Effects of dietary style on adult enterovirus infection risk factors: a retrospective analysis

Background: Gut microbiota serves as a defense against enteric pathogens, whereas dietary intake inuences the composition and function of gut microbiota. We aimed to examine the impact of diet on the enteroviral infection in adult patients of hand, foot, and mouth disease (HFMD). Methods: A total of 266 adult patients of HFMD were recruited in this study, with 80 healthyvolunteers served as the control. Swab samples and clinical characteristics were collected. Enteroviral genotype was further assessed by PCR testing. Social-demographic data and dietary records were obtained through follow-up phone calls. Dietary patterns were derived with PCA analysis. Correlation between dietary patterns and clinical characteristics, enterovirus genotype, and HFMD risk factors were evaluated. Results:Three distinct dietary patterns were identied in the participants, which were modern, "atypical south", and "traditional north", respectively. This study found the dietary pattern of adult HFMD signicantly differed from that of the controls. A vast majority of controls followed the modern pattern, which was a healthy diet. In contrast, the result showed unhealthy dietary patterns ('atypical south' and 'traditional north') were risk factors for adult HFMD. Besides, the dining place was a leading contributor to the dietary pattern. Our data showed eating at a food stall, or take-out is a risk factor of adult HFMD, whereas eating at the dining room is a protective factor. Conclusions:Our study indicated dietary pattern was associated with the incidence of adult HMFD. Improving habit contribute

low-ber diet have a higher risk of bowel infection by degrading the colonic mucus barrier, which serves as a defense against enteric pathogens [27]. Other studies proved that probiotics enhancing intestinal epithelial barrier function and protecting the host from enteric virus infection [28,29]. However, there's no study revealing the impact of diet on the enteroviral infection to date.
This study systematically analyzed the epidemiological characteristics and dietary records of the participates. A total of 266 adult patients of HFMD were recruited in this study, with 80 healthy volunteers served as the control. Virus genotype was con rmed by laboratory testing. Social-demographic data and dietary records were obtained through follow-up phone calls. Dietary patterns were derived with PCA analysis. Correlation between dietary patterns and clinical characteristics, HFMD risk factors were evaluated. Results of the study suggested diet was correlated with adult HFMD pathogenesis and progression.

Results
Demographics of the study projects A total of 299 adult patients were screened from 47,383 reported HFMD cases from August 2014 to June 2018. Results of PCR testing showed 266 patients were positive for the enterovirus. Finally, these 266 cases were recruited according to the inclusion criteria.
The ages of the 266 patients were ranging from 16 to 70 years old (27.5 ± 7.8) (Table 1). Females outnumbered males with a male-to-female ratio of 0.89. The most common occupation was house-hold (62/266, 23.3%), followed by the worker (57/266, 21.4%) and intellectual (54/266, 20.3%). Regarding the dining place, most of the adult patients preferred take-out or food stall (168/266, 63.2%), few of them ate at the dining room or restaurant (6/266, 2.3%). A vast majority of them didn't smoke (224/266, 84.2%), lived in the rural area (151/266, 56.8%), and had low income (167/266, 62.8%).  Aligned with the prior reports [30], EV-A71-associated adult HFMD cases decreased in recent years ( Fig. 2B). Although sporadic cases of CV-A6 occurred since the beginning of this study, the epidemic of CV-A6-associated HFMD started in October 2016. Meanwhile, no pan-enterovirus-associated HFMD case was reported from August 2015 to August 2017. Because a multitude of the pan-enterovirus-associated HFMD cases could be con rmed positive for CV-A6. CV-A10-associated cases were rare in this study, it was rst reported in the year 2016.

Risk Factors For Adult HFMD
This study analyzed the social-demographic characteristics of participants with a single-factor analysis (α = 0.1) to determine the risk factors of adult HFMD ( Principal components analysis (PCA) revealed three distinct dietary patterns: modern, "atypical south",and "traditional north".The three principal components (PC1, PC2, and PC3) accounted for 26% (13.3%, 7.0%, and 5.7%, respectively) of the variance in total food group intake ( Fig. 3A/B). The loadings of food groups of each dietary pattern were shown in Table 3  Association of various dietary patterns with social-demographic characteristics were further investigated (Table 4). Comparing with the control, there was a signi cant difference in the dining place (eating at home, dining room, or take-out) for all three dietary patterns (p-value < 0.01 for all three patterns). In addition, "traditional north" pattern presented a signi cant difference in smoking (p-value < 0.05).

Correlation Between Dietary Type And Enterovirus Strains
We examined the association between enterovirus strains and the dietary patterns of the participants. Correlation analysis showed there was no statistical difference between the enterovirus strains (EV-A71, CV-A16, CV-A6, CV-A10, and Pan-enterovirus) and various dietary patterns (modern, atypical south, traditional north) ( Table 7).

Study Design
This study screened adult patients from reported HFMD cases. Cases were further con rmed by PCR testing. Clinical characteristics data were collected. Social-demographic data and dietary records were obtained through follow-up phone calls. Dietary patterns were derived with PCA analysis. Correlation between dietary patterns and clinical characteristics, HFMD risk factors were evaluated. The purpose of this study was to reveal the association between dietary and HFMD pathogenesis and progression.

Case De nition
Clinical criteria for the diagnosis of HFMD was published by the Chinese Ministry of Health in 2010 [42].
Patients with the following symptoms were de ned as having HFMD: fever, oral ulcers, and vesicular rash on the hands, feet, or buttocks.

