Objective
This study investigated whether birthweight is linked to an increased risk of the development of systemic sclerosis.
Study Design
This was a multicenter case-control study with perinatal data that were obtained from 332 cases with systemic sclerosis and 243 control subjects. Birthweight was treated as a dichotomous variable (<2500 g vs ≥2500 g); low birthweight was defined as a weight <2500 g; small for gestational age was defined as birthweight <10th percentile for gestational age adjusted for sex. The relationship between systemic sclerosis and both low birthweight and small for gestational age was expressed with the crude (univariate analysis) and adjusted (multivariate analysis) odds ratio (OR).
Results
Significantly increased ORs were observed in the univariate analysis for low birthweight (OR, 2.59; 95% confidence interval [CI], 1.39–5.05) and small for gestational age (OR, 2.60; 95% CI, 1.34–5.32) subjects. Similarly increased risks were confirmed for both conditions in the multivariate analysis (OR, 3.93; 95% CI, 1.92–8.07; and OR, 2.58; 95% CI, 1.28–5.19), respectively.
Conclusion
Low birthweight and small for gestational age at birth are risk factors for the adult onset of systemic sclerosis.
Systemic sclerosis (SSc), also known as scleroderma, is a chronic autoimmune disease characterized by vascular obliteration, excessive extracellular matrix deposition, and fibrosis of the connective tissues. Estimates of the disease’s prevalence and incidence range from 50-300 cases per million people and 2.3–22.8 cases per million people per year, respectively. Women are more likely to experience SSc than are men, with reported ratios ranging from 3:1–14:1; a slightly increased susceptibility has also been reported among black patients.
The cause of SSc has remained elusive despite intense investigation, although there is convincing evidence that genetic factors contribute to its onset and development. Genetics studies suggest that SSc is a complex polygenic disease. Candidate gene studies have identified critical immunoregulatory genes and gene regions (in particular, the human leukocyte antigen region) as susceptibility genes for the development of the disease. However, even though the genetic contribution to the disease has been shown, it now seems that environmental agents also play a critical role. The link between genes and environment is represented by the new field of research of epigenetics: the study of heritable changes in genes and gene expression that do not involve DNA nucleotide sequences. Epigenetic modifications include DNA methylation, histone modifications, and microRNA expression. In humans, cytosine methylation and its modifications in response to maternal diet is 1 of the most widely studied epigenetic modifications and is a sign of adverse exposure in utero. Interestingly, differentially methylated regions that are dependent on the mother’s diet have been identified in the liver of female offspring, which represents potential marks of developmental programming that may link the intrauterine environment to metabolic health later in life.
This is an epidemiologic study that was inspired by the robust work of David Barker and others, who postulated that the environment can modify the developmental trajectory of an individual even during the first stages of life, thus laying the foundations for disease in adulthood (the so-called fetal programming theory). In fact, it is possible that adverse environmental conditions during fetal growth could alter developmental processes and explain how a single genotype can give rise to different phenotypes (developmental plasticity). This approach has been shown to be valid particularly for chronic diseases such as cardiovascular disease, metabolic diseases (including diabetes mellitus), osteoporosis, and some forms of cancer. In line with this theory, several studies have shown that low birthweight (LBW), which can be indicative of exposure to an adverse fetal environment during specific stages of gestation (so-called “critical” periods, when rapid cell division takes place), led to an increased risk of chronic diseases in adulthood because of programming of the neuroendocrine setting, antioxidant defenses, inflammation, and the immune system itself.
This study was based on the hypothesis that the immune system is subject to developmental plasticity during its maturation, with a real possibility that disruptors in the early fetal environment may impair its function by epigenetic mechanisms, thereby increasing the onset of chronic autoimmune diseases. The main goal was to evaluate whether a significant correlation exists between birthweight and/or gestational age and the subsequent development of SSc.
