Association of parental obesity with the profile of metabolic-bariatric surgery patients: a cohort study of the German StuDoQ|MBE registry

STRENGTH AND LIMITATIONS OF THIS STUDY

  • The study was based on a large and representative sample of German bariatric patients.

  • We used routine workflow data to generate our model.

  • Data on parental obesity (PO) status as well as onset of obesity are based on patient self-reporting and physician input (recall bias), and granularity of such type of data is limited.

  • Missing or unknown PO values are most likely associated with center-specific routine workflow in data collection, this type of effect is probably completely at random.

Introduction

Obesity and associated comorbidities are a growing medical concern worldwide and far from being only a problem in highly developed countries.1 After an initial steep incline in adult obesity rates in the 70s and 80s, a likewise increase in childhood and adolescent obesity became evident in healthcare after about 20 years.1 2 Children and adolescents with obesity suffer a high risk of perpetuating obesity into adulthood and developing severe obesity.3 4 The aetiologic underpinnings of obesity are diverse and include inherited as well as environmental and lifestyle factors.2 5–7 Parental obesity (PO) has been linked to the risk of offspring obesity via numerous effects, encompassing (epi-)genetic, environmental and lifestyle factors.2 8

Due to its association with numerous comorbidities such as arterial hypertension (AHT), type 2 diabetes mellitus (T2DM), cardiovascular disease (CVD) and cancer, obesity results in increased morbidity and mortality.7 9 10 Moreover, childhood and adolescent obesity is associated with future comorbidities such as T2DM, AHT and CVD in adulthood.11 The direct and indirect effects of obesity and its associated comorbidities place a great burden on healthcare systems.12 Metabolic-bariatric surgery (MBS) is highly effective in achieving long-term weight reduction and is superior to the current nonsurgical treatment options.13 14 This translates into improvement in comorbidities, recovery of functionality and quality of life as well as improved all-cause mortality.15 16 The number of annual MBS procedures performed worldwide has greatly increased over time and can be expected to exceed one million cases by 2023.17 18

Individual patient stratification of obesity is difficult. PO contributes to clinical phenotypes and offspring obesity risk. Furthermore, its therapy has been shown to influence offspring weight trajectory, although overall evidence is scarce and somewhat contradictory.19 20 Quantifying PO as a risk factor in large cohorts is challenging as the underpinning mechanisms of obesity are diverse, granular information on important risk factors is oftentimes unavailable, and specifically, intergenerational aspects of obesity are poorly documented.2–8 19

In a recent registry-based cohort study investigating risk factors of poor response to MBS, we demonstrated an independent association of early-onset obesity (<18 years of age) and poor response to MBS.21 In this current study, we investigated how PO is associated with pre-surgical clinical patient profiles and how this may underpin poor response. Specifically, we hypothesised that (1) PO is associated with early onset (<18 years) of disease in MBS patients, (2) there is an independent association of PO and disease duration as defined by age at onset until age at MBS, (3) PO is associated with body mass index (BMI) at the time of MBS and (4) the results differ with the degree of PO (none, one or both parents).

Methods

This large retrospective cohort study was conducted with data from the German Society of General and Visceral Surgery registry for metabolic and bariatric diseases (DGAV-StuDoQ|MBE), a database focusing on MBS patients in participating German hospitals. The database includes pre-surgical sociodemographic and clinical information as well as follow-up data (3 months and from 1 year after MBS annually). All patients provided written informed consent before registry entry, and each participating German hospital has obtained institutional review board approval before participation. Data extraction was performed after formal application and a positive vote of the registry steering committee.

Patient population

All patients who underwent MBS between February 2015 and August 2020 and with data on age at surgery, PO status as a patient-reported variable (‘Does overweight or obesity exist in direct family members (mother/father/siblings/children)?’, answer: yes/no/unknown/does not apply), age at the onset of obesity, preoperative body weight and at least 1 year of follow-up after MBS were included. Patients with non-primary or combined procedures, less common procedures (non-sleeve-gastrectomy/non-Roux-en-Y gastric bypass/non-one-anastomosis gastric bypass) and incomplete information on PO were excluded (online supplemental figure 1). In addition to sociodemographic as well as surgery-related data, information on the prevalence of AHT and T2DM was available for most patients. The status of PO was transformed into a categorial variable (variable values: ‘obesity in no parent’, ‘obesity in one parent’, ‘obesity in both parents’). For subgroup analyses, ‘obesity in one parent’ was divided into ‘paternal’ and ‘maternal’.

