Association between Number of Siblings and Cardiovascular Risk Factors in Childhood and in Adulthood: The Cardiovascular Risk in Young Finns Study

Objective To determine the association of number of siblings on cardiovascular risk factors in childhood and in adulthood. Study design In total, 3554 participants (51% female) from the Cardiovascular Risk in Young Finns Study with cardiovascular disease risk factor data at baseline 1980 (age 3-18 years) and 2491 participants with longitudinal risk factor data at the 2011 follow-up. Participants were categorized by number of siblings at baseline (0, 1, or more than 1). Risk factors (body mass index, physical activity, hypertension, dyslipidemia, and overweight, and metabolic syndrome) in childhood and in adulthood were used as outcomes. Analyses were adjusted for age and sex. Results In childhood, participants without siblings had higher body mass index (18.2 kg/m 2 , 95% CI 18.0-18.3) than those with 1 sibling (17.9 kg/m 2 , 95% CI 17.8-18.0) or more than 1 sibling (17.8 kg/m 2 , 95% CI 17.7-17.9). Childhood physical activity index was lower among participants without siblings (SD -0.08, 95% CI -0.16-0.00) compared with participants with 1 sibling (SD 0.06, 95%CI 0.01-0.11) or more than 1 sibling (SD -0


C
ardiovascular disease (CVD) is the leading cause of death worldwide and a major portion of these deaths could be prevented. 1In addition to well known risk factors for CVD, family size, described by the number of offspring, has been shown to impact the prevalence of CVD among parents. 2,3owever, the available evidence has been contradictory, with some studies showing that the number of offspring associates with the risk of CVD in mothers, 4,5 or in both parents, 3 and other studies have found no 6,7 or nonlinear 8,9 associations.
Research on offspring in low-or middle-income countries has shown negative effects of a larger family size on child health in childhood/adolescence mainly through nutritional factors. 10,11A Finnish study found no association between the number of inhabitants in the household and death from coronary heart

ORIGINAL ARTICLES
disease. 12 Similarly, the association between growing up in a large family and adulthood mortality was not demonstrated. 13Although, an earlier study found that children without siblings have higher blood pressure in adulthood compared with children with siblings. 14Moreover, children from small families (ie, 1 or 2 child families) are more likely to graduate from high school in the US compared with large families because of intrafamilial resources diluting in larger families 15 and lower education has been shown to increase prevalence of risk behaviors such as smoking, obesity, physical inactivity, and unhealthy diet. 16Most studies have focused on the effects of family size on parent's health, been performed in less developed countries, or studied mortality and the association between family size and the development of CVD in offspring remains unknown.Therefore, we investigated if the number of siblings associates with cardiovascular health in childhood and in adulthood in the longitudinal Cardiovascular Risk in Young Finns Study (YFS).The YFS is a population-based cohort of well-characterized individuals, followed from childhood to adulthood for up to 31 years.We hypothesized that the number of siblings would affect cardiovascular risk factor levels in childhood and adulthood.

Methods
The YFS is an ongoing longitudinal population-based multicenter study of cardiovascular risk factors from childhood to adulthood, conducted in 5 university hospital cities in Finland (Helsinki, Kuopio, Oulu, Tampere, and Turku) and their rural surrounds.The baseline study was conducted in 1980 when 3596 randomly selected children and adolescents age 3, 6, 9, 12, 15, and 18 years participated.Since 1980, the cohort has been regularly followed up in 3-to 9year intervals.A detailed description of the cohort has been published previously. 17Participants or their parents provided written informed consent, and the study was approved by local ethics committees.Participants included in this study had childhood risk factor data available from baseline (n = 3420) and adult risk factor data (n = 2491) from the 2011 follow-up study (n between 1979 and 2441), or in case of missing information from the 2011 follow-up, data from the 2007 follow-up was used (n between 406 and 438).

Family Size, Number of Siblings
Information on the number of children in the family was collected from parents' self-report questionnaires at baseline in 1980.Participants were categorized by the number of children in the family as (1) 1 child/no siblings (n = 536), 15% of total cohort; (2) 2 children/1 sibling (n = 1543), 43% of total cohort; (3) 3 or more children/2 or more siblings (n = 1475), 42% of total cohort.

