Quantifying the health impact of crop breeding: Revisiting the Disability-Adjusted Life Years Approach
======================================================================================================

* Bert Lenaerts

## Abstract

Crop breeding, through methods like biofortification, plays a crucial role in increasing agricultural yields, contributing to a stable and healthy food supply by boosting the micronutrient content of staples, which is pivotal for combating chronic and hidden hunger, especially in rural areas. This shift towards improved nutrition can be sustainably integrated into regular diets, marking a significant stride in public health. The burden of hunger and its consequences on health are increasingly quantified using the Disability-Adjusted Life Years (DALYs) approach, which merges years of life lost and years lived with disability, offering a comprehensive view of health impacts and aiding in resource allocation despite its complexity and subjective nature. A practical formula for quantifying the health impact of biofortification was introduced by Stein et al. (2005) and Zimmermann and Qaim (2004). This entails calculating the efficacy or relative reduction in hunger burden based on the current and post-intervention nutrient intake against the recommended dietary allowances. As data on consumption and recommended intake levels are variable and not readily available, this paper proposes relying on relative estimates to bridge the data gaps and uncertainties, thus streamlining the quantification of crop breeding’s impact on diets and overall health.

## 1 Classic DALY approach

Crop breeding increases yields, enhances disease resistance, and improves climate resilience, ensuring a stable food supply. Besides boosting yields, crop breeding can increase the micronutrient content of staple crops through biofortification (Bouis et al., 2011). These crop improvements can help combat chronic and hidden hunger and provide a sustainable way to improve nutrition through regular diets, especially in rural areas.

The burden of hunger is increasingly being measured in Disability-Adjusted Life Years (DALYs). DALYs measure disease burden by combining years of life lost (YLL) and years lived with disability (YLD). They provide a comprehensive view of health impacts, enabling comparison across diseases and aiding resource allocation. However, DALYs are complex to calculate, involve subjective disability weights, and focus on adverse health aspects. Despite these drawbacks, DALYs are valuable for understanding and addressing public health priorities.

Stein et al. (2005) and Zimmermann and Qaim (2004) pioneered a practical formula for quantifying the health impact of biofortification: ![Formula][1]</img>  where E is the efficacy or relative reduction in the burden of hunger, CI is the current intake, BI is the intake after intervention (such as biofortification) and RDA is the recommended dietary allowance for a specific nutrient (see Table S1 for an index of terms). Although Stein et al. (2005) and Zimmermann and Qaim (2004) applied their method to the biofortification of micronutrients (iron, zinc, vitamin A), equation 1 also applies to macronutrients (calories and proteins).

The efficacy rate can then be used to quantify the (absolute) reduction in the burden of hunger (compared to the baseline or Business As Usual (BUA) scenario): ![Formula][2]</img>  

## 2 Modified DALY approach

Operationalising equation 2 requires estimates of CI, BI, RDA and DALYs. The Institute for Health Metrics and Evaluation provides the health burden (expressed in DALYs) of different nutrient deficiencies (Section 3). Estimates of CI and RDA are not readily available and exhibit considerable variation across sources (Lividini and Masters, 2022; Safiri et al., 2021; Wang et al., 2023; Wessells and Brown, 2012). Quantifying CI requires commodity-specific accounting for (1) the nutrient supply through production, trade, waste and non-food usage, (2) nutrient-specific losses through processing and cooking and (3) bioavailability through chemical composition, food interactions, and human absorption capacity. RDA captures the nutrient intake needed to meet the requirements of healthy individuals and can be obtained from different agencies (US National Institutes of Health, European Food Safety Authority, UN Food and Agriculture Organization). However, different versions exist (minimum dietary requirement, average dietary requirement and adequate intake), complicating comparisons across sources.

We propose to use relative estimates of CI, BI and RDA to overcome existing data gaps and uncertainty: ![Formula][3]</img>  ![Formula][4]</img>  where ϘCI is the relative nutrient intake increase and CI% is the relative nutrient intake gap.