Supplemental material

Statistics

Continuous variables are presented as mean and SD. Categorical variables are presented as frequency and percentage. We performed a separate descriptive cohort analysis regarding PO categories and the onset of obesity, with early onset (<18 years) vs late onset (≥18 years).

The association between PO category and early onset of the disease was assessed in a multivariate logistic regression model, with the output presented as ORs. Similarly, the association of onset of obesity and AHT as well as T2DM was each evaluated using a multivariate logistic regression model. The goodness-of-fit was tested with the Hosmer–Lemeshow test.

The association of PO category and duration of obesity was modelled using a multivariate linear regression model, with the output presented as regression coefficient. The disease duration was defined as the time interval between ‘age at MBS’ minus ‘age at onset’. Multicollinearity was estimated with the variance inflation factor. The effects of the PO category on the onset as well as duration of obesity on BMI at MBS were also assessed with a multivariate linear regression model. All statistical analyses were performed using Stata version 17.0 (StataCorp LLC, Texas, USA).

Public and patient involvement

None.

Results

Patient characteristics

Table 1 lists the patient demographics as well as the prevalence of AHT and T2DM according to the PO category. A total of 11 891 patients were included, of which 4679 (39.4%) had one-sided PO and 3761 (31.6%) two-sided PO (table 1). The mean±SD age at surgery was 43.5±11.5, mean±SD BMI at the time of MBS was 49.1±7.7 kg/m2 and 9922 (74.2%) patients were female. On a descriptive level, increasing degrees of PO (none, one-sided, two-sided) were associated with female sex, younger age at onset as well as at MBS (table 1), longer pre-surgical duration of obesity and higher BMI. The prevalence of the sole paternal PO was considerably lower than any other predisposition value and on a descriptive level correlated with a lower age at onset, as well as age at surgery and a higher prevalence of male sex. Increasing degrees of PO were associated with a lower mean prevalence of T2DM (online supplemental table 1). No clinically relevant differences in the prevalence of AHT were seen. Strict maternal or paternal obesity did not display any difference in the prevalence of AHT or T2DM. Furthermore, no clinically relevant differences in performed surgical procedure types could be observed.

Table 1

Metabolic-bariatric surgery patient characteristics according to parental obesity categories (none, one parent or both parents)

Online supplemental table 2 provides a patient demographics characterised by early-onset obesity (<18 years) vs late onset (≥18 years). Early onset was associated with younger age and higher BMI at MBS (online supplemental table 2). The prevalence of paternal and maternal obesity was also higher in the early-onset group. Despite considerably longer duration of obesity, as well as higher BMI at surgery, the prevalence of T2D and AHT was lower in early-onset patients. This coincides with a higher haemoglobin A1C (HbA1c) at time of surgery in late-onset patients. No clinically relevant differences in performed surgical procedure types were observed.

Regression analysis

We found one-sided and two-sided PO to be independently associated with early-onset obesity (<18 years) (OR 1.61, [95% CI, 1.47 to 1.76], p<0.001 and OR 2.45, [95% CI, 2.22 to 2.71], p<0.001, respectively) (table 2A and figure 1 and online supplemental table 3A). In this multivariate logistic regression model, male sex was associated with a decreased OR for early-onset obesity (OR 0.81 [95% CI, 0.74 to 0.88], p<0.001).

Table 2

Outcomes of multivariate models testing for (A) early-onset, (B) pre-surgical duration of obesity and (C) pre-surgical BMI

Figure 1
Figure 1

Outcome of multivariate models testing for (A) risk factors of early-onset obesity (<18 years), (B) predictors of pre-surgical duration of obesity, (C) predictors of pre-surgical body mass index (BMI). The degree of parental obesity (PO) (one sided, two sided) is associated with an increase of duration of obesity as well as the risk of early onse. The age at onset, disease duration and two-sided PO are independently associated with BMI at time of surgery.

