In brief
In a study of 4,023 nursing home residents across seven European countries and Israel, 32% had depressive symptoms but only about 47% of those residents received antidepressants; overall, 35. 6% of residents used antidepressants, mostly SSRIs, with prescribing linked to both resident diagnoses and facility...
What this article is about
Quick Answer
In a study of 4,023 nursing home residents across seven European countries and Israel, 32% had depressive symptoms but only about 47% of those residents received antidepressants; overall, 35.6% of residents used antidepressants, mostly SSRIs, with prescribing linked to both resident diagnoses and facility characteristics.
Student takeaways
Key Takeaways
- Among 4,023 nursing home residents assessed across seven European countries and Israel, 32% had depressive symptoms, but only 47.2% (n = 598) of those residents were receiving an antidepressant medication.
- Overall antidepressant use in the full sample was 35.6% (n = 1,431 residents), with prevalence varying notably by country.
- Among residents taking an antidepressant, 59.9% were prescribed an SSRI, the most commonly used class.
- Diagnoses of anxiety, depression, or bipolar disorder, along with pain, falls, and higher social engagement, were associated with a greater likelihood of antidepressant use.
- Age 85 years or older, living in a rural facility, and a diagnosis of schizophrenia were each associated with a lower likelihood of being prescribed an antidepressant.
Student summary
Why This Research Matters
Depression is common in nursing homes, but how often is it actually treated with medication, and who gets prescribed an antidepressant? This 2020 study, published in BMC Geriatrics, used data from the SHELTER study (Services and Health for Elderly in Long TERm care), an observational cohort that assessed nursing home residents in seven European countries (Czech Republic, England, Finland, France, Germany, Italy, and the Netherlands) plus Israel. Researchers used the interRAI Long-Term Care Facility instrument, a standardized clinical assessment tool that captures function, mental health, social engagement, and medication use, to build a cross-sectional picture of antidepressant prescribing across 4,023 residents entering the study.
The headline numbers are striking. About 32% of residents had depressive symptoms on assessment, yet only around 47% of those residents were actually receiving an antidepressant. Overall, 35.6% of the full sample (1,431 residents) were taking an antidepressant medication of some kind, whether or not they had documented depressive symptoms. Among residents on an antidepressant, 59.9% were prescribed a selective serotonin reuptake inhibitor (SSRI), the class generally considered safer for older adults compared with older tricyclic antidepressants. Prevalence varied noticeably by country, which suggests prescribing habits are shaped as much by local practice culture as by resident need.
The study also looked at which resident and facility characteristics were linked to antidepressant use. Residents with a diagnosis of anxiety, depression, or bipolar disorder were more likely to be on an antidepressant, as were residents with pain, a history of falls, or higher levels of social engagement. On the facility side, having a pharmacist on staff was associated with higher antidepressant use. Interestingly, the factors that reduced the likelihood of being prescribed an antidepressant were being 85 years or older, living in a rural facility, and having a diagnosis of schizophrenia. The authors framed this pattern as evidence that prescribing decisions are shaped by a mix of clinical presentation and facility-level resources and habits, not simply by whether a resident meets criteria for depression.
Why does this matter for nursing students and future clinicians? First, it's a reminder that depressive symptoms and antidepressant treatment don't line up neatly. Under-treatment (residents with symptoms not receiving medication) and possible over-treatment or off-label use (antidepressants prescribed without a clear mood-disorder rationale) can both be happening in the same population. Second, the finding that oldest-old residents were less likely to receive antidepressants raises an important appraisal question: is this appropriate caution given medication risk in very old age, or is it a sign that depression in the oldest residents is under-recognized and under-treated? The study cannot answer that question because it only captures a snapshot in time; it cannot establish why prescribing happened, only that it was associated with certain characteristics.
As a cross-sectional analysis, the study has real limits worth naming as you critically appraise it. It cannot show whether antidepressants actually helped residents, since there is no follow-up data on outcomes, symptom severity, or side effects. Depressive symptoms were screened using a rating scale rather than a full clinical diagnostic interview, so some misclassification is possible. Facilities were not randomly selected, and the seven countries plus Israel are not necessarily representative of long-term care systems elsewhere, including Canada. Nonetheless, the study is a useful, source-grounded starting point for thinking critically about psychotropic prescribing patterns in long-term care, a topic directly relevant to nurses working in geriatric and long-term care settings who are often the ones monitoring for treatment response, side effects, and unmet mental health needs. The authors themselves call for future longitudinal research to establish the actual efficacy and safety of antidepressant use in this population, which is exactly the kind of evidence gap nursing students should learn to recognize when reading prescribing-pattern research.
