Household Size and Age as Primary Drivers of COVID-19 Infection Among Priority Populations in Australia

This study of a Victorian cohort of priority populations reveals that younger age and larger household size, rather than culturally and linguistically diverse (CALD) status or socioeconomic factors, were the primary drivers of COVID-19 infection risk, highlighting the need for pandemic policies focused on housing density and protections for younger essential workers.

The Gist

Imagine the COVID-19 pandemic as a massive, invisible storm that swept through Australia. While the storm hit everyone, it didn't knock on every door with the same force. Some houses got battered, while others stayed relatively dry.

This study is like a team of detectives looking back at a specific group of people in Victoria, Australia, to figure out why some houses got flooded with infections while others didn't. They wanted to know: Was it because of who you are (your culture or money)? Or was it because of where you live and how old you are?

Here is the story of their findings, broken down into simple terms:

The Main Characters: Who Was in the Study?

The researchers gathered a group of 433 people. They weren't just random people; they were "priority populations." Think of them as the people on the front lines or those most vulnerable:

• Healthcare workers (the heroes in scrubs).
• People with chronic health issues (the ones needing extra care).
• People from Culturally and Linguistically Diverse (CALD) backgrounds (people who speak languages other than English at home).

The Big Question: What Made the Virus Stick?

The researchers asked a simple question: "How many times did you catch COVID?" (They counted confirmed cases from the start of the pandemic until the end of 2023).

They expected to find that being from a specific cultural background or having less money was the main reason people got sick. But the data told a different story.

The Two "Super Villains" of Infection

After crunching the numbers, the study found that two specific factors were the real drivers of infection, acting like a double-edged sword:

1. The "Young and Busy" Factor (Age)

• The Finding: Younger adults (ages 18–34) got infected much more often than older adults.
• The Analogy: Think of the virus as a rumor. Younger people are like the social butterflies at a party; they are out and about, working in shops and restaurants, and hanging out with friends. They can't always work from home. Older adults, on the other hand, were more likely to stay home or work in offices where they could isolate.
• The Result: Being young made you about 37% more likely to catch the virus compared to being over 55.

2. The "Crowded House" Factor (Household Size)

• The Finding: People living in medium-sized houses (2 to 5 people) got infected significantly more than people living alone.
• The Analogy: Imagine a virus is a fire. If you live alone, the fire has nowhere to go once it starts. But if you live in a house with 3 or 4 other people, the fire spreads from room to room instantly. Even if one person brings the virus home, everyone else in the house is likely to catch it too.
• The Result: Living in a house with 2–5 people made you about 42% more likely to get infected than someone living alone.

The Plot Twist: What Didn't Matter?

The researchers thought that Cultural Background (CALD) and Money (Socioeconomic Status) would be the biggest reasons for getting sick.

• The Expectation: They thought that because some cultures live in larger families or because poorer people can't afford to stay home, these groups would be hit hardest.
• The Reality: When they adjusted for the "Crowded House" and "Young Age" factors, culture and money stopped being the main culprits.
• The Explanation: It turns out that the reason some cultural groups had higher infection rates wasn't because of their culture itself, but because they were more likely to live in larger, crowded houses. Once you account for the size of the house, the culture itself didn't make the virus stick any harder.

Why Does This Matter? (The Takeaway)

This study is like a map for future storms. It tells us that if we want to stop the next pandemic from hurting people, we can't just say "protect the poor" or "protect specific cultures." We need to fix the physical conditions that let the virus spread.

The Recommendations:

1. For the Young: We need better protections for young workers in shops and restaurants. They need paid sick leave so they don't have to choose between their paycheck and staying home when they feel sick.
2. For Crowded Homes: We need to help people in big families isolate. If a family of five gets sick, they can't easily separate. We need policies that give them safe places to stay away from each other, or financial help so they don't have to keep working while sick.

In a Nutshell:
The virus didn't care about your accent or your bank account. It cared about how many people you live with and how much you have to go out into the world. To win the next battle, we need to make sure everyone has a safe, spacious place to stay and a job that lets them rest when they are sick.

Technical Summary

1. Problem Statement

The COVID-19 pandemic exacerbated global health disparities, with specific populations in Australia (particularly first-generation migrants and those in high-density housing) experiencing disproportionate disease burdens and mortality. While previous studies identified broad risk factors such as Cultural and Linguistically Diverse (CALD) status, low socioeconomic status (SES), and occupation, the relative contributions of these factors to single versus repeated infections remained incompletely understood.

This study sought to:

1. Characterize the risk factors for COVID-19 infection within a cohort of "priority populations" (healthcare workers, those with chronic conditions, and CALD communities).
2. Determine whether CALD status and SES independently predict infection risk after controlling for other sociodemographic and occupational variables, specifically household size and age.

