Inclusive recruitment strategies for research with older adults: what actually works
A 92-year-old woman agreed to join my study on medication management. She signed the consent form, answered the phone twice, then disappeared. No answer for eleven days. When I finally reached her, she told me her son had decided the study "wasn't worth her time" and had quietly stopped passing on my calls. He never asked her. He just decided.
That single case reshaped how I think about inclusive recruitment strategies for research with older adults. The barriers are rarely about willingness. They are about gatekeepers, logistics, trust, and the assumption, baked into most protocols, that a research participant is a healthy adult with a smartphone, a car, and a clear calendar.
Recruiting older adults is not harder because older adults are difficult. It is harder because most recruitment designs were never built for them.
Key Takeaways
- Older adults are often labelled "hard to reach," but the real obstacle is study design, not the population.
- Five recruitment methods dominate practice, and each has a distinct failure mode you can plan around.
- Inclusion means reaching past the "well-connected, cognitively sharp, digitally fluent" subset of older adults.
- Consent, capacity, and gatekeeper dynamics need to be handled deliberately, not as an afterthought.
- Response rates improve most when you combine at least two recruitment channels and follow up in person.
Why older adults get labelled "hard to reach"
The term shows up in almost every gerontology paper. It is also misleading.
Older adults are not a hidden population in the way that, say, undocumented migrants or people in active addiction are. They are visible. They answer their landlines. They are often home. What makes them appear unreachable is a stack of practical and cognitive barriers that most recruitment protocols ignore.
Consider what typically blocks participation:
- Cognitive load: long consent forms, multi-page questionnaires, abstract instructions
- Physical frailty or chronic pain that makes travel to a research site exhausting
- No smartphone, or a smartphone used only for calls and photos
- A son, daughter, or caregiver who acts as an informal gatekeeper without realising it
- Distrust of institutions, often rooted in a lifetime of being talked over by doctors
- Hearing loss that turns a phone screening into a frustrating guessing game
- Simply not seeing themselves as "the kind of person who joins a study"
What counts as "hard to reach" in this population?
Three profiles come up again and again in my own fieldwork: people over 80 living alone, people with mild cognitive impairment, and people whose primary language differs from the study language. Each requires a different recruitment route.
The person who shows up easily, signs quickly, and completes every follow-up is usually younger, wealthier, and more educated than the population you actually need. If your sample looks like that, you have a recruitment bias problem, not a recruitment success.
What are the five recruitment strategies or methods?
Across the literature and my own projects, five approaches cover almost everything:
1. Community-based and venue recruitment
You go where older adults already are: senior centres, places of worship, retirement communities, pharmacies, day clinics, hair salons. The advantage is trust by association. The limitation is that you only reach people who are already socially active.
2. Clinical and registry-based recruitment
Hospitals, GP practices, and patient registries give you access to people with specific conditions. In my experience this is the fastest route to a defined clinical sample, but it comes with the heaviest gatekeeping. Physicians screen out anyone they judge "too frail," and you never hear about the people they excluded.
3. Snowball and peer referral
One participant introduces you to two more. This works remarkably well in tight-knit communities and among people who distrust formal institutions. The catch is sample homogeneity: you end up with a cluster of friends who share income, neighbourhood, and worldview.
4. Media, mail, and direct outreach
Newspaper ads, radio spots, letters, and posted flyers still pull participants, especially in regions with strong local media. Response rates are low per contact, but the reach is broad. A single well-placed notice in a local paper once got me more calls in a week than three months of email blasts.
5. Digital and telephone recruitment
Email lists, online panels, and phone screening. This is the method most researchers default to because it is cheap. It is also the one that most sharply excludes the exact population you are trying to include.
None of these five works alone. The pattern I keep seeing is that combining two channels, one physical and one remote, roughly doubles the yield compared to using either in isolation.
| Strategy | Best for | Main weakness |
|---|---|---|
| Community venues | Socially active older adults | Misses isolated people |
| Clinical / registry | Condition-specific samples | Heavy gatekeeping by staff |
| Snowball referral | Hard-to-trust communities | Sample becomes too similar |
| Media / mail | Broad geographic reach | Low response per contact |
| Digital / phone | Fast, low cost | Excludes the digitally excluded |
Inclusion means reaching past the easy sample
Inclusive recruitment is not the same as broad recruitment. It is a deliberate decision about who you refuse to leave out.
Language and cultural diversity
If your study materials exist only in the dominant language, you have already excluded a large share of older adults in most countries. Translated consent forms are the minimum. Bilingual recruiters who can sit down and explain the study in person are far better.
Digitally excluded participants
Roughly a third of adults over 75 in many high-income countries never use the internet for anything beyond basic calls. Building your entire recruitment funnel around an online form is a design choice with a predictable outcome. Offer a phone number. Offer a paper reply slip. It costs almost nothing and changes who shows up.
Socioeconomic reach
Recruiting through universities and well-funded clinics skews toward people with transport, time, and health literacy. Partnering with food banks, social housing associations, or mobile health units is slower but pulls in people who would otherwise never hear about your study.
Ethics and consent with older participants
Consent is where inclusive recruitment most often collapses.
Standard consent forms are written for someone with a graduate reading level and twenty uninterrupted minutes. For an 85-year-old with early hearing loss and fatigue, that form is a wall. I now run consent as a conversation, with a one-page plain-language summary, and I read it aloud. The refusal rate did not rise. It dropped.
Capacity is the harder question. Mild cognitive impairment does not automatically mean someone cannot consent, but it does mean you need a clear, documented process for assessing understanding. And when a caregiver is present, watch the dynamics closely. I have watched a daughter answer three consecutive questions on her mother's behalf while her mother sat silent. That is not consent. That is a proxy decision wearing the costume of one.
Confidentiality also needs care. Older adults in small communities are often more identifiable in a dataset than younger participants, simply because there are fewer people who match their profile.
What actually improves response rates
Three things have consistently moved the needle in my projects:
- Personal follow-up calls. A letter alone gets a low return. A letter followed by a call from the same named person converts far better.
- Removing every possible friction point: pre-paid return envelopes, flexible home visits, shorter forms.
- Working through trusted intermediaries, whether that is a GP, a faith leader, or a community nurse.
What did not work, in my experience: glossy brochures, generic email blasts, and anything that made the participant feel processed rather than invited.
Should you adopt a named recruitment framework?
Several structured models exist for recruiting older adults, and they can be useful checklists. My honest view is that frameworks help you plan but do not replace local knowledge. The specifics of your region, your population, and your topic matter more than any template.
What I would take from any model: recruit through more than one channel, design for the least digitally connected participant, and treat every gatekeeper as a factor to be managed rather than an obstacle to be resented.
That 92-year-old woman eventually joined the study. We called her directly, on her terms, at the time she chose. She completed every single follow-up. The barrier was never her. It was the assumption that she needed to come to us.