Blood Pressure and Weight
At baseline, brachial artery blood pressure was measured using a standard mercury sphygmomanometer for participants age ³6 years.In case of missing information, data from the 1983 follow-up was used.Adult blood pressure measurements were collected in the 2011 follow-up using a random-zero sphygmomanometer (Hawksley and Sons Ltd).All measurements were taken from the right arm after the participant had been seated for 5 minutes.Three measurements were taken, and the average of these measurements was used.
At baseline and all follow-up visits, weight was measured without shoes in light clothes with a digital Seca weighing scale to nearest 0.1 kg.A Seca stadiometer was used for the measurement of height.Body mass index (BMI) was calculated as weight (kg) divided by height in meters squared (m 2 ).The baseline measurement was used as the primary indicator of childhood/adolescent BMI.In case of missing information, data from the year 1983 follow-up was used.For adulthood BMI, data were derived from the latest follow-up study in 2011.In case of missing information, data from the 2007 follow-up was used.

Physical Activity Index
At ages 3 and 6 years, a physical activity index was calculated from the parents' ratings of the amount and vigorousness of their child's play time and the child's general level of activity. 18At ages 9-18 years, data on frequency and intensity of leisure-time physical activity, participation in sports club training, participation in sport competitions, and habitual leisure time was acquired with a self-administered questionnaire. 19The values for the physical activity indices in childhood were standardized and combined.Adulthood physical activity index was calculated by assessing the frequency of physical activity, intensity of physical activity, frequency of vigorous physical activity, hours spent on vigorous physical activity, and average duration of physical activity. 19

Blood Biochemistry
Fasting serum lipids such as serum total cholesterol, highdensity lipoprotein (HDL)-cholesterol, and triglycerides were measured in the same laboratory at each follow-up with standard methods.Low-density lipoprotein (LDL)cholesterol concentration was calculated using the Friedewald equation. 20The applied methods have been reported previously. 21,22Serum glucose concentration was determined by the enzymatic hexokinase method (Glucose reagent, Beckman Coulter Biomedical).The concentration of glycated hemoglobin A1c (HbA1c) was assayed with an immunoturbidimetric method (HbA1c assay, Abbot) on an architect ci8200 analyzer (Abbott) in 2011.Serum insulin was measured in 1980 with a modification of the immunoassay method of Herbert et al. 23 Adverse Cardiovascular Health Metrics in Childhood According to pediatric guidelines, we defined abnormal blood pressure in childhood as pediatric hypertension or prehypertension based on either systolic blood pressure being in the uppermost 90th percentile of the age-, sex-, and yearspecific distribution. 24,25Integrated guidelines 26 were used to define high total cholesterol (³5.17 mmol/L), low HDLcholesterol (<1.03 mmol/L), high LDL-cholesterol (³3.36 mmol/L), and high triglycerides (³1.13 mmol/L for children age 3-9 years and ³1.47 mmol/l for children age 10-18 years) in childhood.Centers for Disease Control and Prevention recommendations were used to determine overweight (85th to 95th percentile) and obesity (95th percentile or greater) in childhood. 27verse Cardiovascular Health Metrics in Adulthood Systolic blood pressure >140 mm Hg, diastolic blood pressure >90 mm Hg, or self-reported use of blood pressure medication were used as criteria for hypertension in adulthood.Participants were considered to have type 2 diabetes if they had fasting glucose ³7 mmol/L or HbA1c ³48 mmol/L, or they self-reported diabetes or use of glucose-lowering medication.Participants who had fasting plasma glucose from 5.6 mmol/L to 6.9 mol/L or HbA1c from 39 mmol/L to 46 mmol/L, and no self-reported diabetes were assigned as individuals with prediabetes. 28Classification for hypercholesterolemia was assigned for participants if they had LDL-cholesterol >3 mmol/L or used lipidlowering medication.Classification for hypertriglyceridemia was assigned for participants if they had triglycerides >1.7 mmol/L. 29Participants with BMI from 25 kg/m 2 to 29.9 kg/m 2 were assigned as being overweight and those with BMI ³30 kg/m 2 as being obese. 30

