level of BMI, namely, an interaction or moderating effect.

Figure 1.

Relations among dietary intake, sleep quality and cognitive function

Methods

Sample

The study was carried out between October 2015 and January 2016 after approval by the institutional review board and local ethics committee of Hubei University of Chinese Medicine. A simple random sample was selected from the senior population of Wuhan in central China. Those with speech disorders and severe vision or hearing deficits were excluded from the study. Participants with their written consent were interviewed for data collection. Utmost care was taken to ensure their privacy and confidentiality.

The sample for this study consisted of 339 elderly people (age ≥ 60), with a mean age of 70.8 and a standard deviation of 7.9. About 37.6% were males and 62.4% females. The mean body mass index (BMI) was 23.4 (SD = 3.4). About 66.9% percent of the participants had a spouse while 33.1% did not at the time of the study. About 15.3% of the participants lived alone. Education levels of the participants ranged from elementary to college degrees, 33.0% with high school or higher education and 67.0% with less than high school education. More than half (68.5%) of participants had monthly income < ¥3000 (Chinese currency). The frequencies and percentages that the participants consumed fruits, meat, fish/seafood/aquatic products, mushrooms, and nuts over the past year are listed in Table 1 below.

Table 1.

Frequency & Percentage (%) of Participants’ Dietary Intake

Times/ Week	Meat Products	Fish/Seafood/ Aquatic Products	Fruits	Nuts	Mushrooms/ Algae
0	9.0	17.9	15.2	34.4	16.4
1-2	40.7	41.7	26.7	37.2	58.2
3-4	11.8	17.2	7.5	8.4	14.5
5-6	5.3	7.5	4.7	3.1	5.0
≥ 7	33.2	15.7	46.0	16.9	6.0

Measurements

Questionnaires used in this study were translated into Chinese and checked for semantic accuracy with backtranslation to English.

Dietary Intake

Dietary intake was evaluated using “Health monitoring questionnaire (the diet section) of State Administration of Traditional Chinese Medicine. Including food-intake frequency (meat products, poultry, nuts, mushroom/algae food, fish/other aquatic, vegetable, fruit, bean products, and dairy products). Although “meat products were vaguely defined in the questionnaire, it typically include pork, beef, and mutton cooked at home or processed products from market in China. Participants of the study were inquired about what dietary food they had consumed at what frequency over the past year. Their answers were recorded by selecting one of the following choices: 1 (rarely or never), 2 (1-2 times per week), 3 (3-4 times per week), 4 (5-6 times per day), and 5 (every day).

Cognitive Function

Participants' cognitive function was evaluated by the Mini- Mental State Examination (MMSE), which is a widely used, standardized screening test that takes less than 20 minutes to complete. The sum of items scores weighted at the dimension levels is commonly used for screening or analysis. Specifically, the maximum score of 30 points consists of different cognitive domains: orientation of three words (3 points), subtracting serial sevens from 100 (5 points), language testing by naming objects (2 points), repeating a sentence (1 point), and comprehension tested by complying with a three-step command (3 points) and copying a spatially arranged design of figures (1 point). However, cognitive function in this study is specified as a latent variable in the measurement and structural equation model, such that each item is weighted by its factor loading. The advantage is that measurement errors are partitioned out in the modeling, while allowing for item sensitivity in the measurement.

Pittsburgh Sleep Quality Index (PSQI)

The Pittsburgh Sleep Quality Index (PSQI) was used to assess the sleep quality of the participants. The Chinese PSQI is a 19-item self-report instrument for assessing sleep quality and disturbance over the past month before questionnaire administration. Seven component scores (subjective sleep quality, sleep latency, sleep duration, habitual sleep efficiency, sleep disturbances, sleep medication, and daytime dysfunction) were computed from the items. The scores for each component range from 0 (no difficulty) to 3 (severe difficultly). Sleep quality was specified as a latent variable in the model of this study that is measured with the component scores.

Analysis

The analysis was conducted in two stages. First, the dimensional structure of the three scales was examined with confirmatory factor analysis (CFA). Second, a structural equation model (SEM) was estimated. In this model, cognitive function was specified as the dependent variable, dietary intake as independent variable, and sleep quality and demographic variables as covariates (see Figure 1).

The main analysis and modeling were performed using Mplus 7.4. Reported in the result section is the median of each focal parameter distribution, together with its lower and upper limits of the 95% confidence intervals (CI).

