Abstract
Background
Early-life antibiotic exposure has been shown to disrupt gut microbiota, impair immune function, and perturb metabolic health. These are pathways known to contribute to frailty, yet direct evidence connecting antibiotic use to frailty is notably limited. We aimed to examine the association between extensive antibiotic use during early life and the subsequent risk of frailty among middle-aged and older-aged adults.
Methods
We used data from the UK Biobank, a population-based cohort study that recruited adults aged 40–70 years during 2006 and 2010 across 22 centres in England, Scotland and Wales. “Extensive antibiotic use during early life” was self-reported as long-term or recurrent antibiotic courses (≥ 3 per year) during childhood or adolescence. To comprehensively capture frailty’s multidimensional nature and validate the robustness of associations, we applied three validated measures each targeting distinct frailty dimensions: frailty phenotype for physical frailty, hospital frailty risk score for multimorbidity-related frailty, and frailty index for comprehensive cumulative deficit frailty. Multivariable logistic regression models were used to explore the relationships between extensive antibiotic use during early life and risk of frailty.
Results
A total of 156,401 participants were included in this study, among whom 22,692 participants self-reported extensive antibiotic use during early life. Compared to people without extensive antibiotic use during early life, those with this exposure had increased odds of physical frailty, hospital frailty and comprehensive frailty. The corresponding odd ratio (95% confidence interval were 1.36 (1.32–1.41), 1.42 (1.37–1.48) and 1.86 (1.80–1.92), respectively. Subgroup analyses were consistent with the main results.
Conclusions
Our findings indicate that extensive antibiotic use during early life is associated with elevated risk of frailty in middle and older adulthood. They highlight the importance of prudent antibiotic use in early life and the need for additional studies to explore the underlying causal pathways.
Supplementary Information
The online version contains supplementary material available at 10.1186/s12889-026-26612-0.
Keywords: Extensive antibiotic use during early life, Physical frailty, Hospital frailty, Comprehensive frailty
Introduction
Frailty is a complex of symptoms that is characterized as a state of increased vulnerability due to cumulative decline in multiple physiological systems (sarcopenia, nutritional deficiencies, hormonal changes, and increased inflammation) [1]. Though not a disease in itself, frailty is strongly associated with increased risk of all-cause mortality and other health outcomes, such as disability, falls, fractures, decreasing mobility, loneliness, lower quality of life, depression, cognitive decline, dementia, and hospitalization [2]. As the global population ages, frailty has become an emerging global health burden with major implications for clinical practice and public health [3].
Antibiotics are usually used to treat or prevent bacterial infections, however, the unnecessary use of antibiotics contributes to the emergence of antibiotic resistance. Global antibiotic consumption was projected to rise by up to 200% by 2030 compared to the 42 billion DDDs estimated in 2015 [4]. Bacterial resistance, driven by antibiotic consumption, is seen as a threat to global health, development, and sustainability. It was estimated 4.95 million deaths associated with antimicrobial resistance in 2019, including 1.27 million deaths attributable to bacterial resistance [5]. If no urgent actions are taken, there are an estimated 10 million deaths and 100 trillion USD in economic loss annually from antimicrobial resistance by 2050 [6]. In addition, antibiotics are amongst the most commonly prescribed drugs for children due to their greater vulnerability to infectious diseases and higher associated mortality risks compared with adults. Inappropriate antibiotic prescriptions are also most common in pediatric setting [7]. A review indicated that more than two-thirds of children receive antibiotics before the age of two years, and over half of children receive at least one antibiotic each year in many settings [8].
Mounting evidence indicates antibiotic exposure is associated with adverse short-term and long-term health outcomes. Firstly, a systematic review showed that antibiotic exposure was associated with an increased risk of immunological and metabolic diseases, including allergic diseases, inflammatory bowel diseases, rheumatological diseases, diabetes mellitus, and weight gain [9]. Previous findings have shown that antibiotic exposure was linked to chronic disease and their deaths, such as cardiovascular disease [10] and cancer [11]. In addition, several previous studies have revealed that antibiotic treatment impaired central nervous system [12] and even cognitive impairment [13]. A positive association has been confirmed between antibiotic exposure and the problems related to metacognition, executive function, impulsivity, hyperactivity, attention-deficit hyperactivity disorder, anxiety and emotional problems [12, 14]. However, evidence directing linking antibiotic use to frailty remains limited.
To our knowledge, no study has previously examined the association between extensive antibiotic use during early life and frailty in middle and old age. Frailty is a heterogeneous construct encompassing physical, clinical, and comprehensive deficits, and single-dimensional measures may fail to capture the full spectrum of vulnerability. In this study, we used three complementary validated frailty indicators (physical frailty phenotype (PFP), hospital frailty risk score (HFRS) and frailty index (FI)) in the UK Biobank to test whether early-life extensive antibiotic use is consistently associated with frailty across multiple operationalizations.
Methods
Study population
The UK Biobank is a large population-based cohort study that recruited over 502,000 participants aged 40–70 years during 2006 and 2010. Individuals who were living within one of 22 assessment centres across England, Scotland and Wales were invited to participate for baseline assessment. UK Biobank was approved by the North-West Multicenter Research Ethics Committee and all participants provided written informed consent for collection of questionnaire and biological data [14]. This research was conducted under UK Biobank application number 92,597.
Frailty
To comprehensively assess frailty and address its multidimensionality, we selected three validated measures with distinct focuses. The FI and PFP were available only at baseline, while the HFRS was available both at baseline and during follow-up.
