Meat and vegetable intake is important for human health(1). It has been examined that unhealthy eating of meat and vegetable is significantly associated with the elevated risk of experiencing some non-communicable diseases (NCD), including type 2 diabetes, CVD, cancers and mental health problems(Reference Papier, Fensom and Knuppel2–Reference Głąbska, Guzek and Groele7). Thus, population-based promotion of healthy eating of meat and vegetable is critical for improving and/or maintaining health state for people. Moreover, periodical investigation on and assessment of population-level meat and vegetable consumption patterns is a prerequisite of precision intervention for community-based NCD prevention from the perspective of public health.
For typical individuals, their food intake choices are mainly driven by food supply and price, socio-economic factors, habitual response and health knowledge and thus are usually stable(Reference Groth, Fagt and Brøndsted8–Reference Lindmark, Stegmayr and Nilsson10). When the living environment changed, food supply and residents’ food choice habits would change as well(Reference Marty, de Lauzon-Guillain and Labesse11). The coronavirus disease 2019 (COVID-19) exerted a wide impact on residents’ daily life, including food supply and dietary intake habits, due to confinement measures implemented during the pandemic period(Reference Zheng, Wang and Zhang12–Reference Picchioni, Goulao and Roberfroid14). Hence, the COVID-19 pandemic was a new driving factor of food intake choices for individuals, particularly the vulnerable population like older adults, during and even after the pandemic.
During the COVID-19 pandemic, eating behaviours of meat and vegetable were explored among residents in some countries, including China, showing mixed findings regarding changes in consumption patterns(Reference Zheng, Wang and Zhang12,Reference Mekanna, Panchal and Li13,Reference Skotnicka, Karwowska and Kłobukowski15–Reference González-Monroy, Gómez-Gómez and Olarte-Sánchez19) . However, there was no study that compared meat and vegetable intake before, during and after the COVID-19 pandemic for older adults. China had the largest group of older people in the world, with 264 million (18·7 % of the total population) adults aged 60+ years in 2020(20). Therefore, as a strategy of healthy aging, it is meaningful to assess meat and vegetable intake and then implement precision healthy eating interventions for older residents in China. For this purpose, we developed this study with the aim to compare red meat, white meat and vegetable consumption for adults aged 60+ years before, during and after the COVID-19 pandemic in regional China.
Methods
Study design and participants
The study area was Nanjing municipality, a megacity in eastern China, which had about 8·4 million residents (18·3 % aged 60+ years) registered within twelve administrative districts (including five urban and seven suburban districts) in 2018(21). In this study, urban and suburban districts were defined based on the official classifications released by the China National Bureau of Statistics(22). In 2018, a broad programme, the Healthy Aging Healthy Elders (HAHE) study, was developed for periodically evaluating the prevalence of common NCD (including type 2 diabetes, hypertension, body weight, etc.) and related lifestyle and behaviours (including eating behaviours, physical activity (PA), sedentary behaviour (SB), etc.) among older adults aged 60+ years every 5 years in Nanjing Municipality, aiming to inform sustainably precision intervention of lifestyle and behaviours as well as prevention of targeted NCD(Reference Zhang, Kang and Jing23).
The HAHE study was designed as a series of independent cross-sectional surveys, with the first wave (HAHE-2018) conducted in mid-2018 as the baseline survey(Reference Zhang, Kang and Jing23). The eligible participants of the HAHE study referred to those who (1) were local residents registered in Nanjing municipality, (2) were aged 60+ years, (3) had no physical/psychiatric issues and (4) were without literal/cognitive problems(Reference Zhang, Kang and Jing23). For the HAHE-2018 study, the study design and methodology have been described in detail elsewhere(Reference Zhang, Kang and Jing23). Briefly, to warrant representativeness of overall older residents aged 60+ years, it was estimated that the sample size for the HAHE-2018 study should be approximately 21 000, and participants were randomly selected with a multi-stage sampling approach from each of the twelve districts of Nanjing municipality(Reference Zhang, Kang and Jing23).
