This study was conducted between July and September 2018 at Tokyo area. The participants were healthy 30 older adults (14 men and 16 women, aged ≥65 ), and the inclusion criteria were as follows: (1) no drugs and supplements such as antioxidant, anti-obesity, or anti-diabetic; (2) no diagnosis of sleep apnea syndrome, diabetes, dyslipidemia, or hypertension by a physician; and (3) absence of the use of glucose/insulin-lowering or anti-hypertension related medications. This inclusion criteria was similar to our previous paper . Study protocol conformed to the Declaration of Helsinki and was approved by the Ethics Committee for Humans at Waseda University (approval number: 2018-031). We registered this human study at UMIN system (www.umin.ac.jp/ctr/ number: UMIN000032858). After the experimental details had been explained, informed consent was obtained from all the participants. Participants were answered a questionnaire on dietary habits, lifestyle habits, and health and medication status prior to the start of this study. Eight participants were excluded from the study owing to the submersion of their feces during experiments.
The experiments were conducted for a week, and participants were asked to maintain their lifestyle habits such as diet and exercise throughout experiments. During the experimental period, the physical characteristics of all participants were measured, and all subjects were asked to wear a continuous glucose monitoring (CGM) system as mentioned in our previous paper . Participants were also asked to collect their feces in a tube with 20% glycerol containing phosphate-buffered saline. Tube had been distributed in advance for intestinal microbiota evaluation (morning of Day 8). The collected fecal samples were carried to the university at 4 °C, and were stored at -80 °C until analysis after immediately frozen in liquid nitrogen .
According to previous paper , body mass was measured using a digital balance ( 0.1kg minimum, Inbody 230, Inbody Inc., Tokyo, Japan), and height was measured using a wall-mounted stadiometer (0.1cm minimum, YS-OA, As One Corp., Japan). Calculation of body mass index (BMI) was well known as (Kg/ m2).
Determination of tissue glucose levels
Participants were required to wear a CGM system (FreeStyle Libre Pro; Abbott Laboratories, Chicago, IL, USA) for the continuous monitoring tissue glucose levels throughout experiment. CGM system can continuously measure and store data of tissue glucose levels with 15 min intervals, and this system is considered to be less burdensome than other glucose monitoring systems especially for elderly people. To evaluate CGM, mean (M) and standard deviation (SD), coefficient of variation (CV), and peak glucose levels were calculated for parameters.
Fecal DNA extraction and 16S rRNA gene sequencing
Fecal DNA extraction and 16S rRNA gene sequencing was performed as previously described . 16S rRNA gene sequencing was performed on the Illumina platform. We amplified the V3-V4 variable regions of the 16S rRNA gene with PCR using the following primers:
Forward Primer: 5′-TCGTCGGCAGCGTCAGATGTGTATAAGAGACAGCCTACGGGNGGCWGCAG-3′;
Reverse Primer: 5′-GTCTCGTGGGCTCGGAGATGTGTATAAGAGACAGGACTACHVGGGTATCTAATC-3′.
Amplicon PCR was performed with 2.5 µL of microbial DNA (5 ng/µL), 5 µL of each primer (1 µM), and 12.5 µL of 2× KAPA HiFi HotStart Ready Mix (Kapa Biosystems, Wilmington, MA, USA). We used same protocol of the cycling parameters . The cycling parameters were as follows: denaturation at 95°C for 3 min, followed by 25 cycles of denaturation, annealing, and elongation at 95°C for 30 s, 55°C for 30 s, and 72°C for 30 s, respectively, and a final extension at 72°C for 5 min. We purified PCR amplicons by AMPure XP beads (Beckman Coulter Inc., Brea, CA, USA).
To perform multiplex sequencing, adapters and barcodes were ligated to amplicons by the following kit, Nextera XT Index Kit v2 (Illumina Inc., San Diego, CA, USA) . Index PCR was performed with 5 µL PCR product, 5 µL of each Nextera XT Index primers, 25 µL of 2× KAPA HiFi HotStart Ready Mix, and 10 µL of PCR-grade water. The conditions of PCR were followed: 1 cycle at 95°C for 3 min, 8 cycles of denaturation, annealing, and extension at 95°C for 30 s, 55°C for 30 s, and 72°C for 30 s, respectively, followed by a final extension at 72°C for 5 min. We checked the quality of the purified products by an Agilent 2100 Bioanalyzer with a DNA 1000 kit (Agilent Technologies, Santa Clara, CA, USA). Finally, the DNA library was diluted to a concentration of 4 nM and sequenced using MiSeq Reagent Kit v3 (Illumina Inc.) carrying on an Illumina MiSeq 2×300 bp platform, according to the manufacturer’s instructions and our previous paper .
Analysis of 16S rRNA gene sequences
Analysis of 16S rRNA gene sequences was performed as previously described . We processed the 16S rRNA sequence reads using the help of quantitative insights into the microbial ecology (QIIME) pipeline version 1.9.1 . The quality-filtered sequence reads were assigned to operational taxonomic units (OTUs) using closed-reference OTU picking at 97% identity with the UCLUST algorithm . And then we compared these reads with reference sequence collections in the Greengenes database (August 2013 version). From 44 samples 1,296,946 reads were obtained in total, and therefore, 22,361±2,721 reads were obtained per sample on an average. At the phylum and genus levels, the taxonomy summary was calculated using the QIIME software (version 1.9.1).
Data were analyzed using the Predictive Analytics Software for Windows (SPSS Japan Inc. Tokyo, Japan). Kolmogorov–Smirnov test was applied to check whether data showed normal or non-normal distributions. In this study, the correlation between fasting (1 point before breakfast) and postprandial (breakfast, lunch, and supper) glucose levels and the gut bacteria was investigated. The correlation coefficient was calculated using Pearson's or Spearman’s test for the parameters showing either a normal or non-normal distribution, respectively. Gut bacteria found in more than half of the subjects were selected and used for statistical analysis (50 types of gut bacteria from ≥11 persons were used). Levene's test was used to compare the variance in fasting and peak (after each meal) glucose levels. Statistical significance was set at p < 0.05.