Strict long COVID, symptom persistence, and functional decline among community-dwelling older adults in Quito, Ecuador: a cross-sectional study

Abstract

Background

Long COVID is an increasingly recognized health concern in older adults, yet data from Latin America remain limited. Older adults may be particularly vulnerable to post-COVID sequelae because of frailty, multimorbidity, and reduced physiological reserve. This study aimed to estimate the prevalence of strict long COVID among older Ecuadorian adults and to characterize persistent post-COVID symptoms, clinical correlates, and functional/cognitive worsening.

Methods

We conducted a cross-sectional study of 1,050 community-dwelling adults aged 65 years or older with PCR-confirmed SARS-CoV-2 infection in Quito, Ecuador. Data were collected through structured face-to-face interviews and cognitive screening. Strict long COVID was defined as symptoms persisting for more than 12 weeks after SARS-CoV-2 infection. Symptoms lasting 4–12 weeks were classified as persistent post-COVID symptoms but were not considered strict long COVID. Multivariable logistic regression was used to evaluate factors associated with strict long COVID.

Results

The mean age was 74.2 ± 7.5 years, and 565 participants (53.8%) were women. Overall, 191 participants met criteria for strict long COVID, corresponding to a prevalence of 18.2% (95% CI, 16.0–20.6). Persistent post-COVID symptoms lasting 4–6 weeks and 7–12 weeks were reported by 401 (38.2%) and 458 (43.6%) participants, respectively. Participants with strict long COVID had higher frequencies of dyspnea, cognitive impairment, sleep disturbance, myalgia, depression, anxiety, and difficulty sleeping. Functional/cognitive worsening was more frequent among participants with strict long COVID than among those without strict long COVID (53.4% vs. 36.6%; p < 0.001), particularly for walking or climbing stairs. In multivariable analysis, severe or critical acute COVID-19 was independently associated with strict long COVID (adjusted OR, 2.36; 95% CI, 1.62–3.41; p < 0.001). Female sex, age ≥ 81 years, incomplete vaccination, diabetes with organ involvement, hospitalization, and care dependence were not independently associated with strict long COVID.

Conclusion

Strict long COVID affected nearly one in five older adults with PCR-confirmed SARS-CoV-2 infection in Quito, Ecuador. Severe or critical acute COVID-19 was the main factor independently associated with strict long COVID, and affected participants had greater functional/cognitive worsening. These findings support integrating post-COVID screening, functional assessment, and geriatric rehabilitation into primary care for older adults in Latin America.

Highlights

• Strict long COVID prevalence: Strict long COVID, defined as symptoms persisting for more than 12 weeks, affected 18.2% of older adults with PCR-confirmed SARS-CoV-2 infection.

• Persistent post-COVID symptoms: Symptoms lasting 4–12 weeks were common and should be classified as persistent post-COVID symptoms rather than strict long COVID.

• Acute severity association: Severe or critical acute COVID-19 was the main factor independently associated with strict long COVID.

• Functional and cognitive impact: older adults with strict long COVID had greater functional/cognitive worsening, particularly in mobility-related domains.

• Geriatric care implications: Findings support integrating post-COVID screening, functional assessment, and geriatric rehabilitation into primary care for older adults.

Introduction

Coronavirus disease 2019 (COVID-19), caused by severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2), has generated a large burden of acute and post-acute disease across all age groups worldwide [1]. Although many individuals recover after the acute phase, a clinically relevant proportion develop persistent, relapsing, or newly emerging symptoms after infection [2]. The World Health Organization (WHO) defines post-COVID-19 condition as a syndrome occurring in people with probable or confirmed SARS-CoV-2 infection, usually 3 months from symptom onset, with symptoms lasting at least 2 months and not explained by an alternative diagnosis [3]. The National Institute for Health and Care Excellence (NICE) similarly distinguishes ongoing symptomatic COVID-19 from 4 to 12 weeks and post-COVID-19 syndrome beyond 12 weeks [4]. This temporal distinction is essential for epidemiological studies because prevalence estimates vary substantially depending on whether symptoms after 4 weeks, 8 weeks, 12 weeks, or longer are classified as long COVID [5].

Long COVID is a heterogeneous condition that can affect multiple organ systems and may present with fluctuating clusters of symptoms over time [6]. Frequently reported manifestations include fatigue, dyspnea, cognitive dysfunction, sleep disturbance, musculoskeletal pain, anxiety, depression, and reduced exercise tolerance [7]. Recent systematic reviews and global analyses have shown wide variation in prevalence estimates, reflecting differences in case definitions, symptom ascertainment, study design, infection severity, vaccination status, circulating variants, and follow-up duration [8]. This heterogeneity has led to increasing emphasis on standardized definitions and reproducible operational criteria for post-COVID conditions [9]. For older adults, distinguishing persistent post-COVID symptoms from strict long COVID is particularly important because baseline comorbidity, functional impairment, and geriatric syndromes may overlap with post-viral symptoms [10].

Older adults represent a clinically important population for long COVID research because aging is associated with immune senescence, multimorbidity, frailty, sarcopenia, polypharmacy, and reduced physiological reserve [11]. These factors may increase vulnerability to severe acute COVID-19 and may also delay recovery after infection [12]. In geriatric populations, persistent symptoms can have consequences beyond symptom burden, including reduced mobility, loss of independence, impaired quality of life, and increased need for rehabilitation or caregiving support [13]. Recent studies in older adults have linked post-COVID symptoms with functional decline, lower physical activity, fatigue, cognitive-communication difficulties, and poorer health-related quality of life [14]. Therefore, long COVID in late life should be evaluated not only as a symptom syndrome but also as a potential contributor to disability and care dependence [15].

Several risk factors for long COVID have been described in the literature, including female sex, pre-existing comorbidities, higher body mass index, smoking, socioeconomic disadvantage, and greater severity of acute COVID-19 [16]. However, associations are not uniform across studies and may differ according to age group, case definition, follow-up period, and analytic strategy [17]. Comorbid conditions such as diabetes, cardiovascular disease, and chronic respiratory disease may contribute to prolonged recovery through inflammatory, metabolic, endothelial, and immune-mediated pathways [18]. At the same time, hospitalization and severe acute disease may act as markers of greater systemic injury, deconditioning, and post-acute functional risk, especially in older adults [19]. Because cross-sectional studies cannot establish causality, these factors should be interpreted as correlates or associated factors rather than determinants in the strict epidemiological sense [20].