Statistical Analyses
Baseline characteristics of the study population are reported as mean (SD) or median (25th and 75th percentiles, if skewed distributions) for continuous variables or as proportions for categorical variables.The relationship between number of siblings and continuous outcome variables was assessed using the generalized linear model adjusted with Tukey-Kramer approximation and with logistic regression models for categorical outcome variables.All analyses were adjusted for sex and age.
Sensitivity analyses were conducted for both childhood and adulthood outcomes to study the robustness of our findings.First, we combined data on the number of siblings from baseline and the 1983 and 1986 follow-up surveys to take account for the possible misclassification of participants where the number of children increased after the baseline survey.Second, using combined data from baseline and data collected on the number of siblings from the parents of the participants when they contributed data to the latest YFS field study in 2018-2020 (n = 1274).The parents were enquired how many childbirths they have had.Participants were categorized by the number of children in the family as (1) 1 child/ no siblings (n = 450), 13% of total cohort; (2) 2 children/1 sibling (n = 1438), 40% of total cohort; (3) 3 or more children/2 or more siblings (n = 1670), 47% of total cohort.Third, we evaluated the associations using different cutpoints for the number of siblings as 1 child/no siblings, 1 sibling, 2 siblings, and 3 or more siblings.Fourth, additional adjustments for birth order, childhood/adulthood socioeconomic status, childhood living region categorized as urban or rural, 31 and total years of education were also analysed.Both sex Â exposure and age Â exposure interactions were individually studied to investigate if the associations were similar by sex and age groups.The investigations were made separately for childhood and adulthood outcomes.Except for adult hypertriglyceridemia and LDL-cholesterol concentration in childhood, we observed no interactions between number of siblings with sex or age on risk factor/ outcome (P value >.05 for all).
All statistical analyses were performed using SAS v 9.4 (SAS Institute), and statistical significance was inferred at a 2-tailed P value of <.05.

Results
Characteristics of the participants are show in Table I.The total number of participants with data on the number of siblings and the covariates in childhood was 3554 (51% female).Of these, 2491 had at least 1 adulthood outcome measurement available.The mean age of the participants was 41.6 AE 5 years at the 2011 follow-up.Median number of children in the family was 2.0 (IQR 2.0-3.0,range 0-18).

Childhood Risk Factors
Of the childhood risk factors, the number of siblings was associated with childhood LDL-cholesterol, BMI, and physical activity (Table II III.Compared with participants without siblings, the odds of overweight among those with one sibling (OR 0.66, 95% CI 0.49-0.88),and those with 2 or more siblings (OR 0.44, 95% CI 0.32-0.61)were lower.

Adulthood Risk Factors
The ORs for adulthood outcomes by number of siblings at baseline are shown in Table IV.Compared with participants without siblings, the odds for hypertension among those with 1 sibling (OR 0.73, 95% CI 0.54-0.98),and those with 2 or more siblings (OR 0.71, 95% CI 0.52-0.97)were lower.Compared with participants without siblings, the odds of obesity among those with one sibling (OR 0.72, 95% CI 0.54-0.95), and those with 2 or more siblings (OR 0.75, 95% CI 0.56-1.01)were lower.
Results for the association between the number of siblings and adulthood risk factors are shown in Table V (available at www.jpeds.com).Participants without siblings had higher LDL-cholesterol (3.23 mmol/L, 95% CI 3.14-3.31mmol/L) than those with 1 sibling (3.22 mmol/L, 95% CI 3.17-3.27mmol/L) but lower than those with 2 or more siblings (3.30 mmol/L, 95% CI 3.25-3.35mmol/L) (P for trend .08).No other significant associations between the number of siblings and adulthood risk factors were observed.