Source abstract
Study Overview
Abstract Background Late-life depression is common among older adults living in nursing homes (NHs). Over the last 30 years there has been an increase in the rates of prescription of antidepressant medications across all ages, with the largest rise reported in older adults. This study aimed to describe the pattern of antidepressant medication use among NH residents from 7 European countries and Israel and to examine patient and facilities characteristics that may account for it. Methods We conducted a cross-sectional analysis of data from the SHELTER study, an observational longitudinal cohort study that collected comprehensive resident data using the interRAI Long-Term Care Facility instrument in 7 European Countries and Israel. Descriptive statistics were used to examine sample characteristics. Potential correlates of antidepressant medication use were identified using multiple logistic regression modeling. Results Among 4023 residents entering the study, 32% had depressive symptoms and nearly half of these individuals used antidepressants. Antidepressant medication use varied by country, with a prevalence in the overall sample of 35.6% (n = 1431). Among antidepressant users, 59.9% were receiving selective serotonin reuptake inhibitors (SSRI). The strongest correlates of antidepressant use included reported diagnosis of anxiety, depression, bipolar disorder, pain, falls and high level of social engagement. Age over 85 years, living in facilities located in rural areas and a diagnosis of schizophrenia reduced the likelihood of being prescribed with an antidepressant. Conclusions A large proportion of residents in European long-term care facilities receive antidepressant medications. The decision to prescribe antidepressants to NH residents seems to be influenced by both patient and facility characteristics. Future longitudinal studies should evaluate the efficacy and safety of antidepressant use in NHs thus providing evidence for recommendations for clinical practice.
Evidence appraisal
Main Findings
- Among 4,023 nursing home residents assessed across seven European countries and Israel, 32% had depressive symptoms, but only 47.2% (n = 598) of those residents were receiving an antidepressant medication.
- Overall antidepressant use in the full sample was 35.6% (n = 1,431 residents), with prevalence varying notably by country.
- Among residents taking an antidepressant, 59.9% were prescribed an SSRI, the most commonly used class.
- Diagnoses of anxiety, depression, or bipolar disorder, along with pain, falls, and higher social engagement, were associated with a greater likelihood of antidepressant use.
- Age 85 years or older, living in a rural facility, and a diagnosis of schizophrenia were each associated with a lower likelihood of being prescribed an antidepressant.
Practice transfer
Clinical Relevance
- Nurses in long-term care should be aware that depressive symptoms and antidepressant treatment often do not align, meaning some residents with symptoms may be undertreated and should be flagged for reassessment.
- Because SSRIs were the predominant class prescribed, nurses should stay alert for SSRI-specific concerns in older adults, such as falls risk, hyponatremia, and GI bleeding risk, when monitoring residents.
- The association between falls and antidepressant use suggests nurses should consider medication review as part of falls-risk assessment in residents already prescribed antidepressants.
- Lower antidepressant prescribing in residents 85 and older may reflect appropriate caution, but nurses should still screen this group for undetected or undertreated depressive symptoms rather than assuming absence of need.
- Country- and facility-level variation in prescribing patterns (e.g., pharmacist presence) highlights that practice culture, not just resident presentation, shapes treatment decisions, supporting the case for standardized depression screening protocols in long-term care.
Faculty notes
Educational Relevance
This BMC Geriatrics article (Giovannini et al., 2020) reports a cross-sectional secondary analysis of the SHELTER study, an interRAI-based observational cohort assessing 4,023 nursing home residents across seven European countries (Czech Republic, England, Finland, France, Germany, Italy, Netherlands) and Israel. The paper is well-suited as a teaching case for appraising descriptive/correlational pharmacoepidemiology studies in long-term care, and for discussing the gap between symptom prevalence and treatment prevalence in geriatric mental health.
