The coronavirus disease 2019 (COVID-19) pandemic has affected >770 million people worldwide, causing >7 million deaths as of 31 December 2023.1 The period between July and September 2020 constituted the third wave of the pandemic in Hong Kong, resulted in >1.2 million reported cases between 23 January 2020 and 29 January 2023.2 The containment strategies implemented during the third wave included mandatory mask wear in public places, even when exercising in public outdoor areas; suspensions of public leisure facilities and private gyms; and the initiation of work-from-home arrangements.3 These strategies led to reductions in physical activity and daily movement, with the goal of viral containment. Furthermore, compulsory social distancing and suspension of dine-in services were included among the policies that could affect various dietary and lifestyle habits, although these methods were less stringent than approaches in cities under lockdown. Overall, the unprecedented public health crisis created many challenges for Hong Kong residents attempting to maintain healthy lifestyle habits. Nevertheless, few studies have examined dietary and physical activity habits in the general population during the COVID-19 pandemic.4 5 6

Considering that individuals with chronic diseases are more likely to develop severe cases of COVID-19, this study aimed to measure the prevalences of unhealthy dietary habits and physical inactivity levels in an adult Chinese population, to identify factors associated with their adoption of these dietary and physical activity habits, and to perform a time trend analysis comparing the proportions of the population that adopted healthy dietary habits, physical activity levels, and avoidance of smoking and alcohol consumption during the third wave of the COVID-19 pandemic.

We utilised a methodology similar to a previous population-based, random telephone survey conducted in Hong Kong.7 Two-stage sampling was performed, in which participants were recruited by trained interviewers through a telephone interview system based on telephone calls to landlines identified by random digit dialling. The sample population was randomly selected by the Centre for Health Behaviours Research at The Chinese University of Hong Kong. Calls were made during typical office hours, 9 am to 5 pm, Monday through Saturday between 7 and 31 October 2020. Three attempts were made if the call initially was not answered. Territory-wide, any Chinese adults aged ≥18 years who could communicate in Chinese via telephone were eligible to participate. Assuming an outcome variable rate of 35%, at least 1456 participants were required to achieve a precision level of 2.5% from the following formula:
where ‘p’ stands for proportion and ‘N’ stands for sample size.

The interviews were performed using a fieldwork manual highlighting standard operating procedures by a team of trained interviewers and supervised by an experienced project coordinator throughout the study. The characteristics of survey participants are shown in Table 1.

The survey consisted of five sections: (1) socio-demographic details (age, sex, marital status, education level, job status, household income, and receipt of comprehensive social security assistance); (2) clinical information (eg, presence of chronic diseases); (3) smoking (current daily amount/ex-/non-smoker) and alcohol consumption habits (daily amount in the preceding 7 days); (4) dietary screening via the Hong Kong Diet Score (HKDS), using a validated scale that contained nine items assessing the participant’s daily consumption of nine food groups in the preceding 7 days; and (5) level of physical activity in the preceding 7 days, as determined by a Chinese version of the 7-item International Physical Activity Questionnaire Short Form (IPAQ-C).

The traditional Mediterranean diet is well-defined and has been positively associated with favourable health outcomes.8 9 10 The Mediterranean diet score is used to measure compliance with a traditional Mediterranean diet. This scoring system has been widely utilised in studies that measure Mediterranean diet adherence or adaptation as an indicator of healthy dietary choices. In this study, we developed the HKDS, a dietary screener that contained nine items assessing dietary intake of nine food groups (alcohol, legumes, grains, fruits, vegetables, meats, dairy, red/orange vegetables, and fatty fish) in the preceding 7 days. The screener incorporated key traditional Greek diet characteristics, known as the Mediterranean diet score of de Groot et al,8 which were also used in a study of Hong Kong Chinese by Woo et al (Table 2).11 The original 8-item survey was modified by removing the ratio of monounsaturated fatty acids to saturated fatty acids and replacing ethanol with alcohol. Dietary fatty acids and ethanol are widely distributed among various food groups; they are typically assessed through weighted foods, which are unlikely to be accurately determined using a single question in a telephone interview. Two additional items were included regarding carotenoid-rich and omega-3–rich food intake based on the Hong Kong Centre for Food Safety Recommended Nutrient Intake for vitamin A12 and the World Health Organization recommendation for omega-3. Both nutrients are inversely associated with incidence of non-communicable diseases (NCDs). For each item, consumption at or above the recommended amount was scored as 1 point and 0 points otherwise; however, for ethanol, consumption below the specified amount was scored as 1 point and 0 points otherwise. Each participant received a total score of 0 to 9; a score of ≥5 was considered high. A pilot survey was conducted with a convenience sample of 23 participants. Intraclass correlation coefficient estimates and 95% confidence intervals (CIs) were determined using a two-way mixed-effects model to assess internal consistency regarding the number of serves (ie, serving sizes of the food group consumed) reported. The intraclass correlation coefficient was 0.87, indicating good reliability. Cohen’s κ was calculated to evaluate agreement between test and retest scores. Agreement between the two tests was fair (κ=0.24, 95% CI=-0.15 to 0.63; P=0.239).

