Intake of Docosahexaenoic Acid-Enriched Milk Beverage Prevents Age-Related Cognitive Decline and Decreases Serum Bone Resorption Marker Levels

The pathogenic mechanism of dementia is still unknown, and the fundamental treatment remains to be established. Thus, there is growing interest in preventing dementia through diet. One of the functional ingredients attracting attention is docosahexaenoic acid. We conducted a 12-month, randomized, double-blind, placebo-controlled clinical trial in healthy elderly Japanese individuals with a Mini-Mental State Examination score of 28 or higher at baseline using a docosahexaenoic acid-enriched milk beverage containing 297 mg docosahexaenoic acid and 137 mg eicosapentaenoic acid. Consumption of a docosahexaenoic acid-enriched milk beverage increased the fatty acid levels of docosahexaenoic acid and eicosapentaenoic acid in erythrocyte membranes, which was the primary outcome of this study. Moreover, intake of this beverage prevented age-related cognitive decline and decreased serum bone resorption marker levels. Our data demonstrate that, even at a low dose, long-term daily intake of docosahexaenoic acid prevents dementia and may show beneficial effect on bone health.


Introduction
The issues surrounding population aging have recently become a matter of great concern. Interest in maintenance of cognitive function is particularly high, and various risk factors for dementia and its symptomatic treatment have been reported 1 . Regarding eating habits, the focus is on the physiological functions of various food materials. Omega-3 polyunsaturated fatty acids ω-3 PUFAs exhibit a variety of physiological activities 2,3 . Notably, the ω-3 PUFA docosahexaenoic acid DHA cannot be adequately synthesized in humans. Burdge and colleagues showed that the conversion rates of DHA from ingested alpha-linolenic acid were 0 -4 in young men 4 and 9 in young women 5 . Thus, oral intake of DHA is necessary. DHA is essential for neuronal morphology and numerous neuronal functions, including brain cognition, learning and memory, and neuropsychology 6 . The brain is the organ with the highest amount of DHA in humans; the amount of DHA in the brain is 11 -20 , whereas the corresponding figure for eicosapentaenoic acid EPA is less than 0.1 7 . Previous reports strongly suggest that DHA intake ameliorated cognitive decline in elderly people and Alzheimer s disease patients 8,9 . Yurko-Mauro et al. 10 determined in a meta-analysis that a high intake of DHA, more than 1 g/day, is necessary to improve cognitive function. Although the importance of DHA intake is fully recognized, the incorporation of DHA into food is difficult due to its very low oxidative stability 11 , and its intake has been limited to supplements such as soft capsules.
The absorption efficiency of DHA is lower than that of other fatty acids 12 . Interestingly, ingestion of foods containing DHA is associated with higher DHA absorption than a soft-capsule supplement form 13 . The absorption of longchain PUFAs such as DHA is enhanced by emulsification 14,15 .
The ingestion of foods containing emulsified DHA could improve cognitive function, even at low doses. Therefore, we previously conducted a 12-month clinical trial using a DHA-enriched milk beverage containing emulsified DHA and found that a daily intake of 297 mg DHA increased the level of DHA in erythrocyte membranes and maintained higher cognitive function in healthy elderly Japanese people 16 .
The association of low bone density and osteoporosis onset with the future risk of dementia and Alzheimer s disease AD development has been investigated 17,18 . Loskutova et al. 19 reported that patients with AD showed a bone mineral density BMD reduction prior to the appearance of symptoms. Thus, bone density is believed to be important for maintaining cognitive function, as well as quality of life. Moreover, some reports showed that intake of fish or fish oil had beneficial effects on BMD 20,21 . Thus, daily intake of a DHA-enriched milk beverage might boost not only cognitive function, but also bone health in patients with AD. However, the effects of ω-3 PUFAs on bone health in healthy elderly Japanese people are still unclear. In this study, we investigated the effects of DHA intake on cognitive function and bone metabolic status using serum-based measures of bone resorption and formation markers in healthy elderly individuals with age-related cognitive decline. Our data clearly showed that daily intake of a milk beverage containing 297 mg DHA prevented age-related cognitive decline and decreased serum bone resorption marker levels.