Methodologically, resident-level data were collected using the interRAI Long-Term Care Facility instrument, a standardized, multidimensional assessment covering function, cognition, mood, social engagement, and medication use. Depressive symptoms were identified via an embedded rating scale rather than a structured diagnostic interview, an important distinction for students to note when interpreting 'depressive symptoms' versus 'clinical diagnosis of depression.' Correlates of antidepressant use were modeled with multiple logistic regression, yielding both resident-level (diagnosis of anxiety, depression, bipolar disorder, pain, falls, social engagement) and facility-level (pharmacist presence) predictors of higher antidepressant use, and factors associated with lower use (age 85+, rural facility location, schizophrenia diagnosis).
The core findings worth foregrounding in class discussion: 32% of residents had depressive symptoms, but only about 47% of those residents were receiving an antidepressant, an undertreatment gap. Overall sample prevalence of antidepressant use was 35.6% (n=1,431), with SSRIs comprising 59.9% of prescriptions among users, consistent with general geriatric prescribing guidance favoring SSRIs over tricyclics in older adults. Country-level variation in prevalence was substantial, which the authors interpret as evidence of practice-pattern variation independent of resident case-mix.
For appraisal-focused teaching, several limitations deserve emphasis: (1) the cross-sectional design precludes causal inference about why specific residents were or were not prescribed antidepressants; (2) facilities were a convenience sample rather than randomly selected, limiting generalizability, including to non-European/Israeli long-term care systems such as Canada's; (3) symptom severity and treatment duration were not captured, so the clinical adequacy of prescribed regimens cannot be assessed; (4) non-pharmacological depression management (e.g., therapy, activity programming) was not measured, so the study cannot speak to whether antidepressants were used as first-line or adjunct treatment; and (5) the reliance on a screening scale rather than diagnostic confirmation introduces potential misclassification of the 'depressive symptoms' group.
Discussion prompts that work well with this article: ask students to distinguish association from causation in the odds-ratio findings (e.g., is the association between pharmacist presence and higher antidepressant use a marker of better psychiatric care access, or of over-medicalization?); ask them to consider why oldest-old residents were less likely to be prescribed antidepressants, and whether this reflects appropriate deprescribing caution or undertreatment; and use the undertreatment gap (32% symptomatic vs. ~47% of those treated) as a launch point for discussing screening-to-treatment pathways in long-term care. The paper's own call for longitudinal efficacy/safety research is a useful example of how a descriptive study appropriately scopes its claims and defers causal or outcome questions to future designs, a habit worth modeling for students learning to write their own literature critiques.
Critical appraisal
Limitations
- The cross-sectional design means associations between resident/facility characteristics and antidepressant use cannot be interpreted as causal.
- Depressive symptoms were identified using a screening rating scale rather than a structured clinical diagnostic interview, which may misclassify some residents.
- Participating facilities were not randomly selected, limiting how far findings can be generalized to all nursing homes, including those outside the seven European countries and Israel studied.
Classroom use
Discussion Questions
- Why might only about 47% of residents with depressive symptoms have been receiving antidepressant treatment, and what factors could explain this undertreatment gap?
- What are the clinical trade-offs between SSRIs and other antidepressant classes in nursing home residents, and how might these trade-offs explain the 59.9% SSRI prescribing rate found in this study?
- How should a nurse interpret the finding that residents 85 and older were less likely to be prescribed antidepressants: as appropriate caution, undertreatment, or something else?
- What might explain the association between pharmacist presence in a facility and higher antidepressant prescribing rates?
- Why did residents with a schizophrenia diagnosis have lower odds of being prescribed an antidepressant, and what clinical reasoning might be behind this pattern?
- How might rural versus urban facility location influence access to mental health assessment and treatment for nursing home residents?
- Given that this is a cross-sectional study, what kind of study design would be needed to determine whether antidepressants actually improve outcomes for these residents?
- How could a nurse use the interRAI Long-Term Care Facility instrument's mental health domain to identify residents who may be undertreated for depression?
- What role might falls risk assessment play alongside antidepressant medication review in long-term care settings, given the association found in this study?
- How generalizable do you think these findings are to Canadian long-term care facilities, and what differences in health system structure might affect prescribing patterns here?
Knowledge check
Quiz
1. What data source did this study use to assess nursing home residents?
- Random telephone surveys of residents' families
- The interRAI Long-Term Care Facility instrument as part of the SHELTER study
- National pharmacy dispensing databases only
- Self-reported online questionnaires completed by residents
Rationale: The abstract states the SHELTER study 'collected comprehensive resident data using the interRAI Long-Term Care Facility instrument in 7 European Countries and Israel.'
