Skip to main content
PMC home page
Journal of Antimicrobial Chemotherapy logoLink to Journal of Antimicrobial Chemotherapy
. 2016 Jan 17;71(4):1098–1105. doi: 10.1093/jac/dkv455

Effect of long-term antibiotic use on weight in adolescents with acne

Despina G Contopoulos-Ioannidis 1,2,*, Catherine Ley 3, Wei Wang 2, Ting Ma 3, Clifford Olson 2, Xiaoli Shi 2, Harold S Luft 2, Trevor Hastie 4, Julie Parsonnet 3,5
PMCID: PMC4790625  PMID: 26782773

Abstract

Objectives

Antibiotics increase weight in farm animals and may cause weight gain in humans. We used electronic health records from a large primary care organization to determine the effect of antibiotics on weight and BMI in healthy adolescents with acne.

Methods

We performed a retrospective cohort study of adolescents with acne prescribed ≥4 weeks of oral antibiotics with weight measurements within 18 months pre-antibiotics and 12 months post-antibiotics. We compared within-individual changes in weight-for-age Z-scores (WAZs) and BMI-for-age Z-scores (BMIZs). We used: (i) paired t-tests to analyse changes between the last pre-antibiotics versus the first post-antibiotic measurements; (ii) piecewise-constant-mixed models to capture changes between mean measurements pre- versus post-antibiotics; (iii) piecewise-linear-mixed models to capture changes in trajectory slopes pre- versus post-antibiotics; and (iv) χ2 tests to compare proportions of adolescents with ≥0.2 Z-scores WAZ or BMIZ increase or decrease.

Results

Our cohort included 1012 adolescents with WAZs; 542 also had BMIZs. WAZs decreased post-antibiotics in all analyses [change between last WAZ pre-antibiotics versus first WAZ post-antibiotics = −0.041 Z-scores (P < 0.001); change between mean WAZ pre- versus post-antibiotics = −0.050 Z-scores (P < 0.001); change in WAZ trajectory slopes pre- versus post-antibiotics = −0.025 Z-scores/6 months (P = 0.002)]. More adolescents had a WAZ decrease post-antibiotics ≥0.2 Z-scores than an increase (26% versus 18%; P < 0.001). Trends were similar, though not statistically significant, for BMIZ changes.

Conclusions

Contrary to original expectations, long-term antibiotic use in healthy adolescents with acne was not associated with weight gain. This finding, which was consistent across all analyses, does not support a weight-promoting effect of antibiotics in adolescents.

Introduction

Overweight and obesity in children have become an epidemic in the USA.1 Their rapid increase in prevalence over the past three decades suggests an environmental cause. One hypothesis is that the obesity increase is due to changes in the microbiome.2–9 This has been postulated to be due, in part, to the widespread use of antimicrobial agents, particularly antibiotics.

Since the 1940s, antibiotics and, more recently, probiotics have been used to promote weight gain in farm animals.10–16 Whether antibiotics similarly affect humans, however, has not been well studied. A randomized trial in 1955 of 310 healthy US army recruits showed that a 7 week treatment with antibiotics (chlortetracycline or penicillin) resulted in greater weight gain than in the placebo-control group.17 Several small case series in infants in the 1950s had conflicting results and together were inconclusive regarding the use of antibiotics as a weight gaining intervention for children.18 More recently, two large European paediatric cohort studies19,20 and a large international cohort study21 claimed that early life antibiotic exposure was associated with being overweight later in childhood. Antibiotic exposure during pregnancy was also recently associated with overweight risk later in childhood.22 Randomized trials of probiotics with or without prebiotics in infant formulas have shown weight gain in some, but not all, studies.23–30

In the medical care setting, antibiotics are used primarily in clinically unstable patients, so an observed increase in weight may be skewed by improvement in the underlying disease process; this bias may be true particularly for studies of malnourished children in low income countries.31 Prolonged antibiotics, however, are used in the treatment of acne in otherwise healthy adolescents. Using electronic health record data from patients seen in a large primary healthcare organization, we sought to assess the effect on body mass of long-term antibiotic use in adolescents with acne. Our primary design focused on a cohort of patients with acne receiving prolonged oral antibiotics. To validate our results further, we also performed a case–control study of patients with acne who did and did not receive prolonged antibiotics.

Methods

Subjects

We selected our population of adolescents from individuals seen between 2001 and 2010 at the Palo Alto Medical Foundation (PAMF), an outpatient primary care organization now serving ∼950 000 patients in northern California. Inclusion criteria for the primary cohort were: (i) diagnosis of acne (ICD-9 code 706.1); (ii) filled prescription for prolonged oral antibiotics (≥4 weeks); (iii) age 12–20 years at start of prolonged antibiotics; and (iv) ≥1 weight recorded in the 18 month period prior to the start of antibiotics (pre-antibiotics) and ≥1 weight recorded in the 12 month period after the end of antibiotics (post-antibiotics). Sequential prescriptions of the same antibiotic with gaps <60 days were considered a single course; prescriptions with gaps ≥60 days were considered separate courses and only the first prolonged antibiotic course was considered.