The PFP defines frailty as a clinical syndrome associated with five symptoms: weight loss, fatigue, weakness, slow walking speed, and reduced or absent physical activity [15]. It focuses on overt physical functional deficits, capturing the “classic” frailty phenotype linked to sarcopenia and functional decline. The PFP has previously been created for UK Biobank participants according to five frailty criteria [16]. Weight loss, exhaustion, slowness, and low physical activity were self-reported at baseline. Weakness was assessed by measured grip strength with the higher value of the right- and left-hand measurements. The PFP was categorized into three levels: non-frail ( < = 0), pre-frail (> 0 to < = 2), and frail (> 2) (Table S1).
The HFRS measures the number of multimorbid conditions without additional data collection, using a weighted score based on International Statistical Classification of Diseases and Related Health Problems, Tenth Revision (ICD-10) for inpatients [17]. Derived from inpatient ICD-10 codes, it quantifies frailty related to multimorbidity and chronic disease burden. The HFRS was computed based on ICD-10 codes for hospital records and each of the 109 frailty-related ICD-10 codes were assigned a weight ranging from 0.1 to 7.1. Based on the weighted codes, participants were calculated into low (< 5), intermediate ( > = 5 to < = 15) and high (> 15) risk of frailty (Table S2) [18].
The FI defines as a ratio of accumulated deficits over the total number of deficits considered [19]. It calculates the ratio of accumulated deficits across physical, mental, and functional domains, providing a comprehensive measure of age-related vulnerability. Validated for UK Biobank participants, the FI uses 49 self-reported variables for physical and mental well-being (Table S3) [20]. An FI value was calculated as the sum of deficits present in a person divided by the total number of deficits composing the FI. The FI was categorized into three levels of frailty: non-frail ( < = 0.10), pre-frail (> 0.10 to < 0.21), and frail ( > = 0.21) [2].
Extensive antibiotic use during early life
Extensive antibiotic use during early life was collected in 2017 or 2018 via a self-reported question “During childhood or as a teenager did you receive long-term or recurrent course (3 or more per year) of antibiotics (for example for tonsillitis or acne)?”. Optional responses were “yes” or “no”, which were used to categorize participants into the exposure or non-exposure group.
Covariates
We included the following factors as covariates in the analyses: age at baseline, sex, ethnicity, education level, income level, Townsend deprivation quintile, siblings, smoking status, drinking status, body mass index (BMI) [13]. Ethnicity was categorized as white and non-white. Education level was grouped into < = 10, 10–12, and > 12 years. Income level was divided into four categories of level 1(< £18,000), level 2 (>= £18,000 to <£31,000), level 3 (>= £31,000 to < £52,000), and level 4 (>= £52,000). The Townsend deprivation quintile was derived from national census data regarding unemployment, car ownership, home ownership, and household overcrowding; higher values correspond to higher levels of socioeconomic deprivation. Notably, while income level reflects individual financial resources, the Townsend index captures broader area-level socioeconomic disadvantage. The number of siblings was categorized as 0, 1–2, >=3. Smoking status and drinking status were each classified as current, former, or never. BMI was grouped into underweight/normal weight (< 25.0 kg/m2), overweight ( > = 25.0 to < 30.0 kg/m2) and obese ( > = 30.0 kg/ m2).
Statistical analysis
Baseline characteristics are presented as proportions for categorical data and mean (SD) for continuous variables. We compared the proportions of antibiotic use during early life in different subgroups using Chi-square tests or Fisher’s exact test.
We used multivariable logistic regression models to estimate adjusted odds ratio (OR) and 95% confidence interval (CI) for the associations between extensive antibiotic use during early life and the risk of physical frailty, hospital frailty and comprehensive frailty. We adjusted covariates step by step, i.e., age was adjusted in model 1; sex, ethnicity, education level, income level, Townsend deprivation quintile, and siblings were further adjusted based on model 1; and smoking status, drinking status, BMI were further adjusted based on model 2. To assess potential multicollinearity among covariates, we calculated variance inflation factors (VIF) for all variables included in the multivariable models. A VIF value>10 was considered indicative of significant multicollinearity.
Subgroup analysis and sensitivity analysis
In addition, we evaluated the association between antibiotic use during early life and risk of physical frailty, hospital frailty and comprehensive frailty by levels of covariates. Finally, we performed three sensitivity analyses to test the robustness of the findings. First, we only included participants who were aged 60 years or older in 2022. Second, we performed a separate analysis focusing on participants aged < = 50 years. Third, we repeated the main analysis using frailty as a continuous variable.
We did all statistical analyses using Stata version 17.0. P-values < 0.05 were considered statistically significant.
Results
Participants characteristics
We excluded participants with missing data for extensive antibiotic use during early life (n = 344,002) and missing covariates (n = 1,338). We also excluded those with > 20% missing data used to calculate the FI and PFP. Finally, 156,401 participants were included in this study (Fig. 1). Participants had a mean age at baseline of 55.9 years (SD 7.7), with 88,601 (56.7%) being female, and 151,564 (96.9%) being white. A total of 22,692 (14.5%) participants reported extensive antibiotic use during early life (Table 1).
Fig. 1.
Flow diagram of study participants
Table 1.