Vaccination has been investigated as a potential strategy to reduce the risk of long COVID, mainly by preventing SARS-CoV-2 infection and severe acute disease [21]. Systematic reviews have generally suggested that vaccination before infection is associated with a lower probability of developing long COVID, although effect estimates vary across study designs and populations [22]. More recent evidence also indicates that the relationship between vaccination and long COVID may be influenced by number of doses, timing of vaccination, prior infection, variant period, and health-seeking behavior [23]. Older adults are a priority group for vaccination and booster strategies because they face a higher risk of hospitalization and mortality after acute COVID-19 [24]. Nevertheless, older people are often underrepresented in clinical trials and may have heterogeneous immune responses because of frailty, comorbidity burden, and immune senescence [25].

Evidence from low- and middle-income countries remains comparatively limited, despite the disproportionate clinical, social, and health-system impact of the COVID-19 pandemic in Latin America [26]. Studies from lower-resource settings suggest that long COVID prevalence and consequences may differ from those reported in high-income countries because of differences in healthcare access, diagnostic capacity, vaccination coverage, socioeconomic vulnerability, and availability of rehabilitation services [27]. Latin America was severely affected during the pandemic, and Ecuador experienced major healthcare strain during early waves of SARS-CoV-2 transmission [28]. However, older adults from Ecuador and similar middle-income settings remain underrepresented in long COVID research, particularly in studies using PCR-confirmed infection and geriatric functional outcomes [29, 30]. Local data are needed to inform primary care follow-up, rehabilitation pathways, vaccination strategies, and policies aimed at preserving autonomy in older adults [31, 32].

The present study aimed to estimate the prevalence of strict long COVID among community-dwelling adults aged 65 years and older in Quito, Ecuador, using symptoms persisting for more than 12 weeks as the primary outcome. A secondary aim was to characterize persistent post-COVID symptoms lasting 4–12 weeks and to evaluate clinical, vaccination-related, functional, and cognitive characteristics according to strict long COVID status. By separating persistent post-COVID symptoms from strict long COVID, this analysis seeks to provide a more conservative and clinically interpretable estimate of disease burden. This distinction is especially relevant for responding to geriatric and public health needs, because older adults may experience both prolonged symptoms and functional decline after SARS-CoV-2 infection. The study also evaluates associated factors using noncausal language appropriate for a cross-sectional design.

Methods

Study design and setting

This was a cross-sectional, observational study conducted in Quito, Ecuador, between July 2020 and December 2022. The study was designed to estimate the prevalence of strict long COVID among community-dwelling older adults with confirmed SARS-CoV-2 infection and to evaluate its clinical, functional, cognitive, and vaccination-related correlates. The primary outcome was strict long COVID, defined as symptoms persisting for more than 12 weeks after SARS-CoV-2 infection. Symptoms lasting 4–12 weeks were analyzed separately as persistent post-COVID symptoms and were not classified as strict long COVID. This distinction was used to align the analysis with international post-COVID case definitions and is summarized in the analytic flow diagram and symptom-duration distribution figure.

Participants and recruitment

Participants were community-dwelling adults aged 65 years or older residing in central and peripheral areas of Quito, Ecuador. Recruitment followed a convenience sampling strategy. Potentially eligible individuals were identified through community-based outreach and available records of PCR-confirmed SARS-CoV-2 infection from health services or the national Ministry of Health database. Individuals were contacted and invited to participate in a structured assessment conducted either at their homes or at outpatient centers, according to their mobility status and personal preference. Because no formal random sampling procedure was implemented, the sample should be interpreted as a community-based convenience sample of older adults with documented SARS-CoV-2 infection. This approach allowed recruitment of many older participants during the pandemic period but may limit generalizability to all older adults in Ecuador. A total of 1,218 individuals were approached for eligibility assessment. Of these, 168 were excluded: 102 did not meet eligibility criteria, 41 declined to participate, and 25 had incomplete data. The final analytic sample included 1,050 participants. Participant selection, exclusions, analytic completeness, and classification according to symptom duration are shown in Fig. 1.

Fig. 1.

STROBE-type participant flow and analytic classification. Participant selection flow diagram showing the number of individuals approached for eligibility assessment, excluded participants, and the final analytic sample. Among 1,218 individuals assessed, 168 were excluded because they did not meet eligibility criteria, declined to participate, or had incomplete data. The final analytic sample included 1,050 community-dwelling adults aged 65 years or older with PCR-confirmed SARS-CoV-2 infection. Participants were further classified according to symptom duration into persistent post-COVID symptoms lasting 4–6 weeks, persistent post-COVID symptoms lasting 7–12 weeks, and strict long COVID, defined as symptoms persisting for more than 12 weeks after infection

Eligibility criteria

Inclusion criteria were age 65 years or older at the time of assessment; confirmed SARS-CoV-2 infection by polymerase chain reaction (PCR), verified through hospital records or the national Ministry of Health database; residence in the community at the time of recruitment; ability to provide written informed consent; and ability to complete the interview reliably. Self-reported SARS-CoV-2 infection without PCR confirmation was not accepted. Cognitive screening and clinical assessment were used to determine whether participants could reliably recall and report symptoms before and after COVID-19. Exclusion criteria were receipt of influenza or pneumococcal vaccination within three months before or after SARS-CoV-2 infection, to reduce potential confounding from vaccine-related immune activation on post-COVID symptoms; severe cognitive impairment that precluded reliable self-reporting; inability to complete the structured interview; withdrawal of consent; refusal to participate; or incomplete core data required for the main analysis. When cognitive impairment was identified during screening, participants were evaluated for their ability to understand the study procedures and provide reliable responses. Individuals with severe impairment that prevented valid self-reporting were excluded from the analytic sample. Participants with mild cognitive difficulties who were able to provide informed consent and complete the structured interview were retained.

