Nationwide real-world assessment of the economic impact of obesity in Spanish healthcare settings (ECC-OS)

Abstract

Objective

This study aimed to estimate the direct healthcare costs associated with obesity in primary care and hospital settings in the adult population in Spain, and to describe the sociodemographic and clinical characteristics of individuals diagnosed with obesity.

Methods

Observational, retrospective study using two unlinked nationwide public databases in Spain: the Primary Care Clinical Database (BDCAP) and the Specialized Care Activity Registry-Minimum Basic Data Set (RAE-CMBD), which collects data for hospitalizations. Adults aged ≥20 with obesity were identified based on diagnosis codes in medical records or discharge reports. Data included demographic and clinical characteristics, healthcare resource utilization, and characteristics of hospitalizations. Data for the year 2022 were analyzed independently for each database, using descriptive statistics.

Results

Of 34,338,405 adults aged ≥20 identified in 2022 in the BDCAP, 10.3% had body mass index (BMI) records and, of these, 37.1% presented a BMI ≥30 kg/m². 3,285,946 (9.6%) had a diagnosis code for obesity. These had more visits to primary care in 2022 (22.2 versus 11.6), diagnostic or therapeutic procedures (20.5 versus 12.4 per person), or prescriptions for chronic medication than those without a diagnosis code. Annual cost in primary care for 2022 were €1656 per person diagnosed with obesity, versus €851 per person without a diagnosis. A total of 15,969 admissions to hospitals due to obesity were registered in the RAE-CMBD, with a cost of €79.8 million.

Conclusions

Obesity is under-recorded in public databases in Spain. Individuals diagnosed with obesity have higher estimated per-person costs than those without an obesity diagnosis.

Resumen

Objetivo

Estimar los costes sanitarios directos de la obesidad en atención primaria y hospitalaria en España, y describir las características sociodemográficas y clínicas de las personas diagnosticadas con obesidad.

Métodos

Estudio observacional retrospectivo usando dos bases de datos públicas nacionales no vinculadas: la Base de Datos Clínicos de Atención Primaria (BDCAP) y el Registro de Actividad de Atención Especializada-Conjunto Mínimo Básico de Datos (RAE-CMBD), para hospitalizaciones. Se identificó a los adultos con obesidad mediante códigos diagnósticos en historias clínicas o informes de alta. Se analizaron descriptivamente para cada base de datos las características demográficas y clínicas, la utilización de recursos sanitarios y el detalle de las hospitalizaciones del año 2022.

Resultados

De 34.338.405 adultos ≥20 años identificados en 2022 en BDCAP, el 10,3% contaba con registros del índice de masa corporal (IMC); de estos, el 37,1% presentó un IMC ≥ 30 kg/m². Un total de 3.285.946 (9,6%) tenían código de diagnóstico de obesidad. Las personas con diagnóstico de obesidad tuvieron más visitas a atención primaria (22,2 versus 11,6), procedimientos diagnósticos o terapéuticos (20,5 versus 12,4 por persona) o medicación crónica que aquellas sin él. El coste anual en atención primaria en 2022 fue de 1.656€/persona con diagnóstico de obesidad, frente a 851€/persona sin diagnóstico. Se registraron 15.969 hospitalizaciones por obesidad en RAE-CMBD, con un coste de 79,8 millones de euros.

Conclusiones

La obesidad está infra-registrada en las bases de datos públicas españolas. El coste sanitario anual por persona fue mayor en las personas con diagnóstico de obesidad, en comparación con aquellas sin él.

Introduction

The World Health Organization (WHO) has declared obesity a serious and growing public health challenge globally and a major determinant of disability and death.^1,2 Overweight and obesity account for 7% of total years lived with disability and for more than 1.2 million deaths across Europe every year.³ Obesity is linked to the onset of more than 200 chronic conditions, encompassing metabolic, biomechanical, and psychosocial disorders.^4,5 For many of these obesity-related diseases (ORDs), the presence of obesity is associated with worse outcomes. Consequently, the overall societal burden of obesity is considered to be considerable and is projected to increase further in the coming years.^6,7 The duration of adiposity and obesity is strongly associated with the development of ORD later in life. Therefore, early intervention may yield substantial health benefits, including a significant reduction in cardiovascular risk.^8,9 Primary care physicians play a pivotal role in the diagnosis and management of obesity, as they are often the first point of contact within the healthcare system for the majority of patients.¹⁰ However, although ORDs are frequently addressed and treated, obesity itself often remains unrecognized and undertreated, particularly within primary care settings.¹¹ The clinical and economic burden of obesity on healthcare systems has been extensively studied and remains a focus of ongoing research.^12–16 Across all healthcare levels, studies have found notable increases in healthcare costs for individuals with obesity compared with individuals without obesity.^17–19 The economic impact of obesity could be alleviated by weight management oriented to limiting the effects of ORD development.²⁰