The IPAQ-C score was regarded as a categorical variable indicating exercise level based on the frequency and intensity of physical activity: (1) low (total activity <600 metabolic equivalent of task [MET]–minutes/week), (2) moderate (total activity ≥600 MET-minutes/week), or (3) high (total activity >3000 MET-minutes/week).

Finally, an unhealthy lifestyle score (0 to 4) was assigned to each participant based on a composite outcome involving low HKDS, physical inactivity, current smoking habit, and alcohol consumption; each unhealthy habit contributed 1 point to the score.

We used SPSS software (Windows version 26.0; IBM Corp, Armonk [NY], United States) for data analysis. Descriptive analyses were performed regarding the participants’ socio-demographic details, clinical information (eg, presence of chronic diseases), and the HKDS. The primary outcome variables included: (1) unhealthy dietary habits (low HKDS score); (2) suboptimal physical activity (low IPAQ-C score, indicating low exercise level); and (3) unhealthy lifestyle score (≥2). Univariable logistic regression was performed to examine associations between socio-demographic variables and each of the first two outcome variables. Multivariable logistic regression was modelled by controlling for covariates with P values <0.20 in univariable regression analysis, a cut-off level commonly used in public health research. For example, Torenfält and Dimberg13 utilised this approach when evaluating stroke and death in middle-aged Swedish men. The approach was also used in a French study14 concerning medical features of patients with COVID-19 and influenza. Additionally, linear regression analysis was conducted in the present study to examine associations between socio-demographic variables and the unhealthy lifestyle score. Time trends for various food intake, physical inactivity, current smoking, and alcohol consumption statuses were evaluated; the prevalences of these lifestyle habits were compared with population-wide figures from governmental reports over the past two decades using the Chi squared test for heterogeneity. P values <0.05 were considered statistically significant.

The Centre for Health Protection has been conducting health surveys periodically since 2003 to collect information about health and lifestyle-related behaviours, as well as practices related to the prevention of NCDs among residents aged ≥15 years.15 The resulting reports have presented key findings concerning physical activity, dietary habits, alcohol consumption, and smoking habits, as well as other self-care practices. We gathered relevant findings from 14 governmental reports covering the period from 2003-2004 to 2018-2019 (calendar years with the most updated figures)15 to perform trend analysis of fruit and vegetable consumption, physical activity level, smoking, and alcohol consumption status among Hong Kong residents. These results were compared with the findings of the present study; adjustments were solely performed for sex because the age distribution was limited in all but the most recent reports.

In total, 2551 individuals were contacted for a telephone interview and 1500 participated; the response rate was 58.8%. Most interviewed individuals were older adults (≥65 years, 66.7%), women (65.6%), and married (77.9%). Of the participants, 40.1% were engaged in professional and office work; only 16.3% had attained a tertiary education or higher. About 16% of participants reported a household income ≥HK$30 000, whereas 41.1% reported a household income <HK$10 000. Health status was predominately self-reported as average (42.5%) or above average (51.6%). More than half of the participants (54.5%) were undergoing long-term medical management or were taking medications for chronic diseases; the most common chronic conditions were diabetes mellitus (21.0%) and hypertension (42.7%) [Table 1].

