Secondary Logo

Journal Logo

Original Studies

Childhood Infections and Subsequent School Achievement Among 598,553 Danish Children

Köhler-Forsberg, Ole MD*,†; Sørensen, Holger J. MD, PhD†,‡; Nordentoft, Merete MD, PhD, DrMedSci†,‡; McGrath, John J. MD, PhD§,¶,‖; Benros, Michael E. MD, PhD; Petersen, Liselotte MSc, PhD‡,‖,**

Author Information
The Pediatric Infectious Disease Journal: August 2018 - Volume 37 - Issue 8 - p 731-737
doi: 10.1097/INF.0000000000001869

Abstract

There is a growing body of research linking poorer scholastic achievement (SA) with an increased risk of adverse health and poorer socioeconomic outcomes later in life. For example, those with poorer SA have a slightly increased risk of subsequent mental disorders such as schizophrenia1,2 or depression.3 Educational achievement is influenced by a complex, highly intercorrelated matrix of environmental risk factors on the individual, family and societal levels, which often covary with socioeconomic gradients. These operate against a background of genetic factors—recent genome-wide association studies have provided insights into the polygenic nature of education-related phenotypes.4

Within the risk factors associated with SA,5,6 there has been renewed interest in the possible influence of infectious agents on subsequent brain functioning. Leaving aside the well-described links between specific early life infections and brain functioning (eg, prenatal exposure to rubella or infectious encephalopathies), there is a growing awareness that a wider range of infections may have a more subtle and/or delayed impact on brain function via cytokines and/or inflammatory markers.7 A recent nationwide register-based study among men found a dose–response association between the number of prior severe infections and subsequent cognitive performance in early adult life.8 This evidence strengthens the case for a possible role of infections on subsequent cognition. However, research on the links between childhood infections and subsequent SA has often used hospital-based records and has tended to focus on serologic confirmation of a small number of specific infectious agents.9,10 Furthermore, most studies have been limited by small sample sizes or by only including hospitalizations for severe infections. Hospital-based events may miss many childhood infections, most of which are self-limiting or treated in primary care settings. Finally, the gut microbiome has recently been linked to brain functioning. The bidirectional microbiome-gut-brain axis links the gut and central nervous system activities, and this important axis could be influenced by infections and anti-infective agents.7

We aimed to investigate the association between infections during childhood with later cognitive function as measured by SA, focusing on 2 indicators of infections: (1) hospitalization for infections (a reliable proxy for moderate-to-severe infections) and (2) prescription of anti-infective agents in the primary care sector (a proxy measure for nonsevere infections). We explored the impact on 2 measures of educational achievement: (1) the completion of grade 9 (a dichotomized variable available for all children) and (2) average performance on the final grade 9 school examinations (a continuous variable available for those children who completed grade 9). The analyses were adjusted for factors associated with child educational achievement, such as birthweight,5,11 gestational age,6,11 mental and physical illness12 as well as parental education6 and parental mental illness (as well as for additional factors such as the Apgar score).13

METHODS

Study Population

Using the Danish Civil Registration System, we conducted a nationwide register-based cohort study within the entire Danish population (approximately 5.5 million inhabitants).14 The study was approved by the Danish Data Protection Agency and the Danish National Board of Health. Information on age, gender and parents is provided for each resident in Denmark via the unique personal identification number, enabling complete linkage between different registers. The present study included all individuals born in Denmark between January 1, 1987 and December 31, 1997. Since the majority of individuals in Denmark complete the ninth grade between the age of 15 and 17 years (95% of the children in the present study completed ninth grade between the age of 15.54 and 16.96 years), we excluded individuals who died or emigrated before the age of 15 years.

The Exposure—Assessment of Infections

We included all infections treated in the primary or secondary care sector until completion of ninth grade or the age of 15 among noncompleters. First, we identified all hospitalizations for infections since birth (Table, Supplemental Digital Content 1, http://links.lww.com/INF/C897), utilizing the Danish National Patient Registry that has registered all contacts to somatic hospitals since 1977.15

Second, we identified all prescriptions for anti-infective agents since January 1, 1995 (Table, Supplemental Digital Content 2, http://links.lww.com/INF/C898), through the Danish National Prescription Registry16 that contains detailed information from all Danish pharmacies on each redeemed prescription after January 1, 1995.