Baseline characteristics of participants with physical frailty, hospital frailty, and comprehensive frailty
| Characteristics | Total (N) | Physical frailty (%) | Hospital frailty (%) | Comprehensive frailty (%) | |||
|---|---|---|---|---|---|---|---|
| No | Yes | No | Yes | No | Yes | ||
| Extensive antibiotic use during early life | |||||||
| No | 133,709 | 98.70 | 1.30 | 97.78 | 2.22 | 78.01 | 21.99 |
| Yes | 22,692 | 97.18 | 2.82 | 97.28 | 2.72 | 64.95 | 35.05 |
| Age, years, n(%) | |||||||
| < 60 years | 96,314 | 98.51 | 1.49 | 98.81 | 1.19 | 79.64 | 20.36 |
| >= 60 years | 60,087 | 98.43 | 1.57 | 95.93 | 4.07 | 70.45 | 29.55 |
| Sex, n(%) | |||||||
| Female | 88,601 | 98.17 | 1.83 | 98.03 | 1.97 | 73.20 | 26.80 |
| Male | 67,800 | 98.48 | 1.52 | 97.28 | 2.72 | 79.92 | 20.08 |
| Ethnicity, n(%) | |||||||
| Non-white | 4837 | 97.19 | 2.81 | 97.97 | 2.03 | 77.63 | 22.37 |
| White | 151,564 | 98.52 | 1.48 | 97.70 | 2.30 | 76.06 | 23.94 |
| Education level, years, n(%) | |||||||
| <=10 | 55,331 | 97.80 | 2.20 | 97.10 | 2.90 | 70.33 | 29.67 |
| (10,12] | 21,482 | 98.43 | 1.57 | 97.89 | 2.11 | 76.83 | 23.17 |
| >12 | 79,588 | 98.97 | 1.03 | 97.08 | 2.30 | 79.94 | 20.06 |
| Income level, £, n(%) | |||||||
| <18,000 | 22,363 | 96.20 | 3.80 | 95.40 | 4.60 | 61.71 | 38.29 |
| [18000, 31000) | 35,247 | 98.21 | 1.79 | 97.08 | 2.92 | 70.94 | 29.06 |
| [31000, 52000) | 44,658 | 98.85 | 1.15 | 98.07 | 1.93 | 77.62 | 22.38 |
| >=52,000 | 54,133 | 99.28 | 0.72 | 98.76 | 1.24 | 84.19 | 15.81 |
| Townsend deprivation quintile, n(%) | |||||||
| 1 (least deprived) | 35,654 | 99.16 | 0.84 | 97.92 | 2.08 | 79.43 | 20.57 |
| 2 | 33,570 | 98.84 | 1.16 | 97.91 | 2.09 | 77.72 | 22.28 |
| 3 | 32,296 | 98.64 | 1.36 | 97.93 | 2.07 | 76.56 | 23.44 |
| 4 | 30,638 | 98.28 | 1.72 | 97.67 | 2.33 | 74.82 | 25.18 |
| 5 (most deprived) | 24,243 | 97.03 | 2.97 | 96.85 | 3.15 | 70.06 | 29.94 |
| Siblings, n(%) | |||||||
| 0 | 18,914 | 98.39 | 1.61 | 97.26 | 2.74 | 75.08 | 24.92 |
| <=2 | 94,212 | 98.66 | 1.34 | 97.84 | 2.16 | 77.47 | 22.53 |
| >=3 | 43,275 | 98.13 | 1.87 | 97.61 | 2.39 | 73.60 | 23.89 |
| Smoking status, n(%) | |||||||
| Never | 90,639 | 98.72 | 1.28 | 98.16 | 1.84 | 79.35 | 20.65 |
| Past | 54,575 | 98.36 | 1.64 | 97.16 | 2.84 | 72.05 | 27.95 |
| Current | 11,187 | 97.15 | 2.85 | 96.64 | 3.36 | 69.75 | 30.25 |
| Drinking status, n(%) | |||||||
| Never | 4338 | 96.01 | 3.99 | 96.01 | 3.99 | 64.75 | 35.25 |
| Past | 4126 | 94.18 | 5.82 | 94.74 | 5.26 | 55.91 | 44.09 |
| Current | 147,937 | 98.67 | 1.33 | 97.84 | 2.16 | 77.01 | 22.99 |
| Body mass index, kg/m2, n(%) | |||||||
| Underweight or Normal (< 25.0) | 62,317 | 99.40 | 0.60 | 98.32 | 1.68 | 82.68 | 17.32 |
| Overweight [25.0, 30.0) | 64,677 | 98.87 | 1.13 | 97.88 | 2.12 | 76.82 | 23.18 |
| Obese ( > = 30.0) | 30,407 | 95.80 | 4.20 | 96.09 | 3.91 | 61.36 | 38.64 |
Extensive antibiotic use during early life was more common in younger people, women, current smoker, past drinker, and individuals with high income, high Townsend deprivation index, and obesity. In addition, people with extensive antibiotic use during early life had a higher prevalence of physical frailty, hospital frailty and comprehensive frailty (Table S4). Compared with participants without frailty, physical frailty, hospital frailty and comprehensive frailty were more likely to be experienced by individuals with the following characteristics: over 60 years, low education, less income, high Townsend deprivation index, current smoker, past drinker, and obesity (Table 1).
Association between extensive antibiotic use during early life and different frailty
The VIF values for all covariates were < 2.7, indicating no significant multicollinearity (Table S7). After adjusting for age, participants with extensive antibiotic use during early life had increased odds of physical frailty (OR 1.50, 95% CI 1.45–1.54), hospital frailty (1.47, 1.42–1.53) and comprehensive frailty (2.10, 2.04–2.16) compared to those without such exposure, After demographic and lifestyle factors were further adjusted, the associations remained statistically significant, though the effect sizes were slightly attenuated, with OR (95% CI) of 1.36 (1.32–1.41), 1.42 (1.37–1.48) and 1.86 (1.80–1.92) for physical frailty, hospital frailty, and comprehensive frailty, respectively (Table 2).