Sample size and analytic sample

The final analytic sample included 1,050 participants. Sample size was primarily determined by feasibility during the study period; however, it was also adequate for estimating the prevalence of strict long COVID. Assuming an expected prevalence of approximately 20%, a sample of 1,050 participants allowed estimation of prevalence with a 95% confidence interval and an approximate margin of error of ± 2.5%. The number of strict long COVID cases was 191, providing an adequate number of outcome events for the planned multivariable logistic regression model. The final model included seven prespecified covariates, resulting in more than 20 outcome events per covariate, which was considered sufficient to reduce the risk of model overfitting. Completeness of the core analytic variables was verified before analysis. Age data, symptom-duration data, and linked records across the main database blocks were available for all 1,050 participants included in the final analytic sample.

With 191 strict long COVID events and seven prespecified covariates in the final model, the analysis provided approximately 27 events per covariate, exceeding commonly used minimum recommendations for logistic regression.

Data collection

Data were collected through structured, face-to-face interviews conducted by a trained field team composed of physicians, rather than by a single interviewer. Interviews were performed at participants’ homes or outpatient centers, depending on participant preference, functional status, and mobility limitations. Interviewers received standardized training in symptom assessment based on post-COVID clinical guidance.

Participants were asked about acute COVID-19 symptoms, duration of symptoms, persistent symptoms after infection, perceived recovery, vaccination history, hospitalization, clinical severity, comorbidities, lifestyle-related risk factors, and functional status before and after COVID-19. To reduce recall bias, the time interval between acute SARS-CoV-2 infection and study assessment was documented and considered during interpretation. Clinical information included age, sex, care dependence, hospitalization, acute COVID-19 severity, vaccination status, diabetes with organ involvement, uncomplicated diabetes, hypertension, hypercholesterolemia, multimorbidity, polypharmacy, sedentary lifestyle, and medical complications.

Outcome definitions

The primary outcome was strict long COVID, defined as the presence of symptoms persisting for more than 12 weeks after SARS-CoV-2 infection. Participants with symptoms lasting 12 weeks or less were classified as not having strict long COVID. Symptom duration was categorized as 4–6 weeks, 7–12 weeks, and more than 12 weeks. The 4–6-week and 7–12-week categories were considered persistent post-COVID symptoms but were not counted as strict long COVID. The prevalence of strict long COVID and the distribution of symptom-duration categories are reported in Table 1 and displayed in Fig. 2. A sensitivity definition was also explored using symptoms lasting more than 90 days, based on the days-elapsed symptom-duration variable. This sensitivity analysis was used to evaluate consistency between the categorical symptom-duration variable and the days-based variable.

Table 1.

Symptom duration and prevalence of strict long COVID among older adults

Symptom duration category	n	%	95% CI, %
Persistent post-COVID symptoms, 4–6 weeks	401	38.2	35.3–41.2
Persistent post-COVID symptoms, 7–12 weeks	458	43.6	40.6–46.6
Strict long COVID, > 12 weeks	191	18.2	16.0–20.6
Sensitivity definition, symptoms > 90 days	72	6.9	5.5–8.5

Data are presented as absolute frequencies and percentages. Strict long COVID was defined as symptoms persisting for more than 12 weeks after SARS-CoV-2 infection. Symptoms lasting 4–12 weeks were classified as persistent post-COVID symptoms and were not considered strict long COVID. The >90-day category was used as a sensitivity definition

Fig. 2.

Symptom duration distribution and strict long COVID prevalence. Distribution of symptom duration categories in the final analytic sample. Symptoms lasting 4–6 weeks and 7–12 weeks were classified as persistent post-COVID symptoms and were not considered strict long COVID. Strict long COVID was defined as symptoms persisting for more than 12 weeks after SARS-CoV-2 infection and was observed in 191 participants, corresponding to a prevalence of 18.2% (95% CI, 16.0–20.6). A sensitivity definition based on symptoms lasting more than 90 days identified 72 participants, corresponding to 6.9% (95% CI, 5.5–8.5)

Functional and cognitive assessment

Cognitive status was assessed using the Mini-Cog and standard neurological physical examination. Functional and cognitive worsening were evaluated by comparing participants’ current status with their self-reported pre-COVID-19 status across the following domains: vision, hearing, walking or climbing stairs, remembering or concentrating, grooming or dressing, and communication. For each domain, worsening was defined as a higher level of difficulty at the time of assessment compared with the participant’s pre-COVID-19 status. A composite variable for any functional or cognitive worsening was created based on worsening in walking or climbing stairs, remembering or concentrating, grooming or dressing, or communication. A broader six-domain composite additionally included worsening of vision and hearing. Functional and cognitive worsening according to strict long COVID status is presented in Table 5; Fig. 3.

Table 5.

Post-COVID functional and cognitive worsening according to strict long COVID status

Domain	Total (N = 1,050)	No strict long COVID (n = 859)	Strict long COVID (n = 191)	p value
Worsening of vision	127 (12.1)	98 (11.4)	29 (15.2)	0.185
Worsening of hearing	80 (7.6)	58 (6.8)	22 (11.5)	0.036
Worsening in walking or climbing stairs	304 (29.0)	218 (25.4)	86 (45.0)	< 0.001
Worsening in remembering or concentrating	237 (22.6)	185 (21.5)	52 (27.2)	0.108
Worsening in grooming or dressing	76 (7.2)	54 (6.3)	22 (11.5)	0.018
Worsening in communication	76 (7.2)	56 (6.5)	20 (10.5)	0.080
Any functional or cognitive worsening*	416 (39.6)	314 (36.6)	102 (53.4)	< 0.001
Any worsening across the six domains†	477 (45.4)	363 (42.3)	114 (59.7)	< 0.001

Data are presented as n (%). Worsening was defined as a higher level of difficulty at the time of assessment compared with the participant’s pre-COVID-19 status. Strict long COVID was defined as symptoms persisting for more than 12 weeks after SARS-CoV-2 infection. Percentages are column percentages

*Includes worsening in walking or climbing stairs, remembering, or concentrating, grooming or dressing, or communication

†Includes worsening in vision, hearing, walking or climbing stairs, remembering or concentrating, grooming or dressing, or communication

Fig. 3.