Estimates of healthcare resource use and costs associated with obesity in both primary care and hospital settings in Spain indicate a considerable economic burden.^21,22 Nonetheless, significant gaps remain in understanding the demographic and clinical profiles of adults diagnosed with obesity and the specific economic impact within the Spanish healthcare system. Data in electronic medical records within large, population-based healthcare databases can offer valuable insights into the impact of chronic diseases on the healthcare system. In Spain, the Spanish Ministry of Health administers the Primary Care Clinical Database (BDCAP from its Spanish acronym) and the Specialized Care Activity Registry-Minimum Basic Data Set (RAE-CMBD from its Spanish acronym) for primary care consultations and services, and for hospital admissions, respectively.^23,24 Despite containing real-world clinical data, these databases have not yet been used to assess the impact of obesity in healthcare settings. The primary objective of this study was to estimate the direct healthcare costs associated with obesity in both primary care and hospital settings in the adult population in Spain, using the BDCAP and RAE-CMBD datasets. Additionally, the study aimed to characterize the sociodemographic and clinical characteristics of individuals diagnosed with obesity in primary care and to describe the attributes of hospitalization episodes in which obesity was the principal diagnosis.

Materials and methods

This is an observational, descriptive and retrospective study using secondary data available from two independent, unlinked databases in Spain, the BDCAP and RAE-CMBD. This nationwide study covered data for the entire year of 2022 (1 January to 31 December). As the study included only aggregated data, patients were not identified or contacted. The protocol for this study was approved by the Ethics Committee of the Hospital Clínic de Barcelona, Spain. The study was conducted in accordance with the protocol and ethical principles that have their origin in the Declaration of Helsinki and that are consistent with applicable ICH Good Clinical Practice (GCP) Guidelines, Good Pharmacoepidemiology Practices (GPPs), and applicable laws and regulations of Spain, as appropriate.

Data sources

The BDCAP belongs to the Spanish National Health System (NHS) and was developed for statistical and research purposes.²³ It collects standardized clinical information from primary care using a large random sample of approximately 25% of the population's medical records, representative of each Spanish Autonomous Community. In the 2022 BDCAP update, the database collected data from a sample of 12.7 million Spanish NHS primary care users. The database collects information on health problems, consultations with hospital specialists, procedures (imaging, laboratory, and other diagnostic and therapeutic procedures), parameters, prescribed medications dispensed at pharmacies, and visits to primary care. Each category is contained in separate and unlinked modules, which do not allow for combined analysis across modules, such as assessing medication use (medication module) by BMI category (parameters module), except for health problems. Health problems are identified using diagnosis codes based on the International Classification of Primary Care, 2nd edition (ICPC-2), or mapped from the International Classification of Diseases (ICD-9 or ICD-10) to ICPC-2. For procedures, several systems are used, normalized to SNOMED-CT.

The RAE-CMBD is the main hospitalization database for treated morbidity and for the patient care process in hospitals in Spain, collecting sociodemographic (sex, age), clinical and administrative patient data from discharge reports (admission and discharge date, main diagnosis and a maximum of 20 secondary diagnoses, up to 20 therapeutic or diagnostic procedures, discharge destination, and death during admission).²⁴ This database also collects information on the type of hospitalization (including in-hospital stays, emergency hospitalization, at-home hospitalization, and other outpatient modalities). The RAE-CMBD encodes disease information using the International Classification of Diseases, 10th revision (ICD-10). In 2020, the CMBD included almost 20 million contacts (i.e., units of activity, such as hospital discharge, major outpatient surgery, etc.), 88.2% of them from public hospitals and 11.9% from private centers.