Dietary habits, as measured by the HKDS (score range, 2-9), were classified as high scoring (5-9) or low scoring (0-4). Approximately 51% of participants had a low score (Table 1), suggesting minimal adherence to the traditional Chinese eating pattern; these participants were mainly younger individuals (aged ≤34 years, 66%) and men (58%) [Table 2]. Greater proportions of participants with lower income, current smokers, and current drinkers had low scores according to the HKDS (54%, 69%, and 67%, respectively) [Table 2]. Participants with chronic diseases had various HKDS results; <50% of patients with renal diseases and diabetes had a high score, and this result indicated that they had a poor dietary habits.

A greater risk of practising unhealthy dietary habits (low HKDS) was associated with male sex (adjusted odds ratio [aOR]=1.31, 95% CI=1.03-1.67), non-married status (ie, single/divorced/widowed) [aOR=1.56, 95% CI=1.20-2.03], a diagnosis of diabetes (aOR=1.53, 95% CI=1.15-2.03), and alcohol consumption (aOR=1.76, 95% CI=1.17-2.64) [Table 3].

Among all participants, 35.5%, 54.0% and 10.5% had low, moderate, and high levels of physical activity, respectively (Table 1). Participants who were non-married (aOR=1.66, 95% CI=1.23-2.22), undergoing long-term management of medical diseases (aOR=1.65, 95% CI=1.08-2.54), had diabetes (aOR=1.39, 95% CI=1.02 to 1.89), and had renal diseases (aOR=9.32, 95% CI=2.06-42.25) exhibited greater likelihood of physical inactivity (Table 4).

Few participants were current daily smokers (2.8%) and alcohol drinkers (8.5%) [Table 1].

Factors associated with an unhealthy lifestyle score are presented in Table 5. Male sex (beta coefficient [β]=0.25, 95% CI=0.16-0.34), non-married status (β=0.19, 95% CI=0.08-0.29), manual work (β=0.17, 95% CI=0.07-0.27), self-reported poor or very poor health status (β=0.26, 95% CI=0.07-0.45), a diagnosis of diabetes (β=0.31, 95% CI=0.20-0.41), and a diagnosis of renal disease (β=0.92, 95% CI=0.53-1.31) increased the likelihood of poor lifestyle habits. Housewife or retired status (β=-0.15, 95% CI=-0.25 to -0.04) and a higher household income (≥HK$30 000; β=-0.20, 95% CI=-0.33 to -0.08) decreased the likelihood of poor lifestyle habits.

The Health Behaviour Survey, with a response rate of 70.8% in the 2018/2019 report, is a population-based fieldwork study conducted by the Centre for Health Protection of the Department of Health.16 In that survey, female participants comprised 52.7% of the sample, compared with 65.6% in the present telephone survey. The age-group with the largest proportion of participants in the Survey was 65 to 74 years (36.7% vs 11.0%), which might have influenced the sex ratio.

Time trend analysis showed that the proportion of surveyed Hong Kong residents eating five daily servings of fruits and vegetables declined for both sexes in general (Fig a). Similarly, a significantly smaller proportion of participants reported walking >10 minutes for ≥5 days per week, and this proportion has continued to decline since 2016 (Fig b). There was a gradual decrease in the number of participants with a moderate or high level of physical activity. Despite a notable peak in 2019, there was a decline in 2020 with <60% of participants reportedly engaging in these physical activities. Finally, significantly smaller proportion of the study participants reported not currently smoking or consuming alcohol, compared with the proportions in previous population-based surveys (2010-2019) [Fig c and d].

In this population-based study of 1500 Hong Kong residents during the third wave of the COVID-19 pandemic, we found that the proportion of people with healthy food intake (ie, daily consumption of five servings of fruits and vegetables) has decreased since 2003; although a slight increased was observed in 2020, it was still below the overall average. Additionally, we found that the prevalence of low physical activity has gradually increased. In contrast, the rates of smoking and alcohol consumption were below the rates observed in pre-pandemic population-based surveys. Men and women in various age-groups had dietary habits less adherent to the traditional Chinese eating pattern, as measured by the HKDS, than 20 years prior. Adherence to the traditional eating pattern was significantly lower among male participants, single individuals, low-income participants, alcohol drinkers, individuals with low physical activity, and patients with diabetes mellitus or renal disease. We also found that men, individuals with non-married status, manual workers, individuals with self-perceived poor or very poor health status, and patients with diabetes and renal disease had a greater likelihood of poor lifestyle habits.