Functional/cognitive worsening by strict long COVID status. Comparison of post-COVID functional and cognitive worsening according to strict long COVID status. Worsening was defined as a higher level of difficulty at the time of assessment compared with pre-COVID-19 status. Participants with strict long COVID had a higher frequency of worsening in walking or climbing stairs, hearing, grooming or dressing, and overall functional/cognitive worsening than participants without strict long COVID. The composite outcome of any functional/cognitive worsening included worsening in walking or climbing stairs, remembering or concentrating, grooming or dressing, or communication

Covariates

Covariates were selected based on clinical relevance, prior evidence, and geriatric plausibility rather than relying solely on univariate statistical significance. The main covariates considered for the multivariable model were sex, age group, vaccination status, diabetes with organ involvement, acute COVID-19 severity, hospitalization, and care dependence. Age was analyzed descriptively as both a continuous variable and categorical age groups. For the multivariable model, advanced age was defined as age 81 years or older. Vaccination status was categorized as incomplete vaccination with fewer than three COVID-19 vaccine doses versus three or more doses. Acute COVID-19 severity was dichotomized as severe or critical versus mild or moderate disease. Care dependence was dichotomized as self-sufficient versus partial or permanent care dependence.

Prespecified interaction terms for sex-by-age group and vaccination-by-age group were explored because sex and age differences were clinically relevant to the study objective.

Care status was classified as self-sufficient, partial care dependence, or permanent care dependence according to the participant’s reported need for assistance with basic or instrumental activities of daily living. Multimorbidity was defined as the presence of two or more chronic medical conditions. Diabetes with organ involvement included diabetes associated with documented renal, cardiovascular, neurological, ophthalmologic, or peripheral vascular complications. Acute COVID-19 severity was classified according to clinical presentation and need for hospitalization, oxygen support, or intensive care, and was dichotomized as severe/critical versus mild/moderate for regression analysis.

Statistical analysis

Baseline characteristics were summarized using descriptive statistics. Continuous variables are presented as means and standard deviations or medians and interquartile ranges, according to distributional characteristics. Categorical variables are presented as absolute frequencies and percentages. Group comparisons were performed according to strict long COVID status. Categorical variables were compared using the chi-square test or Fisher’s exact test, as appropriate. Continuous variables were compared using the student’s t test or the Mann–Whitney U test according to distributional assumptions and variable characteristics. The prevalence of strict long COVID was calculated as the number of participants with symptoms persisting for more than 12 weeks divided by the final analytic sample. Exact or binomial 95% confidence intervals were calculated for prevalence estimates. The distribution of symptom duration and the sensitivity definition based on symptoms lasting more than 90 days are reported in Table 1. Demographic, clinical, vaccination-related, and comorbidity characteristics according to strict long COVID status are reported in Table 2. Clinical course and perceived recovery are reported in Table 3. Symptoms reported during the last seven days are reported in Table 4. Functional and cognitive worsening are reported in Table 5. A multivariable logistic regression model was fitted to identify factors independently associated with strict long COVID. The dependent variable was strict long COVID, defined as symptoms persisting for more than 12 weeks after SARS-CoV-2 infection. The model included sex, age ≥ 81 years, incomplete vaccination with fewer than three doses, diabetes with organ involvement, severe or critical acute COVID-19, hospitalization, and partial or permanent care dependence. Results are reported as adjusted odds ratios (aORs) with 95% confidence intervals and p values. Multicollinearity was assessed using the variance inflation factor, with values greater than 5 considered indicative of potential concern. Model performance was evaluated using goodness-of-fit and discrimination diagnostics. The multivariable model is presented in Table 6 and visualized as a forest plot in Fig. 4. Missing data for the core analytic variables were minimal because participants with incomplete core data were excluded before construction of the final analytic sample. Therefore, complete-case analysis was used and no imputation procedures were applied.

Table 2.

Demographic, clinical, and vaccination characteristics according to strict long COVID status

Characteristic	Total (N = 1,050)	No strict long COVID (n = 859)	Strict long COVID (n = 191)	p value
Age, years, mean ± SD	74.2 ± 7.5	74.4 ± 7.5	73.7 ± 7.6	0.286
65–70 years	405 (38.6)	323 (37.6)	82 (42.9)	0.509
71–80 years	421 (40.1)	351 (40.9)	70 (36.6)
81–90 years	194 (18.5)	159 (18.5)	35 (18.3)
≥ 91 years	30 (2.9)	26 (3.0)	4 (2.1)
Male sex	485 (46.2)	383 (44.6)	102 (53.4)	0.033
Female sex	565 (53.8)	476 (55.4)	89 (46.6)
Self-sufficient	805 (76.7)	663 (77.2)	142 (74.3)	0.526
Requires partial care	199 (19.0)	161 (18.7)	38 (19.9)
Requires permanent care	46 (4.4)	35 (4.1)	11 (5.8)
One COVID-19 vaccine dose	35 (3.3)	23 (2.7)	12 (6.3)	0.039
Two COVID-19 vaccine doses	129 (12.3)	108 (12.6)	21 (11.0)
Three or more COVID-19 vaccine doses	886 (84.4)	728 (84.7)	158 (82.7)
Mild acute COVID-19	410 (39.0)	368 (42.8)	42 (22.0)	< 0.001
Moderate acute COVID-19	291 (27.7)	231 (26.9)	60 (31.4)
Severe acute COVID-19	324 (30.9)	244 (28.4)	80 (41.9)
Critical acute COVID-19	10 (1.0)	3 (0.3)	7 (3.7)
Hospitalization	484 (46.1)	393 (45.8)	91 (47.6)	0.693
Diabetes with organ involvement	142 (13.5)	115 (13.4)	27 (14.1)	0.876
Uncomplicated diabetes	87 (8.3)	63 (7.3)	24 (12.6)	0.026
Hypertension	506 (48.2)	412 (48.0)	94 (49.2)	0.816
Hypercholesterolemia	112 (10.7)	94 (10.9)	18 (9.4)	0.627
Multimorbidity	83 (7.9)	66 (7.7)	17 (8.9)	0.678
Polypharmacy	86 (8.2)	64 (7.5)	22 (11.5)	0.088
Sedentary lifestyle	510 (48.6)	414 (48.2)	96 (50.3)	0.662
Medical complications	475 (45.2)	377 (43.9)	98 (51.3)	0.075

Data are presented as mean ± standard deviation or n (%), as appropriate. Strict long COVID was defined as symptoms persisting for more than 12 weeks after SARS-CoV-2 infection. Group comparisons were performed using the chi-square test or Fisher’s exact test for categorical variables and the student’s t test or Mann–Whitney U test for continuous variables, according to distributional assumptions. Percentages are column percentages

Table 3.