2. What percentage of the 4,023 residents entering the study had depressive symptoms?
- 12%
- 32%
- 50%
- 75%
Rationale: The abstract states: 'Among 4023 residents entering the study, 32% had depressive symptoms and nearly half of these individuals used antidepressants.'
3. What was the overall prevalence of antidepressant use in the full sample?
- 15.6%
- 25.9%
- 35.6%
- 45.2%
Rationale: The abstract states: 'Antidepressant medication use varied by country, with a prevalence in the overall sample of 35.6% (n = 1431).'
4. Which class of antidepressant was most commonly used among residents on antidepressants?
- Tricyclic antidepressants (TCAs)
- Selective serotonin reuptake inhibitors (SSRIs)
- Serotonin-norepinephrine reuptake inhibitors (SNRIs)
- Monoamine oxidase inhibitors (MAOIs)
Rationale: The abstract states: 'Among antidepressant users, 59.9% were receiving selective serotonin reuptake inhibitors (SSRI).'
5. Which of the following was associated with an INCREASED likelihood of antidepressant use?
- Diagnosis of schizophrenia
- Rural facility location
- Diagnosis of anxiety
- Age over 85 years
Rationale: The abstract lists diagnosis of anxiety, depression, bipolar disorder, pain, falls, and high social engagement as strongest correlates increasing antidepressant use.
6. Which factor was associated with a DECREASED likelihood of being prescribed an antidepressant?
- History of falls
- Diagnosis of schizophrenia
- High level of social engagement
- Reported diagnosis of pain
Rationale: The abstract states: 'Age over 85 years, living in facilities located in rural areas and a diagnosis of schizophrenia reduced the likelihood of being prescribed with an antidepressant.'
7. What type of study design was used in this analysis?
- Randomized controlled trial
- Cross-sectional analysis of observational cohort data
- Systematic review and meta-analysis
- Qualitative phenomenological study
Rationale: The abstract describes 'a cross-sectional analysis of data from the SHELTER study, an observational longitudinal cohort study.'
8. According to the study's conclusion, what should future research focus on?
- Comparing nursing home costs across countries
- Evaluating the efficacy and safety of antidepressant use in nursing homes through longitudinal studies
- Developing new antidepressant medications specifically for older adults
- Measuring nurse staffing ratios in long-term care facilities
Rationale: The abstract concludes: 'Future longitudinal studies should evaluate the efficacy and safety of antidepressant use in NHs thus providing evidence for recommendations for clinical practice.'
9. Approximately what proportion of residents with depressive symptoms were receiving antidepressants, according to the study?
- Nearly all
- Nearly half
- Less than 10%
- Exactly two-thirds
Rationale: The abstract states that among residents with depressive symptoms, 'nearly half of these individuals used antidepressants.'
10. Why is the cross-sectional design of this study considered a limitation for interpreting its findings?
- It means the sample size was too small to detect any associations
- It prevents the researchers from establishing cause-and-effect relationships between resident/facility characteristics and antidepressant use
- It means the data was collected only from a single facility
- It excluded all residents without a formal depression diagnosis
Rationale: As a cross-sectional, correlational analysis, the study can identify associations (e.g., via logistic regression) but cannot establish causality, which is why the authors call for future longitudinal studies.
Study cards
Flashcards
What is the SHELTER study?
SHELTER (Services and Health for Elderly in Long TERm care) is an observational longitudinal cohort study that collected resident data across long-term care facilities in seven European countries and Israel using the interRAI Long-Term Care Facility instrument.
How many residents were included in this cross-sectional analysis of antidepressant use?
4,023 residents entering the SHELTER study.
What proportion of residents had depressive symptoms at assessment?
32% of residents had depressive symptoms.
What proportion of residents with depressive symptoms were receiving antidepressant treatment?
Nearly half (approximately 47%) of residents with depressive symptoms were receiving antidepressants.
What was the overall prevalence of antidepressant use across the full sample?
35.6% of the sample (1,431 residents) were using antidepressant medications.
Which antidepressant class was most commonly prescribed among users?
SSRIs (selective serotonin reuptake inhibitors), used by 59.9% of antidepressant users.