Concerning hospitalizations, we used individuals without any hospitalization for infections as the reference group. Regarding anti-infective agents, we expected the majority to having redeemed prescriptions for anti-infective agents.17 Therefore, we chose the control group to include individuals who redeemed ≤2 prescriptions for anti-infective agents. Furthermore, the control group included individuals who redeemed ≤2 prescriptions for the abovementioned anti-infective agents and who redeemed prescriptions for penicillin G (Anatomic Therapeutical Chemical code: J01CE01) and/or antibiotics for topical use (Anatomic Therapeutical Chemical code: D06A), since these agents are not expected to have an impact on the gastrointestinal system. Hence, exposed individuals had redeemed ≥3 prescriptions.

To further investigate the importance of the frequency and severity of infections, we calculated the cumulative measures of (1) number of hospitalizations for infections and (2) number of prescriptions for anti-infective agents.

The Outcome—Assessment of School Performance

In Denmark, children usually start school at 6 to 8 years of age, and the compulsory basic school requirement is 9 years of education.5,18 Children have to either attend a public sector school or obtain education at an equal level in a private school system. Most private schools offer an examination after ninth grade (approximately 98%). Some children who attend public sector schools do not take the examination, mainly due to severe learning disabilities or school difficulties for other reasons.

We used the registry of compulsory school completion assessments and test scores compiled by the Ministry of Education from school reports.19 This Register has a high validity and coverage (96.4% in 2008).19 Danish students in the ninth grade (age 15 or 16) are required to complete a general test of academic achievement. The European Credit Transfer and Accumulation System grading scale is used which is scored on a 7-point scale ranging from −3 (lowest) to 0, 2, 4, 7, 10 and 12 (highest). An officially appointed external examiner reviewed each student in written examinations and coexamined the student in verbal examinations. The outcome was calculated as the grade point average per pupil.20 The main tests cover major domains of academic achievement, including Danish and English, mathematics and physics. These test scores are supplemented by teacher ratings on the same 7-point scale. Reporting of test scores and teacher evaluations is a legal requirement for all schools except special schools for those with learning disabilities.

We included the 2 following outcomes:

  1. completion of ninth grade
  2. average grade at completion of ninth grade.

A priori, we defined individuals who had not completed ninth grade by the age of 17 as noncompleters.

Covariates

Using the Danish Psychiatric Central Research Register,21 we identified whether the children had received a diagnosis with any mental disorder [International Classification of Diseases, 8th edition (ICD-8): 290–315 or ICD-10: F00-99] before completion of ninth grade or the age of 17 years among noncompleters. From the Danish Medical Birth Registry,22 we included birthweight, whether the individual was small for gestational age at birth (yes/no), the Apgar score after 1 and 5 minutes, and whether any malformation was present or suspected at birth.

Chronic somatic diseases that are associated with an increased risk of infections can also lead to more days out of school and thus worse SA. Therefore, we identified whether the children before the completion of ninth grade, or the age of 17 years among noncompleters, had been diagnosed with the following general medical disorders: neoplasm, asthma or any atopic disorder, any autoimmune disease (including type 1 diabetes), and cystic fibrosis (Table, Supplemental Digital Content 3, http://links.lww.com/INF/C899). Furthermore, we identified whether the children were the first-born child or not. Among the parents, we identified the highest educational level23 for both parents and any mental disorder (ICD-8: 290–315; ICD-10: F00-99) since 1969.21

Statistical Analysis

The primary analyses investigated the potential association between a hospitalization for infections and the chance of completing ninth grade, compared with individuals without a hospitalization for infections. We conducted logistic regression analyses adjusted for all the abovementioned covariates including the length of hospitalizations for infections. We show odds ratios (ORs) and 95% confidence intervals (95% CIs).