Materials and Methods
The study was carried out in accordance with the principles of the Declaration of Helsinki and Good Clinical Practice. The protocol was approved by the Ethics Committee of Shimane University approval number 2899 . All participants provided written informed consent before participation. This study is registered in the UMIN Clinical Trial Registry registry number UMIN000031699 .

Participants
A 12-month, randomized, double-blind, placebo-controlled, parallel-arm study was conducted among healthy elderly people in Shimane, Japan. The participants were volunteers who underwent a medical examination that included anthropometry, blood biochemical assessment, and cognitive function and mental health tests. Participants were excluded if there was evidence of a medical disorder including cognitive impairment, renal, respiratory, cardiac, or hepatic disease, diabetes mellitus, and endocrine, metabolic, or hematological disturbances or of the use of any psychotropic drug/supplement that might significantly influence the outcomes of the study or if they had hypersensitivity or allergy to milk beverages.
After screening, 87 participants were randomly divided into two groups: a placebo group n 41 and a DHA group n 46 . The placebo group consumed 200 mL of a milk beverage, whereas the DHA group consumed 200 mL of a milk beverage containing 297 mg DHA C22:6, n-3 and 137 mg EPA C20:5, n-3 . Group allocation was performed by stratified random assignment according to age and sex, as described previously 22 24 . Neither the participants nor the researchers knew which milk beverages were being consumed. The placebo and DHA-enriched milk beverages had Ingredient composition was measured by the following methods: subtraction [100 -{moisture (g) + protein (g) + lipid (g) + ash (g)}] to measure carbohydrates; the combustion method (nitrogen conversion factor = 6.38) to measure protein content; the Reese-Gottlieb method to measure the amount of fat; gas chromatography to measure the amount of DHA and EPA; the direct incineration method to determine the amount of ash content; and the air oven method to determine percent moisture content. Energy was calculated by assuming 4 kcal/g protein, 4 kcal/g carbohydrate, and 9 kcal/g fat.
no differences in appearance, flavor, or aroma. Both milk beverages were produced by Omu Milk Products Co., Ltd. Fukuoka, Japan , whereas the DHA-rich oil PRORARE ® and soybean oil were provided by Fuji Oil Co., Ltd. Osaka, Japan . The compositions of the placebo beverage and the DHA-enriched beverage are shown in Table 1. According to the Japan Gerontology Society, older people who score a 28 or higher on the Mini-Mental State Examination MMSE are considered to maintain healthy cognitive function. Accordingly, participants who recorded an MMSE score of 28 or higher at the initial examination were specifically included in the present analysis placebo group: n 27, DHA group: n 26 Supporting Fig. 1 .

Anthropometry, Body Composition, and Dietary
Intake Analysis Body weight, height, waist circumference, and systolic and diastolic blood pressure were measured by hospital nurses. Body composition was determined by bioelectrical impedance analysis WB-150; Tanita Corporation, Tokyo, Japan . To investigate the effects of intake of the placebo and DHA-enriched milk beverages on health status and lifestyle during the intervention, a general questionnaire was administered that posed lifestyle questions, including those related to medical history.
Dietary intake before and after the interventional study was assessed using the validated Brief-type self-administered Diet History Questionnaire BDHQ for elderly Japanese individuals 25 . The BDHQ is a four-page questionnaire with fixed portion sizes that inquires about the consumption frequency of selected food items to estimate the dietary intake of 58 food and beverage items in the preceding month. The food and beverage items listed in the BDHQ are foods commonly consumed in Japan, which were mainly selected from a food list used in the National Health and Nutrition Survey of Japan. Based on the participants responses to the BDHQ, an ad hoc computer algorithm estimated the amounts of 98 nutritional factors consumed during the previous month.

Blood Sampling
Blood samples were collected at baseline before milk beverage intake and after 12 months of the intervention in the morning or afternoon, after it was confirmed that the participants had not consumed breakfast or lunch. The samples were separated into two sets of tubes: tubes containing a serum isolator for blood biochemical analysis and tubes containing ethylenediaminetetraacetic acid EDTA for fatty acid analysis. After the clinical parameters were measured, the samples were separated into serum, plasma, and erythrocyte red blood cell RBC aliquots and, after the clinical parameters were measured, the remaining serum and RBC samples were stored at 80 within 8 h of collection until analysis.