Study Population
The inclusion criteria for the enrollment of this study were as the following: 1) diagnosed as HFMD case; 2) positive for enterovirus by laboratory testing; 3) adult patients who de ned as 16-years-old or older; 4) could be tracked through follow-up phone calls.
A total of 299 adult patients were screened from 47,383reported HFMD cases in Jiaxing from August 2014 to June 2018. Among which, 266 cases were con rmedpositive for enterovirus by laboratory testing. According to the inclusion criteria, this study recruited 266 adult patients of HFMD, with 80 healthy adult subjects served as the controls. All healthy volunteers were negative for enteroviral infection. Socio-demographic data of adult HFMD patients were collected through follow-up phone calls.

Enterovirus Genotype
Throat swab specimens from 299 adult patients of HFMD were collected by trained medical personnel. Samples were preserved at -80℃.RNA was extracted from the specimens by TRIzol (Invitrogen, CA, USA).
The cDNA sample was synthesized by using the PrimeScript TM RT kit (Takara, Dalian, China). One-step RT-PCR assays were performed to detect enterovirus RNA, using EV-A71/CV-A16/Pan-enterovirus commercial kits and CV-A6/ CV-A10 commercial kits (Da An Gene Co. Ltd, China).

Dietary Assessment
A semi-quantitative FFQ (food frequency questionnaires) was used to assess the weekly intake of foods. The FFQ contained 24 food groups from the China Food Composition data, with each food group included several typical food items (Supplementary Table 1). Participants were asked during the followup phone calls how often they had consumed each food group over the last week (Supplementary Table 2).

Discussion
Adult hand, foot, and mouth disease is used to be seen as a mild self-limiting viral infection, and its hazard was underestimated [33][34][35]. Our study showed female patients outnumbered males with a maleto-female ratio of 0.89, and most of the patients were housewives, teachers (intellectual), and students.
These patients had a high chance of close contact with the children. Since close contact with the infected individuals is a primarily transmitting route of human enterovirus, these adult patients might serve as a potential infectious source of HFMD.
Consistent with the previous reports, our study demonstrated CV-A16 is the most prevalent causative agents of adult HFMD, followed by EV-A71, Pan-enterovirus, CV-A6, and CV-A10 [8][9][10][11]. Interestingly, this study found the incidence rate of CV-A6-associated adult HFMD increased since the year 2016. This nding was in accordance with the prior reports that CV-A6-associated HFMD is becoming more prevalent in recent years and having high morbidity in adults [13]. Notably, our study demonstrated the morbidity of adult HFMD dropped in the year 2017 and elevated dramatically in the year 2018. Taken together, these ndings reminded us of the tough situation in HFMD prevention.
Numerous studies reported diet alters the composition and activity of gut microbiota, and in turn in uences gut's resistance to colonization of pathogens [21][22][23]. Our study revealed three dietary patterns (modern, atypical south, and traditional north) of the participates. The modern pattern presented as a healthy diet, which characterized by the main intake of fruit, poultry, aquatic products, milk, vegetable juice, fungi or algae, and cereals. Whereas the "atypical south" and "traditional north" could be classi ed as unhealthy diet, which characterized by low-ber (rarely eating of vegetable and legumes) and high-fat (red meat, organ meat) food groups. A low-ber dietor high-fat diet would impactthe host's health. A study indicated the low-ber diet would degrade the colonic mucus barrier which serves as a defender against enteric pathogens [27]. Other studies proved a high-fat diet promoted pathobiont expansion and caused colitis or even intestinal carcinogenesis [36,37].Consistent with these reports, our study showed dietary patterns "atypical south" and "traditional north" are risk factors of adult HMFD. As mentioned by prior studies, the long-term of low-ber/high-fat diet ("atypical south" and "traditional north" patterns) might affect the composition and function of gut microbiota, and weaken the protective effect of gut against the colonization andreplication of enterovirus. Proceeding studies reported probiotics could enhance intestinal epithelial barrier [38,39]. Commensal bacteria, such as lactobacillus, can protect human intestinal epithelial from enteric virus infection [28,29]. Consistently, our study indicated a vast majority of controls followed the modern diet, which is rich in indigestible bers. The fermentation of indigestible bers produces short-chain fatty acids, which is essential in anti-in ammation and intestinal homeostasis [40,41]. The impact of microbiome on the enterovirus infection required further studies.
Besides what's to eat, where to eat also matters. Our study revealed the dining place is a signi cant contributor to the dietary pattern. Eating at food stall or take-out was a risk factor of adult HFMD, whereas eating at the dining room was a protective factor. Multivariate linear regression analysis showed "eating at home" and "eating at food stall or take-out" were negatively correlated with the modern dietary pattern, whereas these dining places positively contributed to the unhealthy dietary patterns "atypical south" and "traditional north". Most of the food stall or take-out faced poor sanitary condition, especially the "gutter oil" problem. Besides, the best selling on the food-stall or take-out was always low-ber/high-fat food like organ meat and fast food. Notably, "eating at home" did not necessarily mean eating healthy. Imbalanced diet and dietary preference might be the problem of "eating at home".

Conclusions
In conclusion, our study analyzed the dietary pattern and its association with the epidemiological and clinical characteristics of adult HFMD. To the best of our knowledge, this is the rst study focused on the impact of diet on the morbidity of HFMD. However, the dietary records collected in this study can't re ect this seasonal pattern. Food records in the different season should be analyzed in the future study. Availability of data and materials The dataset used in the study are available from the corresponding author on reasonable request.
Ethics approval and consent to participate   Distribution of enterovirus genotype.

Figure 3
Dietary patterns of adult HFMD patients.

Figure 4
Spatial distribution of dietary patterns.