Table 2.
Association between extensive antibiotic use during early life and different frailty: subgroup analysis
| Variables | Model 1 | Model 2 | Model 3 | |||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| OR | 95% CI | P-value | OR | 95% CI | P-value | OR | 95% CI | P-value | ||||
| Physical frailty | 1.50 | 1.45 | 1.54 | < 0.000 | 1.44 | 1.39 | 1.48 | < 0.000 | 1.36 | 1.32 | 1.41 | < 0.000 |
| Hospital frailty | 1.47 | 1.42 | 1.53 | < 0.000 | 1.48 | 1.43 | 1.54 | < 0.000 | 1.42 | 1.37 | 1.48 | < 0.000 |
| Comprehensive frailty | 2.10 | 2.04 | 2.16 | < 0.000 | 1.96 | 1.90 | 2.02 | < 0.000 | 1.86 | 1.80 | 1.92 | < 0.000 |
Model 1: adjusted for age.
Model 2: adjusted for sex, ethnicity, education level, income level, Townsend deprivation quintile, siblings based on model 1.
Model 3: adjusted for smoking status, drinking status, BMI based on model 2.
OR: odds ratio; CI: confidence interval
Subgroup analyses and sensitivity analyses
After we stratified analyses by levels of covariates, we found the effect of antibiotic use during early life on risk of physical frailty was higher in past drinkers, and people with less income and underweight. The association between antibiotic use and hospital frailty was higher in people less than 60 years, those with a high Townsend deprivation index, underweight individuals, and never drinker. The association of antibiotic use during early life with comprehensive frailty was higher in never drinker, and people with less income, a high Townsend deprivation index, and obesity (Table 3).
Table 3.
Extensive antibiotic use during early life and different frailty: subgroup analysis
| Variables | Physical frailty | Hospital frailty | Comprehensive frailty | ||||||
|---|---|---|---|---|---|---|---|---|---|
| OR | 95% CI | OR | 95% CI | OR | 95% CI | ||||
| Age, years old | |||||||||
| < 60 | 1.38 | 1.33 | 1.43 | 1.50 | 1.43 | 1.58 | 1.85 | 1.78 | 1.92 |
| >= 60 | 1.32 | 1.24 | 1.40 | 1.30 | 1.22 | 1.38 | 1.82 | 1.72 | 1.93 |
| Sex | |||||||||
| Female | 1.38 | 1.33 | 1.43 | 1.50 | 1.43 | 1.58 | 1.89 | 1.82 | 1.96 |
| Male | 1.32 | 1.25 | 1.40 | 1.27 | 1.18 | 1.36 | 1.83 | 1.73 | 1.94 |
| Ethnicity | |||||||||
| Non-white | 1.48 | 1.26 | 1.75 | 1.32 | 1.06 | 1.65 | 1.93 | 1.62 | 2.28 |
| White | 1.36 | 1.32 | 1.40 | 1.43 | 1.37 | 1.48 | 1.85 | 1.80 | 1.91 |
| Education level, years | |||||||||
| <=10 | 1.40 | 1.33 | 1.47 | 1.44 | 1.36 | 1.53 | 1.91 | 1.82 | 2.01 |
| (10,12] | 1.28 | 1.18 | 1.39 | 1.49 | 1.35 | 1.66 | 1.77 | 1.63 | 1.92 |
| ＞12 | 1.36 | 1.30 | 1.42 | 1.38 | 1.30 | 1.47 | 1.84 | 1.76 | 1.92 |
| Income level, £ | |||||||||
| <18000 | 1.47 | 1.36 | 1.59 | 1.48 | 1.35 | 1.61 | 1.97 | 1.82 | 2.13 |
| [18000, 31000) | 1.35 | 1.27 | 1.45 | 1.41 | 1.30 | 1.52 | 1.90 | 1.78 | 2.03 |
| [31000, 52000) | 1.35 | 1.27 | 1.43 | 1.40 | 1.30 | 1.51 | 1.82 | 1.72 | 1.93 |
| >=52000 | 1.32 | 1.25 | 1.39 | 1.41 | 1.31 | 1.52 | 1.78 | 1.69 | 1.88 |
| Townsend deprivation quintile | |||||||||
| 1 (least deprived) | 1.37 | 1.28 | 1.47 | 1.37 | 1.26 | 1.50 | 1.87 | 1.75 | 2.00 |
| 2 | 1.35 | 1.26 | 1.44 | 1.33 | 1.21 | 1.45 | 1.76 | 1.65 | 1.89 |
| 3 | 1.35 | 1.26 | 1.45 | 1.43 | 1.31 | 1.56 | 1.85 | 1.72 | 1.98 |
| 4 | 1.31 | 1.23 | 1.41 | 1.54 | 1.41 | 1.68 | 1.88 | 1.75 | 2.01 |
| 5 (most deprived) | 1.43 | 1.33 | 1.54 | 1.44 | 1.32 | 1.57 | 1.95 | 1.81 | 2.09 |
| Siblings | |||||||||
| 0 | 1.32 | 1.21 | 1.44 | 1.41 | 1.26 | 1.57 | 1.85 | 1.69 | 2.02 |
| <=2 | 1.34 | 1.29 | 1.40 | 1.41 | 1.34 | 1.49 | 1.83 | 1.75 | 1.90 |
| >=3 | 1.43 | 1.35 | 1.51 | 1.45 | 1.35 | 1.56 | 1.93 | 1.82 | 2.05 |
| Smoking status | |||||||||
| Never | 1.36 | 1.30 | 1.42 | 1.45 | 1.37 | 1.53 | 1.89 | 1.82 | 1.97 |
| Past | 1.36 | 1.29 | 1.43 | 1.37 | 1.29 | 1.46 | 1.82 | 1.73 | 1.92 |
| Current | 1.43 | 1.28 | 1.58 | 1.50 | 1.32 | 1.71 | 1.79 | 1.60 | 1.99 |
| Drinking status | |||||||||
| Never | 1.66 | 1.40 | 1.97 | 1.62 | 1.33 | 1.97 | 2.29 | 1.92 | 2.73 |
| Past | 1.69 | 1.44 | 1.98 | 1.44 | 1.21 | 1.73 | 1.83 | 1.55 | 2.15 |
| Current | 1.34 | 1.30 | 1.38 | 1.41 | 1.36 | 1.47 | 1.85 | 1.79 | 1.91 |
| Body mass index, kg/m2 | |||||||||
| Underweight or Normal (<25.0) | 1.37 | 1.30 | 1.45 | 1.40 | 1.30 | 1.50 | 1.80 | 1.71 | 1.90 |
| Overweight [25.0, 30.0) | 1.36 | 1.30 | 1.43 | 1.38 | 1.29 | 1.47 | 1.87 | 1.78 | 1.96 |
| Obese (>=30.0) | 1.33 | 1.25 | 1.41 | 1.51 | 1.40 | 1.62 | 1.94 | 1.82 | 2.06 |