Materials and Methods
A multicenter case-control study was conducted from June 2012 to November 2013; 332 consecutive prevalent cases of SSc were enrolled from the rheumatologic outpatient clinics of the following hospitals: Careggi University Hospital of Florence, La Sapienza University Hospital of Rome, the IRCCS Foundation and San Matteo University Hospital of Pavia, and the University Hospital of Ancona. The study was approved by the Ethics Committee of the Meyer Children’s Hospital, University of Florence. Cases were defined as patients affected by SSc according to the recent classification developed by the American College of Rheumatology/European League. Two hundred forty-three consecutive control subjects were recruited from the surgical outpatient clinic of the Careggi University Hospital of Florence during the same period. Control subjects were matched to cases with the use of a frequency-matching method to obtain a similar age/sex distribution in both groups.
To collect demographic and perinatal information, a standardized questionnaire was created, and the patients were interviewed by trained medical personnel. The following information was collected: patient identification, sex, age, birthweight, gestational age at birth, their mother’s age at birth, whether they were breastfed, and the mother’s smoking habit. Birthweight and gestational age were recorded as continuous variables; when subjects did not recall exact values, they were asked to categorize birthweight as <2500, 2500-3999, or ≥4000 g and gestational age as <37, 37-41, or >41 weeks. With regard to the occupational histories of patients and control subjects, particular attention was given to exposure to crystalline silica, organic solvents, welding fumes, epoxy resins, and pesticides. To minimize recall bias and ensure the validity of the approach, the data that were collected were compared with the data recorded in the clinical charts, even if this was possible only in 40 of the patients (12%) and 24 of the control subjects (10%) because of a lack of available records, especially in the case of older subjects. Moreover, the interview was repeated 1 month later in a randomized group of 40 patients and 40 control subjects to assess the level of uncertainty that had been attributed to recall.
The following disease characteristics were obtained directly from the hospital databases for all patients with a diagnosis of SSc: auto-antibody pattern (antinuclear antibodies, anticentromere antibodies, anti-Scl-70), age at disease onset, type of disease (diffuse cutaneous, limited cutaneous), organ involvement (articular, cardiac, gastrointestinal, pulmonary), and the presence of pulmonary hypertension and digital ulceration. The exclusion criteria for both patients and control subjects were refusal to participate, the presence of chronic diseases (such as coronary heart disease and related disorders, stroke, hypertension, and type 2 diabetes mellitus ), and occupation-related SSc risk factors (in particular organic solvents, silica, white spirit, welding fumes, and epoxy resins). Furthermore, the presence of other autoimmune diseases was an exclusion criterion for patients with SSc; the presence of any autoimmune disease was an exclusion criterion for control subjects. The Regional Center of Rare Diseases of the Meyer Children’s Hospital, University of Florence, developed the research protocol, supervised the data collection and performed the statistical data analysis.
LBW was defined as a weight at birth of <2500 g, as per the International Statistical Classification of Diseases and Related Health Problems, 10th revision. The term small for gestational age (SGA) refers to infants whose birthweights and/or lengths are at least 2 standard deviation units less than the mean for gestational age. In clinical practice, SGA commonly is defined as a birthweight <10th percentile for gestational age and sex relative to the population standard; it is used as a measurable proxy for intrauterine growth restriction and later health risks.
The study population was divided according to birthweight (<2500, ≥2500, or ≥4000 g) and gestational age (preterm, <37 weeks; at term, 37-41 weeks; postterm, >41 weeks). The association between LBW and SSc and between SGA and SSc was expressed with the odds ratio (OR) calculated with a univariate analysis that considered 2 birthweight groups (<2500 and ≥2500 g). Two multivariate analyses were then performed to test the associations among all 3 birthweight groups (<2500, 2500-3999, and ≥4000 g) and both SGA conditions, with the risks for SSc adjusted for other confounding factors. A sample size of 318 cases and 212 control subjects was estimated, with the assumption of a 5% exposure among control subjects and a minimum appreciable OR of 3 (alpha error, 5%; power, 90%). The statistical analysis was performed with the Stata software (version 10; StataCorp LP, College Station, TX). Eleven percent of the subjects in both groups (66/575) were excluded because of uncertainty about their birthweight (15.4% of case studies and 6.2% of control subjects); the gestational age was uncertain in 4.5% of case studies and 0.0% of the control group.