In China, after the initial outbreak of COVID-19 in late December of 2019, emergency containment approaches against the COVID-19 epidemic lasted till January 8 of 2023(24). Considering that residents’ lifestyle and behaviour patterns would change during the period of the COVID-19 pandemic, we created an idea in 2020 to compare lifestyle and behaviour patterns among older adults before, during and after the COVID-19 pandemic in Nanjing municipality. During the COVID-19 pandemic, population movement was strictly restricted, particularly in urban areas, due to confinement measures(Reference Liu, Yue and Tchounwou25). Therefore, a special HAHE study (HAHE-2021 study) was designed and then conducted in two districts randomly determined from all five urban districts of Nanjing in mid-2021. About 4000 participants would be sufficient for investigating red meat, white meat and vegetable intake levels and exploring the associated factors. The methodology and instruments/questionnaire, as well as the eligibility of participants involved in this special HAHE-2021 survey, were adopted from the HAHE-2018 study.
Fortunately, the emergency response to the COVID-19 epidemic was terminated on 8 January 2023 in China(24), which allowed us to conduct the second wave of HAHE survey in 2023 as initially scheduled (HAHE-2023). Of course, the methodology, instruments, participants’ eligibility and sampling approach in the HAHE-2023 survey were the same as those used in the HAHE-2018 study. Participants were also randomly recruited from all twelve districts of Nanjing municipality. As for the sample size for HAHE-2023 survey, considering that information gathered in HAHE-2023 study would be used as potential post-pandemic reference data for future NCD and lifestyle/behaviours surveillances and a conservative response rate was expected in the first year after COVID-19 epidemic, the overall sample size estimated for HAHE-2023 survey would be larger than that estimated for HAHE-2018 study. Therefore, the finally determined sample size would be approximately 30 000 for the HAHE-2023 survey.
All five urban districts were involved in the HAHE-2018 and HAHE-2023 studies, while only two urban districts were included in the special HAHE-2021 survey. To maximise the data comparability between these three surveys, participants in the same two urban districts involved in the HAHE-2021 survey were derived from the HAHE-2018 and HAHE-2023 surveys, respectively, for the purpose of comparing meat and vegetable consumption among urban older adults before, during and after the COVID-19 pandemic in Nanjing municipality. Finally, participants from the same two urban districts analysed in this comparison study were 4355, 4622 and 4815 adults aged 60+ years for HAHE-2018, HAHE-2021 and HAHE-2023 surveys, respectively. Obviously, the data analysed in the present comparative study were from three independent cross-sectional surveys conducted in the same two urban districts in Nanjing municipality. Figure 1 displayed the selection flowchart of participants for these three surveys.
Flow chart of participant’s selection in this comparative study.
Written informed consents were obtained from all participants within each of these three HAHE surveys. The HAHE studies for original data collection were reviewed and approved by the Ethics Committee of Nanjing Medical University Affiliated Nanjing Municipal Center for Disease Prevention and Control. The methods employed in each HAHE survey were in line with the Declaration of Helsinki. Moreover, as only second-hand de-identified data from HAHE study were analysed in the present comparative study, it was granted an exemption of ethical approval by the Ethics Committee of the Geriatric Hospital of Nanjing Medical University.
Data collection
Dada collected in HAHE studies included participant’s socio-demographic attributes (age, educational attainment, marital status, health insurance), history of main NCD (hypertension, diabetes and abnormal lipid profile), food consumption (red meat, white meat and vegetable intake, etc.), PA, SB, drinking and smoking, family history of diabetes and hypertension, which were self-reported through a standardised questionnaire(Reference Yao, Wang and Xiong26). Body weight and height were objectively assessed with light clothing and no shoes. Weight was recorded to the nearest 0·1 kg, while height was measured to the nearest 0·01 metre. Each of them was assessed two times, and then the mean values were used to compute BMI for each participant.