Clinical course and perceived recovery according to strict long COVID status

Variable	Total (N = 1,050)	No strict long COVID (n = 859)	Strict long COVID (n = 191)	p value
Symptoms for 1–10 days	91 (8.7)	58 (6.8)	33 (17.3)	< 0.001
Symptoms for 11–14 days	46 (4.4)	33 (3.8)	13 (6.8)
Symptoms for 15–20 days	52 (5.0)	44 (5.1)	8 (4.2)
Symptoms for 21–30 days	118 (11.2)	106 (12.3)	12 (6.3)
Symptoms for 31–60 days	557 (53.0)	492 (57.3)	65 (34.0)
Symptoms for 61–90 days	114 (10.9)	91 (10.6)	23 (12.0)
Symptoms for > 90 days	72 (6.9)	35 (4.1)	37 (19.4)
Isolation < 10 days	197 (18.8)	147 (17.1)	50 (26.2)	0.006
Isolation 10–14 days	355 (33.8)	304 (35.4)	51 (26.7)
Isolation > 14 days	498 (47.4)	408 (47.5)	90 (47.1)
No hospitalization	566 (53.9)	466 (54.2)	100 (52.4)	0.474
Hospitalization < 10 days	134 (12.8)	114 (13.3)	20 (10.5)
Hospitalization 10–14 days	118 (11.2)	98 (11.4)	20 (10.5)
Hospitalization 15–30 days	169 (16.1)	134 (15.6)	35 (18.3)
Hospitalization > 30 days	62 (5.9)	46 (5.4)	16 (8.4)
Perceived recovery: completely disagree	60 (5.7)	40 (4.7)	20 (10.5)	< 0.001
Disagree	208 (19.8)	156 (18.2)	52 (27.2)
Neither agree nor disagree	206 (19.6)	169 (19.7)	37 (19.4)
Agree	374 (35.6)	318 (37.0)	56 (29.3)
Completely agree	200 (19.0)	175 (20.4)	25 (13.1)

Data are presented as n (%). Strict long COVID was defined as symptoms persisting for more than 12 weeks after SARS-CoV-2 infection. Percentages are column percentages. Group comparisons were performed using the chi-square test or Fisher’s exact test, as appropriate. Perceived recovery was assessed using an ordinal agreement scale

Table 4.

Symptoms reported during the last 7 days according to strict long COVID status

Symptom	Total (N = 1,050)	No strict long COVID (n = 859)	Strict long COVID (n = 191)	p value
Dyspnea	223 (21.2)	160 (18.6)	63 (33.0)	< 0.001
Cough	420 (40.0)	345 (40.2)	75 (39.3)	0.883
Sputum production	155 (14.8)	124 (14.4)	31 (16.2)	0.603
Cognitive impairment	298 (28.4)	231 (26.9)	67 (35.1)	0.029
Sleep disturbance	412 (39.2)	319 (37.1)	93 (48.7)	0.004
Vision changes	147 (14.0)	110 (12.8)	37 (19.4)	0.024
Hearing loss	72 (6.9)	57 (6.6)	15 (7.9)	0.657
Myalgia	389 (37.0)	303 (35.3)	86 (45.0)	0.015
Arthralgia	420 (40.0)	335 (39.0)	85 (44.5)	0.186
Reduced mobility	196 (18.7)	155 (18.0)	41 (21.5)	0.320
Depression	234 (22.3)	180 (21.0)	54 (28.3)	0.036
Anxiety	197 (18.8)	146 (17.0)	51 (26.7)	0.003
Difficulty sleeping	351 (33.4)	261 (30.4)	90 (47.1)	< 0.001

Data are presented as n (%). Strict long COVID was defined as symptoms persisting for more than 12 weeks after SARS-CoV-2 infection. Symptoms correspond to those reported during the 7 days before assessment. Percentages are column percentages. Group comparisons were performed using the chi-square test or Fisher’s exact test, as appropriate

Table 6.

Multivariable logistic regression model for strict long COVID

Variable	Adjusted OR	95% CI	p value
Female sex	0.74	0.54–1.02	0.070
Age ≥ 81 years	0.89	0.59–1.34	0.583
Incomplete vaccination, < 3 doses	1.10	0.72–1.69	0.662
Diabetes with organ involvement	1.04	0.66–1.66	0.855
Severe or critical acute COVID-19	2.36	1.62–3.41	< 0.001
Hospitalization	0.75	0.52–1.08	0.121
Partial or permanent care dependence	1.17	0.80–1.72	0.412

The dependent variable was strict long COVID, defined as symptoms persisting for more than 12 weeks after SARS-CoV-2 infection. The model included sex, age group, vaccination status, diabetes with organ involvement, acute COVID-19 severity, hospitalization, and care dependence. Results are presented as adjusted odds ratios with 95% confidence intervals. 

OR  odds ratio, CI  confidence interval

Fig. 4.

Forest plot of adjusted odds ratios. Multivariable logistic regression model showing factors associated with strict long COVID, defined as symptoms persisting for more than 12 weeks after SARS-CoV-2 infection. Adjusted odds ratios are shown with 95% confidence intervals. The vertical dashed line represents an adjusted odds ratio of 1.0. Severe or critical acute COVID-19 was the only factor independently associated with strict long COVID, whereas female sex, age ≥ 81 years, incomplete vaccination, diabetes with organ involvement, hospitalization, and partial or permanent care dependence were not independently associated with the outcome

The revised analysis is presented in six tables and four figures. Table 1 reports symptom duration and strict long COVID prevalence. Table 2 compares demographic, clinical, and vaccination characteristics according to strict long COVID status. Table 3 presents clinical course and perceived recovery. Table 4 summarizes symptoms reported during the last seven days. Table 5 reports functional and cognitive worsening. Table 6 presents the multivariable logistic regression model. Figure 1 shows the STROBE-type participant flow and analytic classification. Figure 2 displays symptom-duration distribution and strict long COVID prevalence. Figure 4 shows the forest plot of adjusted odds ratios for strict long COVID. Figure 3 displays functional and cognitive worsening according to strict long COVID status.