Results
The clinical characteristics of the study population are presented in Table 1 . The expected sex distribution of disease and frequency matching that was adopted for the enrolled control subjects demonstrated a high prevalence of female cases in the study population compared with male cases. The age comparison between case studies and control subjects showed that the mean age of patients with SSc was slightly higher in the control group. Mothers of patients with SSc were slightly older than the control subjects at the time of delivery.
| Variable | Cases (n = 332) | Control subjects (n = 243) | P value | Total (n = 575) |
|---|---|---|---|---|
| Mean age, y ± SD | 59.1 ± 13.2 | 56.1 ± 6.2 | < .05 | 58.1 ± 11.5 |
| Maternal age at delivery, y ± SD | 29.3 ± 6.0 | 27.9 ± 5.8 | < .05 | 28.7 ± 5.9 |
| Sex, n (%) | .07 | |||
| Male | 15 (4.5) | 18 (7.4) | 33 (5.7) | |
| Female | 317 (95.5) | 225 (92.6) | 542 (94.3) | |
| Birthweight, n (%) | < .05 | |||
| Missing | 51 (15.4) | 15 (6.2) | 66 (11.5) | |
| <2500 g | 46 (13.9) | 16 (6.6) | 62 (10.8) | |
| ≥2500 g | 235 (70.8) | 212 (87.2) | 447 (77.7) | |
| Gestational age at birth, n (%) | .66 | |||
| Missing | 15 (4.5) | 2 (0.9) | 15 (2.6) | |
| Preterm | 9 (2.7) | 4 (1.6) | 13 (2.3) | |
| Term | 299 (90.1) | 230 (94.6) | 531 (92.3) | |
| Postterm | 9 (2.7) | 7 (2.9) | 16 (2.8) | |
| Small for gestational age, n (%) | < .05 | |||
| Missing | 57 (17.2) | 15 (6.2) | 72 (12.5) | |
| Yes | 40 (12.0) | 14 (5.8) | 54 (9.4) | |
| No | 235 (70.8) | 214 (88.1) | 449 (78.1) | |
| Breastfeeding, n (%) | .41 | |||
| Missing | 108 (32.5) | 4 (1.7) | 112 (19.5) | |
| Yes | 202 (60.8) | 217 (89.3) | 419 (72.9) | |
| No | 22 (6.7) | 22 (9.0) | 44 (7.6) | |
| Mother’s smoking habit, n (%) | .29 | |||
| Missing | 107 (32.2) | 7 (2.9) | 114 (19.8) | |
| Yes | 9 (2.7) | 12 (4.9) | 21 (3.6) | |
| No | 216 (65.1) | 224 (92.2) | 440 (76.5) | |
| Disease subtype, n (%) | ||||
| Missing | 13 (3.9) | |||
| Limited cutaneous systemic sclerosis | 246 (74.1) | |||
| Diffuse cutaneous systemic sclerosis | 73 (22.0) | |||
| Antibodies, n (%) | ||||
| Missing | 96 (28.9) | |||
| Anticentromere antibodies | 133 (40.1) | |||
| Scl70 | 100 (30.1) | |||
| Anticentromere antibodies + Scl70 (topoisomerase I) | 3 (0.9) |
It was observed that LBW and SGA were more prevalent in subjects with SSc than in the control subjects (13.9% vs 6.9% and 12.0% vs 5.8%, respectively), with a less pronounced difference regarding preterm births (2.7% vs 1.6%). Table 2 contains the ORs of SSc that were evaluated as a univariate analysis for LBW, SGA, maternal age at delivery, gestational age, breastfeeding, and the mother’s smoking behavior during pregnancy. SSc was more prevalent in LBW infants than in non-LBW infants (OR, 2.59; 95% confidence interval [CI], 1.39–5.05). When the analysis was limited to female subjects only, the OR rose to 2.90 (95% CI, 1.51–5.89). A similar result was found when exposure to SGA was considered (OR, 2.60; 95% CI, 1.34–5.32, increasing to OR, 2.73; 95% CI, 1.38–5.73 in female subjects). Both estimated risks were statistically significant ( P < .05). The results showed that advanced maternal age (>34 years) was associated with an increased risk of SSc (OR, 2.28; 95% CI, 1.34–3.88). The mother’s age at the time of delivery therefore can be considered an additional, independent risk factor for the disease. Conversely, the breastfeeding and smoking habits of the mothers of patients with SSc were not associated with a significantly increased risk of SSc.