Exclusion criteria included: (i) any current or prior diagnosis of chronic immunosuppressive, inflammatory or autoimmune diseases, cystic fibrosis or endocrine diseases associated either with weight gain or failure to thrive; (ii) medications known to be associated with weight change (e.g. neuropsychiatric medications); and (iii) immunosuppressive medications.

Data extraction

All information was extracted from the electronic health record and the pharmacy-fill databases. Missing race and ethnicity information was imputed using methodology previously described.32 All somatometric measurements were reviewed by independent investigators (J. P., C. L. and D. G C.-I.) blinded to the timing of measurements relative to the prolonged antibiotic course. Physiologically implausible outlier values were excluded (<1% of measurements). For the calculation of weight-for-age Z-score (WAZ) and BMI-for-age Z-score (BMIZ), we used the SAS program from CDC (Supplementary Methods 1, available as Supplementary data at JAC Online).33 The STROBE guidelines for reporting of observational studies were followed.34

Analyses

Cohort study

As adolescents are expected to have physiological weight gain as part of their normal growth, we used as our study endpoints changes in standardized WAZ and BMIZ measures rather than absolute changes in weight and BMI. Our study endpoints were the within-individual change in WAZ and BMIZ in the period pre-antibiotics compared with the period post-antibiotics. With this design, each adolescent served as his/her own control, eliminating between-individual variability in other factors that could affect weight. Changes in WAZs and BMIZs were studied in three ways: (i) last measurement pre-antibiotics versus first measurement post-antibiotics was compared by paired t-test (primary endpoint); (ii) means of measurements pre-antibiotics versus post-antibiotics were compared by piecewise-constant-mixed models; and (iii) WAZ and BMIZ trajectory slopes pre-antibiotics versus post-antibiotics (Z-score changes/6 month interval) were compared by piecewise-linear-mixed models (Supplementary Methods 2). A positive value in these pre-post analyses indicates a weight gain post-antibiotics compared with pre-antibiotics. We also compared the proportion of adolescents with either WAZ or BMIZ increases or decreases ≥0.2 Z-scores (between last measurement pre-antibiotics versus first post-antibiotics) by χ2 test. Models are described in detail in Supplementary Methods 2. All endpoints were pre-specified.

We measured the within-individual baseline longitudinal variability of the WAZs in the pre-antibiotic period for adolescents who had ≥3 WAZs within this period by using the (i) slope of the regression line fit to at least three serial WAZs, and (ii) root-mean-square-error (RMSE) associated with the regression line (WAZ-RMSE). Within-individual variability exists if (i) baseline WAZ slope is non-zero, indicating deviation of the growth trajectory over time from the original Z-score trajectory, or (ii) WAZ-RMSE is non-zero. Variability was considered high if (i) absolute value of the baseline WAZ trajectory slope was ≥0.2 Z-score change over 1 year, and/or (ii) baseline WAZ-RMSE was ≥0.25 (Figure S1).

To assess the consistency of our results we also examined results for the following subgroups: (i) adolescents with low baseline WAZ variability pre-antibiotics (considered post hoc); (ii) adolescents with last WAZ (or BMIZ) pre-antibiotics close to zero (i.e. between −0.5 and 0.5 Z-scores), to capture those with standardized measurements near the 50th percentile of the population growth curve (i.e. 0 Z-score); (iii) adolescents prescribed only tetracyclines; and (iv) adolescents with WAZ measurements recorded within 60 days pre- and post-antibiotics (considered post hoc).

We decided a priori to adjust the piecewise-constant model and piecewise-linear-mixed model analyses for the following covariates: (i) age at start of antibiotic course; (ii) sex; (iii) race; (iv) ethnicity; (v) antibiotic type (tetracyclines versus other); and (vi) antibiotic course duration. For the subgroup of adolescents who had their baseline WAZ variability assessed (i.e. with ≥3 baseline WAZs), we included an additional covariate for the size of the within-individual baseline WAZ variability (low versus high, see criteria above).

Case–control study

To validate our results, we performed a case–control study comparing patients with acne on prolonged antibiotics (cases) with similar patients with acne on topical acne treatments only (controls). Controls were adolescents with acne who received topical acne therapies after their acne diagnosis; they were 12–20 years of age at the start of the first topical treatment and had ≥1 weight recorded in the 18 months before the start and ≥1 weight in the 12 months after the end of a dummy period, a period of the same duration as the prolonged antibiotic course of their matching case. The start date of the dummy period was the date of the control's first topical acne treatment. Controls were matched 1 : 1 to cohort–cases by age (birth year), gender, race and ethnicity. For each cohort–case we selected one control with the largest number of weight measurements. We calculated the mean change pre-post antibiotics in cases and pre-post dummy period in controls (last WAZ pre-antibiotics or pre-dummy period versus first WAZ post-antibiotics or post-dummy period in cases and controls, respectively) and compared the between-groups mean changes by paired t-test. Significance was considered at P < 0.05. In further exploratory analysis, we also matched each age of the control at the start of topical treatment to their matching age of the case at the start of antibiotics.