Results

Confirmatory Factor Analysis

The factor structure of the cognitive function, diet intake, and sleep quality were tested using confirmatory factor analysis (CFA) and weighted least square estimation. The CFA showed that dietary intake, sleep quality and cognitive function were measured well with the items in this study, and the model fit the data well with comparative fit index (CFI= .966), Tuker- Lewis Index (TLI = .962), and Root Mean Square Error of Approximation (RMSEA = .027). The posterior medians of the standardized factor loadings are listed in Table 2 below. As all the factor loadings were statistically different from zero (p < .01), we have not indicated such significance in the table and in Figure 1.

Table 2.

Standardized Factor Loadings of the CFA and Confidence Intervals

Construct	Subdomains	Item Content	Factor Loading	Upper Limit	Upper Limit
Cognitive Function	Orientation	B1: Tell time and place	.553	.426	.680
	Memory	B2: Immediate memory	.711	.569	.852
		B4: Delayed recall	.485	.376	.595
	Calculation & Attention	B3: Subtract by seven	.661	.578	.744
	Language	B6: Verbal repetition	.706	.565	.848
		B8: Verbal comprehension	.738	.578	.897
		B9: Write	.824	.734	.914
		B7: Read a sentence	.767	.621	.914
	Constructional Praxis	B10: Design Copying	.714	.617	.810
Dietary Intake		D3: Meat products	.403	.270	.536
		D5: fish/aquatic products/Seafood	.408	.270	.546
		D7: Fruits	.565	.431	.699
		D10: Nuts	.562	.430	.694
		D11: Mushroom /algae food	.557	.424	.690
Sleep Quality		SA: Subjective sleep quality	.802	.677	.926
		SB: Sleep latency	.669	.549	.789
		SC: Sleep duration	.448	.307	.588
		SD: Habitual sleep efficiency	.535	.391	.678
		SE: Sleep disturbances	.530	.401	.659

Relations of the Key Variables

The full SEM model fit the data well (χ2 = 394.595, df = 282, p < .001, Cfi= .923, TLI = .912, and RMSEA = .035). The whole model explained 72% the total variance of MMSE, while dietary intake did 18.3% (γ2). Figure 1 depicts the relations among dietary intake and cognitive functions, while controlling for sleep quality and other demographic variables. In addition, SES had an indirect effect on cognition function via dietary intake. An asterisk indicates a significance of p < .05 of single tail test. The D3 and D5 with low factor loadings indicated that smaller variances in those two variables. Dietary intake was positively related to cognitive function (γ = .428, z = 5.67, p < .001, 95% CI [.280, .576]), implying that one unit increase of overall intake of the nutrients were associated with .428 unit increase in cognitive function.

Some covariates were also found to be significant. An increase in age was associated with .179 unit decrease in cognitive function (γ = -.179, z = -3.35, p < .01, CI [-.283, -.074]). Social economic status (SES) was specified as a latent variable in this model, with income and education as formative indicators (31). The effect of education on SES doubled the effect of income, suggesting that education contributed mainly to the SES construct. One unit increase in SES was associated with .324 unit increase in cognitive function (γ = .324, z = 4.68, p < .01, CI [.188, .459]). Participants living with spouses averaged .503 unit higher in cognitive function than those without (γ = .503, z = 7.24, p < .01, CI [.367, .639]). Participants living alone averaged .234 unit lower in cognitive function than those living with a family member or others (γ = -.234, z = -2.95, p < .01, CI [-.390, -.079]). SES also indirectly affected cognitive function through dietary intake (γ = .460, z = 6.44, p < .01, CI [.320, .601]). However, sleep quality and BMI were not related to cognitive function in this sample. In addition, there were no gender difference in the cognitive function. When the interaction between BMI and dietary intake was added to the model, neither their main nor interaction effect remained significant.

Discussion

The present study applied SEM to evaluate the associations among dietary intake, sleep quality, and cognitive function, while controlling for BMI and other socio demographic confounders in a sample of older adults in Wuhan, China. The main finding was that more consumption (dietary intake) of a cluster of meat, fish/seafood/aquatic products, fruits, nuts, and mushroom/algae food was associated with higher cognitive function, and that cognitive function was adversely affected by other factors, such as older age, living without a spouse or family member, and lower SES. In addition, SES had an indirect effect on cognitive function through dietary intake. However, gender, BMI and sleep quality did not make any difference in cognitive function in this study.