Sensitivity Analyses
In sensitivity analyses that additionally adjusted for family annual income and childhood living region (urban/rural), we observed no alterations in the main results (data not shown).In sensitivity analyses that adjusted further for birth order, the results for the association of adulthood obesity among participants with 2 or more siblings was diluted (OR 0.82, 95% CI 0.57-1.18),but adjustment had little effect on  ).There were no significant differences between the groups for other risk factors.For the adult outcomes, we additionally adjusted the analyses for participant's years of education, but the results (data not shown) remained consistent with our main findings.We also analyzed the association of number of siblings and adulthood hypertension additionally adjusting for childhood and adulthood BMI and the results remained essentially similar.Compared with participants without siblings, the odds for hypertension among those with one sibling (OR 0.76, 95% CI 0.55-1.04),and those with 2 or more siblings (OR 0.71, 95% CI 0.52-0.98)were lower.Results that used different cut-points for the number of siblings (ie, 1 child/no siblings, 1 sibling, 2 siblings, and 3 or more siblings) were similar to the main analyses.As we used number of siblings based on data collected at the 1980 baseline survey, misclassification of the number of siblings was possible.First, we combined data on the number of siblings from baseline and the 1983 and 1986 follow-up surveys to take account for the possible misclassification of participants, we observed no alterations in the main results (data not shown).Second, we used data collected on the number of siblings from the parents of the participants when they contributed data to the latest YFS field study in 2018-2020, the results were similar in cohort (data not shown).Because of the significant interaction observed between sex and adulthood hypertriglyceridemia, we conducted the analyses separately for women and men.In women, the number of siblings was not associated with the odds for hypertriglyceridemia.However, in men, the participants with 2 or more siblings had lower odds for hypertriglyceridemia (OR 0.63, 95% CI 0.