Analyses were performed in STATA/SE 12 (StataCorp LP, College Station, TX, USA) and/or SAS 9.3 (Cary, NC, USA). For the case–control matching we used the gmatch SAS macro.35 Graphs were generated in STATA/SE 12 and R.36

Power calculations

Our cohort had >93% power to detect a WAZ or BMIZ change of 0.15 Z-scores between the last measurement pre-antibiotics and first post-antibiotics with α = 0.05.

Ethics

Approval was given by the Institutional Review Board of the PAMF Research Institute (no. 11–02-054).

Results

Cohort study

Of the 9817 children and adolescents followed at the PAMF who received long-term antibiotics, 1012 eligible adolescents (Figure S2) with 5755 WAZs were analysed. Among these 1012 adolescents, 542 (54%) also had height measurements, yielding 1712 BMIZs. The median age at the start of the antibiotic course was 16 years. Subjects were primarily of white race (74%) and non-Hispanic ethnicity (87%); the majority of prescriptions were for tetracyclines (82%) and the median antibiotic course duration was 60 days (Table 1). WAZ measurements were obtained a median of 3.9 months pre-antibiotics and 3.7 months post-antibiotics; BMIZ measurements were obtained a median of 7.0 months pre-antibiotics and 5.5 months post-antibiotics. Baseline WAZ longitudinal variability could be assessed in 42% (426 of 1012) of adolescents; high baseline variability was found in 22% of these (Table S1).

Table 1.

Cohort study subject characteristics

Characteristic Category No. of adolescents (%) or median values (IQR; range) (n = 1012)
Agea (years) 16 (15–17; 12–20)
Gender male 519 (51%)
female 493 (49%)
Raceb white 745 (74%)
Asian 166 (16%)
African American 11 (1%)
other 15 (1%)
Ethnicityb non-Hispanic/Latino 880 (87%)
Hispanic/Latino 128 (13%)
Type of antibiotic tetracyclines 831 (82%)
penicillins 66 (7%)
macrolides 56 (6%)
cephalosporins 44 (4%)
other 18 (1%)
Antibiotic course duration of prolonged antibiotic course (days) 60 (30–99.5; 9–744c)
number of adolescents with duration of antibiotic course <180 days 885 (88%)
WAZ number of adolescents with ≥1 WAZ pre-antibiotics and ≥1 WAZ post-antibiotics 1012
number of adolescents with ≥2 WAZs pre-antibiotics 671 (66%)
number of adolescents with ≥2 WAZs post-antibiotics 545 (54%)
number of adolescents with ≥2 WAZs, each pre-antibiotics and post-antibiotics 384 (38%)
number of WAZs pre-antibiotics 2 (1–4; 1–13)
number of WAZs post-antibiotics 2 (1–3; 1–15)
mean WAZ pre-antibioticsd +0.50 (−0.018 to 1.086; −2.22 to +2.50)
trajectory slopee of WAZs pre-antibiotics −0.06 (95% CI: −0.016 to +0.003)
BMIZ number of adolescents with ≥1 BMIZ pre-antibiotics and ≥1 BMIZ post-antibiotics 542
number of adolescents with ≥2 BMIZs pre-antibiotics 170 (31%)
number of adolescents with ≥2 BMIZs post-antibiotics 124 (23%)
number of adolescents with ≥2 BMIZs, each pre-antibiotics and post-antibiotics 53 (10%)
number of BMIZs pre-antibiotics 1 (1–2; 1–10)
number of BMIZs post-antibiotics 1 (1–1; 1–8)
mean BMIZ pre-antibiotics +0.32 (−0.19 to +0.85; −2.99 to +2.36)
trajectory slopee of BMIZs pre-antibiotics +0.005 (95% CI: −0.017 to +0.026)
Time of measurements (days) time of last WAZ pre-antibiotics to antibiotic start date 118 (41–252; 1–540)
time from antibiotics end date to first WAZ post-antibiotics 111 (46–203; 1–363)
time from last BMIZ pre-antibiotics to antibiotics start date 209 (90–352; 1–537)
time from antibiotics end date to first BMIZ post-antibiotics 165 (81–266; 1–365)
Pre-antibiotics WAZs number of adolescents with last WAZ pre-antibiotics close to zerof 379/1012 (37%)
Pre-antibiotics BMIZs number of adolescents with last BMIZ pre-antibiotics close to zerof 248/542 (46%)
Baseline WAZ variabilityg number of adolescents with baseline measurement variability assessed 426
number of adolescents with high baseline measurement variability 92 (22%)
Open in a new tab

aAge at start of prolonged oral antibiotics.

b34% and 43% of individuals had missing/not reported racial or ethnicity information, respectively. Using previously validated methods for imputation of missing race/ethnicity data we imputed the race and ethnicity for 78% and 99% of missing values, respectively.

cDuration of antibiotics represents the days of antibiotics only (without considering any gaps between sequential prescriptions). Antibiotic courses ≥365 days were found for only 2.5% (25 of 1012) of adolescents.

dMean WAZs (Z-scores) in the 18 months pre-antibiotics.

eRegression slopes of measurements (WAZ or BMIZ, respectively) (Z-score changes/6 month interval) in the 18 months pre-antibiotics (by piecewise-linear-mixed model; Table S3B).

fSubgroup of adolescents with WAZ (or BMIZ) close to zero pertains to adolescents with last WAZ pre-antibiotics (or BMIZ, respectively) close to zero (close to the 50th percentile of the population growth curve) (close to zero = between −0.5 and +0.5 Z-scores).

gHigh baseline WAZ variability was defined as WAZ slope ≥0.2 Z-score change over 12 months and/or WAZ-RMSE ≥0.25 (baseline variability was assessed in those adolescents with ≥3 WAZs in the baseline period pre-antibiotics; to allow for meaningful calculation of slope and RMSE).