Ethical considerations

The study was approved by the Ethics Committee for Research in Human Beings of Carlos Andrade Marín Specialty Hospital (CEISH-HCAM) on July 6, 2020, under protocol code IESS-HCAM-CEISH-2020-200-DF. Written informed consent was obtained from all participants before data collection. All procedures involving human participants were performed in accordance with relevant national and institutional ethical guidelines and with the principles of the Declaration of Helsinki. Data were anonymized before analysis to protect participant confidentiality.

Results

Participant flow and analytic sample

A total of 1,218 individuals were approached for eligibility assessment. Of these, 168 were excluded: 102 did not meet eligibility criteria, 41 declined to participate, and 25 had incomplete data. The final analytic sample included 1,050 community-dwelling adults aged 65 years or older with PCR-confirmed SARS-CoV-2 infection. Complete data for age, symptom duration, and core analytic variables were available for all participants included in the final analysis. The participant selection process and analytic classification are shown in Fig. 1.

Symptom duration and prevalence of strict long COVID

Symptom duration was categorized into three clinically relevant groups: 4–6 weeks, 7–12 weeks, and more than 12 weeks. Overall, 401 participants (38.2%; 95% CI, 35.3–41.2) reported persistent post-COVID symptoms lasting 4–6 weeks, and 458 participants (43.6%; 95% CI, 40.6–46.6) reported symptoms lasting 7–12 weeks. These two categories were considered persistent post-COVID symptoms but were not classified as strict long COVID. Strict long COVID, defined as symptoms persisting for more than 12 weeks after SARS-CoV-2 infection, was identified in 191 participants, corresponding to a prevalence of 18.2% (95% CI, 16.0–20.6). In a sensitivity analysis using the days-elapsed variable, symptoms lasting more than 90 days were reported by 72 participants, corresponding to 6.9% (95% CI, 5.5–8.5). The distribution of symptom duration and strict long COVID prevalence is presented in Table 1; Fig. 2.

Demographic, clinical, and vaccination characteristics

The mean age of participants was 74.2 ± 7.5 years. Most participants were between 65 and 80 years of age: 405 (38.6%) were aged 65–70 years, and 421 (40.1%) were aged 71–80 years. A total of 194 participants (18.5%) were aged 81–90 years, and 30 (2.9%) were aged 91 years or older. The cohort included 485 men (46.2%) and 565 women (53.8%). Participants with strict long COVID had a mean age of 73.7 ± 7.6 years, compared with 74.4 ± 7.5 years among those without strict long COVID. The distribution of age groups did not differ significantly between groups (p = 0.509). Strict long COVID was observed in 102 men (53.4%) and 89 women (46.6%), with a statistically significant difference by sex in the unadjusted comparison (p = 0.033). Regarding care status, 805 participants (76.7%) were self-sufficient, 199 (19.0%) required partial care, and 46 (4.4%) required permanent care. Care dependence did not differ significantly according to strict long COVID status (p = 0.526). Demographic, clinical, and vaccination characteristics according to strict long COVID status are shown in Table 2. Vaccination coverage was high in the final analytic sample. Overall, 886 participants (84.4%) had received three or more COVID-19 vaccine doses, 129 (12.3%) had received two doses, and 35 (3.3%) had received one dose. The distribution of vaccine doses differed between groups in the unadjusted analysis (p = 0.039), mainly because one-dose vaccination was more frequent among participants with strict long COVID. However, incomplete vaccination did not remain independently associated with strict long COVID in the multivariable model.

Acute COVID-19 severity, hospitalization, and perceived recovery

Acute COVID-19 severity differed significantly according to strict long COVID status (p < 0.001). Among participants with strict long COVID, 42 (22.0%) had mild acute disease, 60 (31.4%) had moderate disease, 80 (41.9%) had severe disease, and 7 (3.7%) had critical disease. In contrast, among participants without strict long COVID, 368 (42.8%) had mild disease, 231 (26.9%) had moderate disease, 244 (28.4%) had severe disease, and 3 (0.3%) had critical disease. Hospitalization was reported in 484 participants (46.1%) and did not differ significantly between those with and without strict long COVID (47.6% vs. 45.8%; p = 0.693). Duration of hospitalization also did not show a statistically significant difference between groups. Perceived recovery differed significantly by strict long COVID status (p < 0.001). Participants with strict long COVID were less likely to report agreement or complete agreement with perceived recovery and more likely to report disagreement. Specifically, among participants with strict long COVID, 20 (10.5%) completely disagreed and 52 (27.2%) disagreed that they had recovered, compared with 40 (4.7%) and 156 (18.2%), respectively, among participants without strict long COVID. Clinical course and perceived recovery are summarized in Table 3.

Symptoms reported during the last 7 days

Participants with strict long COVID had a higher burden of several symptoms reported during the 7 days before assessment. Dyspnea was more frequent among participants with strict long COVID than among those without strict long COVID (33.0% vs. 18.6%; p < 0.001). Cognitive impairment was also more frequent in the strict long COVID group (35.1% vs. 26.9%; p = 0.029), as were sleep disturbance (48.7% vs. 37.1%; p = 0.004), vision changes (19.4% vs. 12.8%; p = 0.024), myalgia (45.0% vs. 35.3%; p = 0.015), depression (28.3% vs. 21.0%; p = 0.036), anxiety (26.7% vs. 17.0%; p = 0.003), and difficulty sleeping (47.1% vs. 30.4%; p < 0.001). Cough, sputum production, hearing loss, arthralgia, and reduced mobility did not differ significantly between groups in the symptom assessment. Symptoms reported during the last 7 days according to strict long COVID status are shown in Table 4.

Functional and cognitive worsening after COVID-19

Post-COVID functional and cognitive worsening was common in the final analytic sample and was more frequent among participants with strict long COVID. Worsening in walking or climbing stairs was reported by 86 participants with strict long COVID (45.0%) compared with 218 participants without strict long COVID (25.4%; p < 0.001). Worsening in hearing was also more frequent in the strict long COVID group (11.5% vs. 6.8%; p = 0.036), as was worsening in grooming or dressing (11.5% vs. 6.3%; p = 0.018). Worsening in remembering or concentrating was reported by 52 participants with strict long COVID (27.2%) and 185 participants without strict long COVID (21.5%), although this difference did not reach statistical significance (p = 0.108). Worsening in communication was also more frequent among participants with strict long COVID, but the difference was not statistically significant (10.5% vs. 6.5%; p = 0.080). The composite outcome of any functional or cognitive worsening was significantly more frequent among participants with strict long COVID than among those without strict long COVID (53.4% vs. 36.6%; p < 0.001). When all six domains were considered, including vision and hearing, any worsening was observed in 59.7% of participants with strict long COVID and 42.3% of those without strict long COVID (p < 0.001). These findings are presented in Table 5; Fig. 3.