We repeated the main analysis among participants aged > = 60 years (Table S5) or < = 50 years (Table S8), and the OR remained essentially unchanged. Table S6 shows the positive associations between extensive antibiotic use during early life and risk of physical frailty, hospital frailty and comprehensive frailty when frailty was treated as a continuous variable.
Discussion
Our findings indicate that extensive antibiotic use during early life is associated with a higher risk of physical frailty, hospital frailty and comprehensive frailty, although minor discrepancies existed among the different frailty indicators. This consistency across complementary measures strengthens our conclusion, as it rules out the possibility that the association is limited to a single operationalization of frailty and instead suggests a broad impact of early antibiotic exposure on the pathways underlying multiple frailty dimensions. Our results, based on a large general population sample, reinforce the observed association between early-life extensive antibiotic use and late-life frailty, highlighting the need for prudent antibiotic use in childhood and adolescence.
Frailty represents a decline of the function of multiple systems. In this study, we found that people who reported extensive antibiotic use during early life had a 36%, 42%, and 86% increased risk of physical frailty, hospital frailty and comprehensive frailty, respectively. Although no direct studies have proven that antibiotics were associated with frailty, there are several studies that have shown the effects of antibiotics on physical health. A systematic review and meta-analysis found that antibiotic use was positively associated with adverse health outcomes, such as excessive weight gain, asthma, allergies, inflammatory bowel disease, neurodevelopmental disorders, and autism spectrum disorders [8]. Evidence had shown that antibiotic use is an independent risk factor for poor cognitive function [13, 21] and mental health issues [14]. Several population-based observational studies revealed positive correlations between antibiotic use and anxiety and depression. A cohort study observed that long-term antibiotic exposure increased the risk of adverse mental traits and decreased the memory and intelligence [14]. It was shown that children who had received antibiotics in the first 6 months of life would have increased risk of anxiety and emotional problems [12]. Antibiotic exposure in the first 2 years of life was associated with increased risk of attention-deficit [22], and prenatal antibiotic exposure increased the risk of cerebral palsy and epilepsy [23]. There were some animal studies exploring the effects of antibiotics on cognition. Microbiota depletion by antibiotic exposure during early adolescence in mice had impact on anxiety and cognitive behaviors [24]. It was revealed that low-dose penicillin in late pregnancy and early postnatal life induced impaired anxiety-like and social behaviors in mice [25]. Wang et al. also showed that ampicillin increased anxiety-like behavior and impaired spatial memory in rats [26].
In our analyses, sociodemographic factors, such as older age, less income and high Townsend deprivation index, were related to a greater risk of frailty. These results are supported by another study [27]. Lower income and a high Townsend deprivation index mean that people may not have sufficient financial resources to cover living expenses, engage in leisure activities (such as physical activity), access medicine or even afford healthy food, all of which render people more susceptible to frailty.
In addition, we found that never or past drinkers were linked to an increased risk of frailty compared with current drinkers, which was in line with previous results [28, 29]. However, Guo et al. found that there was no significant causal association between alcohol use and the risk of frailty [30]. A meta-analysis of 44,051 participants indicated that compared with non-drinkers, people with heavier alcohol consumption predicted a lower incidence of frailty with nominal statistical significance [31]. Several potential explanations exist. On the one hand, alcohol is often consumed socially and moderate consumption was shown to facilitate social bonding, and might possibly help construct or reinforce social support or network and prevent social isolation [32]. On the other hand, it should be noted the potential explanations contributing to the discrepant results could be the differences in alcohol consumption quantification, demographic characteristics and types of alcohol, “sick quitter” effects and other confounding factors [31].