WAZs

The change between the last WAZ pre-antibiotics and the first WAZ post-antibiotics was −0.041 Z-scores (P < 0.001; paired t-test) (Figure 1a and Table S2). Subgroup analyses, restricted to adolescents with (i) low baseline WAZ variability (n = 334), (ii) baseline WAZ pre-antibiotics close to zero (n = 379), (iii) prescribed tetracyclines (n = 831), and (iv) measurements within 60 days pre- and post-antibiotics, showed decreases of −0.031 Z-scores (P = 0.017), −0.024 Z-scores (P = 0.134), −0.051 Z-scores (P < 0.001) and −0.024 Z-scores (P = 0.211), respectively. In all subgroup analyses, even when we considered the uncertainty around the estimated effect sizes, the upper 95% CIs excluded a weight gain effect larger than +0.014 Z-scores (Figure 1a and Table S2).

Figure 1.

Figure 1.

Open in a new tab

(a) Changes between last WAZ pre-antibiotics and first WAZ post-antibiotics (paired t-test). Shown are the point estimates of the changes and 95% CIs thereof. ‘Whole cohort’ = primary cohort; ‘Low variability’ = subgroup of 334 adolescents with low baseline measurement variability; ‘Close to zero’ = subgroup of 379 adolescents with last WAZ pre-antibiotics close to zero; ‘Tetracyclines’ = subgroup of 831 adolescents prescribed tetracyclines; ‘Within 60 days’ = subgroup of 123 adolescents with WAZs within 60 days pre-antibiotics and within 60 days post-antibiotics. Negative values indicate that values post-antibiotics are lower than values pre-antibiotics. (b) Changes between mean of WAZs pre-antibiotics and mean of WAZs post-antibiotics (piecewise-constant-mixed model). Shown are the point estimates of the changes and 95% CIs thereof. ‘Whole cohort’ = primary cohort; ‘Low variability’ = subgroup of 334 adolescents with low baseline measurement variability; ‘Close to zero’ = subgroup of 379 adolescents with last WAZ pre-antibiotics close to zero; ‘Tetracyclines’ = subgroup of 831 adolescents prescribed tetracyclines; ‘Within 60 days’ = subgroup of 123 adolescents with WAZs within 60 days pre- and within 60 days post-antibiotics. Negative values indicate that values post-antibiotics are lower than values pre-antibiotics.

In our secondary analyses, piecewise-constant-mixed modelling showed a decrease in the mean WAZ of −0.050 Z-scores post-antibiotics (P < 0.001) (Figure 1b and Table S2). A statistically significant decrease post-antibiotics, was also seen in three of four subgroup analyses (Table S2). Additionally, the trajectory slopes pre- versus post-antibiotics using a piecewise-linear-mixed model showed a decrease post-antibiotics of −0.025 Z-scores/6 month interval (P < 0.001) (Figure 2 and Table S2). The decrease was statistically significant in two of four subgroups (Table S2). In the primary cohort, 26% (267 of 1012) of the adolescents had a clinically significant WAZ decrease post-antibiotics ≥0.2 Z-scores; in contrast, 18% (180 of 1012) had an increase ≥0.2 Z-scores (P < 0.001) (Table S2). Results were similar for adolescents with: (i) low baseline measurement variability; (ii) last WAZs pre-antibiotics close to zero; and (iii) prescribed tetracyclines (Table S2). Adjusted analyses using piecewise-constant-mixed model and piecewise-linear-mixed model gave results similar to unadjusted analyses (Table S3).

Figure 2.

Figure 2.

Open in a new tab

Change in WAZ trajectory slopes pre-antibiotics and post-antibiotics (piecewise-linear-mixed model). Open circles = WAZ measurements for the 1012 adolescents (grey = for WAZs during pre- or post-antibiotics period; green = for WAZs during the antibiotic course period); the x-axis shows the time each WAZ measurement was obtained relative to a time (time 0) in the middle of the antibiotic course, for each individual; red line = WAZ trajectory slope pre-antibiotics; blue line = WAZ trajectory slope post-antibiotics; broken red line = reference line/extension of the WAZ trajectory slope pre-antibiotics; the vertical broken line corresponds to time 0 in the middle of the antibiotic course.

BMIZs

In the subset of 54% of adolescents with height measurements, analyses showed a decrease in BMIZ post-antibiotics similar to the decrease in WAZ post-antibiotics; results, however, were only statistically significant in the subgroup of adolescents receiving tetracyclines (Table S2 and Table S3).