Multivariable analysis

In the multivariable logistic regression model, severe or critical acute COVID-19 was the only factor independently associated with strict long COVID. Participants with severe or critical acute disease had more than twice the odds of strict long COVID compared with those with mild or moderate disease (adjusted OR, 2.36; 95% CI, 1.62–3.41; p < 0.001). Female sex was not independently associated with strict long COVID after adjustment (adjusted OR, 0.74; 95% CI, 0.54–1.02; p = 0.070). Age ≥ 81 years was also not independently associated with the outcome (adjusted OR, 0.89; 95% CI, 0.59–1.34; p = 0.583). Similarly, incomplete vaccination with fewer than three doses (adjusted OR, 1.10; 95% CI, 0.72–1.69; p = 0.662), diabetes with organ involvement (adjusted OR, 1.04; 95% CI, 0.66–1.66; p = 0.855), hospitalization (adjusted OR, 0.75; 95% CI, 0.52–1.08; p = 0.121), and partial or permanent care dependence (adjusted OR, 1.17; 95% CI, 0.80–1.72; p = 0.412) were not independently associated with strict long COVID.

The multivariable model is shown in Table 6 and visualized in Fig. 4. No statistically significant sex-by-age or vaccination-by-age interaction was observed; therefore, interaction terms were not retained in the final multivariable model. Model diagnostics showed no evidence of relevant multicollinearity, with all variance inflation factor values below 2. The model had modest discrimination, with an area under the receiver operating characteristic curve of approximately 0.61. The number of outcome events was adequate for the number of prespecified covariates included in the model, reducing concern for overfitting.

Discussion

This cross-sectional study provides evidence on strict long COVID and persistent post-COVID symptoms among community-dwelling older adults in Quito, Ecuador. Using a strict definition based on symptoms persisting for more than 12 weeks after SARS-CoV-2 infection, the prevalence of long COVID was 18.2%. In addition, a large proportion of participants reported symptoms lasting 4–12 weeks, which were classified as persistent post-COVID symptoms but not as strict long COVID. This distinction is important because it avoids overestimating long COVID prevalence while still recognizing the substantial burden of post-acute symptoms in older adults.

The prevalence of strict long COVID observed in this study is within the range reported in international studies, although comparisons should be interpreted cautiously because definitions, study populations, follow-up periods, variants, vaccination status, and methods of symptom ascertainment vary widely across studies. Earlier estimates of long COVID prevalence have differed substantially depending on whether symptoms were assessed after 4 weeks, 8 weeks, 12 weeks, or longer follow-up periods. By applying a stricter > 12-week definition, our analysis identified a smaller but clinically meaningful group of older adults with prolonged symptoms. A sensitivity analysis using symptoms lasting more than 90 days yielded a lower estimate, suggesting some inconsistency between categorical and days-based symptom-duration variables. This discrepancy reinforces the importance of clearly defining the outcome and standardizing symptom-duration assessment in future studies.

The most consistent independent factor associated with strict long COVID in this cohort was the severity of acute COVID-19. Participants who experienced severe or critical acute disease had more than twice the adjusted odds of strict long COVID compared with those who had mild or moderate disease. This finding is clinically plausible, as severe acute infection may reflect a higher inflammatory burden, greater pulmonary or systemic injury, longer recovery time, and greater physiological stress in older adults. In geriatric populations, acute illness may also accelerate loss of reserve, precipitate deconditioning, and unmask underlying vulnerability. Therefore, the association between acute severity and strict long COVID should be interpreted not only as a marker of viral disease burden but also as a potential indicator of post-acute functional risk.

In contrast with the original analysis, female sex, advanced age, diabetes with organ involvement, incomplete vaccination, hospitalization, and care dependence were not independently associated with strict long COVID after adjustment. This does not necessarily mean that these factors are unimportant. Rather, their effects may depend on the definition of long COVID, the timing of assessment, residual confounding, survival bias, and the interplay between frailty, access to care, and baseline health status. Previous studies have frequently reported higher long COVID risk among women and people with comorbidities, but our strict-definition analysis did not confirm these associations in the adjusted model. These differences highlight the sensitivity of risk-factor estimates to outcome definition and underscore the need for cautious, noncausal language when interpreting findings from cross-sectional data.

No statistically significant sex-by-age or vaccination-by-age interaction was observed; therefore, interaction terms were not retained in the final multivariable model. The lack of an independent association between age ≥ 81 years and strict long COVID deserves consideration. Older adults are biologically vulnerable to severe acute COVID-19, complications, and functional decline, but this does not necessarily translate into a higher measured prevalence of strict long COVID in cross-sectional analyses. Several explanations are possible. The oldest participants may be less likely to report symptoms in the same way as younger older adults, may have competing health conditions that obscure attribution to COVID-19, or may be underrepresented among survivors available for community assessment. In addition, symptom persistence and functional decline are overlapping but distinct constructs. Age may be more strongly related to disability, dependence, and recovery trajectory than to the binary classification of strict long COVID itself.

The vaccination findings should also be interpreted cautiously. Although incomplete vaccination showed an association in unadjusted comparisons, it was not independently associated with strict long COVID in the multivariable model. This may reflect high vaccination coverage in the sample, limited variability in vaccine exposure, timing of vaccination in relation to infection, or residual confounding by health-seeking behavior and access to care. Vaccination remains a major public health strategy for preventing severe COVID-19 and its complications, but the present analysis does not support a strong independent association between incomplete vaccination and strict long COVID after adjustment. Therefore, the revised manuscript should avoid stating that vaccination was independently protective against long COVID in this cohort.