Another important finding of our study was that BMI was associated with a higher likelihood of frailty, which was consistent with previously published studies [33, 34]. There are several potential explanations. First, obesity, as defined by a high BMI, has been associated with adipocyte accumulation and sarcopenia, including reduced physical function, which are key components of frailty. Second, high BMI is associated with more adipose tissue, which can promote the development of inflammation, metabolism and transmission of metabolic information between different organs by secreting cytokines such as adiponectin, interleukin-6, tumour necrosis factors, etc [34]. Our study also indicated that underweight was associated with an increased risk of frailty because of malnutrition, decreased muscle mass and decreased physical function [34].
Possible mechanisms that might contribute to the associations between extensive antibiotic use in the early life and frailty in the late life are as follows, first, antibiotics induce a decrease in gut microbial diversity and an increase in pro-inflammatory cytokines, altering basic physiological equilibria. Further, antibiotics influence brain health through the microbiome-gut-brain-axis [35]. Antibiotics also influence insular cortex functional connectivity that can involve cognitive flexibility and memory processing [36].
One important strength of our study is the large sample size of middle-aged and older adults and the cohort study with long follow-up time. Notably, we used three validated, complementary frailty measures to capture physical, clinical, and comprehensive dimensions of frailty. This approach addressed the inherent heterogeneity of frailty and ensures our findings are not dependent on a single definition. Also, frailty was measured by three indicators: FI and PFP measured only at baseline, and HFRS measured at both baseline and during follow-up, reducing bias associated with over-reliance on self-reported data. Some sensitivity analyses were performed to confirm the validity of the findings. To the best of our knowledge, this is the first study to demonstrate the associations of extensive antibiotic use during early life with frailty.
Our study has several limitations. First, and most notably, the assessment of extensive antibiotic use during early life relied on retrospective self-report data collected in 2017–2018, which is inherently prone to recall bias and potential misclassification. Participants were asked to report antibiotic use that occurred up to four decades prior, a timeframe that poses significant challenges to memory accuracy. Previous research on retrospective reporting had consistently demonstrated that such data may be affected by underreporting (e.g., forgetting infrequent or less impactful courses), overreporting (e.g., overestimating frequency due to recall of recurrent illnesses), or misclassification (e.g., confusing antibiotics with other medications or misremembering the number of courses) [37–39]. For instance, individuals with a history of chronic childhood illness may be more likely to recall antibiotic use, while those with milder or less frequent infections may underreport it. If extensive antibiotic use was underreported, this would tend to dilute the observed associations, leading to an underestimation of the true effect size. Conversely, if overreporting was more common among participants who later developed frailty, this could introduce overestimation. The touchscreen questionnaire remains the only available tool to assess early-life antibiotic exposure in this cohort as linked primary care or hospital records primarily capture disease diagnoses, not medication prescribing data. Additionally, UK biobank’s retrospective questionnaire focused specifically on childhood/adolescent antibiotic use and did not capture adult exposure, we adjusted for key adult lifestyle and health covariates (e.g., smoking, drinking, BMI) that may correlate with both adult antibiotic use and frailty risk. However, residual confounding from unmeasured adult antibiotic exposure cannot be fully ruled out, and future studies with longitudinal data on antibiotic use across the lifespan would help disentangle the relative contributions of early-life vs. adult exposure.
Second, due to the lack of detailed data on early-life antibiotic use (e.g., specific antibiotic classes, cumulative dose, or exact timing of exposure), our study was unable to assess dose-response relationships or differential effects of antibiotic types (e.g., broad-spectrum vs. narrow-spectrum) on frailty risk. Further studies with access to granular prescribing data are needed to clarify these nuances. Third, as an observational study, residual confounding from unmeasured factors cannot be ruled out. For example, underlying childhood illnesses that necessitated antibiotic use (e.g., recurrent infections, chronic inflammatory conditions) could themselves be independent risk factors for late-life frailty, potentially mediating or confounding the observed association. While we adjusted for a comprehensive set of sociodemographic and lifestyle covariates, we cannot entirely account for unmeasured childhood health conditions or genetic factors. Last, the UK Biobank population is predominantly white (96.9%) with higher education and income levels relative to the general UK population, which may limit the generalizability of our findings to more ethnically diverse or socioeconomically disadvantaged groups. Despite these limitations, the large sample size, consistent associations across three validated frailty measures (physical frailty phenotype, hospital frailty risk score, and frailty index), and robust subgroup/sensitivity analyses add credibility to our findings.
Conclusions
Compared with individuals without extensive antibiotic use during early life, those who reported such exposure were associated with a higher risk of frailty in middle and older adulthood, and the associations were consistently observed across three validated frailty indicators, including physical frailty, hospital frailty and comprehensive frailty. Given the observational nature of this study, these findings do not establish causation. Well-designed long-term cohort or nested case-control studies are needed to clarify the potential role of antibiotic duration, dose and type of antibiotics for the prevention of frailty, as well as to disentangle confounding by underlying childhood health conditions. Our findings highlight the potential relevance of early-life antibiotic use to health aging trajectories, underscoring the value of prudent antibiotic use in childhood and adolescence while emphasizing that further research is required to elucidate the underlying mechanisms and validate whether targeted interventions on early-life antibiotic use could potentially contribute to frailty prevention.
Supplementary Information
Acknowledgements
This research has been done using the UK Biobank resource under application number 92597, and we are grateful to UK Biobank participants.
Abbreviations
- AMR
antibiotic resistance
- FP
frailty phenotype
- HFRS
hospital frailty risk score
- FI
frailty index
- ICD-10
International Statistical Classification of Diseases and Related Health Problems, Tenth Revision
- BMI
body mass index
- SD
standard deviation
- OR
odds ration
- CI
confidence interval
Authors’ contributions
XZ and DZ conceived of and designed the analyses. XZ, XS and DZ collected the data. XZ, XS, BL, JW, JP, QW and DZ performed the data analyses. XZ, XS, BL, JW, JP, QW, DZ and CSL interpreted the data results. XZ and DZ wrote the first draft of the work. All authors revised the work critically for intellectual content, approved the final version of the work to be published, and are accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved.