A standardised procedure was released for data collection in community-based population surveys on NCD by the National Center for Chronic Non-communicable Disease Control and Prevention of the Chinese Center for Disease Control and Prevention in 2018(Reference Wang, Zhang and Li27). This recommended data collection procedure and variable definition were followed strictly in our HAHE studies since 2018(Reference Zhang, Kang and Jing23,Reference Yao, Wang and Xiong26) . Briefly, each participant was appointed with a specific date to visit the designated local community health service centre without breakfast in the morning(Reference Wang, Zhang and Li27). Within a quiet room at the community health service centre, the investigated information was gathered via an interview conducted by research team members, while body weight and height were measured within the next room by community health service centre staff(Reference Wang, Zhang and Li27). Each participant would be provided with a free breakfast when completing the entire survey. As for the HAHE-2021 special survey during the COVID-19 epidemic, due to the population mobility restriction, participants were asked to do COVID nucleic acid testing on the day before the appointed date of survey. Only those with a negative nucleic acid test were invited to take part in the survey on the appointed date. For those with a positive nucleic acid test, they were suggested to do a nucleic acid test again 14 d later. And, after nucleic acid testing was negative, they were reappointed a new survey date. Surely, they also needed to redo the nucleic acid test before the newly appointed survey day.
Study variables
Outcome variable
Red meat, white meat and vegetable intake was measured with a validated Chinese version of FFQ(Reference Liu, He and Zhang28). The question item in the FFQ used to gather information on food intake was ‘In last year, how often did you consume the following foods under a typical situation? Please respond to all items of food in frequencies one by one.’(Reference Liu, He and Zhang28). The consumption frequencies of red meat, white meat and vegetable were computed as times per week for analysis. The Chinese Nutrition Society recently recommended intake frequencies of meat and vegetable especially for Chinese older adults in 2016(29). In this recommendation, red meat and white meat were not differentiated from each other(29). This means that the meat intake recommendation is just for red meat plus white meat, and no separate recommendation is available for red meat or white meat. Therefore, based on these specific Chinese Nutrition Society recommendations for consumption frequencies of red plus white meat and vegetable, participants were accordingly categorised as ‘recommendation reached’ or ‘recommendation not reached’ in this study.
Explanatory variables
Selected socio-demographic variables were analysed in the study, including age (60–69, 70–79 or 80+ years old), sex (men or women), education (≤ 9, 10–12 or 13+ years of schooling), marital status (single or married/having a partner) and health insurance (employee health insurance scheme or resident health insurance scheme). Participants were also classified as ‘smokers’ or ‘non-smokers’, ‘drinkers’ or ‘non-drinkers’ according to the definitions used by the China Centers for Disease Control and Prevention(Reference Wang, Zhang and Li27). PA and SB were measured with the validated Chinese version of international physical activity questionnaire (IPAQ-CHN)(Reference Qu and Li30). The moderate and vigorous PA time in the past 7 d was measured separately. Subsequently, based on the weekly total time of moderate PA time plus doubled vigorous PA time, a participant was categorised as having ‘sufficient PA (≥ 150 min/week)’ or ‘insufficient PA (< 150 min/week)’(31). Meanwhile, daily SB time was used to classify participants into ‘shortened SB (< 2 h/d)’ or ‘prolonged SB (≥ 2 h/d)’ based on the SB time recommended for Chinese adults(31).
Participants’ weight status was measured using BMI, which was defined as body weight in kilograms divided by height in metres squared. Participants were classified into ‘underweight (BMI < 18·5)’, ‘normal weight (18·5 ≤ BMI < 24·0)’, ‘overweight (24·0 ≤ BMI < 28·0)’ or ‘obesity (BMI ≥ 28·0)’ based on BMI cutoffs recommended specifically for Chinese adults(32).
Family histories of diabetes and hypertension were self-reported by participants and defined as ‘positive’ if either of their parents had been diagnosed having diabetes or hypertension. Otherwise, the family histories were categorised as ‘negative’. Moreover, a subject was recorded as a diabetic or hypertensive patient if he/she was ever diagnosed as such by a physician. Otherwise, the participant was categorised as a non-diabetic or non-hypertensive individual. Additionally, participants were classified as ‘having normal lipid profile’, only when all of their cholesterol, TAG and HDL-cholesterol /LDL-cholesterol levels were normal(Reference Wang, Zhang and Li27). Otherwise, they were categorised as ‘having abnormal lipid profile’(Reference Wang, Zhang and Li27).