Blood Biochemical Analysis and Apolipoprotein E4
Measurement Changes in erythrocyte membrane fatty acid composition were the primary outcome and other changes were secondary outcomes of this study UMIN000031699 . DHA and EPA levels in plasma and in the RBC plasma membrane RBC-PM were determined by direct transmethylation 22,23 . Briefly, 100 μL of RBCs was suspended by vigorous mixing of the cells in 1.2 mL of 4 mM EDTA solution containing 0.005 2,6-di-t-butyl-4-methylphenol BHT and centrifugation at 10,000 g for 15 min at 4 . After the supernatant was discarded, the resulting pellet was re-suspended by vigorous mixing in 1 mL Dulbecco s phosphate-buffered saline PBS containing 0.005 BHT and centrifugation at 10,000 g for 15 min at 4 . This procedure was repeated once. The final pellet was homogenized in 250 μL Dulbecco s PBS containing 0.5 saponin and 0.005 BHT by using an ultrasonic homogenizer Bioruptor ® ; Cosmobio Co., Tokyo, Japan and its fatty acid and protein concentrations were measured. The fatty acid concentrations in the plasma and the RBC-PM suspension were measured by capillary gas chromatography of the corresponding methyl esters prepared by transesterification with acetyl chloride. Chromatography was performed using an Agilent 6850A gas chromatograph Agilent Technologies, Santa Clara, CA with a flame ionization detector and an automatic sampler using a 25-m fused-silica column with an internal diameter of 0.25 mm DB-WAX P/N 122-7032; J&W Scientific, Folsom, CA . The peaks were identified by comparison with those of reference compounds and partially by gas chromatography-mass spectrometry.
Plasma total cholesterol, low-and high-density lipoprotein cholesterol, triglyceride, and glucose levels were determined using an automatic analyzer BiOLiS 24; Tokyo Boeki Medical System Ltd., Tokyo, Japan . HbA1c was measured using a kit from TFB, Inc. Tokyo, Japan .
Apolipoprotein E APOE gene mutations were measured as previously described 26 . APOE single nucleotide polymorphism genotyping was carried out by real-time polymerase chain reaction 7900HT; Applied Biosystems, Foster City, CA using the TaqMan single nucleotide polymorphism genotyping assay for rs429358 and rs7412 with identification numbers C___3084793_20 and C____904973_10, respectively Applied Biosystems .

Bone Area Ratio and Serum Bone Metabolic Status
Marker Measurement The bone area ratio was evaluated by quantitative ultrasound performed with an ultrasound bone densitometer Venus-α; Nihon Kohden, Tokyo, Japan . The bone area ratio measured with this instrument is significantly correlated with the BMD of the lumbar vertebrate r 0.77, p 0.01 and of the calcaneus r 0.83, p 0.01 by dual-energy X-ray absorptiometry 25 . The levels of the serum bone metabolic status markers tartrate-resistant acid phosphatase-5b TRACP-5b and bone alkaline phosphatase BAP were measured with a chemiluminescent enzyme immunoassay and an enzyme-linked immunosorbent assay, respectively. Serum calcium and phosphate concentrations were measured with an ion-selective electrode method.

Cognitive Evaluation
Cognitive function was assessed using the MMSE 27 , Hasegawa s Dementia Scale-Revised HDS-R 28 , and the Japanese version of the Montreal Cognitive Assessment MoCA-J 29 , which were listed on the website of the Japan Gerontology Society.

Statistical Analysis
Results are expressed as the mean standard deviation SD . Participant characteristics, cognitive test scores, and erythrocyte plasma membrane fatty acid levels at baseline and 12 months were analyzed by two-way group time repeated-measures analysis of variance ANOVA followed by Tukey s test. Changes ∆ in cognitive test scores, fatty acid levels, and participants parameters between baseline and after 12 months of the intervention were analyzed using Levene s test for equality of variance. Independent uniformly distributed data were compared using Student s t-test and independent non-uniformly distributed data were compared using Mann-Whitney U test. All analyses were performed using SPSS ver. 23.0 IBM, Armonk, NY , and p-values 0.05 were considered significant.