We observed both one-sided and two-sided PO to correlate with a highly significant increase of disease duration (regression coefficient 2.39 [95% CI, 1.93 to 2.83], p<0.001 and regression coefficient 4.27, [95% CI, 3.80 to 4.75], p<0.001, respectively) (table 2B and figure 1 and online supplemental table 3B). Moreover, strict paternal obesity had a stronger effect on duration of obesity compared with strict maternal obesity (regression coefficient: strict paternal 2.58, [95% CI, 1.97 to 3.18], p<0.001 and regression coefficient: strict maternal 2.28, [95% CI, 1.78 to 2.77], p<0.001). Each additional year of age at time of surgery correlated with an increase of disease duration (regression coefficient 0.61, [95% CI, 0.60 to 0.63], p<0.001). In this model, male sex independently correlated with a significant decrease of disease duration (regression coefficient −0.87, [95% CI, −1.29 to −0.44], p<0.001).

A higher age at onset and longer duration of obesity both correlated negatively with the BMI at time of MBS (regression coefficient 0.13, [95% CI, −0.14 to −0.11], p<0.001 and regression coefficient −0.05, [95% CI, −0.07-(-)0.04], p<0.001, respectively) (table 2C and figure 1 and online supplemental table 3C). In this linear regression model, only two-sided PO was associated with a significant increase in BMI at the time of surgery (regression coefficient – two parent PO 0.49, [95% CI, 0.14 to 0.85], p=0.006 and regression coefficient – one parent PO −0.19 [95% CI, −0.53–0.14], p=0.26).

Finally, we analysed the association of early onset with AHT and T2DM, adjusting for sex, age at surgery, and BMI in multivariate logistic regression models (table 3 and figure 2 and online supplemental table 4). Early-onset obesity was independently associated with an OR of 1.20 (95% CI, 1.05 to 1.30, p=0.004) for AHT. We found no independent association of early onset, age at onset or pre-surgical duration of obesity and T2DM prevalence. Investigating the effect of early onset and duration of obesity on the HbA1c value at the time of surgery using a multivariate linear regression model, we did not find a significant correlation. Age and BMI at surgery as well as male sex were positively correlated with HbA1c value at the time of MBS.

Table 3

Outcomes of multivariate models testing for risk factors for (A) arterial hypertension, (B) type 2 diabetes mellitus and (C) haemoglobin A1C

Figure 2
Figure 2

Outcome of multivariate models testing for: (A) risk factors of arterial hypertension at MBS, (B) risk factors of type-2 diabetes mellitus at MBS, (C) predictors of pre-surgical haemoglobin A1C (HbA1c) value (%) adjusting for age at surgery, sex and BMI. Early-onset obesity (<18 years) carried an OR of 1.17 for the prevalence of arterial hypertension at the time of surgery, adjusting for age and BMI at surgery, as well as sex. No statistically significant influence of early onset of obesity on the prevalence of type 2 diabetes at the time of surgery was seen. Early-onset and duration of obesity did not have a significant influence on the HbA1c-value at the time of surgery, but age and BMI at MBS as well as male sex. AHT, arterial hypertension; BMI, body mass index; MBS, metabolic-bariatric surgery; T2DM, type-2 diabetes mellitus.

Discussion

In this large registry-based cohort study of MBS patients, we investigated how PO associates with pre-surgical patient risk factors. PO was found to be associated with early onset as well as longer disease duration despite significantly younger mean age at surgery. Obesity in both parents further increased these risks and was additionally associated with higher pre-surgical BMI. Early onset but not PO was associated with an increased risk of AHT at time of MBS, and no significant association was seen for T2DM or HbA1c values.

The present literature on the onset of obesity is equivocal. In line with previous publications, early-onset obesity was associated with a younger age at MBS despite significantly longer duration of obesity.22 Also, longitudinal data from the USA show a clear link between maternal obesity and onset of offspring obesity across ethnic groups.23 Yet, a longitudinal study from Sweden did not find a correlation between parental predisposition and onset of obesity.24 However, the Swedish study comprised less than 2% of cases of the US study and thus was potentially underpowered to detect small to moderate effect sizes. Similar to our findings, previous data demonstrated earlier onset of overweight to be associated with increased BMI as well as a steeper BMI trajectory in the general population.25 Also, a study focusing on MBS patients illustrated early-onset obesity, defined as <20 years, to carry a 1.84 times risk of a BMI of >40 kg/m2 at MBS.22 However, data on the impact of duration of obesity as well as degree of PO on the BMI at MBS is lacking.