Population

An initial assessment was performed on the complete adult cohort (individuals over 20 years of age) within BDCAP to characterize the database and evaluate the availability of obesity-related data. This evaluation examined the presence of obesity diagnostic codes, the availability and values of recorded body mass index (BMI), and the alignment between BMI measurements and diagnostic codes. The BDCAP population for the analysis (sociodemographic characteristics, healthcare utilization, costs) included adults with and without a documented diagnosis of obesity. Patients with obesity were identified based on the presence of an active obesity diagnosis code in their electronic medical records, including both newly diagnosed cases in the year of analysis and historical occurrences. The population without obesity was estimated by subtracting the number of individuals with obesity from the total number of people with documented health problems in the database. The list of diagnosis codes is shown in Supplementary Table 1. For the RAE-CMBD, hospitalizations with an obesity diagnosis listed in the discharge report as the main diagnosis associated with the hospitalization episode were analyzed.

Variables and measurements

For BDCAP, the variables assessed included sociodemographic characteristics (gender, age, income level, and employment status), clinical characteristics (BMI, comorbidities, HbA1c levels, cholesterol and triglyceride levels, eGFR, blood pressure, and medication type), and four distinct categories of resource use (visits to primary care, medical referrals, medical procedures, and medications). The BDCAP defines ‘visit’ as any contact between the user and any of the health center's professionals, and, if applicable, several contacts may be counted on the same day. Because BDCAP does not capture the reasons for visits, referrals, procedures, or medications, it is not possible to link healthcare resource use to a documented diagnosis. Therefore, resource utilization is reported for individuals with a given diagnosis but cannot be definitively attributed to that diagnosis.

For the RAE-CMBD, hospitalizations of people with obesity recorded as a primary diagnosis were selected and stratified by type of hospitalization (in-hospital hospitalization, hospitalization at home, emergencies, and other outpatient modalities). For in-hospital admissions, the number of episodes, length of hospital stays, number of ICU admissions, and length of ICU stays were recorded.

Costs were estimated by multiplying the number of units of each healthcare resource used by the corresponding unitary cost. Unit costs were obtained from the eSalud database (http://oblikue.com/bddcostes/), a specific source of healthcare costs in Spain. Costs were updated with the consumer price index to 2024 euros. For primary care, only visits to the PCP and referrals to specialists were considered evaluable for costs. The medical procedures collected all activities related to the process of addressing a health problem, and includes categories such as imaging tests, therapeutic procedures, and laboratory tests. Because this is available in the form of broad categories, it is not possible to assign a specific unit cost to directly translate their use into cost estimates. The BDCAP collects data on the prescribed medication dispensed to the individuals in pharmacies grouped at Anatomical Therapeutic Chemical (ATC) level-4 (pharmacological or therapeutic class), and their specific costs could not be calculated.

Statistical methods

The data from each database were analyzed and reported separately for each database. Since patient level data is not accessible in the BDCAP and the CMBD, the summary statistics for these databases comprised absolute frequency (N) and relative frequency (%) for categorical variables and means for continuous variables, when available. Analysis was conducted using SAS Enterprise Guide Software, version 7.14.

Results

Recorded prevalence of obesity in the primary care database

A total of 34,338,405 adults aged ≥20 with a health problem were identified in the BDCAP in 2022 (Fig. 1). Only 10.3% of them had BMI records in the year of analysis, and of these, 37.1% presented a BMI ≥30 kg/m² (3.8% with respect to the total number of adults in the database). Among patients with obesity according to a BMI record, 45.7% did not have a diagnosis code for obesity in their medical records. A total of 3,285,946 adults aged ≥20 had a diagnosis code for obesity, accounting for 9.6% of the population with a health problem in 2022. Among those with a diagnosis code for obesity, only 27.2% had a BMI record in the year of analysis.

Figure 1.

Recording of obesity in the BDCAP database in Spain.

The percentage of people with a record of BMI in primary care ranged from 4.9% in Madrid to 23.5% in Catalonia (Table S2). Fig. 2A shows the proportion of patients with a BMI ≥30 kg/m² (obesity) and <30 kg/m² (not obesity) in each Autonomous Community and Fig. 2B the proportion of individuals with a diagnosis code for obesity among those with BMI ≥30 kg/m². Castilla la Mancha presented the largest difference between the rate of obesity and diagnosis (42.7% with BMI ≥30 kg/m²; of these, 1.5% with diagnosis), and Catalonia the smallest (32.6% with BMI ≥30 kg/m²; of these, 82.7% with diagnosis). The distribution of individuals diagnosed with obesity in primary care varied significantly across regions in Spain, ranging from 0.2% in Castilla-La Mancha to 18.5% in Catalonia (Fig. 2C).

