In this study, we established reference equations for lung and lobar volumes, based on inspiratory chest CT in a large sample of a general population without lung disease, using linear regression analysis. Age and height were positively associated with lung and lobar volumes, except for the LLL in women. The RLL was consistently the largest lobe, followed by the LUL and LLL, then the RUL and lastly the RML. Upper lobes showed a stronger association with age, indicating they comprise more of the total lung volume as individuals age. Men have higher lung volumes than women at the same height. Overall, age and height explained 7.8–19.9% of the variation in volumes.

Lung volume measurement by CT has the benefit of allowing regional volume assessment, such as lobar volumes, possibly beneficial for treatments like EBV and lung volume reduction. The only other modality capable of achieving this is VP/SPECT, which is much less available [19]. Prior studies, specifically related to lobar volumes, have mostly focused on ventilation or collapsibility of the lobes by comparing scans in in- and expiration [20, 21]. In EBV studies, attention regarding lobar volumes centers on pre- and post-treatment changes [22]. A lower limit for important difference in lobar volume reduction between pre- and post-treatment has been established at 563 mL [23]. This metric does not take into account variations influenced by patient height, age, and sex and may therefore be refined by an approach similar to that employed in the present study. In terms of reference equations for lobar volumes, to our knowledge, only one study previously established such equations at full inspiration based on a cohort of 469 COPDgene participants without COPD (92 never smokers and 377 current or former smokers), incorporating height, sex, and ethnicity [5]. Unlike our study and the GLI reference equations for TLC, they did not observe a significant relationship between age and any of the volumes, possibly attributed to the small cohort size. They did find similar results regarding the relationship to height and a similar intercept [5]. In our study, ethnicity stratification was not performed, given that 97.5% of the participants identified as having Caucasian origins. Similarly to this prior study, we stratified for sex, justified by our finding that women had consistently smaller lobar volumes for the same height compared to men.

The observation that upper lobes had a greater increase in volume with age, as opposed to lower lobes, suggests a potential susceptibility of upper lobes to damage, even among the lung-healthy population. This aligns with research indicating a higher incidence of emphysema and lung cancer in the upper lobes [24, 25]. Therefore, it would be expected that smokers and those with COPD would exhibit relatively larger upper lobes compared to their lower lobes when compared to healthy individuals. In this study, we observed a significant difference between the lung-healthy population and current smokers for every lung or lobar volume, except for the LLL and RML in men and the RLL in men and women. This further supports the idea that the upper lobes may have a greater susceptibility to damage. Another possible explanation, especially in non-smoking individuals, is dysanapsis, which is the mismatch of airway caliber to lung volume [26]. An abnormally high lung volume in combination with a relatively normal caliber airway may result in a measured obstruction, as defined by FEV1/FVC.

Regarding CT lung volumes, previous studies primarily compared TLV measurements obtained at inspiration and/or expiration to plethysmography-derived volumes, showing strong correlations [6,7,8,9,10,11,12,13,14]. Typically, TLC was slightly higher than measured TLV, while RV was lower at expiration [27]. In line with this, a previous study established that there is a substantial discrepancy between the GLI-predicted TLC and TLV, with significantly larger estimated TLC than measured TLV [15]. This difference can only partially be explained by measurement position variations between seated plethysmography and supine CT scans, with a 9.9% smaller volume in the supine position [28].

This study’s strengths lie in its extensive lung-healthy sample of 7306 individuals. Additionally, the current study included a standardized CT scan protocol on third-generation dual-source CT, with very short acquisition time due to high-pitch scanning mode. Previous research has shown consistent CT-based measurements for lobar volumes, with TLV being more reproducible than TLC as measured by body plethysmography [11], which further validates our approach. The reference equations are applicable for assessing lobe-specific hyperinflation in COPD, which is specifically potentially useful in treatments like EBV and lung volume reduction. Furthermore, these equations may be of potential value in assessing other conditions like restrictive pulmonary diseases or surgical planning for lobe removal in lung cancer patients. The application of the derived equations to smokers and subjects with airway obstruction indeed showed higher volumes (especially in the upper lobes). This shows that the equations are capable of pinpointing differences between groups.

The study focused on a regionally specific population from the north of the Netherlands, characterized by above-average height, approximately 5 cm above the WHO growth chart median [29], and predominantly Caucasian ethnicity. Additionally, this cohort consisted of individuals between the ages of 45 and 80. Given that EBV is typically used for severe emphysema patients aged around 50–70, these limitations are acceptable for our purposes. However, it is crucial to recognize that these reference equations may therefore not be universally applicable. It would be important to calibrate these reference equations in other populations.

The explained variance for the lobar and lung volumes was relatively low, ranging from 7.8 to 19.9%. This suggests that other factors not included in the regression analysis play a role in the size of the lobes and lungs. However, these factors are unknown, and no other explanatory factors are used in similar approaches to reference equations, such as the GLI reference equations for TLC, which reported coefficients of variation of over 10% for their models [2]. These unknown factors are likely related to the shape and size of the chest. While sex and height are now used as crude approximations for these factors, they do not fully capture the complexity of individual variations. Genetic factors could also play a role, as well as aspects like fissure integrity.

In Lifelines, plethysmography measurements are not available, which prevents the comparison or correlation of participants’ TLC with their TLV measurements.