Introduction

Respiratory tract infections (RTI) are common infections in childhood, with an average of six to eight episodes a year [1,2]. Children’s infections lead to absences from day care and school, as well as parental absences from work. RTIs cause discomfort, but in most cases, the symptoms will resolve within 1–2 weeks without treatment. Viral RTIs, such as the common cold, should not be treated with antibiotics. However, ear infections, sinusitis, and pneumonia are bacterial complications of RTI, in which antibiotics are often used in children.

Antimicrobial medicines are the cornerstone of modern medicine [3]. In Finland, guidelines exist for rational antibiotic prescribing in children, yet deviations, particularly in outpatient settings where most antibiotics are prescribed, have been reported [4,5]. This may increase the risk of antimicrobial resistance (AMR), which is an increasingly serious threat to global public health. AMR occurs when bacteria, viruses, fungi, and parasites no longer respond to antimicrobial medicines and infections become difficult, or even impossible, to treat, which increases the risk of disease spread, severe illness, disability, and death [3]. AMR is directly related to how antibiotic medications are used in society. In Finland, previous studies have shown a decreasing trend in the prescribing of antibacterial medicines for children before the COVID-19 (coronavirus disease) pandemic in 2020 [4,6,7]. A further reduction was observed after the COVID-19 pandemic [8], paralleling the findings from the whole European Economic Area (EEA) [9]. However, despite the decrease seen in some countries, the worldwide consumption of antibiotics has been increasing [10].

In addition to population-level risks related to the rise of AMR, inappropriate use of antimicrobials in children can be associated with other increased risks. A recent systematic review and meta-analysis showed that antibiotic exposure in children is associated with, for example, an increased risk of atopic dermatitis, allergic symptoms, food allergies, asthma, obesity, psoriasis, autism spectrum disorders, and neurodevelopmental disorders [11]. There are also studies that have reported a possible association between early-life antibiotic exposure and type 1 diabetes [12,13]. Studies suggest that antibiotic exposure plays an important role in the development of childhood immune disorders, likely through disruption of the microbiota during early life [14]. It has been estimated that 10% of children account for 25% of the total antibiotic use [15].

Previous studies have shown that socioeconomic and other family-related factors can be related to the levels of respiratory tract infections and antibiotic exposure in children. Thrane et al. (2003) [16] and Jensen et al. (2016) [15] showed that higher parental educational level was associated with a lower number of antibiotic treatments, whereas a study by Hatakka et al. (2010) [17] found that mother’s academic education increased the risk of recurrent acute respiratory illnesses. Using a socioeconomic measure based on the region of residence, Larsen et al. (2021) [18] found a declining association between socioeconomic municipality scores and antibiotic treatments, indicating that regions with a lower socioeconomic status were associated with higher number of antibiotic treatments. Regarding other family-related factors, Hatakka et al. (2010) [17] found that furry pets and older child age reduced the risk of recurrent acute respiratory illnesses, and Toivonen et al. 2016 [19] showed that having older siblings increase the risk of recurrent infections. Thus, although some previous findings have been contradictory regarding the direction of the association, it nevertheless appears that the child’s background is associated with antibiotic exposure.

Day care attendance is known to be an important source of infection transmission in children [20,21]. Previous studies have observed a rapid increase in respiratory infections and antibiotic use after starting day care [16,22,23], followed by a subsequent decrease [22]. Thus, factors affecting day care attendance also play a role in the number of respiratory infections and antibiotic use in children.

In Finland, the day care attendance rate has been lower than in the other Nordic countries and also below the OECD (The Organisation for Economic Co-operation and Development) average, especially for younger children – apart from the most recent years [24,25]. The lower attendance rate is likely linked to the Finnish family benefit system, in which families are entitled to child home care allowance (HCA) if the child does not attend day care. This, in turn, enables parents to take care of the child at home. The HCA is widely used (90% of mothers), and most of the families use it at least for some time after the parental leave period (mean duration 13 months) [26]. At the time of the study data, in the Finnish parental leave scheme, children were approximately 10 months old when the earnings-based parental leave period ended.

In this paper, we use comprehensive register data to describe the antibiotics used by the study group, and to study if the cumulative antibiotic use differs between children by the duration of home care before entering day care. We hypothesise that longer duration of home care (i.e., delayed entry into day care) is associated with lower cumulative antibiotic use. We use descriptive methods and Poisson models to explore the hypothesis, accounting also the socioeconomic and other family-related factors. To be able to account for the potential impact of school entry at age 7 in Finland, the follow-up lasted until the children were 12 years old. As antibiotic exposure and strategies for taking care of young children are likely to be linked, it is important to study their associations. The findings may provide tools for enhancing the rational use of antibiotics.