Data analysis
First, the distribution was examined for each of red meat, white meat and vegetable consumption frequencies using the Kolmogorov–Smirnov test, showing all of them fit a normal distribution. Then, using mean value (standard deviation, mean (sd)) or percentage (%), the distribution of meat and vegetable consumption was described among participants by socio-demographic attributes. Next, differences in meat and vegetable intake frequencies between categories of participants and different survey years were examined with χ 2 or ANOVA tests. Finally, with adjustment (where applicable) for age, gender, education, marital status, health insurance, body weight status, smoking, drinking, PA, SB, self-reported diabetes, self-reported hypertension, self-reported abnormal lipid profile, family history of hypertension, family history of diabetes, red plus white meat intake, vegetable intake and survey year, multivariate logistic regression models were employed to estimate OR and 95 % CI to identify potential influencing factors of red meat plus white meat and vegetable consumption. P < 0·05 (two-sided) was treated as a significant level. EpiData 3·1 (The EpiData Association 2008) was employed to enter data, while SPSS version 20·0 for Windows (SPSS Inc.) was used for data analysis.
Results
There were 13 792 participants totally included in this comparative study, with 4355 (31·6 %), 4622 (33·5 %) and 4815 (34·9 %) from the HAHE survey before, during and after the COVID-19 pandemic, respectively. The response rates were 90·6, 85·9 and 91·4 % for surveys 2018, 2021 and 2023, respectively. The main reasons for non-response were ‘not accessible’ and ‘refuse’. The oldest older residents (aged 80+ years) tended to refuse to take part in the survey in 2021 during the COVID-19 pandemic period. Table 1 presents the distribution of participants within age, gender, education, marital status and body weight status by survey year. Among the overall participants analysed, 5·7 % were 80+ years old, while 47·1 % were men. There were 14·2 % with 13+ schooling-year education and 9·7 % without a spouse/partner. Moreover, 18·0 and 1·8 % of participants were obese and underweight, respectively. Age, educational attainment, marital status or body weight status differed significantly among participants in HAHE-2018, HAHE-2021 and HAHE-2023 surveys, while no difference in gender was identified between participants from the three surveys.
Selected characteristics of participants aged 60+ years in surveys 2018, 2021 and 2023 in urban areas of Nanjing, China (n 13 792)
* χ 2 test.
† Body weight status was categorised based on recommendations for Chinese adults using BMI.
Table 2 shows the comparison of red meat, white meat and vegetable intake levels among participants by characteristics between different survey years in this study. For overall participants, the mean weekly intake frequencies of red meat, white meat and vegetable were 3·94 (sd 3·22), 2·09 (sd 2·06) and 10·43 (sd 5·00), respectively. Red meat and white meat intake frequencies were different among participants by age, gender or education, while there was no difference in vegetable consumption frequency for participants by age or sex.
Consumption level of red meat, white meat and vegetable by selected characteristics of participants aged 60+ years in surveys 2018, 2021 and 2023 in urban areas of Nanjing, China (n 13 792)
* sd.
† ANOVA test was used to examine the differences in food intake frequency between the three years.
‡ Body weight status was categorised based on recommendations for Chinese adults using BMI.
When looking at consumption frequencies of red meat, white meat and vegetable among participants before, during and after the COVID-19 pandemic, significant differences were examined. Specifically, the mean weekly intake frequencies among participants in 2018, 2021 and 2023 were, respectively, 3·85 (sd 2·83), 3·21 (sd 2·90) and 4·71 (sd 3·94) for red meat; 1·38 (sd 1·21), 2·08 (sd 1·90) and 2·73 (sd 2·55) for white meat; and 10·98 (sd 4·84), 10·00 (sd 5·04) and 10·34 (sd 5·04) for vegetable. Moreover, the mean consumption frequencies of red plus white meat were 5·23 (sd 3·41), 5·29 (sd 4·03) and 7·44 (sd 5·23) times per week, respectively, among older residents before, during and after the COVID-19 pandemic in this study. Furthermore, consumption of red meat, white meat and vegetable also significantly differed within each sub-group of age, gender or educational attainment.