Case–control study

A total of 497 case–control pairs were analysed (Table S2, Table S4 and Figure S3). The between-group difference of the mean change pre-post antibiotics in cases and pre-post dummy period in controls was −0.042 Z-scores (P = 0.027). Further exploratory analysis of 151 case–control pairs, with additional matching of age of the control at start of topical treatment to their matching age of the case at start of antibiotics yielded a between-group difference of mean changes of 0.001 Z-scores (P = 0.97) (Table S2).

Discussion

To our knowledge, our large cohort study is the first to evaluate the effect of antibiotics on weight in generally healthy adolescents. Contrary to our initial literature-based hypothesis, a variety of tests failed to demonstrate any weight gain effect associated with prolonged antibiotic use. Instead, antibiotics were associated with a statistically, although not clinically, significant decrease in WAZs. The results were robust and consistent in all primary and secondary analyses and in the validation case–control study. Parallel analyses for BMIZs in the cohort showed consistent decreases in BMIZ post-antibiotics, but results were statistically significant only in adolescents who received tetracyclines.

Our cohort study had the statistical power to detect small positive or negative changes in WAZs. There are plausible explanations for a decrease in weight post-antibiotic therapy, however, including decreased appetite while on antibiotics due to gastrointestinal distress, catabolic effects of tetracyclines37 and/or alteration of gut microbiota towards non-obesogenic bacteria.4

Our results contrast with observations in animals where weight gain post-antibiotics has been described.10–16 The clinical significance of the growth promoting effect of in-feed antibiotics in animals is under debate, with the true effect likely smaller than originally shown.38 Yet, obesity in mice has been associated with alterations in gut microbiota, with obesogenic gut microbiota harvesting energy from the diet more efficiently.2,4,7,39–41 Several large-scale sequencing studies in humans showed gut microbiota differences in obese subjects,3,5,42 in obese individuals losing weight43–46 and in individuals with severe malnutrition.47,48 These studies support the idea that, by influencing colonizing flora, antibiotics could alter weight.49,50 We were unable, however, to corroborate these findings in our cohort of healthy adolescents.

To our knowledge, the only previous trial17 to explore the antibiotic–weight gain effect in healthy individuals randomized 310 US army recruits to 7 weeks of chlortetracycline, penicillin or placebo and identified a statistically, but not clinically, significant weight gain with antibiotics versus placebo (3.4% and 2.9% body-weight increase in the chlortetracycline and penicillin group versus 2.0% in the placebo group; P < 0.05). Another randomized trial in children showing the positive effect of antibiotics on weight studied malnourished children and any antibiotic–weight gain effect could be due to treatment of underlying infections.47,48

Our findings conflict with those of prior paediatric cohort studies claiming an association of early life antibiotic exposure with body mass later in childhood (Table S5). These studies, however, varied considerably on: (i) exposure periods probed; (ii) timepoints when body mass changes were analysed; (iii) body mass outcomes analysed; (iv) subgroups considered; and (v) adjustments made in their analyses. Ajslev et al.19 showed that antibiotic exposure in the first 6 months of life was associated with increased overweight risk at 7 years, but only in unadjusted analyses and only in children of normal weight mothers. Trasande et al.20 found that early life antibiotic exposure in the first 6 months, but not in other early life periods, was associated with increased weight-for-length Z-scores at 10 and 10–20 months and increased BMIZs at 38 months but not at 7 years. Only exposure during 15–23 months was associated with increased BMIZ score at 7 years. Murphy et al.21 claimed that antibiotic exposure in the first 12 months was associated with increased BMI at 5–8 years, but this appeared only in boys and not in girls. Azad et al.51 also showed that antibiotic exposure during the first 12 months was associated with increased overweight risk at 9 and 12 years and increased risk of high-central adiposity at 12 years, but the risk with specific early life exposure periods varied significantly. Bailey et al.52 claimed that cumulative exposure only to more than four antibiotic courses and exposure only during the first 0–5 and 6–11 months (to broad-spectrum antibiotics only), was associated with increased risk of obesity between 2 and 5 years. Recently, Mueller et al.22 also showed that antibiotic exposure during pregnancy was associated with increased child's overweight risk at 7 years. Harmonized testing and replication of analyses across different research teams would be helpful in clarifying reported findings.

We acknowledge, though, that our study was not a replication of the prior early life antibiotic exposure studies. Timing of antibiotic administration may be important for body-habitus development, with adolescents less susceptible to an antibiotic–weight gain effect relative to younger children. Cox et al.53,54 proposed the theory that different dosage, timing and type of antibiotic regimen may have qualitatively different effects on post-antibiotic treatment growth. Perturbations to gut microbiome during critical windows early in life, when the microbiome is still in its formative stages, may have long-term metabolic consequences; the effects of antibiotics after gut microbiota have matured may be markedly different.55