A key contribution of this study is the geriatric characterization of functional and cognitive worsening after COVID-19. Participants with strict long COVID had a substantially higher frequency of overall functional or cognitive worsening than those without strict long COVID. The difference was particularly notable for walking or climbing stairs, suggesting that mobility impairment may be one of the most clinically relevant consequences of prolonged post-COVID symptoms in older adults. Worsening in grooming or dressing and hearing was also more common among participants with strict long COVID. These findings support the view that long COVID in older adults should not be evaluated only as a symptom syndrome but also as a condition with potential implications for autonomy, daily functioning, rehabilitation needs, and caregiver burden.

The symptom profile observed among participants with strict long COVID was consistent with a multidimensional post-COVID syndrome. Dyspnea, cognitive impairment, sleep disturbance, difficulty sleeping, myalgia, depression, anxiety, and vision changes were more frequent among those with strict long COVID. This pattern suggests that strict long COVID in older adults may involve respiratory, neurocognitive, psychological, sleep-related, and musculoskeletal dimensions. In clinical practice, this supports the need for comprehensive assessment rather than single-symptom screening. In primary care settings, older adults with prolonged symptoms should be evaluated for mobility decline, cognitive complaints, sleep problems, mood symptoms, and respiratory limitation.

The public health implications are relevant for Ecuador and similar middle-income settings in Latin America. Even when strict long COVID was defined conservatively, nearly one in five older adults met the > 12-week criterion, and a much larger proportion experienced persistent symptoms lasting 4–12 weeks. This indicates that post-COVID care pathways should distinguish between persistent post-COVID symptoms and strict long COVID while ensuring that both groups receive appropriate follow-up. Community-based screening, functional assessment, rehabilitation referral, chronic disease management, and vaccination counseling can be integrated into primary care systems. Given limited availability of specialized long COVID clinics in many Latin American settings, geriatric-focused primary care may be the most feasible platform for identifying and managing persistent symptoms and functional decline.

This study has several strengths. It included a relatively large sample of older adults, focused on a Latin American population that remains underrepresented in long COVID research, and required PCR-confirmed SARS-CoV-2 infection rather than relying solely on self-reported infection. Data were collected through structured interviews by trained health personnel, and the revised analysis applied a stricter and more transparent outcome definition. The study also incorporated functional and cognitive worsening, which is particularly relevant in geriatric research and adds clinical depth beyond symptom prevalence alone.

Several limitations should be acknowledged. First, the cross-sectional design precludes causal inference and does not allow assessment of symptom trajectories over time. Second, the convenience sampling strategy may introduce selection bias and limits generalizability to all older adults in Quito or Ecuador. Third, the study relied in part on retrospective recall of symptoms and pre-COVID functional status, which may be affected by memory limitations, symptom perception, and the long interval between infection and assessment. Although cognitive screening was used to improve reliability, recall bias cannot be excluded. Fourth, functional and cognitive worsening were based on reported difficulty rather than comprehensive objective geriatric assessments. Fifth, the absence of biological markers, inflammatory profiles, pulmonary function tests, and standardized frailty measures limits mechanistic interpretation. Sixth, residual confounding by frailty, baseline disability, comorbidity burden, health-seeking behavior, socioeconomic status, and access to care may have influenced the observed associations. Finally, the discrepancy between the > 12-week categorical definition and the > 90-day sensitivity definition indicates that symptom-duration measurement may not have been fully harmonized across database variables.

Future research should prioritize longitudinal studies of older adults after SARS-CoV-2 infection to better characterize recovery trajectories, persistence of symptoms, incident disability, healthcare utilization, and mortality. Studies should use standardized long COVID definitions, objective functional and cognitive measures, frailty assessment, and clear timing of infection, vaccination, and symptom evaluation. Comparative studies across Latin America are also needed to understand how social determinants, healthcare access, vaccination strategies, and rehabilitation availability influence long COVID outcomes in aging populations. Interventional studies should evaluate whether geriatric rehabilitation, physical activity programs, cognitive support, sleep interventions, and integrated primary care follow-up can reduce disability and improve quality of life among older adults with strict long COVID and persistent post-COVID symptoms.

Conclusion

In this cross-sectional study of community-dwelling older adults in Quito, Ecuador, strict long COVID, defined as symptoms persisting for more than 12 weeks after SARS-CoV-2 infection, affected nearly one in five participants. In addition, a substantial proportion of older adults reported persistent post-COVID symptoms lasting 4–12 weeks, highlighting a broader post-acute symptom burden that should be distinguished from strict long COVID. Severe or critical acute COVID-19 was the main factor independently associated with strict long COVID. In contrast, advanced age, female sex, diabetes with organ involvement, incomplete vaccination, hospitalization, and care dependence were not independently associated with strict long COVID after adjustment. Participants with strict long COVID also showed greater functional and cognitive worsening, particularly in mobility-related domains. These findings support the need to integrate post-COVID screening, functional assessment, and geriatric rehabilitation into primary care for older adults. Future longitudinal studies using standardized long COVID definitions and objective geriatric measures are needed to clarify recovery trajectories, identify modifiable risk factors, and guide interventions aimed at preserving independence and quality of life in aging populations.

Abbreviations

aOR

Adjusted odds ratio

CEISH

Comité de Ética en Investigación en Seres Humanos

CI

Confidence interval

COVID-19

Coronavirus disease 2019

PCR

Polymerase chain reaction

SARS-CoV-2

Severe acute respiratory syndrome coronavirus 2

SDG

Sustainable development goal

VIF

Variance inflation factor

WHO

World health organization

Authors’ contributions

F.G-A. developed the research protocol and design, data collection, statistical analysis, evaluation, data interpretation, critical analysis, discussion, writing, and final approval of the manuscript.

Funding

The author received no external funding for this study.

Data availability

The data supporting the findings of this study are available from the corresponding author upon reasonable request, subject to ethical and privacy restrictions.

Declarations

Ethics approval and consent to participate

The study received approval from the Committee on Ethics in Human Research (CEISH) from Carlos Andrade Marín Specialty Hospital (HCAM) on July 6, 2020, under protocol code IESS-HCAM-CEISH-2020-200-DF, in compliance with national and international research regulations.

Consent for publication

Not applicable. This manuscript does not contain identifying images or personal clinical details that could compromise participant anonymity.

Competing interests

The authors declare no competing interests.