Table 3 demonstrates the proportion of subjects who met the recommendation of red plus white meat and vegetable consumption in this study. Overall, 36·5 and 49·9 % of participants met recommendations of red plus white meat and vegetable consumption, respectively. The proportion of participants who met the recommendation of red plus white meat intake was different in age, sex and education, while a significant difference in the proportion of subjects meeting the recommendation of vegetable consumption was identified by age and education but not sex.
Proportion of participants meeting intake recommendation of meat and vegetable for residents aged 60+ years in surveys 2018, 2021 and 2023 in urban areas of Nanjing, China (n 13 792)
* The criteria of meat consumption were recommended by the Chinese Nutrition Society in 2016.
† Meat refers to either red meat or white meat.
‡ χ 2 test was used to examine the differences in the proportion of participants meeting recommendations between the three years.
§ Body weight status was categorised based on recommendations for Chinese adults using BMI.
Additionally, in 2018, 2021 and 2023, respectively, the proportion of participants who met recommendation was 23·2 % (1010/4355), 32·6 % (1509/4622) and 52·3 % (2518/4815) for red plus white meat intake and 53·7 % (2339/4355), 46·8 % (2161/4622) and 49·6 % (2387/4815) for vegetable consumption, showing significant differences in surveys conducted before, during and after COVID-19. Moreover, similar scenarios on the proportion of subjects meeting food consumption recommendations were identified for participants stratified by age, sex or education in the study.
Table 4 displays survey year and potential influencing factors associated with the odds of meeting the recommendation of red plus white meat and vegetable consumption among older adults in this study. After control for potential confounders, survey year was positively associated with the likelihood of meeting red plus white meat consumption recommendation (OR = 1·60, 95 % CI = 1·45, 1·77; OR = 3·61, 95 % CI = 3·29, 3·97 for the year 2021 and 2023 compared with 2018, respectively), but in negative relation to the odds of meeting vegetable intake recommendation (OR = 0·94, 95 % CI = 0·83, 0·99; OR = 0·80, 95 % CI = 0·73, 0·87 for the year 2021 and 2023 relative to 2018, respectively) among participants in the study. Interestingly, compared with during COVID-19 pandemic, subjects were more likely to meet consumption recommendation of either red plus white meat (OR = 2·24, 95 % CI = 2·06, 2·44) or vegetable (OR = 1·16, 95 % CI = 1·07, 1·27) after COVID-19 pandemic. Additionally, age, education, marital status and drinking were each positively associated with the odds of meeting the recommendation of red plus white meat intake, while education, marital status, smoking and drinking were significantly positively associated with the likelihood of meeting the vegetable consumption recommendation.
The association of survey year and socio-demographic characteristics with likelihood of meeting intake recommendation of meat and vegetable among residents aged 60+ years in urban areas of Nanjing, China (n 13 792)
* OR; CI.
† Meat refers to either red meat or white meat.
‡ Model 1 was a univariate logistic regression analysis.
§ Model 2 was a multivariate logistic regression analysis with adjustment for age (where applicable), gender (where applicable), education (where applicable), marital status (where applicable), body weight status (where applicable), smoking (where applicable), drinking (where applicable), physical activity, sedentary behaviour, self-reported diabetes, self-reported hypertension, self-reported abnormal lipid profile, family history of hypertension, family history of diabetes, health insurance, meat intake (where applicable), vegetable intake (where applicable) and survey year in multivariate logistics regression models.
|| Body weight status was categorised based on recommendations for Chinese adults using BMI.