Secondary analyses explored the association between individuals with ≥3 prescriptions for anti-infective agents compared with those with ≤2 prescriptions regarding the chance of completing ninth grade.

In addition, we explored whether a shorter time since the last hospitalization for infections was associated with a higher risk for noncompletion of ninth grade, and these time-depending analyses were further adjusted for the number of infectious episodes. Furthermore, we investigated whether the number of hospitalizations for infections and amount of anti-infective agents, respectively, was associated with a lower average grade score at completion of the ninth grade. We estimated mean differences in average grade scores according to the number of infections using linear regression and present results as coefficients for the change in average grade score. Dose response between the cumulative measures and the outcomes was tested with the Wald test.

All statistical work was performed in Stata 14 (Stata, College Station, TX).

Sensitivity Analyses

First, we performed all the abovementioned analyses among term-born children, that is, born during or after gestational week 37 weighing ≥2500 g to exclude preterm births as a potential mediating factor in the association between infections and SA. Second, we performed additional analyses where we excluded individuals with a chronic general medical disorder (Table, Supplemental Digital Content 3, http://links.lww.com/INF/C899 which were adjusted for in the main analyses). Third, we performed analyses on each specific birth year to rule out differences in prescribing practice and vaccination recommendations during the study period. Fourth, because information on anti-infective agents is only available since 1995, we performed sensitivity analyses among those born 1987–1994 and those born 1995–1997. Fifth, we explored whether the risks differed depending on the age at the infection. Finally, we performed analyses where we extended follow-up until the age of 17 among noncompleters (instead of until the age of 15).

RESULTS

We identified 598,553 children (301,247 boys and 297,306 girls) born between 1987 and 1997, of which 547,805 (91.5%) completed the ninth grade during our study period. Table 1 depicts the baseline characteristics among completers and noncompleters. Furthermore, Table 1 shows the associations between all covariates with the outcomes noncompletion of ninth grade and average grade scores.

TABLE 1
TABLE 1:
Baseline Characteristics of 598,553 Individuals Born in Denmark 1987–1997, Divided into Completers and Noncompleters of Ninth Grade

Separate Effects of Covariates

A total of 47,035 (7.9%) and 37,406 (6.3%) had a maternal or paternal history of any mental disorder, respectively, and 36,164 children (6.0%) had been diagnosed with any mental disorder, with all these factors being associated with significantly lower odds of completing ninth grade (Table 1). Also, being hospitalized with infections for a longer time, parental educational level, suffering from a chronic general medical disorder, a low Apgar score after 1 and after 5 minutes, having a malformation, being underweight at birth or short for gestational age, and not being a first-born child were all associated with lower odds for completing ninth grade. All the above covariates were adjusted for in the subsequent analysis.

Prior Infections and Completion of Ninth Grade

Individuals with any hospital contact for infections had lower odds of completing ninth grade (OR = 0.82; 95% CI = 0.79–0.85) compared with individuals without any hospitalization for infections, with the odds decreasing in those with longer cumulative time spent in hospital (Table 1). These findings were found in both boys and girls (Fig., Supplemental Digital Content 4, http://links.lww.com/INF/C900). Figure 1 demonstrates that the association with the number of infections and completing ninth grade showed a linear dose–response relationship (P < 0.001). Children with ≥5 hospitalizations for infections had the lowest odds (OR = 0.62; 95% CI = 0.57–0.66) of completing ninth grade. Furthermore, we found that the probability of completing ninth grade was the lowest 1–2 years after the last hospitalization for infections (OR: 0.63; 95% CI = 0.58–0.69), but the probability was still significantly decreased more than 10 years after the last hospitalization of the child (Fig. 2). We found no difference between bacterial, viral and other types of infections with respect to completing ninth grade (Fig., Supplemental Digital Content 5, http://links.lww.com/INF/C901). To investigate whether mental disorders affected the results, we divided the children into those with and without a prior mental disorder. The association between hospitalizations for infections and completion of ninth grade was OR = 0.80 (95% CI = 0.77–0.83) among those without a mental disorder (n = 562,389) and OR = 0.88 (95% CI = 0.81–0.95) among those with a mental disorder (n = 36,164).