Participants Characteristics
No significant differences were observed between the placebo and DHA groups in any of the participants characteristics at baseline Table 2 . Additionally, no significant differences were found for any values at 12 months compared with baseline in either group Table 2 . Although there was a significant difference between the groups in diastolic blood pressure at 12 months, no significant difference was found in the mean change from baseline to 12 months Table 2 . No significant differences were found between the two groups in the mean change from baseline to 12 months Table 2 .
There were no significant differences between the two groups in dietary intake based on responses to the BDHQ, including all fats and fatty acids, regarding the mean changes from baseline to 12 months Supporting Table 1 , indicating that DHA-enriched milk beverage intake for 12 months did not affect the dietary intake and nutritional values.

Fatty Acid Profiles of the Erythrocyte Plasma Mem-
brane There were no significant differences in any of the values at baseline between the two groups Table 3 . At 12 months, DHA and EPA concentrations and n-6/n-3 and DHA/arachidonic acid AA ratios were significantly higher compared with baseline Table 3 . DHA and EPA concentrations and n-6/n-3 and DHA/AA ratios were significantly higher in the DHA group than in the placebo group Table  3 . Additionally, 12-month intake of a DHA-enriched milk beverage was associated with significantly better mean changes delta 12 months baseline in the DHA group than in the placebo group for AA, EPA, n-6/n-3, DHA/AA, and EPA/AA Table 3 .

Cognitive Function Test Scores
No significant differences in cognitive function test scores were found at baseline between the two groups Table 4 . Because the MMSE total score of the placebo group at 12 months was lower than that of baseline, there were no significant differences in the MMSE total score of the DHA group between 12 months and baseline Table 4 . On the other hand, the MMSE and HDS-R test scores were significantly higher in the DHA group at 12 months than in the placebo group Table 4 . However, only the mean change delta 12 months baseline of the MMSE total score was significantly higher in the DHA group than in the placebo group Table 4 .
In addition to the total score results, analysis of subitems revealed significant greater mean changes delta 12

Bone area ratio and serum bone metabolism markers
At baseline, there were no significant differences between the groups Table 5 . In addition, there were no Values are mean±standard deviation (SD). Placebo group, n = 27; DHA group, n = 26. #, significant difference compared with baseline in each group at p < 0.05. *, significant difference between groups at p < 0.05. †, significant difference between groups by Student' s t-test at p < 0.05. MMSE, Mini-Mental State Examination; HDS-R, Hasegawa' s Dementia Scale-Revised; MoCA-J, the Japanese version of the Montreal Cognitive Assessment. significant differences between the two groups at 12 months Table 5 . Additionally, no significant differences were observed in any values at 12 months compared with baseline in either group Table 5 . Because the BAP of the placebo group showed significant greater mean changes delta 12 months baseline compared with the DHA group, the mean TRACP-5b change of the DHA group was significantly greater than that of the placebo group.