Discussion
This population study aimed primarily to compare red meat, white meat and vegetable consumption before, during and after the COVID-19 pandemic among older urban adults aged 60+ years in regional China and then to identify the potential influencing factors associated with the odds of meeting the recommendation of red plus white meat and vegetable consumption. Compared with the corresponding intake frequencies before the COVID-19 outbreak, frequencies of red meat and vegetable consumption declined during the COVID-19 pandemic and then went up after the COVID-19 emergency. However, white meat consumption frequency during the COVID-19 pandemic was higher than that before the COVID-19 outbreak but became lower after the COVID-19 pandemic. Moreover, participants were more likely to reach the consumption recommendation of red plus white meat but less likely to meet the vegetable intake recommendation during and after the COVID-19 pandemic compared with that before the COVID-19 outbreak. And, meaningfully, after the COVID-19 pandemic, the older residents tended to meet recommendations of both red plus white meat and vegetable consumption relative to that during the COVID-19 pandemic.
The main findings observed in our study were not strictly consistent with those reported in existing studies, as mixed findings were documented on changes in red meat and white meat and vegetable consumption influenced by the COVID-19 pandemic in previous studies(Reference Zheng, Wang and Zhang12,Reference Mekanna, Panchal and Li13,Reference Skotnicka, Karwowska and Kłobukowski15–Reference González-Monroy, Gómez-Gómez and Olarte-Sánchez19) . Among these published studies, the majority compared dietary patterns among participants during and before the COVID-19 pandemic(Reference Zheng, Wang and Zhang12,Reference Skotnicka, Karwowska and Kłobukowski15–Reference Montero López, Mora Urda and Martín Almena18) , and a few reported changes in food consumption after the COVID-19 pandemic(Reference Mekanna, Panchal and Li13,Reference González-Monroy, Gómez-Gómez and Olarte-Sánchez19) . The previous studies comparing dietary patterns between during and before the COVID-19 pandemic were mainly from Europe and Latin America(Reference Skotnicka, Karwowska and Kłobukowski15–Reference Montero López, Mora Urda and Martín Almena18), and they reported that vegetable intake increased significantly(Reference Skotnicka, Karwowska and Kłobukowski15,Reference Lombardo, Guseva and Perrone16) but meat (not specified red or white meat) consumption remained unchanged in selected European countries during the pandemic(Reference Skotnicka, Karwowska and Kłobukowski15). Another two cross-sectional surveys conducted among adults aged 18+ years in Latin America and Spain observed that most of the participants reported no changes in dietary habits and only a few subjects increased their consumption of healthy foods during the COVID-19 lockdown period(Reference Enriquez-Martinez, Martins and Pereira17,Reference Montero López, Mora Urda and Martín Almena18) . Additionally, two studies compared dietary patterns among participants before, during and after the COVID pandemic(Reference Mekanna, Panchal and Li13,Reference González-Monroy, Gómez-Gómez and Olarte-Sánchez19) . In a study documenting dietary patterns before and after the COVID-19 lockdown among adults aged 20+ years in Italy, vegetable intake increased, but either red or white meat consumption remained unchanged after the COVID-19 lockdown(Reference González-Monroy, Gómez-Gómez and Olarte-Sánchez19). In a systematic review, vegetable intake was observed increasing during lockdown but decreasing after lockdown, while white meat consumption decreased during and after the COVID-19 pandemic relative to before lockdown among participants aged 18+ years(Reference Mekanna, Panchal and Li13).
In the present study, men tended to eat more both red meat and white meat compared with women. However, the difference in vegetable intake between men and women was not significant among participants either before, during or after lockdown. Interestingly, among overall participants with adjustment for potential confounders, the odds for meeting recommendations of red plus white meat and vegetable intake were not different between men and women. This implied that, although men consumed more red and white meats, they were at a similar likelihood of achieving the intake recommendation in the study.
Regardless of the mixed findings, a basic mechanism can be used to explain changes in meat and vegetable consumption influenced by the COVID-19 pandemic. The driving forces for food intake choices mainly include food supply and price, individuals’ socio-economic factors, habitual response and health knowledge(Reference Groth, Fagt and Brøndsted8–Reference Lindmark, Stegmayr and Nilsson10). These driving forces will change as the living environment changes(Reference Marty, de Lauzon-Guillain and Labesse11). The COVID-19 pandemic imposed a wide and unexpected impact on the living environment and food supply due to confinement measures implemented during the period of emergency(Reference Zheng, Wang and Zhang12–Reference Picchioni, Goulao and Roberfroid14). Therefore, individuals’ eating habits might differ from those before and after the COVID-19 pandemic. And, further, due to the diversity of living environment and food supply worldwide, it is plausible that the specific changes in residents’ dietary patterns in different areas might also be diverse during the COVID-19 pandemic.