2. Methodology

Study Design and Population

• Design: Cross-sectional analysis of data from the Optimise longitudinal cohort study.
• Setting: Victoria, Australia.
• Timeframe: Recruitment occurred between September 2020 and December 2021; data analysis covered the period from December 2019 to December 2023.
• Participants: The study focused on priority groups deemed at risk of acquiring COVID-19, experiencing severe illness, or suffering negative pandemic impacts.
  • Initial Recruitment: 779 participants.
  • Final Analysis Sample: 433 participants (after exclusions for missing baseline data, missing outcome data, and outliers reporting >20 infections).
  • Demographics: Median age 51 years; 75% female; 24% CALD status.

Data Collection

• Primary Outcome: Self-reported count of confirmed COVID-19 infections (PCR or Rapid Antigen Test positive) from Dec 2019 to Dec 2023. A confirmed infection required a positive test with at least one month between successive positives.
• Covariates: Baseline sociodemographic data including age, sex, CALD status (defined as non-Australian/UK/Irish birth or non-English home language), SES (based on SEIFA IRSD deciles), household size, chronic health conditions, occupation, and work environment.
• Validation: Self-reported counts were cross-validated against prospective monthly surveys and diaries to mitigate recall bias.

Statistical Analysis

• Model: Multivariate Poisson regression was used to analyze the relationship between baseline characteristics and the count of infections.
• Metrics: Unadjusted and Adjusted Incidence Rate Ratios (IRRs) with 95% Confidence Intervals (CIs) were calculated.
• Significance: Statistical significance was set at $p < 0.05$.
• Sensitivity Analysis: Work environment was reclassified based on longitudinal patterns (always home, always work, or mixed) to test stability against the baseline snapshot.

3. Key Results

Descriptive Statistics

• Infection Distribution: 25% reported zero infections, 48% reported one infection, and 27% reported two or more infections.
• Univariate Findings:
  • CALD Status: Associated with a 24% higher infection rate (IRR 1.24, 95% CI: 1.02–1.50).
  • Household Size: Strong positive association. Households of 2–5 people had a 53% higher rate (IRR 1.53), and 6+ people had a 71% higher rate (IRR 1.71) compared to those living alone.
  • Age: Strong inverse association. Older age groups (35–54 and 55+) had significantly lower infection rates compared to the 18–34 age group.
  • Chronic Conditions: Associated with a 27% lower infection rate (IRR 0.73), likely due to protective behaviors.

Multivariate (Adjusted) Analysis
After adjusting for all covariates, the following independent predictors remained significant:

• Age: Younger age (18–34 years) was a strong predictor of higher infection rates. Participants aged 55+ had a 37% lower infection rate compared to the 18–34 group (aIRR 0.63, 95% CI: 0.48–0.82).
• Household Size: Living in a household of 2–5 people was associated with a 42% higher infection rate compared to living alone (aIRR 1.42, 95% CI: 1.11–1.83).
• Non-Significant Factors: CALD status and Socioeconomic Status (SES) lost statistical significance after adjustment for age and household size. This suggests that the observed disparities in CALD and low-SES groups were largely mediated by their tendency to live in larger households and be younger, rather than cultural or economic factors acting independently.

4. Key Contributions

1. Disentangling Risk Factors: The study provides evidence that while CALD and low-SES groups suffered higher infection burdens, these were not independent risk factors when controlling for household density and age. The primary drivers were structural (housing) and demographic (age).
2. Focus on Repeated Infections: By analyzing the count of infections rather than just binary presence/absence, the study highlights that repeated infections are driven by the same structural factors (household size) as single infections.
3. Refutation of Cultural Determinism: The findings challenge the notion that cultural background (CALD status) is an intrinsic risk factor for transmission, suggesting instead that the risk is driven by the material conditions (overcrowding) often associated with specific migration patterns.
4. Policy-Relevant Stratification: The study identifies that "medium-sized" households (2–5 people) carry a significant risk, a nuance often missed in binary "overcrowded" vs. "not overcrowded" analyses.

5. Significance and Implications

Public Health Policy
The study argues that future pandemic preparedness must move beyond broad demographic categorizations (e.g., "CALD" or "Low SES") and target the material conditions that facilitate transmission:

• Housing Policy: Interventions should focus on enabling safe isolation for individuals in larger households. This includes providing financial support and facilities for quarantine away from the household to prevent intra-family transmission.
• Workplace Protections: Younger essential workers (retail, hospitality) are at highest risk due to inability to work remotely. Policies must ensure paid sick leave and economic security to allow for effective isolation without financial ruin.
• Targeted Messaging: Public health strategies should prioritize age-specific and household-size-specific interventions rather than solely focusing on cultural or linguistic barriers.

Limitations

• Sample Bias: The cohort was 75% female, which may have obscured associations with other factors (as women often exhibited more protective behaviors).
• Recall Bias: Infection counts were self-reported, though cross-validated.
• Generalizability: The cohort was a "priority population" sample and not fully representative of the general Victorian population.

Conclusion
The study concludes that in the Australian context, housing density and age are the dominant drivers of COVID-19 infection patterns. Addressing structural inequalities in housing and labor markets is more critical for future pandemic resilience than focusing solely on cultural or socioeconomic demographics.