Worldwide, insufficient intake of fruits and vegetables and inadequate physical activity have been attributed to 34% and ≥20% of NCDs, respectively.17 18 The incidence of chronic diseases (eg, cancers, diabetes, and cardiovascular disease) in Hong Kong has increased by 60% in the past two decades.19 Although processed food intake was not assessed in this study, the HKDS results indicated very low intakes of legumes and red/orange vegetables among Hong Kong residents; a previous study showed that fruit intake in Hong Kong is among the lowest levels worldwide.20 The report of Health Behaviour Survey 2018/2019 revealed that 95.6% of surveyed participants had inadequate daily fruit and vegetable intake, based on World Health Organization recommendations.16 The present study showed further reduction, such that the proportion of Hong Kong residents consuming the recommended (≥5) daily servings of fruits and vegetables declined by 3.9% in the past two decades. Additionally, 10% of the population reported consuming more than one daily serving of processed meat, and the frequency of processed meat consumption increased by 3.1%.16 Another known risk factor for NCDs, overweight or obesity, affected approximately half of the Hong Kong population aged 15 to 84 years in 2015; this value was slightly (2%) below the global average.21

Disruptions to usual routines can lead to new dietary behaviours. For example, dine-in restrictions at restaurants and bars during the pandemic led to greater use of food delivery services, and take-away food is often less healthy.22 However, we observed increased fruit and vegetable consumption after the third wave of the COVID-19 outbreak. This finding is supported by the work of Wang et al,23 which showed that fruit and vegetable consumption increased. Collectively, the disease prevention and control policies that prohibited group gatherings and shortened the operating hours of bars and clubs (if such facilities were not entirely closed) could have significantly reduced smoking and alcohol consumption in Hong Kong during the pandemic. Additionally, the lower level of physical activity could be related to the closure of public sports avenues and restricted access to sports facilities.

The dietary habits of Hong Kong residents have changed from the traditional Chinese diet to a fast-paced dining experience involving convenient, processed foods with limited diversity. Hong Kong has a population of >7.5 million people, and 90% of its food supply is imported from other countries.24 Woo et al11 concluded that, despite geographical and cultural differences, traditional Chinese dietary habits were conceptually similar to the health-promoting Mediterranean diet. In 2001, high overall adherence to the Mediterranean diet was observed across all age-groups in Hong Kong, except for younger populations and men.11 In contrast, the present study showed that a smaller proportion of participants in all age and sex groups had high overall adherence to the Mediterranean diet. In this study, intriguingly, women consistently exhibited higher Mediterranean diet score (71%) and HKDS (53%; P<0.005) results, compared with their male counterparts. Moreover, the present study revealed lower prevalences of some chronic diseases among women: diabetes, cardiovascular disease, liver disease, and renal disease. The contemporary diets in modern Hong Kong and many developed regions have low fibre content and high processed food content; they also include food additives, refined sugar, and hydrogenated fats.25 The subtle but consistent westernisation of dietary habits appears to be detrimental for residents who consistently consume these food items.

A lack of colourful vegetables and fruits may reduce the diversity of beneficial gut microbiota,26 although leafy greens such as pak choy (Chinese cabbage), choy sum (Chinese flowering cabbage), and Chinese kale are staple foods in the Hong Kong diet throughout the year. Moreover, in a review of literature concerning exercise and gut microbial composition, Mitchell et al27 found that exercise alters gut microbiota; however, the direction of apparent change has varied among studies. Increases in butyrate-producing bacteria and faecal butyrate concentrations, with protective anti-inflammatory effects and the potential to enhance anti-infection immunity, have been observed among physically active adults.28 The mechanisms are unclear but the benefits of adopting a lifestyle with a diverse diet and physical activity consistently create an optimal environment for gut microbiota.