In our cohort, paternal obesity had a greater impact on early onset and duration of obesity than maternal obesity. Current literature is inconclusive on this aspect. A cross-sectional study analysing parental genetic, lifestyle and social parameters and their impact on offspring overweight illustrated only maternal BMI as being significantly associated with offspring overweight,8 while the Fels longitudinal study demonstrated a stronger association of maternal BMI with infant and early childhood weight, which in turn affects adult obesity.4 26 However, amidst diverse (epi)genetic, lifestyle and environmental factors, specific paternal mechanisms for intergenerational obesity and discrete phenotype heritability have been repeatedly demonstrated.5 6 27–29 Notably, a systematic review and meta-analysis showed a stronger association of paternal than maternal obesity on offspring obesity.30 Whether there is a specific paternal risk for obesity and how this might be associated with pre-surgical MBS profiles and postoperative outcomes remains unclear.

In a recent study, we found early onset to be associated with poor 1 year response to surgery (total weight loss<20%).19 Combined with our current results as well as the already demonstrated positive effects of maternal MBS on offspring health, including improved factors of metabolic function, the timeliness of intervention for overweight and obesity is emphasised.31–33

In addition, we found a clear association between early-onset obesity (<18 years) and the prevalence of AHT at the time of surgery, but not for T2DM. Current literature on this aspect is equivocal. A recent study of MBS patients demonstrated early-onset obesity to be associated with a lower prevalence of AHT and T2DM.22 Yet, one study focusing on patients with overweight (BMI>25 kg/m2) demonstrated younger age at the onset of overweight onset to be associated with a significantly increased risk of AHT, while another study illustrated a positive association between early-onset as well as duration of obesity and the risk for T2DM.34 35 Crucially, the last two studies involved mostly patients with overweight or moderate obesity and did not focus on MBS patients. Thus, despite illustrating an association between the onset of obesity and AHT and T2DM in the general population, the significance in the subpopulation of MBS patients, and especially those at higher BMI values such as in our cohort, remains unclear.

Future research should focus on the association between PO and its underlying mechanisms for offspring obesity and onset of obesity as well as its treatment. This could assist in a more personalised approach as well as pre-surgical prediction of likely treatment outcome.

Limitations

One limitation of our study is the possible selection bias. However, missing or unknown PO values are most likely associated with center-specific routine workflow in data collection. This type of effect is probably completely at random. Moreover, the data on PO as well as onset of obesity is based on patient self-reporting and physician input (recall bias). Granularity of such type of data is limited. Strikingly, compared with most studies investigating age, onset, duration and BMI-specific effects, the median BMI in our cohort is significantly higher. Results may differ compared with lower BMI ranges. Because no information on ethnic and sociodemographic status is reported in this registry, the generalizability of our findings to other populations may be reduced. Yet, our study population is a large, representative cohort of German MBS patients and seems comparable to other published data in terms of other demographic factors as well as performed surgical procedures.18

Conclusion

PO improves pre-surgical stratification of MBS patients and is linked to known risk factors for poor response to surgery. Future studies should focus on the effects of parental risk factors for obesity and their impact on treatment outcome in offspring of MBS patients and how parental weight loss before conception might affect offspring response to therapy.

Data availability statement

Data may be obtained from a third party and are not publicly available. Original data can be provided on request and after a positive vote by the StuDoQ steering committee through the StuDoQ|MBE registry.

Ethics statements

Patient consent for publication

Ethics approval

All patients provided written informed consent prior to registry entry, and each participating German hospital has obtained institutional review board approval before participation. Data extraction and analysis were cleared by the StuDoQ steering committee and performed in accordance with StuDoQ data protection guidelines. Due to the retrospective nature and analysis of anonymized registry data, the ethics committee of the Albert-Ludwigs-University Freiburg deemed a further ethics vote unnecessary.

This post was originally published on https://bmjopen.bmj.com