Materials and methods

Methods

First, we used descriptive methods to study antibiotic purchases for children during the whole study period (2000–2017). We analysed the shares of 0–12-year-old children with antibiotic purchases in one-year age categories and provided a description of the purchased antibiotics by ATC category. For the child’s first 12 years, we calculated and described the average cumulative number of antibiotic purchases based on the duration of HCA.

In addition to providing descriptive evidence, we employed statistical models to study the association between the number of HCA months and antibiotic exposure. Specifically, we examined the total exposure to antibiotics until the age of 12. In adjusted models, we accounted child’s birth order and mother’s age and education at child’s birth. We used Poisson regressions to estimate the following model:

where our response of interest was cumulative purchases of antibiotics between the ages of 0 and 12. Our main explanatory variable was that is the sum of HCA months when the child was 0–6 years (0–72 months) of age. We set the reference category as 13–24 months since this corresponds to the maximum duration of home care for a single-child family. In fully adjusted models, we included a vector for control variables (child’s birth order and mother’s immigrant background, age and education at child’s birth). Since prescription patterns may have evolved over time, we used child’s birth year in all models, both unadjusted and adjusted models. Furthermore, we use separate models to examine the association by the number of siblings at age 12.

We conducted all statistical analyses using R (version 4.3.1).

Results

First, we show descriptive evidence and statistics from the data, then we control for sociodemographic background variables using a Poisson regression model.

Descriptive evidence

Use of antibiotics in all 0–12-year-old children.

Fig 1 shows the number on antibiotic purchases in each of the one-year age categories and covers the entire study period. Each age category shows the cumulative number of antibiotic purchases, as they include all antibiotic purchases made thus far. By the age of 12, less than 5% of children had no outpatient antibiotic purchases and one fifth had more than 15 purchases, while the majority of the children had 1–10 antibiotic purchases.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 1. Share of children with cumulative number of antibiotic purchases from 0 to 12 years (2000–2017).

https://doi.org/10.1371/journal.pone.0335354.g001

The six most commonly prescribed antibiotics for 0–12-year-old children between 2000 and 2017 were amoxicillin, azithromycin, cefalexin, amoxicillin/beta-lactamase inhibitor, combination of sulfadiazine and trimethoprim, and phenoxymethylpenicillin. These pharmaceutical substances cover approximately 94% of the issued antibiotic prescriptions and are typically used to treat respiratory tract infections. Fig 2 shows the 10 most commonly purchased antibiotics by child’s age.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 2. Ten most commonly purchased antibiotics by child’s age in 2000–2017.

https://doi.org/10.1371/journal.pone.0335354.g002

Use of antibiotics in 0–12-year-old children by the duration of HCA.

In Table 1, we show the descriptive sample characteristics by the duration of HCA, examined until the child’s 6th birthday. The average cumulative HCA was 21.4 months (just under 2 years). Thus, as the average parental leave is approximately 10 months, these children entered day care around the age of 2.5 years. The average age of the mother was 29.8 years at the child’s birth. Mothers were younger and lower educated in groups where the average duration of HCA was the highest. The average number of cumulative antibiotic purchases by the age of 12 was 10 purchases. In the groups of 1–12 months and 37 + months of HCA, the average numbers of cumulative purchases by the age of 12 were 10.3 and 8.6, respectively.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 1. Descriptive statistics of cumulative antibiotic purchases and background variables by duration of home care allowance.

https://doi.org/10.1371/journal.pone.0335354.t001

In Fig 3, we show the average cumulative number of antibiotic purchases during the child’s first 12 years by the duration of HCA. The descriptive findings show that the higher the number of months in home care, implying later entry to day care, the lower the number of average cumulative antibiotic purchases. During the entire examined period, the average number of cumulative antibiotic purchases remains lowest in the group of children with the longest home care periods. In children who were in home care for 1–12 months (until approximately 1–2 years of age) or entered day care right after the parental leave period ended (approximately 10 months of age), the cumulative antibiotic use remains at the highest level throughout the follow-up period.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 3. The average cumulative number of antibiotic purchases 0–12 years after birth by the duration of home care allowance.

https://doi.org/10.1371/journal.pone.0335354.g003

In Fig 4, we show the average cumulative number of antibiotic purchases 0–12 years after birth by the duration of HCA (until the child turns 6 years old) and number of siblings. The number of cumulative antibiotic purchases remains lowest in the group of children with the longest HCA duration, regardless of the number of siblings. However, in all HCA duration categories, the average cumulative number of antibiotic purchases was lowest in children with three or more siblings.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 4. The average cumulative number of antibiotic purchases 0–12 years after birth by the duration of home care allowance and number of siblings.