Funding sources
This work was supported by the National Natural Science Foundation of China (72304174), Key R&D program (Major Scientific and Technological Innovation Projects) of Shandong Province, China (2025CXGC020108), National Science Foundation of Shandong Province, China (ZR2025MS1150), Young Talent of Lifting engineering for Science and Technology in Shandong Province, China (SDAST2024QTA097), Youth Innovation Team Development Program of Colleges and Universities in Shandong Province (2025RWG027). The funder of the study had no role in study design, data collection, data analysis, data interpretation, or writing of the report.
Data availability
The datasets analysed during the current study are available in the UK biobank repository (application number 92597), http://biobank.ndph.ox.ac.uk/showcase/.
Declarations
Ethics approval and consent to participate
UK Biobank was approved by the North-West Multicenter Research Ethics Committee and all participants provided written informed consent for collection of questionnaire and biological data.
Consent for publication
Not applicable.
Competing interests
The authors declare no competing interests.
Footnotes
Publisher’s Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
References
- 1.Proietti M, Cesari M. Frailty: what is it? Adv Exp Med Biol. 2020;1216:1–7. [DOI] [PubMed] [Google Scholar]
- 2.Fan J, Yu C, Guo Y, Bian Z, Sun Z, Yang L, et al. Frailty index and all-cause and cause-specific mortality in Chinese adults: a prospective cohort study. Lancet Public Health. 2020;5:e650–60. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Hoogendijk EO, Afilalo J, Ensrud KE, Kowal P, Onder G, Fried LP. Frailty: implications for clinical practice and public health. Lancet. 2019;394:1365–75. [DOI] [PubMed] [Google Scholar]
- 4.Klein EY, Van Boeckel TP, Martinez EM, Pant S, Gandra S, Levin SA, et al. Global increase and geographic convergence in antibiotic consumption between 2000 and 2015. Proc Natl Acad Sci U S A. 2018;115:E3463–70. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Murray CJL, Ikuta KS, Sharara F, Swetschinski L, Robles Aguilar G, Gray A, et al. Global burden of bacterial antimicrobial resistance in 2019: a systematic analysis. Lancet. 2022;399:629–55. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.O’Neill J. Tackling drug-resistant infections globally: final report and recommendations. 2016. https://wellcomecollection.org/works/thvwsuba. Accessed 11 September 2016.
- 7.Guo S, Sun Q, Zhao X, Shen L, Zhen X. Prevalence and risk factors for antibiotic utilization in Chinese children. BMC Pediatr. 2021;21:255. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Duong QA, Pittet LF, Curtis N, Zimmermann P. Antibiotic exposure and adverse long-term health outcomes in children: A systematic review and meta-analysis. J Infect. 2022;85:213–300. [DOI] [PubMed] [Google Scholar]
- 9.Lambrecht BN, Hammad H. The immunology of the allergy epidemic and the hygiene hypothesis. Nat Immunol. 2017;18:1076–83. [DOI] [PubMed] [Google Scholar]
- 10.Heianza Y, Zheng Y, Ma W, Rimm EB, Albert CM, Hu FB, et al. Duration and life-stage of antibiotic use and risk of cardiovascular events in women. Eur Heart J. 2019;40:3838–45. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Lofling L, Bahmanyar S, Kieler H, Lambe M, Wagenius G. Antibiotic use prior to a lung cancer diagnosis: a population-based study. Cancer Causes Control. 2021;32:597–607. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Slykerman RF, Coomarasamy C, Wickens K, Thompson J, Stanley TV, Barthow C, et al. Exposure to antibiotics in the first 24 months of life and neurocognitive outcomes at 11 years of age. Psychopharmacology. 2019;236:1573–82. [DOI] [PubMed] [Google Scholar]
- 13.Liu Z, Wei S, Chen X, Liu L, Wei Z, Liao Z, et al. The effect of long-term or repeated use of antibiotics in children and adolescents on cognitive impairment in middle-aged and older person(s) adults: a cohort study. Front Aging Neurosci. 2022;14:833365. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Liang X, Ye J, Wen Y, Li P, Cheng B, Cheng S, et al. Long-term antibiotic use during early life and risks to mental traits: an observational study and gene-environment-wide interaction study in UK biobank cohort. Neuropsychopharmacology. 2021;46:1086–92. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Fried LP, Tangen CM, Walston J, Newman AB, Hirsch C, Gottdiener J, et al. Frailty in older adults: evidence for a phenotype. J Gerontol. 2001;56:M146–56. [DOI] [PubMed] [Google Scholar]
- 16.Hanlon P, Nicholl BI, Jani BD, Lee D, McQueenie R, Mair FS. Frailty and pre-frailty in middle-aged and older adults and its association with Multimorbidity and mortality: a prospective analysis of 493 737 UK biobank participants. Lancet Public Health. 2018;3:e323–32. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Ushida K, Shimizu A, Hori S, Yamamoto Y, Momosaki R. Hospital frailty risk score predicts outcomes in chronic obstructive pulmonary disease exacerbations. Arch Gerontol Geriatr. 2022;100:104658. [DOI] [PubMed] [Google Scholar]