Figure 2.

Distribution of obesity in the BDCAP among Autonomous Communities in Spain. (A) Distribution of people attending primary care with BMI ≥30 kg/m² (orange) or <30 kg/m² (green), among those with BMI records (these graphs were made from data from the parameters module of BDCAP, which did not include BMI data for Asturias and Galicia). (B) Percentage of people with a diagnosis code for obesity among those with BMI ≥30 kg/m². (C) Rate of diagnosis of obesity (individuals with a diagnosis code) in each region of Spain (data from the health problems module of BDCAP). This percentage was calculated as the proportion of individuals aged >20 years with a diagnosis of obesity in primary care over the total number of individuals aged >20 years attending primary care. BDCAP, Primary Care Clinical Database; BMI, body mass index.

People with obesity in primary care

The demographic and clinical characteristics of people diagnosed with obesity in primary care (N = 3,285,946) are shown in Table 1. In this population, women were overrepresented with respect to men (57.6% versus 42.5%). Among those with a BMI record (N = 894,152; 27.2%), 67.5% had a BMI of 30–39 kg/m², and 12.6% had a BMI ≥40 kg/m². Notably, 19.9% of individuals diagnosed with obesity had a BMI <30 kg/m² at the time of the study in 2022, suggesting successful weight loss. A substantial proportion of individuals identified in the BDCAP were pensioners (40.6%), and the majority (70.7%) had an annual income of less than 18,000 €.

Table 1.

Demographic and clinical characteristics of people with a diagnosed with obesity in the BDCAP.

Variable	People aged ≥20 with a diagnosis code for obesity(N = 3,285,946)
Sex
Men	1,394,746 (42.5)
Women	1,891,200 (57.6)
Age (years)
20–34	323,831 (9.9)
35–64	1,623,071 (49.4)
≥65	1,339,044 (40.8)
BMI (kg/m^2)^a
<25^b	17,260 (1.9)
25–29^b	160,908 (18.0)
30–39	603,397 (67.5)
≥40	112,533 (12.6)
Income level (€/year)
Very low	783,502 (23.8)
<18.000	1,541,556 (46.9)
18,000–99,999	939,230 (28.6)
≥100,000	15,910 (0.5)
Unknown	5,748 (0.2)
Employment status
Active	1,128,178 (34.3)
Unemployed/non-active	607,180 (18.5)
Pensioners	1,334,131 (40.6)
Other^c	216,457 (6.6)
Main comorbidities^d
K86 – uncomplicated hypertension	1,633,568 (49.7)
T93 – lipid metabolism disorders	1,543,911 (47.0)
T90 – non-insulin-dependent diabetes mellitus	808,624 (24.6)
D82 – dental/gum diseases	750,435 (22.8)
A23 – unspecified risk factor	714,701 (21.8)
L87 – unspecified bursitis/tendinitis/synovitis	708,372 (21.6)
A77 – other unspecified viral diseases	698,930 (21.3)
P74 – anxiety disorders/anxiety state	676,819 (20.6)
P06 – sleep disorders	671,180 (20.4)
L86 – lumbar/thoracic radiating pain syndromes	584,406 (17.8)
Most common drug classes used^b
Analgesics and antipyretics	1,652,080 (56.3)
Drugs for peptic ulcer and gastro-esophageal reflux disease	1,311,515 (44.7)
Non-steroid anti-inflammatory and antirheumatic drugs	1,204,684 (41.0)
Lipid-modifying agents	1,064,946 (36.3)
Anxiolytics	793,779 (27.0)
Opioids	764,588 (26.0)
Blood glucose lowering drugs, excluding insulins	759,476 (25.9)
Antithrombotic agents	741,440 (25.3)
Antidepressants	720,590 (24.5)
Vitamin A and D, and combinations of the two	576,658 (19.6)

All values are given as N (%).

Abbreviations: BDCAP, Primary Care Clinical Database; BMI, body mass index.

a

In patients with a BMI record, N = 894,152.

b

These are BMI values for the year 2022, and the diagnosis of obesity may be current or past. Thus, there may be people with the obesity diagnosis code who lost weight and who, in the year of analysis, have a BMI <30.

c

Individuals who cannot be assigned any of the previous situations, such as mutual societies.

d

This includes the 10 most common comorbidities. Please see a full list in Table S3.