https://doi.org/10.1371/journal.pone.0335354.g004

Results from the Poisson regression model

Next, we show that accounting for the mother’s characteristics or the presence of older siblings using statistical methods does not greatly affect the findings of our descriptive analyses. Table 2 presents findings from the unadjusted model (Model 1) as well as from the model accounting for mothers’ education and age at childbirth and a dummy if the child is the second or subsequent child (Model 2). Furthermore, Table 2 (unadjusted model, Panel A, constant line) shows that between the ages 0–12, children in the sample have 10.7 antibiotic purchases, on average.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 2. Cumulative antibiotic purchases between the ages of 0 and 12 by the duration of home care allowance.

https://doi.org/10.1371/journal.pone.0335354.t002

The results from the Poisson model, expressed in incidence rate ratios (IRRs), further affirm that the cumulative use of antibiotics differs between the HCA duration groups and that the cumulative use is lower in children with longer HCA duration.

Compared to the reference group (HCA for 13–24 months), higher expected number of antibiotic purchases is associated with groups with shorter HCA duration (IRR > 1), and lower cumulative use with groups with longer HCA duration (IRR < 1), regardless of maternal characteristics or the presence of older siblings.

Children with an HCA duration of 37 months or longer had, on average, 1.6 fewer antibiotic purchases than the children in the reference group in the adjusted model.

Lastly, we examine the association between the number of siblings and HCA duration, as older children are often in home care alongside their younger sibling. Table 3 divides the sample by the number of biological siblings at age 12. The results remain consistent with those in Table 2 and Fig 3. Table 3 also shows that children who enter day care earlier, i.e., whose HCA duration was shorter, have more antibiotic purchases between the ages of 0 and 12 when compared to children who entered day care later, i.e., have longer home care periods. Again, accounting for maternal characteristics or the existence of older siblings does not change the direction of the association.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 3. Cumulative antibiotic purchases between the ages of 0 and 12 by number of siblings and duration of home care. Fully adjusted models.

https://doi.org/10.1371/journal.pone.0335354.t003

Supplementary results by gender, region and time period are provided in the S1-S3 Tables. Briefly, boys have more antibiotics purchases than girls (9.96 purchases vs. 9.07 purchases), and children born in cities had larger cumulative purchases when compared children born in rural areas. However, the results on HCA duration remain consistent with Table 2 and Table 3, with a few minor exceptions. Children who reside in rural areas and spend long periods in home care made fewer purchases compared to children living in urban areas with the same duration in home care. There were no large differences in cumulative antibiotic use at age 12 between different birth cohorts (cohorts 2000–2006 in this study) but the connection between long HCA duration and cumulative purchases grew stronger for those in later birth cohorts.

Discussion

The results of our study show that almost all children were exposed to prescribed antibiotics within the first 12 years of their life. One fifth of the children had more than 15 antibiotic purchases by the age of 12. The three most commonly purchased antibiotics for 0–12-year-old children between 2000 and 2017 were amoxicillin, azithromycin, and cefalexin. These medicines are often prescribed for children to treat bacterial ear and chest infections and tonsillitis, for example. It is noteworthy that azithromycin, the second most commonly used antibiotic, belongs to the macrolide class, which is generally not among the most recommended first-line treatment options according to current guidelines. The overuse of broad-spectrum macrolide antibiotics in children has also been observed in previous studies [5,39].

On average, the children had seven antibiotic purchases by the end of the year they turned 6 years and ten antibiotic purchases by age 12. The cumulative number of purchases was lowest in the groups of children whose home care period (before school age) was the longest. The differences between the groups in the average cumulative number of antibiotic purchases largely persisted until the age of 12. These findings suggesting that lower cumulative antibiotic exposure is associated with longer home care duration remained robust even after controlling for maternal characteristics and the number of siblings.

Earlier studies have shown that the risk of respiratory tract infections is higher among children in day care because they are more recurrently exposed to acute respiratory infections than children in home care [21,40]. Moreover, it is known that there is a rapid increase in sick days during the first few months after entering day care [22] and that earlier day care entry is associated with more antibiotic use in childhood and adolescence [23]. However, to our knowledge, this is one of the first register-based study investigating the association between antibiotic purchases and the duration of home care over an extended period of time, i.e., until the children are 12 years old.