The proportions of subjects who met consumption recommendation of red plus white meat (36·5 % v. 22·2 %) and vegetable (49·9 % v. 33·4 %) among overall participants in this study were higher than the corresponding figures reported in the recent China Health and Nutrition Survey 2015 (CHNS-2015 survey)(Reference Wang, Zhang and Wang33,Reference Ouyang, Zhang and Wang34) . Such an inconsistency might be explained by at least two reasons. The first and most important might be that the recommendation of red plus white meat and vegetable consumption was different in our study and the CHNS-2015 survey(29,35) . In the present study, weekly intake frequency was used to measure an older resident meeting recommendations of red plus white meat and vegetable consumption according to dietary guidelines specifically released for older Chinese residents by Chinese Nutrition Society in 2016(29), whereas, in the CHNS-2015 survey, intake amount was employed to assess an adult meeting recommendations based on guidelines issued for overall Chinese adults issued by Chinese Nutrition Society in the same year(35). The second reason might be that data were collected in different years and areas between our study and the CHNS-2015 survey(Reference Wang, Zhang and Wang33,Reference Ouyang, Zhang and Wang34) . Data analysed in our study were gathered from Nanjing municipality in the eastern region of China in 2018, 2021 and 2023, while data in CHNS-2015 were collected from fifteen provinces of China in 2015 (3 years before the first wave HAHE survey in our study)(Reference Wang, Zhang and Wang33,Reference Ouyang, Zhang and Wang34) .
As for findings between our study and other similar surveys implemented in China, an interesting consistency was observed regarding changes in meat and vegetable consumption due to the COVID-19 pandemic. One study using household-based national-level panel data collected in 2020 found a decrease in red meat intake, an increase in white meat consumption and a non-significant increase in vegetable intake(Reference Zheng, Wang and Zhang12). However, information on participants’ age and sex was not included in this panel study, resulting in no sex-stratified analysis being available(Reference Zheng, Wang and Zhang12). Moreover, an online survey conducted in 2022 showed a decrease in vegetable consumption among adults after the COVID-19 pandemic in Shanghai(Reference Qiu, Li and He36). These consistent findings between our study and those in China could be, at least partly, explained easily by that Chinese people shared the same/similar living environment, food supply system and emergency measures against the COVID-19 pandemic.
Meaningfully, the consumption level of either red plus white meat or vegetable went up among older residents after the COVID-19 pandemic in this study. And, moreover, participants were more likely to meet the intake recommendations of red plus white meat and vegetable after the pandemic. This implied that the impact of the COVID-19 pandemic on residents’ eating behaviours vanished or at least diminished immediately when the emergency ended in China. When the living environment and food supply became as normal as that before the disease outbreak, the driving forces for residents to choose food might, again, largely depend on their socio-economic factors, habitual response and health knowledge(Reference Groth, Fagt and Brøndsted8–Reference Lindmark, Stegmayr and Nilsson10). This might explain the increase in meat and vegetable intake patterns for residents after the COVID-19 pandemic.
This study has some significant implications from the perspective of public health. Unexpected emergency events can exert a sudden impact on residents’ eating behaviours. Importantly, individuals’ eating behaviours would return to a normal pattern when the emergency ended. This study provided evidence that there was no long-term impact of COVID-19 on residents’ eating behaviour after the pandemic. For older adults aged 60+ years in China, they typically retire and do not need to do any job. The retirees, particularly the urban residents, are more dependent upon their living environment relative to those of working-age in China. On the other hand, they have much discretionary time that allows them to choose foods without consideration of the time consumed. Hence, these suggest that specific attention shall be paid to older residents’ eating behaviours in making preparedness plans for potential emergency events in the future.