A high proportion of individuals diagnosed with obesity had comorbidities (Table 1 and Table S3). The most common conditions were hypertension (49.7%), lipid metabolism disorders (47.0%), and diabetes mellitus (24.6%). When grouped by disease categories, the most prevalent were cardiovascular disease (63.9%), hypertension (51.4%), biomechanical (48.6%), and lipid metabolism disorders (47.0%) (Table S4). Psychiatric disorders were observed in 28.3% of individuals with obesity, as reflected also in the prescription rates of anxiolytics and antidepressants (27.0% and 24.5%, respectively).

In individuals diagnosed with obesity who had available records of glycemic control, serum lipids, kidney function, and blood pressure, the data indicate a high proportion of patients with poor control of these parameters (Supplementary Fig. S1). For example, in individuals with a record of total cholesterol levels (29.9% of those with obesity), 34.4% presented a value ≥200 mg/dL. Additionally, among those with recorded systolic blood pressure values (38.1%), 61.2% had >130 mmHg, and in those with eGFR records (15.3%), 63.4% had values <80 mL/min/1.73 m².

Use of healthcare resources and costs in primary care

BDCAP data revealed that individuals diagnosed with obesity had twice as many primary care visits as those without the diagnosis (22.2 visits versus 11.6 in 2022) (Table 2). Additionally, individuals diagnosed with obesity underwent more diagnostic or therapeutic procedures, 20.5 per person, compared to 12.4 per person among those without an obesity diagnosis. Regarding medication use, a higher proportion of individuals with obesity had prescriptions for chronic medication compared to non-chronic medication. The use of chronic medication was notably higher among individuals with obesity compared to those without obesity (68.5% versus 42.2%). This difference was more evident in people with multiple (>3) chronic prescriptions (Fig. S2).

Table 2.

Use of healthcare resources and costs in primary care (BDCAP) in Spain in 2022.

Variable	People aged ≥20 years diagnosed with obesity(N = 3,285,946)	People aged ≥20 years without an obesity diagnosis code(N = 31,052,459)
Primary care visits
Total number of visits	68,550,302	328,198,251
Individuals with a visit, N (%)	3,084,017 (93.9)	28,308,988 (91.2)
Visits/person	22.2	11.6
Cost (unit cost: 76.15 €)	5,220 M€	24,992 M€
Referrals to the specialist^a
Total number of referrals	2,070,316	13,274,023
Individuals with referral, N (%)	1,242,901 (37.8)	8,830,658 (28.4)
Referrals/person	1.7	1.5
Cost (unit cost: 108.12 €)	224 M€	1435 M€
Annual total cost for 2022 (visits and referrals)^b	5444 M€	26,427 M€
Annual total cost/person (visits and referrals)^c	1655 €	851 €
Procedures^d
Total number of procedures	51,628,406	251,003,880
Individuals with procedure, N (%)	2,522,491 (76.8)	20,241,668 (65.2)
Procedures/person	20.5	12.4
Medications,^eN (%)
Individuals with medication	2,936,021 (89.4)	24,324,105 (78.3)
Individuals with chronic medication	2,252,265 (68.5)	13,102,848 (42.2)
Individuals with non-chronic medication^f	683,756 (20.8)	11,221,257 (34.1)

Abbreviations: BDCAP, Primary Care Clinical Database; M€, million euros.

a

A list of specialties and percentages is shown in Table S5.

b

Annual total cost in 2022, adjusted to 2024 euros.

c

Calculated over the total number of people with or without a diagnosis of obesity.

d

Activities related to the care process for health problems, including imaging tests, diagnostic procedures, therapeutic procedures, and laboratory tests.

e

Prescribed medication dispensed in pharmacies to the individuals.

f

Chronic medications were those grouped in level 4 of the ATC classification that accumulated dispensations for at least 6 months throughout the year.

In 2022, the total estimated cost of primary care visits and specialist referrals was €5444 million for individuals diagnosed with obesity and €26,427 million for those without an obesity diagnosis (Table 2 and Fig. S3). This represents €1656 per person diagnosed with obesity, versus €851 per person without an obesity diagnosis. The cost associated with primary care visits and referrals for individuals diagnosed with obesity accounted for 17.1% of the total primary care expenditure (Fig. 2). These estimates did not account for any diagnostic or therapeutic procedures or medication costs.