Discussion
In this sub-analysis, we evaluated how 12-month ingestion of a DHA-enriched milk beverage affected cognitive function, the RBC-PM fatty acid profile, the bone area ratio and serum bone metabolism markers, and serum hepatic and renal function biomarkers in healthy elderly Japanese people who recorded an MMSE score of 28 or higher at baseline. Although DHA levels in RBC-PM were not significantly increased, the results showed that 12-month intake of the DHA-enriched milk beverage significantly increased the DHA/AA ratio and protected against age-related cognitive impairment and osteolysis.
Our previous report showed that 12-month intake of a DHA-enriched milk beverage increased DHA levels in the RBC-PM and prevented age-related short-term memory decline, which was categorized as recall. DHA can be efficiently absorbed from DHA-containing foods 13 . Garaiova et al. 30 and Raatz et al. 15 determined that fatty acid absorption was improved by pre-emulsification of ω-3 fatty acids. Although the amount of DHA was lower than previously reported 31, 32 , the efficiency of DHA absorption might be improved by the consumption of an emulsified food that showed high absorption, such as a DHA-enriched milk beverage. Compared with the placebo group, significant mean changes in the MMSE total score, MMSE subitem 5 Recall , HDS-R subitem 2, and MoCA-J subitem 8 Orientation from baseline to 12 months were clearly indicated in the DHA group Figs. 1A-1C . Previous studies with high DHA dosages 860 mg/day or 1720 mg/day also showed that MMSE subitem scores were improved with 12month intake of DHA 26,32 . Additionally, intake of 1700 mg DHA/day for 6 months had positive effects on the recall evaluation items of the MMSE and Alzheimer s Disease Assessment Scale-cognitive component 33 . Orientation evaluation items are included in a variety of test batteries and they are particularly impaired in elderly people and patients with Alzheimer s disease 34,35 . Age-related impairments in these evaluation items might be suppressed by DHA intake.
Cognitive function is positively correlated with BMD 36 . In addition, cognitive decline is associated with BMD. Sohrabi et al. 37 reported a significant association between BMD and memory abilities such as episodic verbal learning. It has also been reported that patients with early-stage AD 19 have lower BMD values than cognitively healthy individuals. These findings indicate that BMD is important to maintain cognitive function. In this study, the serum levels of the bone resorption marker TRACP-5b, which could be applied to elderly people because of its small diurnal validation and low susceptibility to impaired renal function 38,39 , was significantly decreased with intake of a DHA-enriched milk beverage for 12 months by elderly Japanese people Table 5 . This finding may indicate beneficial effects of DHA on bone density in elderly people. On the other hand, the serum levels of the bone formation marker BAP were also decreased with intake of this DHA-enriched beverage for 12 months Table 5 . It is unknown at this time why DHA intake exerted opposing effects on bone resorption and bone formation. As a result, it is not possible to clarify from this study the effects of 12-month intake of a DHAenriched milk beverage on BMD. However, Bonnick and Shulman 40 showed that suppression of bone turnover markers was strongly associated with reduced fracture risk. Therefore, inhibition of bone resorption may be more important than upregulation of bone formation in elderly people, for whom fracture is one of the biggest risks of a bedridden condition. Although bone mass begins to rapidly reduce at around 50 years old, some functional food components such as ω-3 fatty acids exert positive effects on bone health 41 . Indeed, the risk of osteoporosis in Japanese and Greenland Inuit who consume large amounts of fish containing ω-3 fatty acid is lower than that of people in other countries 42 . Kis et al. 43 showed that intake of tuna oil containing a large amount of DHA increased bone calcium content more efficiently than that of common fish oil containing a large amount of EPA. Weiler and Fitzpatrick-Wong 44 also reported that feeding of piglets with a low n-6/n-3 ratio diet increased serum DHA concentrations and decreased bone resorption and the urinary N-telopeptide/creatinine ratio. Moreover, ω-3 fatty acids reduce bone resorption marker levels via the anti-inflammatory effects of ω-3 fatty acid mediators 45 47 . Focusing on female participants in this study, the mean change in TRACP-5b from baseline to 12 months was negatively correlated with the mean change in the DHA/AA ratio from baseline to 12 months r 0.377, p 0.044, data not shown . The risk of fracture and osteoporosis in women increases sharply with increased osteoclastic resorption activity after menopause 48 . Because it was unclear why the effect was stronger in women than in men, this result suggests that DHA intake might contribute to bone health in women. Further work is required to establish why DHA plays a major role in decreasing TRACP-5b and shows a strong beneficial effect in women.

Conclusion
The present study shows that the daily intake of a DHAenriched milk beverage containing DHA 297 mg positively affects the fatty acid profiles of erythrocyte plasma membranes, cognitive functions, and bone resorption marker levels in healthy elderly individuals with MMSE scores of 28 or higher at the initial examination. Although there are some limitations i.e., a small sample size and unexplained events why a small amount of DHA suppressed age-related cognitive impairment and bone resorption marker levels , this study is the first 12-month, randomized, double-blind, placebo-controlled trial to evaluate the effects of a DHAenriched milk beverage on cognitive function and bone resorption in elderly people. Further investigation is needed to elucidate the mechanism of these effects.

Data Availability
The data underlying this article cannot be shared publicly for the privacy of individuals that participated in the study. The data will be shared on reasonable request to the corresponding author.