| Variable | Cases, n | Control subjects, n | Odds ratio (95% confidence interval) |
|---|---|---|---|
| Birthweight, g | |||
| <2500 | 46 | 16 | 2.59 (1.39–5.05) |
| ≥2500 | 235 | 212 | 1.00 |
| Birthweight (female only), g | |||
| <2500 | 46 | 14 | 2.90 (1.51–5.89) |
| ≥2500 | 224 | 198 | 1.00 |
| Small for gestational age | |||
| Yes | 40 | 14 | 2.60 (1.34–5.32) |
| No | 235 | 214 | 1.00 |
| Small for gestational age (female only) | |||
| Yes | 40 | 13 | 2.73 (1.38–5.73) |
| No | 224 | 199 | 1.00 |
| Maternal age at delivery (female only), y | |||
| <25 | 69 | 68 | 1.00 |
| 25-29 | 100 | 79 | 1.25 (0.80–1.95) |
| 30-34 | 74 | 46 | 1.58 (0.96–2.61) |
| >34 | 74 | 32 | 2.28 (1.34–3.88) |
| Gestational age (female only), wk | |||
| 37-41 | 284 | 213 | 1.0 |
| <37 | 9 | 3 | 2.28 (0.61–8.53) |
| >41 | 9 | 6 | 1.14 (0.40–3.25) |
| Breastfeeding | |||
| Yes | 202 | 217 | 0.93 (0.50–1.73) |
| No | 22 | 22 | 1.0 |
| Breastfeeding (female only) | |||
| Yes | 189 | 202 | 1.07 (0.54–2.14) |
| No | 20 | 23 | 1.0 |
| Mother’s smoking habit | |||
| Yes | 9 | 12 | 0.78 (0.28–2.06) |
| No | 216 | 224 | 1.0 |
| Mother’s smoking habit (female only) | |||
| Yes | 8 | 11 | 0.74 (0.25–2.07) |
| No | 203 | 207 | 1.0 |
Table 3 shows the multivariate analysis limited to female subjects that, through 2 different mathematical models, shows the independent effect of birthweight grouped into different classes: birthweight (overweight, ≥4000 g; underweight, <2500 g; and normal weight, 2500-4000 g), gestational age (term, preterm, and postterm), age of cases and control subjects, and maternal age at delivery. The multivariate analysis shows that the odds of disease in the female LBW group are approximately 4 times higher than in the normal birthweight group. No additional risk was observed in subjects with birthweight >4000 g as compared with subjects of normal weight at birth (2500-3999 g). SGA status increases the OR of disease approximately 2.6-fold in the multivariate analysis that includes the age of cases and control subjects and maternal age at delivery. In both models, the patient’s age is an independent risk factor for disease, with a risk of 4% (model 1) and 5% (model 2) for each additional year ( Table 3 ).
| Parameter | Odds ratio | 95% confidence interval |
|---|---|---|
| Model 1 | ||
| Birthweight, g | ||
| 2500-3999 | 1.00 | |
| <2500 | 3.93 | 1.92–8.07 |
| ≥4000 | 1.06 | 0.57–1.98 |
| Age (by year) | 1.04 | 1.02–1.06 |
| Gestational age, wk | ||
| 37-41 | 1.00 | |
| <37 or >41 a | 2.22 | 0.88–5.59 |
| Maternal age at delivery, y | ||
| <25 | 1.00 | |
| 25-29 | 1.23 | 0.71–2.14 |
| 30-34 | 1.53 | 0.83–2.83 |
| >34 | 2.23 | 1.16–4.26 |
| Model 2 | ||
| Small for gestational age | ||
| No | 1.00 | |
| Yes | 2.58 | 1.28–5.19 |
| Age (by year) | 1.05 | 1.03–1.07 |
| Maternal age at delivery, y | ||
| <25 | 1.00 | |
| 25-29 | 1.18 | 0.72–1.93 |
| 30-34 | 1.46 | 0.84–2.56 |
| >34 | 2.01 | 1.11–3.63 |
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