The large sample size and random sampling design of this territory-wide survey were strengths that enhanced the validity of the findings. However, this study had several limitations that should be addressed. First, cause-and-effect relationships between the COVID-19 pandemic and changes in lifestyle habits could not be established because of the cross-sectional approach. Individuals in home quarantine during the third wave of the pandemic might have experienced temporary changes in lifestyle habits; a prospective observational study and longer trend analysis are needed to facilitate long-term comparisons. Factors other than the pandemic (eg, mental wellness) could also have affected lifestyles among the study population. Although the present study utilised a random sampling strategy, non-response and selection biases were possible because younger segments of the Hong Kong population did not use landlines during the study period. Furthermore, response and social desirability biases may have been present in this telephone survey. Nevertheless, the high response rate and anonymous nature of this survey may have minimised these potential biases. Additionally, generalisation of the study findings should be performed with caution because the survey only included a single Chinese population. Considering that the participants’ characteristics differed from the general population, the findings might not be directly applicable to the general public. Moreover, the survey was conducted in 2020, and lifestyle habits in the general population might have changed throughout the pandemic. Finally, the low case numbers of some self-reported diseases, such as renal disease and cancer, may have resulted in type II error.

This representative population-based survey revealed that larger proportions of the general population had unhealthy lifestyles, including dietary habits and physical inactivity, during the COVID-19 pandemic than during pre-pandemic years. There is an urgent need to formulate and implement effective public health strategies at both individual and organisational levels. The encouragement of healthy lifestyles through evidence-based health promotion programmes is essential. This encouragement could be conveyed to communities through organised and concerted efforts by the government and relevant stakeholders. Future studies should evaluate the effectiveness of various interventions and approaches to achieve these important goals.

Concept or design: WYY Lin, MCS Wong.
Acquisition of data: WYY Lin, J Huang, Y Bai.
Analysis or interpretation of data: WYY Lin, J Huang, Y Bai.
Drafting of the manuscript: WYY Lin, MCS Wong.
Critical revision of the manuscript for important intellectual content: All authors.

All authors had full access to the data, contributed to the study, approved the final version for publication, and take responsibility for its accuracy and integrity.

As editors of the journal, MCS Wong and J Huang were not involved in the peer review process. Other authors have disclosed no conflicts of interest.

This research received no specific grant from any funding agency in the public, commercial, or not-for-profit sectors.

This research was approved by the Survey and Behavioural Research Ethics Committee of The Chinese University of Hong Kong (Ref No.: SBRE-20-099). The interviewees provided informed consent after they were briefed on the study purpose and being assured of the confidentiality measures in place.

1. World Health Organization. WHO Coronavirus Disease (COVID-19) Dashboard. Available from: https://covid19.who.int/. Accessed 19 Jan 2024.

2. Centre for Health Protection, Department of Health, Hong Kong SAR Government. Situation of COVID-19 (23 January 2020 to 29 January 2023). 2023 July 27. Available from: https://www.chp.gov.hk/files/pdf/local_situation_covid19_en.pdf. Accessed 2 Feb 2024.

3. Chief Executive’s Office, Hong Kong SAR Government. Fighting the virus for six months battling another wave of the epidemic together. 2020 Jul 25. Available from: https://www.ceo.gov.hk/archive/5-term/eng/pdf/article20200725.pdf. Accessed 6 Apr 2022.

4. Mattioli AV, Sciomer S, Cocchi C, Maffei S, Gallina S. Quarantine during COVID-19 outbreak: changes in diet and physical activity increase the risk of cardiovascular disease. Nutr Metab Cardiovasc Dis 2020;30:1409-17. Crossref

5. Bin Mahfoz TM, Shaik RA, Ahmad MS, Alzerwi NA, Almutairi R. A nationwide survey to assess COVID-19’s impact on health and lifestyle in Saudi Arabia. Eur Rev Med Pharmacol Sci 2022;26:4092-101. Crossref

6. Mitchell JJ, Bu F, Fancourt D, Steptoe A, Bone JK. Longitudinal associations between physical activity and other health behaviours during the COVID-19 pandemic: a fixed effects analysis. Sci Rep 2022;12:15956. Crossref

7. Wang JX, Huang J, Cheung CS, Wong WN, Cheung NT, Wong MC. Adoption of an electronic patient record sharing pilot project: cross-sectional survey. J Med Internet Res 2020;22:e13761. Crossref

8. de Groot LC, van Staveren WA, Burema J. Survival beyond age 70 in relation to diet. Nutr Rev 1996;54:211-2. Crossref

9. Fung TT, Rexrode KM, Mantzoros CS, Manson JE, Willett WC, Hu FB. Mediterranean diet and incidence of and mortality from coronary heart disease and stroke in women. Circulation 2009;119:1093-100. Crossref