- 18.Gilbert T, Neuburger J, Kraindler J, Keeble E, Smith P, Ariti C, et al. Development and validation of a hospital frailty risk score focusing on older people in acute care settings using electronic hospital records: an observational study. Lancet. 2018;391:1775–82. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Mitnitski AB, Mogilner AJ, Rockwood K. Accumulation of deficits as a proxy measure of aging. Sci World J. 2001;1:323–36. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Mak J, Kuja-Halkola R, Wang Y, Hagg S, Jylhava J. Frailty and comorbidity in predicting community COVID-19 mortality in the U.K. Biobank: the effect of sampling. J Am Geriatr Soc. 2021;69:1128–39. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Mehta RS, Lochhead P, Wang Y, Ma W, Nguyen LH, Kochar B, et al. Association of midlife antibiotic use with subsequent cognitive function in women. PLoS ONE. 2022;17:e264649. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Aversa Z, Atkinson EJ, Schafer MJ, Theiler RN, Rocca WA, Blaser MJ et al. Association of infant antibiotic exposure with childhood health outcomes. Mayo Clinic Proceedings. 2021;96:66–77. [DOI] [PMC free article] [PubMed]
- 23.Meeraus WH, Petersen I, Gilbert R. Association between antibiotic prescribing in pregnancy and cerebral palsy or epilepsy in children born at term: a cohort study using the health improvement network. PLoS ONE. 2015;10:e122034. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Desbonnet L, Clarke G, Traplin A, O’Sullivan O, Crispie F, Moloney RD, et al. Gut microbiota depletion from early adolescence in mice: implications for brain and behaviour. Brain Behav Immun. 2015;48:165–73. [DOI] [PubMed] [Google Scholar]
- 25.Leclercq S, Mian FM, Stanisz AM, Bindels LB, Cambier E, Ben-Amram H, et al. Low-dose penicillin in early life induces long-term changes in murine gut microbiota, brain cytokines and behavior. Nat Communnications. 2017;8:15062. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Wang T, Hu X, Liang S, Li W, Wu X, Wang L, et al. Lactobacillus fermentum NS9 restores the antibiotic induced physiological and psychological abnormalities in rats. Beneficial Microbes. 2015;6:707–17. [DOI] [PubMed] [Google Scholar]
- 27.Gomes C, Guerra RO, Wu YY, Barbosa J, Gomez F, Sousa A, et al. Social and economic predictors of worse frailty status occurrence across selected countries in North and South America and Europe. Innov Aging. 2018;2:igy37. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.Shin J, Kim KJ, Choi J. Smoking, alcohol intake, and frailty in older Korean adult men: cross-sectional study with nationwide data. Eur Geriatr Med. 2020;11:269–77. [DOI] [PubMed] [Google Scholar]
- 29.Shah M, Paulson D, Nguyen V. Alcohol use and frailty risk among older adults over 12 years: the health and retirement study. Clin Gerontol. 2018;41:315–25. [DOI] [PubMed] [Google Scholar]
- 30.Guo X, Tang P, Zhang L, Li R. Tobacco and alcohol consumption and the risk of frailty and falling: a Mendelian randomisation study. J Epidemiol Community Health. 2023;77:349–54. [DOI] [PubMed] [Google Scholar]
- 31.Kojima G, Liljas A, Iliffe S, Jivraj S, Walters K. A systematic review and meta-analysis of prospective associations between alcohol consumption and incident frailty. Age Ageing. 2018;47:26–34. [DOI] [PubMed] [Google Scholar]
- 32.Sayette MA, Creswell KG, Dimoff JD, Fairbairn CE, Cohn JF, Heckman BW, et al. Alcohol and group formation: a multimodal investigation of the effects of alcohol on emotion and social bonding. Psychol Sci. 2012;23:869–78. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 33.Jayanama K, Theou O, Godin J, Mayo A, Cahill L, Rockwood K. Relationship of body mass index with frailty and all-cause mortality among middle-aged and older adults. BMC Med. 2022;20:404. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 34.Yuan L, Chang M, Wang J. Abdominal obesity, body mass index and the risk of frailty in community-dwelling older adults: a systematic review and meta-analysis. Age Ageing. 2021;50:1118–28. [DOI] [PubMed] [Google Scholar]
- 35.Karakan T, Ozkul C, Kupeli AE, Bilici S, Sobarzo-Sanchez E, Capasso R. Gut-brain-microbiota axis: antibiotics and functional Gastrointestinal disorders. Nutrients. 2021;13:389. [DOI] [PMC free article] [PubMed]
- 36.Dinan K, Dinan T. Antibiotics and mental health: the good, the bad and the ugly. J Intern Med. 2022;292:858–69. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 37.Krall EA, Valadian I, Dwyer JT, Gardner J. Recall of childhood illnesses. J Clin Epidemiol. 1988;41:1059–64. [DOI] [PubMed] [Google Scholar]
- 38.Sundermann AC, Hartmann KE, Jones SH, Torstenson ES, Velez Edwards DR. Validation of maternal recall of early pregnancy medication exposure using prospective diary data. Ann Epidemiol. 2017;27:135–9. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 39.Coombes L, Bristowe K, Ellis-Smith C, Aworinde J, Fraser LK, Downing J, et al. Enhancing validity, reliability and participation in self-reported health outcome measurement for children and young people: a systematic review of recall period, response scale format, and administration modality. Qual Life Res. 2021;30:1803–32. [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Data Availability Statement
The datasets analysed during the current study are available in the UK biobank repository (application number 92597), http://biobank.ndph.ox.ac.uk/showcase/.