FIGURE 1
FIGURE 1:
Association between the number of hospitalizations for infections and the OR* of completing ninth grade among 598,553 children born in Denmark 1987–1997.
FIGURE 2
FIGURE 2:
Association between the time since the last hospitalization for infections and the OR* of completing ninth grade (top) and the mean average grade score (bottom).

Regarding the association between redeemed prescriptions and completion of ninth grade, individuals who redeemed ≥3 prescriptions for anti-infective agents had a slightly higher chance for completing ninth grade (OR = 1.04; 95% CI = 1.01–1.07; P = 0.01) when compared with individuals with ≤2 prescriptions (Table 1). In fact, only in the more extreme end of the distribution (≥30 prescriptions), there was a negative association with lower completion of ninth grade suggesting a possible deleterious effect (Fig. 3). This pattern persisted when prescriptions were grouped into different strata and with different reference groups (ie, when using the reference group with 0 prescriptions or the group with 0–1 prescriptions instead of the main reference group with 0–2 prescriptions; results not shown).

FIGURE 3
FIGURE 3:
Association between the amount of prescriptions for anti-infective agents and the chance of completing ninth grade (top) and the mean average grade score (bottom).

Prior Infections and Average Grade Score

Among the 547,805 children who completed ninth grade, the mean average grade score was 6.44 [standard deviation (SD) = 2.46]. The average score decreased by a coefficient of −0.069 (95% CI = −0.076 to −0.062) for each hospitalization for infections (Fig., Supplemental Digital Content 6, http://links.lww.com/INF/C902), that is, by 2.8% of the SD, and by a coefficient of −0.011 (95% CI = −0.013 to −0.010) for each prescription for anti-infective agents (Fig. 3), that is, by 0.44% of the SD. Both the number of hospitalizations for infections and number of prescriptions for anti-infective agents predicted poorer educational outcomes in a dose–response relationship (P < 0.001).

Furthermore, we found that the average grade score was the lowest among individuals who had been hospitalized within the previous year, but the average grade score was still a little lower when the last hospitalization was more than 10 years ago (Fig. 2).

Sensitivity Analysis

When restricting to term-born children (n = 490,254), the findings were consistent with the main analysis, displaying an OR = 0.75 (95% CI = 0.71–0.79) of completing ninth grade after a hospitalization for infection. The average grade score decreased in a dose–response association with the amount of hospitalizations for infections and prescriptions for anti-infective agents (all P < 0.001).

Second, we performed analyses on individuals without the chronic general medical conditions that were adjusted for in the main analyses (n = 537,864). The results in this subgroup displayed that hospitalizations for infections showed a significantly lower odds of completing ninth grade (OR = 0.84; 95% CI = 0.78–0.91) with the average grade score decreasing depending on the amount of infections (all P < 0.001).

Third, separate analyses on each birth year supported the primary results (results not shown).

Fourth, among children born 1987–1994 (n = 400,325) and 1995–1997 (n = 198,228), hospitalizations for infections showed a significantly lower odds of completing ninth grade [OR = 0.85 (95% CI = 0.81–0.89) and OR = 0.72 (95% CI = 0.68–0.76), respectively] and with the average grade score decreasing depending on the amount of infections (all P < 0.001).

Fifth, Table 2 shows the results depending on age at infection. Concerning hospitalizations for infections, the lowest chance for completion of ninth grade and the largest decrease in mean average grade were observed among those individuals who were hospitalized during the first 12 months of life and among those older than 10 years. Regarding prescriptions for anti-infective agents, the odds for completing ninth grade and mean average grade scores were decreased among those with ≥3 prescriptions in the age-span 5–10 years and 10+ years, whereas those 6 months old until 5 years were associated with significantly increased odds.

TABLE 2
TABLE 2:
Association Between the Age at Infection* and the OR† of Completing Ninth Grade and Average Grade Score

Finally, we included follow-up until the age of 17 among noncompleters, and these analyses did not change the results (not shown).