Healthcare resource utilization and costs in hospital care

A total of 15,969 admissions to hospitals of adults aged ≥20 with obesity as the main diagnosis were identified in the CMBD in 2022 (Table 3). Most hospitalizations (76.1%) were in-hospital hospitalization episodes, 4.3% were emergency, and 0.2% were home hospitalizations (Table 4). Of these admissions, 71.4% involved female patients, and 77.8% were for individuals aged 35–64. Most of these admissions (66.7%) were coded as E66.01, indicating morbid obesity due to excess calories. According to the type of procedure related to the admission, 70.4% of in-hospital hospitalization episodes were surgical, and 88.2% of cases were ‘scheduled’. In 2022, 12.6% of in-hospital stays involved admission to an intensive care unit (ICU), and 314 individuals diagnosed with obesity (2.0%) died during hospitalization.

Table 3.

Sociodemographic characteristics and hospitalization details for episodes with obesity as the primary diagnosis in 2022 in the CMBD.

Variable	Individuals aged ≥20 admitted to hospital with obesity as a primary diagnosis(N = 15,969)
Sex
Male	4,554 (28.5)
Female	11,401 (71.4)
Not specified	14 (0.1)
Age (years)
20–34	2,550 (15.7)
35–64	12,419 (77.8)
≥65	999 (6.3)
Diagnosis code (ICD-10)
E66.01 – morbid (severe) obesity due to excess calories	10,644 (66.7)
E66.09 – other obesity due to excess calories	527 (3.3)
E66.1 – drug-induced obesity	20 (0.1)
E66.2 – morbid (severe) obesity with alveolar hypoventilation	688 (4.3)
E66.8 – other obesity type	398 (2.5)
E66.9 – obesity, unspecified	3,692 (23.1)
Type of procedure
Medical	4,729 (29.6)
Surgical	11,240 (70.4)
Type of admission
Urgent	1,463 (9.2)
Scheduled	14,083 (88.2)
Unknown	423 (2.7)
Funding scheme
Public	8,712 (54.6)
Private	7,257 (45.4)
Mortality in hospital	314 (2.0)

Values are given as N (%).

Abbreviations: CMBD, Conjunto Mínimo Básico de Datos; ICD-10, International Classification of Diseases, 10th revision.

Table 4.

Use of healthcare resources and costs for people admitted to hospital with a main diagnosis of obesity in 2022 (data from CMBD).

Variable	Hospital admissionsN (%)	Mean length of hospital stays, days	Mean length of ICU stays, days	Mean unit cost^a€	Total costs€ (%)
General hospitalization	12,151 (76.1)				75,582,664 (94.7)
With ICU	1526 (12.6)	5.4	1.6	7277	11,105,364 (13.9)
No ICU	10,502 (86.4)	2.9		6068	63,732,274 (79.8)
Invalid	123 (1.0)	3.3		6057	745,089 (0.9)
Emergency	683 (4.3)			257	175,531 (0.2)
Home hospitalization	30 (0.2)			204	6149 (0.008)
Other	3105 (19.4)			1309	4,067,263 (5.1)
Total	15,969	–	–	–	79,831,606

Abbreviation: ICU, intensive care unit.

a

In-hospital hospitalization and outpatient procedure unit cost extracted from CMBD and updated with price consumer index to 2024 euros. At-home hospitalization and emergency visits extracted from eSalud.

General hospitalizations without ICU use accounted for 79.8% of total hospitalization costs, while hospitalizations involving ICU, averaging 1.6 days, represented 13.9%; other types of hospitalizations made up the remaining 6.3%. The total estimated costs for these patients were €79.8 million (Table 4).

Discussion

This study revealed significant discrepancies between obesity rates based on BMI and the rates of coded obesity diagnoses across many Autonomous Communities in Spain, suggesting undercoding of the condition and varying attitudes among primary care physicians toward recognizing and documenting obesity. Also, the study suggested that individuals diagnosed with obesity had higher rates of visits to primary care and higher rates of referrals to specialists. Although individuals diagnosed with obesity accounted for only 9.6% of all primary care patients, they were responsible for 17.1% of the total primary care costs, considering visits and referrals to specialists alone. Additionally, individuals diagnosed with obesity underwent diagnostic and therapeutic procedures at a higher rate and were more frequently prescribed chronic medications compared to those without obesity.