Our findings are consistent with some previous studies that examined the relation of single nutrients with cognitive function, such as fish/aquatic products (32, 33), fruits (34, 35), nuts (36, 37), and mushrooms (38). Meat consumption and adequate dietary protein are recognized as important contributors to the health and function among older adults in Japan (39, 40). The underlying mechanism could be that such nutrients are essential for human bodies to function cognitively and physically at normal levels, while malnourishment in these nutrients could undermine these normal functions. The participants had a normotrophic BMI on average, compared to the normal BMI of 18.5~23.9 of this age group. The effect of Dietary intake on cognitive function was controlled for BMI and not moderated by BMI. This non-significant moderation suggested that dietary intake did not affect participants of a certain range of BMI that might be related to insufficient nutrition. It seemed to be the nutrients that were associated with the MMSE, rather than the amount of calories they reflected.

Inconsistent with other findings was that we did not find a relation between sleep quality and cognitive function. There are some possible reasons for this. First, the sample size of this study was too small to have enough power to detect the relation. Second, sampling error could have occurred such that this sample happened to miss the relationship. Third, the participants might have believed that their sleep problems were caused by age or natural aging, so they might not have been aware or reported that they were experiencing sleep problems. We relied on retrospective chart reviews rather than an objective screening such as polysomnography to collect reliable sleep quality data. Finally, overall study design, covariate adjustments, and statistical methods might have also contributed to the inconsistency.

In this study, we also found that SES had an indirect effect on cognitive function via dietary intake. It may imply that dietary eating habits may also be influenced by SES, education more so than income in this study as indicated by education effect that was twice as much as income. Wall et al. found that higher standard of living, education and nutrition literacy were associated with a healthier dietary eating pattern (41). From Araujo et al. reported that the prevalence of inadequate intake of nutrients declined with increasing income and education levels (42). Therefore, emphasis should be laid on enhancing dietary knowledge that might allow individuals to adopt a healthy diet.

Other variables association with cognitive function in our study included age, being single, living alone, and low SES. The greatest risk factor for cognitive decline is age, but the exact cause is still unknown (43). In recent years, it has become increasingly apparent that SES is an important variable in aging (44, 45). People with higher cumulative SES demonstrated an advantage in cognition (46). Many studies have reported that living as a couple and the quality of the relationship are both positively correlated with some cognitive abilities (47, 48). A rich social network has also been linked to a lower risk of cognitive impairment and dementia (49, 50).

This study has several limitations. First, the small sample size may limit the generalizability of the findings due to potential sampling error. Second, the unclear information on type of meat consumed, portion size, and method of cooking in our sample did not allow for exploring the further relationship. Third, the design of this study was cross-sectional, so causality cannot be firmly established. Further longitudinal studies of larger samples with comprehensive dietary assessment may increase our understanding how dietary intake affects cognitive function over time. Given the correlation between dietary intake and sleep quality, future sleep interventions in controlled studies may attempt to establish whether there is a causal relation between sleep quality and changes in dietary intake in older adults with poor sleep.

This study makes several contributions to previous research, in spite of the above-mentioned limitations. It is one of few studies to explore a dietary intake-cognition relation in individuals of an urban community of China. Moreover, our results have practical implications. Elderly with higher SES, living with a companion and better adherence to a dietary intake of meats, fish/aquatic, seafood, fruits, nuts, and mushroom/algae demonstrate better cognitive function. With this knowledge, community health care professionals should provide appropriate dietary guidance. Moreover, health care providers also need to be mindful of factors contributing to a couple relationship, and intervene to limit the impact of living alone on cognitive function.

Author Contributions

Kaili Sun, Hui Hu, and Chongming Yang conceived and designed the study, and Kaili Sun, Ling Wang and Yating Ai developed the protocol for this study. Kaili Sun, Xinxiu Dong, Yingrong Shi and Mengying Li reviewed that literature. Kaili Sun, Qiaoling Yao and Yu Zhang analyzed the data and reported the results. Kaili Sun, Ling Wang and Yating Ai drafted the manuscript, while Chongming Yang and Hui Hu critically reviewed and revised it. All authors approved the final version for publication and agreed to be accountable for all aspects of the work.

Conflicts of Interest

The authors declare no conflict of interest.

Ethical standards

The study protocol was approved by Institutional Review Board and medical ethics committee of Hubei University of Chinese Medicine. The study was conducted according to the guidelines of the Declaration of Helsinki and with written consents from all the participants.