DISCUSSION

To our knowledge, this study is the largest to date investigating infections as a predictor of scholastic outcome in basic school. In this nationwide study of 598,553 Danish children, we found that the number of hospitalizations for infections was significantly associated with a reduced probability of completing ninth grade. We additionally found a moderate association between the number of hospitalizations for infections with lower average school marks and a weaker association between a proxy measure of community-based treatment of nonsevere infections and average school marks. Also, a lower probability of completing ninth grade was noted relative to shorter time since last hospitalization for infection. Concerning age at infection, we found that particularly hospitalizations for infections during the first 12 months of life and after the age of 10 years were associated with the lowest odds of completing ninth grade and the lowest mean average grade scores. Our findings may indicate direct effects of infections on the developing brain but might also be caused by genetic and/or socioeconomic factors leading to a higher rate of infections and worse scholastic outcomes.

We confirmed associations of SA with birthweight,5,11 gestational age,6,11 mental or physical illness12 as well with parental education6 and parental mental illness,13 and we additionally found associations with Apgar score at birth and 5 minutes after birth; however, all these important confounders were adjusted for in the analysis.

Our findings extend our understanding regarding the association between particularly severe infections during childhood and adolescence and cognitive achievement. A previous study found a dose–response association between severe infections and subsequent cognitive performance in early adult life among men.8 Hence, the present results suggest a potential role of infections for later cognitive abilities; however, future studies need to investigate potential mechanisms.

This study is the first large scale study to investigate a proxy measure (ie, prescriptions) for community-based treatment of nonsevere infections and subsequent SA. We found a small protective association between nonsevere infections and completion of ninth grade (OR = 1.04). Furthermore, we found a weak association between the number of nonsevere infections and lower school marks among children with many prescriptions (≥30), but the majority of children had ≤10 redeemed prescriptions for anti-infective agents.

Strengths and Limitations

The major strengths of the present study are the large size, the population-based design and the almost complete follow-up. The Danish Civil Registration System and The Danish Educational Attainment Registry made it possible to obtain almost complete basic school educational status. To account for confounding, we were able to adjust for important predictors, including parental education, parental psychiatric morbidity and individuals’ own registration with either psychiatric diagnosis or somatic diagnosis.

The study has limitations: SA is a broad outcome that is influenced by a complex, highly intercorrelated matrix of environmental factors that act on the individual, family and societal levels. Further, residual confounding by unmeasured factors could influence the findings. For example, we could not examine if family-related factors (other than education) might explain the lower SA in children that are exposed to severe infections (ie, the possibility that lower SA could in some cases be family-related).9 Other possible factors that are associated with infections and SA, such as genetics, could not be assessed. Other limitations include that individuals with some disorders (such as intellectual disability) were underrepresented in the analyses concerning mean grade scores because these children are more likely to attend special schools5 and that we were not able to assess vaccination status. Also, actual brain functioning was unmeasured in our study. This implies that the association between childhood infections and SA might have been partly due to more days out of school. For instance, we found a differential and stronger effect depending on a shorter time since the last infection on SA and a stronger impact of hospitalizations among those 10 years old or older, which could have been mediated via more days being absent from school close to the examination. However, the effect of the time since the last infection was still significant after more than 10 years after the last hospitalization, and hospitalizations during the first year of life was also associated with lower odds of completing ninth grade. In addition, we do not know the number of children who did not complete ninth grade due to severe learning disabilities or school difficulties for other reasons, which may impact on the generalizability of our findings. Furthermore, we had no serologic confirmation of the infectious agents. Finally, the present study was not able to include infections that were not treated with anti-infective agents and did not result in a hospitalization. Hence, the actual number of infections within the present study population may be higher, and the control group may have suffered other infections which were not treated with anti-infective agents.