Crop breeding improves the nutrient supply by increasing yields and micronutrient contents of commodities, closing the nutrient intake gap in the diet. The extent to which crop breeding can improve diets depends on (1) the crop improvement, (2) the importance of the respective crop in the diet, (3) the degree of home consumption and (4) the reach of the new variety regarding adoption. These factors collectively contribute to the increase in relative nutrient intake: ![Formula][5]</img>  where NC% is the relative increase in commodity nutrient intake, DIET% is the proportion of total nutrient intake provided by a specific commodity, PROD% is the proportion of total nutrient intake provided by the production of a specific commodity and AREA% is the proportion of total area harvested that is targeted for a specific commodity. The entry point for crop breeding, NC%, depends on biofortification and yield improvement: ![Formula][6]</img>  where ΔY% is the relative increase in yield (%) and ΔBF% is the relative increase in micronutrient content due to biofortification (%).

We can now rewrite Equation 1 to incorporate the relative parameters introduced earlier: ![Formula][7]</img>  Equation 7 enables quantifying the health impacts of increases in nutrient supply following crop breeding, where ϘCI>1, NC% represents the relative increase in nutrient intake and E represents the relative decrease in the burden of hunger. However, adverse development trends like climate change, volatility in food prices, conflict and economic recessions might reduce the supply of nutrients available to consumers in lower-income countries. Therefore, we adapt equation 7 to incorporate decreases in yield: ![Formula][8]</img>  where ϘCI<1, NC% represents the relative decrease in nutrient intake and E represents the relative increase in the burden of hunger. All parameters used in equations 5-8, except ΔY, are positive.

## 3 Data

### 3.1 Baseline health burden of hunger

The Institute for Health Metrics and Evaluation (IHME, 2024) provides the health burden (expressed in DALYs) of different nutrient deficiencies (calories/proteins, iron, zinc, vitamin A, iodine) disaggregated by country, year, gender and age (Table 1). Nutritional deficiencies are captured as cause and/or risk factors. Cause factors here are hunger-related diseases or injuries (that is, nutritional deficiencies) that directly cause death or disability; risk factors (per cause-risk attribution) are hunger-related factors that are causally associated with the probability of a disease or injury (Gödecke et al., 2018; Lenaerts and Demont, 2021). We recommend risk factors (if available) since they capture multiple cause factors, including indirect ones (like maternal death linked to iron deficiency). Following Gödecke et al. (2018), we consider child underweight a proxy for macronutrient (calories and proteins) deficiency.

View this table:
[Table 1](http://medrxiv.org/content/early/2024/08/26/2024.08.26.24312574/T1)

Table 1 
Overview of global hunger-related risk and cause factors by age group

### 3.2 Relative nutrient intake increase

The proportion of total nutrient intake provided by a specific commodity, DIET%, can be obtained by combining different databases from FAOSTAT (FAOSTAT, 2024). The calorie supply by commodity is available from FAOSTAT’s “Supply Utilization Accounts”, which can be converted into proportions. The supply of protein, iron, zinc and vitamin A by commodity can be obtained as follows:

1.  obtain the respective nutrient supply from FAOSTAT’s “Availability (based on supply utilization accounts)” by commodity group (contains gaps);

2.  obtain the “food supply” (in volume) from FAOSTAT’s “Supply Utilization Accounts” by commodity;

3.  convert the food supply by commodity into proportions across commodity groups;

4.  multiply these commodity-by-commodity group food supply proportions (step 3) with the nutrient supply by commodity group (step 1);

5.  fill gaps by correcting the food supply for processing losses and then multiplying these values with the nutrient content from GeNUS (Smith, 2018; Smith et al., 2016).

The proportion of total nutrient intake provided by the production of a specific commodity, PROD%, can be calculated as the ratio of “production quantity” (in kg) from FAOSTAT’s “Crops and livestock products” over “food supply” (in kg) from FAOSTAT’s “Supply Utilization Accounts”. If needed, the ratio can be capped at 100%.

### 3.3 Relative nutrient intake gap

Zimmermann and Qaim (2004) hypothesised a rectangular hyperbolic relationship between health outcomes and relative nutrient intake (Figure 1A). We propose reversing the quadratic relationship between the prevalence of undernourishment and relative calorie intake identified by Fischer et al. (2005) (Figure 1B) to obtain estimates of the relative nutrient intake gap. Although Fischer et al. (2005) only identified the relationship for calorie intakes, the functional form can be applied to other nutrients after normalisation, assuming the same trend based on chronic hunger per capita versus calorie intake is the same for micronutrient deficiencies and intakes. Concretely:

![Figure 1](http://medrxiv.org/http://medrxiv.stage.highwire.org/content/medrxiv/early/2024/08/26/2024.08.26.24312574/F1.medium.gif)

[Figure 1](http://medrxiv.org/content/early/2024/08/26/2024.08.26.24312574/F1)

Figure 1 
(A) Conceptual relationship between adverse health outcome and relative nutrient intake (source: Stein et al. (2005)). (B) Empirical relationship between the prevalence of undernourishment and relative calorie intake (source: own calculation following Fischer et al. (2005)).