Discussion
We observed that children without siblings tended to have, on average, higher BMI and LDL-cholesterol, lower physical Adjusted for age and sex.Systolic blood pressure >140 mm Hg, diastolic blood pressure >90 mm Hg, or self-reported use of blood pressure medication were used as criteria for hypertension in adulthood.Participants were considered to have type 2 diabetes if they had fasting plasma glucose ³7 mmol/L or HbA1c ³48 mmol/L or they self-reported diabetes or use of glucose-lowering medication.Participants who had fasting plasma glucose from 5.6 mmol/L to 6.9 mol/L or HbA1c from 39 mmol/L to 46 mmol/L and no self-reported diabetes or use of glucose-lowering medication were assigned as individuals with prediabetes.Hypercholesterolemia was assigned for participants if they had LDL-cholesterol >3 mmol/L or use of lipid-lowering medication.Hypertriglyceridemia was assigned for participants if they had triglycerides >1.7 mmol/L.Participants with BMI from 25 kg/m 2 to 29.9 kg/m 2 were assigned as overweight and as obese if BMI was ³30 kg/m 2 .
activity, and higher odds for overweight in childhood compared with those with siblings.In addition, children without siblings were also more likely than their counterparts with siblings to have obesity and hypertension as adults.
3][34] Moreover, research on the influence of the number of siblings for adulthood morbidity has suggested that persons without siblings might be more likely to be hypertensive in adulthood, 35 which supports our observations of higher odds for having adult hypertension and obesity in those without siblings.
9][40] Although a review underlined that although childhood BMI is strongly associated with higher risk of adult obesity, it is not a good predictor of adult obesity or morbidity as most of the adult obesity and obesity-related adult morbidity occurs in adults who had a healthy childhood weight. 41We found that compared with the children with at least 1 sibling, the children without siblings had higher childhood BMI, increased odds for overweight, and they were physically less active, all characteristics that impact adult health. 42Because we did not observe associations between the number of siblings and other CVD risk factors in childhood, the mechanisms behind higher childhood BMI among children without siblings might be due to the increased amount of shared physical activity between siblings, such as sibling-to-sibling interactions, co-operative play, and shared interest in sports. 33onversely, in this study the association of the number of siblings and childhood BMI was independent of childhood physical activity index, suggesting only part of the effect was directed through physical activity index and the mechanism remains vague.However, in the absence of time and resource dilution, parents with less children could have more resources for helping with educational attainment 33 and, for instance, providing transportation for offspring to hobbies, allowing children's easier participation in sports club training, 13 and, thus, could prevent offspring's weight gain.Incidentally, a higher risk for adult morbidities (type 2 diabetes, hypertension, high risk HDL-and LDLcholesterol levels, hypertriglyceridemia) induced by childhood overweight or obesity can largely be avoided or limited by resolving overweight or obesity between childhood and adulthood. 43,44n earlier study found that children without siblings have higher blood pressure in adulthood compared with children with siblings. 14We observed that participants without siblings had higher odds of obesity in adulthood which is a known risk factor for type 2 diabetes and glycemic disorders, dyslipidemia, and hypertension. 42Moreover, compared with the children with 1 or more siblings those without siblings had higher odds for developing hypertension which is known to increase the risk for CVD and coronary heart disease mortality over a long-term follow-up in young and middleage adults with isolated systolic hypertension. 45Although a systematic review and meta-analysis found that childhood obesity is directly associated with adult systolic and diastolic blood pressures, serum triglycerides, and inversely with adult serum HDL-cholesterol concentration, 38 in the present study we observed increased odds for adult hypertension in participants without siblings and the effect was not substantially mediated by BMI in childhood or adulthood.Therefore, knowing factors associated with childhood and adulthood obesity is important.
Lower education has been shown to increase prevalence of risk behaviors such as smoking, obesity, physical inactivity, and unhealthy diet. 16Also, children from small families (ie, 1 child or 2 children families) have been shown to be more likely to graduate from high school in the US compared with large families because of intrafamilial resources diluting in larger families. 15In addition, birth order has been speculated to influence child's education level and mortality in adulthood, especially among women.However, earlier studies suggest that the effect is modest in children with less than 4 siblings. 46,47In this study, the majority of participants had 4 or less siblings and also birth order or additional adjustments for participant's years of education did not alter the results substantially.
Regardless of many studies providing arguments for the negative effects of having many siblings, 13 it is possible that siblings might be beneficial for health outcomes in adulthood because siblings provide a source of emotional support and practical aid. 48In addition, results from a recent study from Sweden based on a register data demonstrated that individuals with no siblings had an elevated risk for mortality in adulthood compared in comparison with men and women from multichild families. 13Finally, our results demonstrate the number of siblings associates with childhood overweight which is, as well as childhood obesity, associated with adverse long-term outcomes 43 and overweight and obesity in childhood/adolescence increases the risk to become overweight or obese adult. 37Indeed, those who sustain overweight or obesity from childhood to adulthood have higher risk of hypertension in adulthood compared with individuals who were overweight or obese in childhood but nonobese as adults. 43Because number of siblings is a nonmodifiable risk factor, at-risk individuals (ie, those without siblings) could benefit from an early intervention and support to tackle the issue.
The main strength of this study is the large study population with comprehensive data on lifestyle, biochemistry, and anthropometric measurements as well as on socioeconomic status starting from childhood and extending into adulthood with over 30 years follow-up.However, this study has limitations.As in all observational studies, an apparent limitation is that causality cannot be established based on our findings.However, using the existing population-based studies with extensive data on established major risk factors from childhood to adulthood is the only possibility to study the associations between the number of siblings and cardiovascular risk October 2021