10. Bamia C, Lagiou P, Buckland G, et al. Mediterranean diet and colorectal cancer risk: results from a European cohort. Eur J Epidemiol 2013;28:317-28. Crossref

11. Woo J, Woo KS, Leung SS, et al. The Mediterranean score of dietary habits in Chinese populations in four different geographical areas. Eur J Clin Nutr 2001;55:215-20. Crossref

12. Centre for Food Safety, Hong Kong SAR Government. Nutrient and health—maintain optimal nutrient intake. 2008. Available from: https://www.cfs.gov.hk/english/multimedia/multimedia_pub/multimedia_pub_fsf_28_02.html. Accessed 30 Sep 2020. Crossref

13. Torenfält I, Dimberg L. Stroke and death — findings from a 25-year follow-up of a cohort of employed Swedish middle-aged men of the Coeur study. J Public Health 2022;30:1713-24. Crossref

14. Faury H, Courboulès C, Payen M, et al. Medical features of COVID-19 and influenza infection: a comparative study in Paris, France. J Infect 2021;82:e36-9. Crossref

15. Center for Health Protection, Department of Health, Hong Kong SAR Government. Non-communicable Disease Branch. Publications. Available from: https://www.chp.gov.hk/en/resources/29/100057.html. Accessed 31 Jan 2024.

16. Centre for Health Protection, Department of Health, Hong Kong SAR Government. Report of Health Behaviour Survey 2018/19. Available from: https://www.chp.gov.hk/files/pdf/report_of_health_behaviour_survey_2018_en.pdf. Accessed 29 Jan 2021.

17. World Health Organization. Global health risks: mortality and burden of disease attributable to selected major risks. 2009. Available from: https://apps.who.int/iris/handle/10665/44203. Accessed 29 Jan 2021.

18. World Health Organization. Physical activity. 2022. Available from: https://www.who.int/news-room/fact-sheets/detail/physical-activity. Accessed 22 Jan 2024.

19. Hospital Authority. Overview of Hong Kong Cancer Statistics of 2018. 2020. Available from: https://www3.ha.org.hk/cancereg/pdf/overview/Overview%20of%20HK%20Cancer%20Stat%202018.pdf. Accessed 30 Sep 2020.

20. Micha R, Khatibzadeh S, Shi P, et al. Global, regional and national consumption of major food groups in 1990 and 2010: a systematic analysis including 266 country-specific nutrition surveys worldwide. BMJ Open 2015;5:e008705. Crossref

21. World Health Organization. Obesity and overweight. 2021. Available from: https://www.who.int/news-room/fact-sheets/detail/obesity-and-overweight. Accessed 23 Jun 2021.

22. Nguyen TH, Vu DC. Food delivery service during social distancing: proactively preventing or potentially spreading coronavirus disease-2019? Disaster Med Public Health Prep 2020;14:e9-10. Crossref

23. Wang J, Yeoh EK, Yung TK, et al. Change in eating habits and physical activities before and during the COVID-19 pandemic in Hong Kong: a cross-sectional study via random telephone survey. J Int Soc Sports Nutr 2021;18:33. Crossref

24. Research Office, Legislative Council Secretariat. Fact sheet: regulation of imported food in Hong Kong. 2016. Available from: https://www.legco.gov.hk/research-publications/english/1516fsc14-regulation-of-imported-food-in-hong-kong-20160226-e.pdf. Accessed 22 Jan 2024.

25. Jew S, AbuMweis SS, Jones PJ. Evolution of the human diet: linking our ancestral diet to modern functional foods as a means of chronic disease prevention. J Med Food 2009;12:925-34. Crossref

26. Garcia-Mantrana I, Selma-Royo M, Alcantara C, Collado MC. Shifts on gut microbiota associated to Mediterranean diet adherence and specific dietary intakes on general adult population. Front Microbiol 2018;9:890. Crossref

27. Mitchell CM, Davy BM, Hulver MW, Neilson AP, Bennett BJ, Davy KP. Does exercise alter gut microbial composition? a systematic review. Med Sci Sports Exerc 2019;51:160-7. Crossref

28. Allen JM, Mailing LJ, Niemiro GM, et al. Exercise alters gut microbiota composition and function in lean and obese humans. Med Sci Sports Exerc 2018;50:747-57. Crossref