Some additional limitations of our study should be acknowledged. All weight and height measures were taken in the setting of routine clinical care. Weight can also vary from day to day, potentially dwarfing any real, but small, increase post-antibiotics. However, inaccurate weight measurement is unlikely to result in a consistent bias towards a negative effect of antibiotics on weight. Large within-individual baseline variability in weight was observed in some subjects, possibly masking an antibiotic effect smaller than +0.2 Z-scores. However, subgroup analyses restricted to adolescents with low baseline variability failed to alter our findings. The timing of weight measurements relative to antibiotic use was not consistent; different results might have been observed had weight measurements been routinely obtained closer to the prolonged antibiotic course. We did not detect weight gain effect, however, among adolescents with measurements taken only within 2 months pre- and post-antibiotics. Subgroup analyses restricted to adolescents prescribed tetracyclines did not show an antibiotic–weight gain effect. Although tetracyclines can cause gastrointestinal distress and decreased appetite, they have been previously associated with weight gain in humans, farm animals and experimental animals.37,38,56 As part of our study design, we did not have available data on any antibiotic-associated side effects or dietary habits of adolescents; so the impact of these factors on the studied effect could not be explored. Moreover, long-term doxycycline treatment has been shown to be associated with changes in the gut microbiome.56 However, as part of our retrospective cohort study design, we were not able to explore such an impact of antibiotic use on the gut microbiome of these adolescents with acne. It is also possible that the effect of antibiotics on weight might have been different in adolescents who were prescribed more than one prolonged antibiotic course. We studied the effect on weight immediately after antibiotic use and up to 1 year post-antibiotics, but not more long-term effects. It is also possible that adolescents particularly conscious about their appearance, hence seeking treatment for acne, might also be seeking to control their weight. Nevertheless, in the case–control study, the weight trajectory slopes pre-antibiotics (for cases) or pre-dummy period (for controls) were both close to zero. Further exploratory analysis with additional case–control matching of age at first topical treatments to age at start of prolonged antibiotics, also failed to detect any weight gain effect.

In conclusion, contrary to our original expectations and prior relevant literature, long-term antibiotic use in healthy adolescents with acne was not associated with weight gain. Our findings were robust and consistent across all analyses and did not support a growth promoting effect of antibiotics in healthy adolescents. Nevertheless, generalizability of our study findings to antibiotic exposures in other patient groups or time periods outside adolescence should be cautious. The antibiotic–weight gain hypothesis should be further pursued in additional large-scale human studies. Standardization across research teams of the probed antibiotic exposure periods, outcomes of interest, timepoints for outcome ascertainment and statistical analysis plan is needed to provide evidence as to whether changes in gut microbiota could translate to clinically significant changes in human weight.

Funding

This work was supported by the Eunice Kennedy Shriver National Institute of Child Health & Human Development of the National Institutes of Health under Award Number R03HD072400.

Transparency declarations

None to declare.

Supplementary data

Supplementary Methods 1 and 2, Figures S1 to S3 and Tables S1 to S5 are available as Supplementary data at JAC Online (http://jac.oxfordjournals.org/).

Supplementary Data
supp_71_4_1098__index.html (804B, html)

Acknowledgements

We would like to thank Ms Pragati Kenkare (MS), Information Management Analyst at the PAMF Research Institute, for her kind contribution in the identification of the control group of individuals for the validation case–control study.