ORIGINAL ARTICLES
Association between Number of Siblings and Cardiovascular Risk Factors in Childhood and in Adulthood: The Cardiovascular Risk in Young Finns Study factors and health outcomes as it is not possible to acquire a life-long trial on CVD progression in humans.Admittedly, findings from longitudinal studies might suffer from bias because of differential loss to follow-up of participants over the course of the study.However, the YFS study population has been dynamic, meaning that participants lost to followup at some point have re-joined the study in the later follow-ups. 49Thus, the cohort remains largely representative of the original population. 17In addition, we do not have information on the onset of the diseases (hypertension, type 2 diabetes, prediabetes, dyslipidemia, overweight/obesity, and metabolic syndrome), and, thus, logistic regression was used for longitudinal analysis.Moreover, given the overall relatively low number of siblings in our cohort, the generalizability of our findings is limited to populations where the typical family sizes are similar to our cohort.Equally, our findings might not apply in less developed or in poorer countries than Finland.Furthermore, data collected from selfreported questionnaires (diabetes, smoking, and physical activity index) are subject to recall bias.However, we also evaluated diabetes with objective factors such as blood biochemistry and national prescription database.Nevertheless, we acknowledge that this is subject to underestimation of smoking habits and possibly to overestimation of physical activity.Finally, we recognize the possibility of misclassification of our main exposure measure of number of siblings at baseline in 1980, especially for those of younger age.In agestratified analyses, we observed more associations among adolescents (baseline age 12-18 years).This is in line with our prior report showing that associations of childhood and adulthood risk factors improve with advancing age. 50Our findings concerning younger age groups at baseline should be interpreted with some caution, as their family size has been more likely to change after the baseline investigation.
Although we prioritized the use of these data as it maintained the largest sample size, our sensitivity analyses that used information collected from the parents of participants in the recently completed (2018-2020) 3-generation YFS, and also, combined data on the number of siblings from the baseline and 2 subsequent follow-up studies, confirmed our findings.
In our representative sample of Finnish children and adolescents, we found that those without siblings had lower physical activity levels and higher BMI and LDL-cholesterol levels in childhood, and higher odds for hypertension and obesity in adulthood than those with 1 or more siblings.Number of siblings could be a simple and useful tool for identifying children at increased risk that might benefit from early intervention and prevention aimed at improving or maintaining cardiovascular health.n

Table I .
, adjusted for sex and age).Participants without siblings had higher adjusted mean LDL-cholesterol level (3.43 mmol/L, 95% CI 3.36-3.49mmol/L) compared with those with 1 sibling (3.38 mmol/L, 95% CI 3.34-3.42mmol/L) but lower than Participant characteristics in childhood (1980) and adulthood (2011) according to the number of siblings at baseline in 1980 Association between Number of Siblings and Cardiovascular Risk Factors in Childhood and in Adulthood: The Cardiovascular Risk in Young Finns Study {Fasting plasma glucose from 5.6 mmol/L to 6.9 mmol/L.**Harmonizingdefinition included waist ³102 cm in men and ³88 cm in women, fasting plasma glucose ³5.6 mmol/L or treatment, hypertriglyceridemia ³1.7 mmol/L and HDL-cholesterol levels <1.0 mmol/L in men and <1.3 in women and blood pressure ³130/³85 mm Hg or treatment.A diagnosis requires ³3 of the 5 criteria.Adult hypercholesterolemia was assigned for participants if they had LDL-cholesterol >3 mmol/L or use of lipid-lowering medication. 2 ) (P for trend .004).Participants without siblings had the lowest physical activity index (À0.08SD,95% CI -0.16 to 0.00) than those with 1 sibling (0.06 SD, 95% CI 0.01-0.11)orthose with 2 or more siblings (À0.02SD,À0.07-0.03)(Pfor trend .01).There were no significant differences between the groups for other risk factors.ORs for adverse metrics in childhood are shown in Table

Table II .
Childhood risk factors according to the number of siblings Standardized mean difference.†N varied between 392 and 534 in participants with no siblings, 1260 and 1534 in participants with 1 sibling, and 1336-1467 in participants with 2 or more siblings.Adjusted for age and sex. *

Table III .
OR and their 95% CIs for childhood smoking, hypertension, and adverse lipid profile in childhood according to the number of siblings n/N, case number/total number.Adjusted for age and sex.*90th percentile or greater.†85th to less than the 95th percentile.‡95thpercentile or greater.ourresults reported in the main analysis (data not shown).As number of siblings was associated with both childhood BMI and physical activity index, we further examined these associations by mutually adjusting for each in the same multivariable model.The associations observed for both childhood BMI and physical activity index remained statistically significant and the effect for BMI and the effect for physical activity index remained consistent.Participants without siblings had higher BMI (18.2 kg/m 2 , 95%CI 18.0-18.4kg/m 2 ) than those with 1 sibling (17. 9 kg/m 2 , 95% CI 17.7-18.

Table IV .
ORs and their 95% CI for adulthood outcomes in 2011 according to the number of siblings at baseline(1980)