Our findings show that the shorter the HCA period, i.e., the earlier the child enters day care, the more antibiotic purchases the child has at the end of the follow-up when she or he is 12 years old compared to the reference group. For example, those with HCA duration of 37 months or longer, had on average, 1.6 fewer antibiotic purchases than those in the reference group. Moreover, most of the differences in antibiotic use between different HCA groups can be attributed to the first 6 years after birth; antibiotic use remains relatively stable between ages 6 and 12 among children with different HCA duration. Though the clinical relevance of 1.6 purchases during 12 years of follow-up can be debated, it may still reflect meaningful variation in exposure to infections and health care use over time. Even relatively small differences in antibiotic use at the population level can be important from a public health perspective, particularly in the context of antimicrobial resistance and stewardship efforts [3].

We found the overall cumulative number of antibiotic purchases is lower among children with a higher number of siblings. The cumulative antibiotic exposure was lowest among children who spent 25 months or longer in home care and who had three or more siblings and highest among those with no siblings and no home care period. In the Poisson regression analyses, we also found that the presence of older siblings is associated with a lower cumulative use of antibiotics. Toivonen et al. (2016) [19] have shown that having older siblings could increase the risk of recurrent infections in small children. However, the severity of infections may vary and not all infections require antibiotic treatment.

The strength of our study is that we used high-quality register data based on a nationally representative sample of children and families covering 80% of all live births in Finland. Our follow-up time was exceptionally long compared to most of the earlier studies on the association of childcare and antibiotics, as we followed the children’s antibiotic purchases from birth to when they were 12 years old. We were also able to apply individual-level information on HCA use to deduce a reverse measure for day care attendance [34,35], as national individual-level information on day care attendance before the 2020s are unavailable in many countries. Finally, many previous studies on day care attendance and antibiotic use have relied on small samples, limited geographical coverage, or questionnaire-based data, rather that large-scale register data.

Despite the considerable strengths, our study also has limitations. Using HCA as a proxy for day care attendance may oversimplify children’s care arrangements and potentially introduce misclassification bias. For instance, some children who do not attend formal day care may receive informal care, which could increase their exposure to respiratory tract infections. However, such informal care arrangements are relatively uncommon in Finland [33], which strengthens the validity of our measure. In addition, families with multiple children are slightly over-represented in our sample. There may also be confounding factors, such as parental behaviour, affecting both day care entry and antibiotic utilisation, that we were not able to account for. During the study period 2000–2017, the outpatient prescribing practices for antibiotics changed, leading to a decrease in the overall prescribing of antibiotics in Finland [6,7]. Though we use child’s birth year in our models, some differences related to the observation time might elude this correction. In addition, the COVID-19 pandemic, known to impact antibiotic prescribing also in Finland [8], is beyond the time frame of our study. This may hinder the generalisability of our findings. Furthermore, our data on antibiotics did not cover all antibiotics used in outpatient care because not all of them were included in the list of reimbursable medicines at the time of the study [6]. This means that our estimate for the total number of antibiotic purchases is likely an underestimation, as it lacks information on non-reimbursed purchases. However, we assume that the range of prescribed and purchased antibiotics was similar in all studied groups, and that the differences in reimbursed antibiotic purchases between the studied groups reflect the differences in all antibiotic purchases.

Future research should examine the total number of antibiotic purchases, including those not reimbursed, with particular emphasis on the types and spectrum of antibiotics prescribed. Special attention should be given to broad-spectrum antibiotics, which are more likely to contribute to antimicrobial resistance and therefore warrant closer monitoring in both research and policy contexts.

In addition to further research, strategies to minimise the spread of infectious diseases in day care settings should be prioritised. Measures such as good hand hygiene, regular cleaning and disinfecting, and proper ventilation, and smaller group sizes have been shown to reduce transmission, especially during and after the COVID-19 pandemic [41,42]. Lowering the incidence of infections and reducing antibiotic use in early childhood care environments could yield both clinical and environmental benefits [43], and support broader public health and antimicrobial stewardship goals.

Conclusions

We show that the average cumulative number of antibiotic purchases from birth to age 12 is lowest among children with the longest duration of home care. The number of siblings does not appear to influence the outcome, and the results remain robust even after controlling for socioeconomic status. Although the differences between groups are moderate, they may still have meaningful implications at the population level, particularly in the context of antimicrobial resistance and public health. These findings, based on large-scale register data, contribute to a broader understanding of how early childhood care arrangements may be associated with long-term patterns of antibiotic use.

Supporting information

Acknowledgments

The authors would like to thank Emilia Norlamo for the language revision.