According to the National Health Survey in Spain, the prevalence of obesity in the adult population in Spain in 2022 was 16.0%, and it was slightly higher in men (16.5%) than in women (15.5%).²⁵ The recent IBERICAN study showed that 33.7% of the population attending primary care in Spain has a BMI ≥30 kg/m².¹⁰ In contrast, in the BDCAP only 9.6% of adults had a diagnosis code for obesity, suggesting a significant under-recognition of obesity in primary care in Spain. Moreover, only 10.3% of individuals (notably, only 3.8% of those with BMI >30 kg/m²) had recorded BMI values, highlighting a significant under-recording of this important clinical parameter. When considering available BMI records in primary care, only 54.3% of adults with obesity (BMI ≥30 kg/m²) had an obesity diagnosis documented in their medical records. Our findings are consistent with previous studies showing low rates of obesity diagnosis coding and BMI recording in primary care settings,^26–29 as evidenced by a recent database-based study in Spain, which found that only 25.0% of active subjects had at least one BMI measurement recorded, and among the 11.8% identified has having overweight or obesity based on BMI and/or diagnosis codes, just 36.5% had a corresponding diagnosis code.³⁰ The reasons for underrecording are unclear, but could be mainly motivated by the lack of recognition of obesity as a disease,³¹ the lack of national-level policies towards obesity prevention and treatment, and the limited healthcare resources dedicated to treating obesity.³²

Recording of obesity was especially relevant in 2022, the year this study was conducted, as obesity was considered a risk factor for COVID-19. This under-recognition has been linked to poorer health outcomes and delayed interventions.^31,33–35 Proper documentation of obesity could support early treatment, the prevention of obesity-related diseases, and more effective allocation of healthcare resources.³⁶

Several factors may influence obesity recording in primary care. Despite similar obesity prevalence between men and women in national surveys, our study found that women are more frequently diagnosed in primary care. This pattern aligns with findings from the Basque Country³⁷ and a meta-analysis showing that female sex was linked to higher rates of weight recording in some countries.³⁸ The higher obesity diagnosis rate in women may be linked to comorbidities more prevalent in women, such as osteoarthritis or anxiety. Women with obesity also report poorer health, higher morbidity, and greater healthcare use compared to men.^39–41

In this study, the rates of diagnosis of obesity in the 17 Autonomous Communities of Spain presented remarkable differences. Catalonia and Galicia presented high percentages of individuals with a diagnosis of obesity in primary care. In contrast, Andalusia showed the lowest recorded obesity diagnosis rate in primary care (0.2%) despite having the highest estimated obesity prevalence in Spain.⁴² These disparities suggest regional differences in the perception and management of obesity in Spanish primary care. Variations in diagnosis rates may be influenced by demographic and socioeconomic factors, such as aging populations (e.g., Galicia) or lower income levels (e.g., Andalusia). Higher healthcare use and comorbidity rates among older adults in regions like Galicia may also contribute to increased diagnosis.⁴³ Additionally, regional differences in weight-control discussions with healthcare providers have been linked to factors like age, income, education, and comorbidities. In children and adolescents, strong regional disparities exist but follow no clear pattern.⁴⁴

This study showed that the use of healthcare resources in primary care was higher in people diagnosed with obesity compared with those without a diagnosis of obesity. The total healthcare cost in primary care derived from visits and referrals in people with an obesity diagnosis amounted to 5444 M€ per year. People diagnosed with obesity also used more diagnostic and therapeutic procedures and were prescribed more chronic medications. It should be noted that this study included individuals across all grades of obesity with only a small proportion classified as grade III (Table 1), suggesting that obesity in all stages, and not only the highest, contributes to increased expenses. Additionally, this study likely underestimated the true societal cost of obesity, as it only included patients with a diagnosis code. Evidence shows that obesity, especially at younger ages or in severe cases, is linked to increased use of healthcare services and medications. Studies in Spain and abroad confirm that higher BMI is associated with significantly greater medical costs, and weight loss could substantially reduce the risk and cost of obesity-related conditions.^21,22 For example, in this study, about one-quarter of patients had diabetes and were taking anti-diabetic medication. As obesity is a major risk factor for diabetes and weight reduction is a main focus of some diabetes therapies, it is likely that early treatment of obesity could substantially reduce costs of treating diabetes later in life.