Moreover, this is the only study that assessed the changes in population-level meat and vegetable consumption before, during and after the COVID-19 pandemic among older residents selected from the same areas in China. This study made some unique contributions to this research area. First, this specific study, with its novelty, focused on older people, the vulnerable population in dealing with emergency events. Second, three main crucial foods (red meat, white meat and vegetable) were investigated in a single study. Third, changes in food consumption before, during and after the COVID-19 pandemic were examined. Fourth, participants were representative community-dwelling older residents from regional China. Lastly, it has meaningful public health implications for both older individuals and government bodies to implement targeted approaches to maintain healthy eating and health conditions for older people during future potential emergency events, particularly those lasting for a long period of time, like COVID-19 pandemic.
Several strengths are worthy of being mentioned. First, participants for each of the three surveys were randomly determined from the same urban areas, and the sample size of each survey could warrant sufficient statistical power. Second, the same study design and instruments were applied in each HAHE survey before, during and after the COVID-19 pandemic. Third, recommendations of food consumption specifically developed for Chinese older people, not the general adults, were employed to investigate participants’ meat and vegetable intake. Finally, interesting and important findings are observed and may be used to inform the making of preparedness plans on residents’ eating behaviours in case of potential emergency events in the future.
This study also has some limitations. First, participants self-reported the information on meat and vegetable consumption, which potentially implied recall bias. Participants might have over-/underestimated the food intake frequencies systematically or occasionally. For a participant who systematically either over- or underestimated food intake frequencies, the recall bias would produce a systematic error and exert little influence on the analysis in this study. However, if food intake frequencies were over-/underestimated occasionally by participants, the recall bias would be an accidental error, which might produce a potential negative influence on the findings. Fortunately, participants tended systematically to over-/underestimate food intake frequencies by self-reporting(Reference Li, He and Zhai37). Thus, only a little influence of recall bias might imply in the findings in this study. Second, frequencies of meat and vegetable consumption were used in the study. Thus, we could not estimate the intake amount of meat and vegetable for participants. Lastly, participants of the survey during the COVID-19 pandemic tended to be younger, as only 4·0 % of them were 80+ years old. Therefore, the specific findings in this study shall still be interpreted prudently with full consideration of participants’ characteristics.
In conclusion, for older urban residents in regional China, they consumed less red meat and vegetable but more white meat during the period of the COVID-19 pandemic relative to the time before COVID-19, and their consumption levels of red plus white meat and vegetable went up after the COVID-19 emergency. These findings highlight the need for targeted interventions to support older adults’ dietary habits during emergency events. Specific implications of this study are helpful for both older residents and government bodies to tackle potential population-level emergencies in the future. For older individuals, they need to do their best not only to maintain their usual eating habits but also to improve healthy eating, for example, a balance between meat, vegetable and other foods consumption. For the government sectors, they shall do their best to ensure a sufficient supply of diverse foods and provide tailored information on healthy eating to older people.
Acknowledgements
Our special thanks go to all workers from local communities for their kind assistance in field data collection.
This present work was supported by the National Natural Science Foundation of China (52078254, recipient: FX). The funder had no role in this work and the decision of submission.
G. Z., H. W., J. K. and F. X. conceived and designed the present study. G. Z. and F. X. are responsible for data acquisition. F. X. analysed the data and obtained financial support. G. Z., H. W., T. D., H. X., Y. X., G. A., F. X. and J. K. wrote and critically reviewed the manuscript. Each author approved the final version for submission and was also responsible for all aspects of the work presented in this manuscript.
No conflict of interest declared.
Written informed consents were obtained from all participants within each of these three HAHE surveys. The HAHE studies for original data collection were reviewed and approved by the Ethics Committee of Nanjing Medical University Affiliated Nanjing Municipal Center for Disease Prevention and Control. The methods employed in each HAHE survey were in line with the Declaration of Helsinki. Moreover, as only second-hand de-identified data from the HAHE study were analysed in the present comparative study, it was granted an exemption of ethical approval by the Ethics Committee of the Geriatric Hospital of Nanjing Medical University.
The data involved in this work are available upon reasonable request to corresponding author FX.