References

  • 1.Ogden CL, Carroll MD, Kit BK et al. Prevalence of childhood and adult obesity in the United States, 2011–2012. JAMA 2014; 311: 806–14. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 2.Ley RE, Backhed F, Turnbaugh P et al. Obesity alters gut microbial ecology. Proc Natl Acad Sci USA 2005; 102: 11070–5. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 3.Ley RE, Turnbaugh PJ, Klein S et al. Microbial ecology: human gut microbes associated with obesity. Nature 2006; 444: 1022–3. [DOI] [PubMed] [Google Scholar]
  • 4.Turnbaugh PJ, Ley RE, Mahowald MA et al. An obesity-associated gut microbiome with increased capacity for energy harvest. Nature 2006; 444: 1027–31. [DOI] [PubMed] [Google Scholar]
  • 5.Turnbaugh PJ, Hamady M, Yatsunenko T et al. A core gut microbiome in obese and lean twins. Nature 2009; 457: 480–4. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 6.Gill SR, Pop M, Deboy RT et al. Metagenomic analysis of the human distal gut microbiome. Science 2006; 312: 1355–9. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 7.Vijay-Kumar M, Aitken JD, Carvalho FA et al. Metabolic syndrome and altered gut microbiota in mice lacking Toll-like receptor 5. Science 2010; 328: 228–31. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 8.Raoult D. Obesity pandemics and the modification of digestive bacterial flora. Eur J Clin Microbiol Infect Dis 2008; 27: 631–4. [DOI] [PubMed] [Google Scholar]
  • 9.Sanz Y, Santacruz A, Gauffin P. Gut microbiota in obesity and metabolic disorders. Proc Nutr Soc 2010; 69: 434–41. [DOI] [PubMed] [Google Scholar]
  • 10.Acar J, Casewell M, Freeman J et al. Avoparcin and virginiamycin as animal growth promoters: a plea for science in decision-making. Clin Microbiol Infect 2000; 6: 477–82. [DOI] [PubMed] [Google Scholar]
  • 11.Collignon P, Wegener HC, Braam P et al. The routine use of antibiotics to promote animal growth does little to benefit protein undernutrition in the developing world. Clin Infect Dis 2005; 41: 1007–13. [DOI] [PubMed] [Google Scholar]
  • 12.Gorbach SL. Antimicrobial use in animal feed—time to stop. N Engl J Med 2001; 345: 1202–3. [DOI] [PubMed] [Google Scholar]
  • 13.Diarra MS, Malouin F. Antibiotics in Canadian poultry productions and anticipated alternatives. Front Microbiol 2014; 5: 282. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 14.Castanon JI. History of the use of antibiotic as growth promoters in European poultry feeds. Poult Sci 2007; 86: 2466–71. [DOI] [PubMed] [Google Scholar]
  • 15.Khan M, Raoult D, Richet H et al. Growth-promoting effects of single-dose intragastrically administered probiotics in chickens. Br Poult Sci 2007; 48: 732–5. [DOI] [PubMed] [Google Scholar]
  • 16.Patterson JA, Burkholder KM. Application of prebiotics and probiotics in poultry production. Poult Sci 2003; 82: 627–31. [DOI] [PubMed] [Google Scholar]
  • 17.Haight TH, Pierce WE. Effect of prolonged antibiotic administration of the weight of healthy young males. J Nutr 1955; 56: 151–61. [DOI] [PubMed] [Google Scholar]
  • 18.Rosenberg IH, Beisel WR, Gordon JE et al. Infant and child enteritis-malabsorption-malnutrition: the potential of limited studies with low-dose antibiotic feeding. Am J Clin Nutr 1974; 27: 304–9. [DOI] [PubMed] [Google Scholar]
  • 19.Ajslev TA, Andersen CS, Gamborg M et al. Childhood overweight after establishment of the gut microbiota: the role of delivery mode, pre-pregnancy weight and early administration of antibiotics. Int J Obes (Lond) 2011; 35: 522–9. [DOI] [PubMed] [Google Scholar]
  • 20.Trasande L, Blustein J, Liu M et al. Infant antibiotic exposures and early-life body mass. Int J Obes (Lond) 2013; 37: 16–23. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 21.Murphy R, Stewart AW, Braithwaite I et al. Antibiotic treatment during infancy and increased body mass index in boys: an international cross-sectional study. Int J Obes (Lond) 2014; 38: 1115–9. [DOI] [PubMed] [Google Scholar]
  • 22.Mueller NT, Whyatt R, Hoepner L et al. Prenatal exposure to antibiotics, cesarean section and risk of childhood obesity. Int J Obes (Lond) 2015; 39: 665–70. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 23.Al-Hosni M, Duenas M, Hawk M et al. Probiotics-supplemented feeding in extremely low-birth-weight infants. J Perinatol 2012; 32: 253–9. [DOI] [PubMed] [Google Scholar]
  • 24.Sazawal S, Dhingra U, Hiremath G et al. Effects of Bifidobacterium lactis HN019 and prebiotic oligosaccharide added to milk on iron status, anemia, and growth among children 1 to 4 years old. J Pediatr Gastroenterol Nutr 2010; 51: 341–6. [DOI] [PubMed] [Google Scholar]
  • 25.Vendt N, Grünberg H, Tuure T et al. Growth during the first 6 months of life in infants using formula enriched with Lactobacillus rhamnosus GG: double-blind, randomized trial. J Hum Nutr Diet 2006; 19: 51–8. [DOI] [PubMed] [Google Scholar]
  • 26.Mohan R, Koebnick C, Schildt J et al. Effects of Bifidobacterium lactis Bb12 supplementation on body weight, fecal pH, acetate, lactate, calprotectin, and IgA in preterm infants. Pediatr Res 2008; 64: 418–22. [DOI] [PubMed] [Google Scholar]
  • 27.Kanamori Y, Sugiyama M, Hashizume K et al. Experience of long-term synbiotic therapy in seven short bowel patients with refractory enterocolitis. J Pediatr Surg 2004; 39: 1686–92. [DOI] [PubMed] [Google Scholar]
  • 28.Uchida K, Takahashi T, Inoue M et al. Immunonutritional effects during synbiotics therapy in pediatric patients with short bowel syndrome. Pediatr Surg Int 2007; 23: 243–8. [DOI] [PubMed] [Google Scholar]