REFERENCES

1. Kendler KS, Ohlsson H, Mezuk B, et al. A Swedish National Prospective and Co-relative Study of School Achievement at age 16, and risk for schizophrenia, other nonaffective psychosis, and bipolar illness. Schizophr Bull. 2016;42:77–86.
2. MacCabe JH, Lambe MP, Cnattingius S, et al. Scholastic achievement at age 16 and risk of schizophrenia and other psychoses: a national cohort study. Psychol Med. 2008;38:1133–1140.
3. Lorant V, Deliège D, Eaton W, et al. Socioeconomic inequalities in depression: a meta-analysis. Am J Epidemiol. 2003;157:98–112.
4. Okbay A, Beauchamp JP, Fontana MA, et al; LifeLines Cohort Study. Genome-wide association study identifies 74 loci associated with educational attainment. Nature. 2016;533:539–542.
5. Christensen K, Petersen I, Skytthe A, et al. Comparison of academic performance of twins and singletons in adolescence: follow-up study. BMJ. 2006;333:1095.
6. Mathiasen R, Hansen BM, Andersen AM, et al. Gestational age and basic school achievements: a national follow-up study in Denmark. Pediatrics. 2010;126:e1553–e1561.
7. Wang Y, Kasper LH. The role of microbiome in central nervous system disorders. Brain Behav Immun. 2014;38:1–12.
8. Benros ME, Sørensen HJ, Nielsen PR, et al. The association between infections and general cognitive ability in young men—a nationwide study. PLoS One. 2015;10:e0124005.
9. Roed C, Omland LH, Skinhoj P, et al. Educational achievement and economic self-sufficiency in adults after childhood bacterial meningitis. JAMA. 2013;309:1714–1721.
10. Ellman LM, Yolken RH, Buka SL, et al. Cognitive functioning prior to the onset of psychosis: the role of fetal exposure to serologically determined influenza infection. Biol Psychiatry. 2009;65:1040–1047.
11. Kirkegaard I, Obel C, Hedegaard M, et al. Gestational age and birth weight in relation to school performance of 10-year-old children: a follow-up study of children born after 32 completed weeks. Pediatrics. 2006;118:1600–1606.
12. Martinez YJ, Ercikan K. Chronic illnesses in Canadian children: what is the effect of illness on academic achievement, and anxiety and emotional disorders? Child Care Health Dev. 2009;35:391–401.
13. Shen H, Magnusson C, Rai D, et al. Associations of parental depression with child school performance at age 16 years in Sweden. JAMA Psychiatry. 2016;73:239–246.
14. Pedersen CB. The Danish civil registration system. Scand J Public Health. 2011;39(7 suppl):22–25.
15. Lynge E, Sandegaard JL, Rebolj M. The Danish National Patient Register. Scand J Public Health. 2011;39(7 suppl):30–33.
16. Kildemoes HW, Sørensen HT, Hallas J. The Danish National Prescription Registry. Scand J Public Health. 2011;39(7 suppl):38–41.
17. Pottegård A, Broe A, Aabenhus R, et al. Use of antibiotics in children: a Danish nationwide drug utilization study. Pediatr Infect Dis J. 2015;34:e16–e22.
18. Ministry of Education. Welcome to the Danish Folkeskole. 2010. Available at: http://eng.uvm.dk/primary-and-lower-secondary-education/the-folkeskole/about-the-folkeskole. Accessed January 10, 2018.
19. Shewbridge C, Jang E, Matthews P, et al. OECD Reviews of Evaluation and Assessment in Education: Denmark: OECD; 2011. 2011.
20. Moberg JY, Magyari M, Koch-Henriksen N, et al. Educational achievements of children of parents with multiple sclerosis: a nationwide register-based cohort study. J Neurol. 2016;263:2229–2237.
21. Mors O, Perto GP, Mortensen PB. The Danish Psychiatric Central Research Register. Scand J Public Health. 2011;39(7 suppl):54–57.
22. Knudsen LB, Olsen J. The Danish Medical Birth Registry. Dan Med Bull. 1998;45:320–323.
23. Jensen VM, Rasmussen AW. Danish Education Registers. Scand J Public Health. 2011;39(7 suppl):91–94.
Keywords:

infections; cognition; scholastic achievement; epidemiology

Supplemental Digital Content

Copyright © 2018 Wolters Kluwer Health, Inc. All rights reserved.