Given the high costs associated with obesity, its early detection should be a high priority in public health. In this regard, the current Spanish guide to the comprehensive and multidisciplinary management of obesity in adults (GIRO) highlights the essential role of primary care, as PCPs are the first point of contact for most patients with overweight or obesity.⁴⁵ Therefore, their attitude during the evaluation and subsequent treatment is crucial to the success of the weight-reduction treatment.⁴⁵

In this study, about 71% of obesity-related hospitalizations were linked to higher-grade obesity, costing €79.8 million annually. This study considered only hospitalizations with obesity as the primary diagnosis, excluding cases where it was secondary; therefore, the reported costs likely underestimate the true burden of obesity-related hospitalizations. Many admissions may have been for bariatric surgery, which has risen sharply in Spain, particularly among patients with type 2 diabetes. Similar trends and rising hospital costs related to obesity have been reported in other countries, often driven by procedures like knee replacements and diabetes care.⁴⁶

As with most studies using secondary data, this study is limited by the quality and consistency of recorded information. Its retrospective design may lead to non-standardized data collection, missing data, and inaccuracies. Moreover, the analysis was not designed to be comparative, and therefore no statistical adjustments for potential confounders variables, such as age or sex, were performed within either primary care and hospitalization datasets or obesity and non-obesity subgroups. For this reason, all differences highlighted are based on observations. The BDCAP database captures all healthcare contacts for individuals with obesity as an active diagnosis, regardless of whether the visits are obesity-related. Only primary care visits and referrals costs could be estimated, as medication and procedure data lacked sufficient detail. The hospital database (RAE-CMBD) also records admissions, not unique patients, possibly counting individuals multiple times. As the databases used in this study were not linked, it was not possible to integrate information across data sources. However, each database captured a distinct dimension of healthcare, with results analyzed and presented separately, offering complementary but independent insights into the burden of disease and associated costs within each care setting. A strength of our study lies in its representative sample and population-based design, which enables the estimation of obesity at the regional level. This approach allowed for an analysis of the geographical distribution of these conditions and facilitated the examination of key sociodemographic factors. To our knowledge, this is the first study to explore obesity recording nationally using a data source that offers representative information at the Autonomous Community level.

In conclusion, this study revealed significant discrepancies between obesity rates based on BMI and the rates of coded obesity diagnoses across Spain. The study also suggests that individuals with an obesity diagnosis incur numerically higher healthcare resource utilization and per-person costs than those without, in line with previous evidence in Spain and other countries. The multifaceted over-costs due to obesity can be observed throughout the whole range of healthcare services, from primary care consultations to hospitalizations, or consumption of medication. Critically, the under-recording of obesity detected in the public databases suggests that the costs estimated are vastly undervalued. Further, these costs only partially consider the morbidity derived from other chronic diseases in which obesity has a major negative impact. From a public health perspective, there is an urgent need to implement policies that address the obesity pandemic and to invest in preventative measures and initiatives to reduce long-term healthcare costs.

CRediT authorship contribution statement

JRA contributed to the conception and design of the study, to data acquisition, analysis, and interpretation, and drafting of the manuscript. MRdS contributed to the conception and design of the study, interpretation of the data, and drafting of the manuscript. SDC contributed to the conception and design of the study, interpretation of the data, and drafting of the manuscript. PPM contributed to the interpretation of the data and critical revision of the manuscript. GC contributed to the interpretation of the data and critical revision of the manuscript. All authors approved the final version of the manuscript.

Ethical considerations

The protocol for this study was approved by the Ethics Committee of the Hospital Clínic de Barcelona, Spain. The study was conducted in accordance with the protocol and ethical principles that have their origin in the Declaration of Helsinki and that are consistent with applicable ICH Good Clinical Practice (GCP) Guidelines, Good Pharmacoepidemiology Practices (GPPs), and applicable laws and regulations of Spain, as appropriate.

Funding

This study was funded by Eli Lilly and Company.

Conflicts of interest

JRA, MRdS and SDC are full-time employees and minor shareholders of Eli Lilly and Company. PPM has received honoraria for scientific consulting, lectures or educational activities from Eli Lilly and Company, Ultragenyx, Ferrer, Boehringer Ingelheim, Daiichi-Sankyo, Novo-Nordisk, Amgen, and Esteve. GC has received honoraria for lectures, presentations, speakers bureaus, or educational events from Eli Lilly and Company, Novo Nordisk, and Abbott, and for participation on a data safety monitoring board or advisory board from Eli Lilly and Company and Abbott.

Acknowledgments

The authors would like to acknowledge Álvaro Silleras for his contributions to the interpretation and utilization of the BDCAP database, and Francisco López de Saro, PhD, for medical writing assistance with the preparation of this manuscript, funded by Eli Lilly and Company.