Name three diagnoses associated with increased likelihood of antidepressant use.
Anxiety, depression, and bipolar disorder were among the diagnoses associated with increased antidepressant use.
Besides diagnoses, what two other resident characteristics were linked to increased antidepressant use?
A history of pain and a history of falls were associated with increased antidepressant use.
What level of social engagement was associated with higher antidepressant use?
A high level of social engagement was associated with increased likelihood of antidepressant use.
What age threshold was associated with a reduced likelihood of antidepressant prescribing?
Age over 85 years was associated with a reduced likelihood of being prescribed an antidepressant.
What facility location characteristic reduced the likelihood of antidepressant prescribing?
Living in a facility located in a rural area was associated with reduced antidepressant prescribing.
What diagnosis was associated with a reduced likelihood of antidepressant prescribing?
A diagnosis of schizophrenia was associated with reduced likelihood of antidepressant prescribing.
What instrument was used to collect resident data in the SHELTER study?
The interRAI Long-Term Care Facility instrument, a standardized comprehensive assessment tool.
What statistical method was used to identify correlates of antidepressant use?
Multiple logistic regression modeling was used to identify potential correlates of antidepressant use.
What study design limits this analysis from establishing causation?
The cross-sectional design limits the study to identifying associations rather than causal relationships.
In which journal was this study published, and when?
The study was published in BMC Geriatrics in 2020.
Did antidepressant prevalence vary by country in this study?
Yes, antidepressant medication use varied by country among the participating European countries and Israel.
What did the study's authors recommend for future research?
They recommended future longitudinal studies to evaluate the efficacy and safety of antidepressant use in nursing homes, to inform clinical practice recommendations.
What is a key clinical takeaway about undertreatment from this study?
A substantial gap exists between residents who have depressive symptoms and those actually receiving antidepressant treatment, suggesting possible undertreatment of late-life depression in long-term care.
Why is monitoring for SSRI-related risks important given this study's findings?
Because SSRIs were the most commonly prescribed antidepressant class among nursing home residents, nurses should be alert to age-related risks such as falls, hyponatremia, and GI bleeding.
Search-ready answers
Frequently asked questions
What is the SHELTER study and where was it conducted?
SHELTER (Services and Health for Elderly in Long TERm care) is an observational cohort study that assessed nursing home residents in seven European countries and Israel using the interRAI Long-Term Care Facility instrument.
How common is antidepressant use among nursing home residents in Europe, according to this study?
The study found that 35.6% of the 4,023 residents assessed (1,431 residents) were using antidepressant medications.
What percentage of nursing home residents with depressive symptoms actually receive antidepressants?
According to this study, only about 47% (nearly half) of residents with depressive symptoms were receiving antidepressant treatment, indicating a treatment gap.
Which antidepressant medications are most commonly used in nursing homes?
Among residents taking an antidepressant, 59.9% were prescribed an SSRI (selective serotonin reuptake inhibitor), the most common class identified in this study.
What factors increase the likelihood that a nursing home resident is prescribed an antidepressant?
The study found that diagnoses of anxiety, depression, or bipolar disorder, along with pain, a history of falls, and high social engagement, were associated with a greater likelihood of antidepressant use.
Are older nursing home residents less likely to be prescribed antidepressants?
Yes, this study found that residents aged 85 and older had a reduced likelihood of being prescribed an antidepressant compared to younger residents.
Does living in a rural nursing home affect antidepressant prescribing?
The study found that residents in facilities located in rural areas were less likely to be prescribed antidepressants than those in non-rural facilities.
Why were residents with schizophrenia less likely to receive antidepressants in this study?
The abstract does not explain the reason, only that a schizophrenia diagnosis was one of the factors associated with reduced likelihood of antidepressant prescribing; this would need further clinical or research explanation.
What are the main limitations of this study on antidepressant use in nursing homes?
As a cross-sectional study, it cannot establish cause and effect, relied on a screening scale rather than diagnostic interviews for depressive symptoms, used non-randomly selected facilities, and did not measure symptom severity, treatment duration, or non-drug treatments.
What did the researchers recommend for future studies on this topic?
They recommended future longitudinal studies to evaluate the efficacy and safety of antidepressant use in nursing homes, in order to build an evidence base for clinical practice recommendations.