  • 29.Underwood MA, Salzman NH, Bennett SH et al. A randomized placebo-controlled comparison of 2 prebiotic/probiotic combinations in preterm infants: impact on weight gain, intestinal microbiota, and fecal short-chain fatty acids. J Pediatr Gastroenterol Nutr 2009; 48: 216–25. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 30.Chou IC, Kuo HT, Chang JS et al. Lack of effects of oral probiotics on growth and neurodevelopmental outcomes in preterm very low birth weight infants. J Pediatr 2010; 156: 393–6. [DOI] [PubMed] [Google Scholar]
  • 31.Gough EK, Moodie EE, Prendergast AJ et al. The impact of antibiotics on growth in children in low and middle income countries: systematic review and meta-analysis of randomised controlled trials. BMJ 2014; 348: g2267. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 32.Wong EC, Palaniappan LP, Lauderdale DS. Using name lists to infer Asian racial/ethnic subgroups in the healthcare setting. Med Care 2010; 48: 540–6. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 33.CDC. A SAS Program for the CDC Growth Charts. http://www.cdc.gov/nccdphp/dnpao/growthcharts/resources/sas.htm.
  • 34.von Elm E, Altman DG, Egger M et al. The Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) statement: guidelines for reporting observational studies. Lancet 2007; 370: 1453–7. [DOI] [PubMed] [Google Scholar]
  • 35.Kosanke J, Bergstralh E. gmatch SAS Macro. October 2003. http://www.mayo.edu/research/documents/gmatchsas/doc-10027248.
  • 36.R Core Team. R: A Language and Environment for Statistical Computing. Vienna, Austria: R Foundation for Statistical Computing, 2013. [Google Scholar]
  • 37.Lees CD, O'Neill M, Steiger E. The effect of intravenous doxycycline hyclate on total parenteral nutrition in protein malnourished rats. JPEN J Parenter Enteral Nutr 1981; 5: 510–2. [DOI] [PubMed] [Google Scholar]
  • 38.Graham JP, Boland JJ, Silbergeld E. Growth promoting antibiotics in food animal production: an economic analysis. Public Health Rep 2007; 122: 79–87. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 39.Bäckhed F, Ding H, Wang T et al. The gut microbiota as an environmental factor that regulates fat storage. Proc Natl Acad Sci USA 2004; 101: 15718–123. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 40.Turnbaugh PJ, Ridaura VK, Faith JJ et al. The effect of diet on the human gut microbiome: a metagenomic analysis in humanized gnotobiotic mice. Sci Transl Med 2009; 1: 6ra14. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 41.Dumas ME, Barton RH, Toye A et al. Metabolic profiling reveals a contribution of gut microbiota to fatty liver phenotype in insulin-resistant mice. Proc Natl Acad Sci USA 2006; 103: 12511–6. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 42.Armougom F, Henry M, Vialettes B et al. Monitoring bacterial community of human gut microbiota reveals an increase in Lactobacillus in obese patients and Methanogens in anorexic patients. PLoS One 2009; 4: e7125. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 43.Duncan SH, Lobley GE, Holtrop G et al. Human colonic microbiota associated with diet, obesity and weight loss. Int J Obes (Lond) 2008; 32: 1720–4. [DOI] [PubMed] [Google Scholar]
  • 44.Santacruz A, Marcos A, Wärnberg J et al. Interplay between weight loss and gut microbiota composition in overweight adolescents. Obesity 2009; 17: 1906–15. [DOI] [PubMed] [Google Scholar]
  • 45.Nadal I, Santacruz A, Marcos A et al. Shifts in clostridia, bacteroides and immunoglobulin-coating fecal bacteria associated with weight loss in obese adolescents. Int J Obes (Lond) 2009; 33: 758–67. [DOI] [PubMed] [Google Scholar]
  • 46.Zhang H, DiBaise JK, Zuccolo A et al. Human gut microbiota in obesity and after gastric bypass. Proc Natl Acad Sci USA 2009; 106: 2365–70. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 47.Trehan I, Goldbach HS, LaGrone LN et al. Antibiotics as part of the management of severe acute malnutrition. N Engl J Med 2013; 368: 425–35. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 48.Smith MI, Yatsunenko T, Manary MJ et al. Gut microbiomes of Malawian twin pairs discordant for kwashiorkor. Science 2013; 339: 548–54. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 49.DiBaise JK, Zhang H, Crowell MD et al. Gut microbiota and its possible relationship with obesity. Mayo Clin Proc 2008; 83: 460–9. [DOI] [PubMed] [Google Scholar]
  • 50.Angelakis E, Merhej V, Raoult D. Related actions of probiotics and antibiotics on gut microbiota and weight modification. Lancet Infect Dis 2013; 13: 889–99. [DOI] [PubMed] [Google Scholar]
  • 51.Azad MB, Bridgman SL, Becker AB et al. Infant antibiotic exposure and the development of childhood overweight and central adiposity. Int J Obes (Lond) 2014; 38: 1290–8. [DOI] [PubMed] [Google Scholar]
  • 52.Bailey LC, Forrest CB, Zhang P et al. Association of antibiotics in infancy with early childhood obesity. JAMA Pediatr 2014; 168: 1063–9. [DOI] [PubMed] [Google Scholar]
  • 53.Cox LM, Yamanishi S, Sohn J et al. Altering the intestinal microbiota during a critical developmental window has lasting metabolic consequences. Cell 2014; 158: 705–21. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 54.Cox LM, Blaser MJ. Antibiotics in early life and obesity. Nat Rev Endocrinol 2015; 11: 182–90. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 55.Yatsunenko T, Rey FE, Manary MJ et al. Human gut microbiome viewed across age and geography. Nature 2012; 486: 222–7. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 56.Angelakis E, Million M, Kankoe S et al. Abnormal weight gain and gut microbiota modifications are side effects of long-term doxycycline and hydroxychloroquine treatment. Antimicrob Agents Chemother 2014; 58: 3342–7. [DOI] [PMC free article] [PubMed] [Google Scholar] [Retracted]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Supplementary Data
supp_71_4_1098__index.html (804B, html)
supp_dkv455_dkv455supp.doc (1.6MB, doc)

Articles from Journal of Antimicrobial Chemotherapy are provided here courtesy of Oxford University Press

RESOURCES