1.  Max-normalise the prevalence of undernourishment (from FAOSTAT) across all countries so that the maximum equals 1 (no limit to the minimum).

2.  Obtain the relative calorie intake as the ratio of calorie supply over average dietary energy requirement (from FAOSTAT), and rescale so that the minimum (of the quadratic relationship) occurs at 1.

3.  The resulting quadratic form can then be inverted and be used to calculate CI/RDA for given normalised levels of the burden of nutrient deficiencies per capita (from IHME): ![Formula][9]</img>  where ![Graphic][10]</img> is the max-normalised disease levels (burden of nutrient deficiency), and a, b and c are the parameters of the quadratic relationship between the prevalence of undernourishment and the relative calorie intake (Figure 1B; Robinson et al., 2015, P. 28).

## Discussion

The method proposed here, using the relative nutrient intake gap and the relative nutrient intake growth, helps overcome data gaps and increases robustness and accuracy. First, relative parameters maintain consistent measurements within a dataset even with systematic errors or uncertainties by reducing the impact of (absolute) outliers, leading to more robust results. For example, Lividini and Masters (2022) report that the apparent intake of vitamin A exceeds dietary reference values in Argentina, Mexico and West Africa. In contrast, IHME (2020) data indicate that the burden of vitamin A deficiency in West Africa ranks among the fifth (highest) quintile globally, and Argentina and Mexico rank in the third (middle). Similarly, Lividini and Masters (2022) report that the apparent intake of bioavailable iron exceeds dietary reference values in Gabon and Papua New Guinea. In contrast, IHME (2020) data indicate that the burden of iron deficiency in Gabon and Papua New Guinea ranks among the fifth (highest) quintile globally.

Second, relative parameters enable comparison across different nutrients (calories, proteins, iron, zinc, vitamin A). Third, relative parameters ensure that repeated measurements are consistent, which is crucial for modelling the impact of crop breeding across multiple crops and geographies. Last, relative parameters can be adjusted and fine-tuned more easily without recalibrating the entire system.

Considering the respective conditions, Equation 7 yields relative reductions in the health burden of hunger between 0 and 1. Figure 2 plots the efficacy rates versus the relative nutrient intake increase (subplot A) and versus the relative nutrient intake gap (subplot B). For high intake gaps, the efficacy rate increases almost linearly with the relative nutrient intake increase, whereas for smaller gaps, the efficacy converges slowly to 1, showing asymptotic behaviour. The efficacy rate drops to zero when the intake gap is closed. This reflects the higher scope of biofortification for reducing hunger in communities with considerable existing deficiencies.

![Figure 2](http://medrxiv.org/http://medrxiv.stage.highwire.org/content/medrxiv/early/2024/08/26/2024.08.26.24312574/F2.medium.gif)

[Figure 2](http://medrxiv.org/content/early/2024/08/26/2024.08.26.24312574/F2)

Figure 2 
(A) Coplot of efficacy (E) versus the relative nutrient intake increase (Q_CI) controlling for the relative nutrient intake gap (CI%). (B) Coplot of efficacy (E) versus the relative nutrient intake gap (CI%) controlling for the relative nutrient intake increase (Q_CI).

Plot 2B ranges from 1 to 2, although most relative nutrient intake increases are not expected to exceed 1.125. The efficacy rate increases uniformly with the nutrient intake gaps, indicating it is easier to close smaller gaps, although the absolute burden of hunger avoided will be smaller. Like subplot A, subplot B evidences asymptotic behaviour and correlates zero efficacy rates with zero intake gaps. The asymptotic behaviour towards 1 reflects the conditions in Equation 7, without which the underlying formula might turn negative for CI > RDA or BI > RDA.

## Data Availability

All data used are available online from the sources listed in the paper.

## Supplementary materials

View this table:
[Table S1](http://medrxiv.org/content/early/2024/08/26/2024.08.26.24312574/T2)

Table S1 
Index

## Acknowledgements

I sincerely thank my colleagues, Erick Boy and Victor Taleon (HarvestPlus), for their invaluable feedback and generous data sharing. Their support has greatly enhanced this research.

*   Received August 26, 2024.
*   Revision received August 26, 2024.
*   Accepted August 26, 2024.


*   © 2024, Posted by Cold Spring Harbor Laboratory

This pre-print is available under a Creative Commons License (Attribution-NonCommercial-NoDerivs 4.0 International), CC BY-NC-ND 4.0, as described at [http://creativecommons.org/licenses/by-nc-nd/4.0/](http://creativecommons.org/licenses/by-nc-nd/4.0/)

## References

1.  Bouis, H. E., Hotz, C., McClafferty, B., Meenakshi, J. V., and Pfeiffer, W. H. 2011. Biofortification: a new tool to reduce micronutrient malnutrition, Food and Nutrition Bulletin, vol. 32, no. 1 (supplement), S31–40 FAOSTAT. 2024. <[http://www.fao.org/faostat/en/#data](http://www.fao.org/faostat/en/#data)>
    
    [CrossRef](http://medrxiv.org/lookup/external-ref?access_num=10.1177/15648265110321S105&link_type=DOI) 
    
    [PubMed](http://medrxiv.org/lookup/external-ref?access_num=21717916&link_type=MED&atom=%2Fmedrxiv%2Fearly%2F2024%2F08%2F26%2F2024.08.26.24312574.atom) 

2.  Fischer, G., Shah, M., N. Tubiello, F., and Van Velhuizen, H. 2005. Socio-economic and climate change impacts on agriculture: an integrated assessment, 1990–2080, Philosophical Transactions of the Royal Society B: Biological Sciences, vol. 360, no. 1463, 2067–83
    
    [CrossRef](http://medrxiv.org/lookup/external-ref?access_num=10.1098/rstb.2005.1744&link_type=DOI) 
    
    [PubMed](http://medrxiv.org/lookup/external-ref?access_num=16433094&link_type=MED&atom=%2Fmedrxiv%2Fearly%2F2024%2F08%2F26%2F2024.08.26.24312574.atom) 

3.  Gödecke, T., Stein, A. J., and Qaim, M. 2018. The global burden of chronic and hidden hunger: Trends and determinants, Global Food Security, vol. 17, 21–29
    
    

4.  IHME. 2020. Global Burden of Disease Study 2019 (GBD 2019) Results. Available from [http://ghdx.healthdata.org/gbd-results-tool](http://ghdx.healthdata.org/gbd-results-tool)., Seattle, United States, IHME
    
    

5.  IHME. 2024. Global Burden of Disease Study 2019 (GBD 2021) Results. Available from [http://ghdx.healthdata.org/gbd-results-tool](http://ghdx.healthdata.org/gbd-results-tool)., Seattle, United States, IHME
    
    

6.  Lenaerts, B. and Demont, M. 2021. The global burden of chronic and hidden hunger revisited: New panel data evidence spanning 1990–2017, Global Food Security, vol. 28, 100480
    
    

7.  Lividini, K. and Masters, W. A. 2022. Tracing global flows of bioactive compounds from farm to fork in nutrient balance sheets can help guide intervention towards healthier food supplies, Nature food, vol. 3, no. 9, 703–15
    
    

8.  Robinson, S., Mason-D’Croz, D., Islam, S., Sulser, T., Robertson, R., Zhu, T., Gueneau, A., Pitois, G., and Rosegrant, M. W. 2015. The international model for policy analysis of agricultural commodities and trade (IMPACT): model description for version 3, IFPRI Discussion Paper 01483, Washington, DC, IFPRI
    
    

9.  Safiri, S., Kolahi, A.-A., Noori, M., Nejadghaderi, S. A., Karamzad, N., Bragazzi, N. L., Sullman, M. J., Abdollahi, M., Collins, G. S., and Kaufman, J. S. 2021. Burden of anemia and its underlying causes in 204 countries and territories, 1990–2019: results from the Global Burden of Disease Study 2019, Journal of hematology & oncology, vol. 14, 1– 16
    
    

10. Smith, M. 2018. Food composition tables for GENuS. Available from: doi:10.7910/DVN/GNFVTT
    
    [CrossRef](http://medrxiv.org/lookup/external-ref?access_num=10.7910/DVN/GNFVTT&link_type=DOI) 

11. Smith, M. R., Micha, R., Golden, C. D., Mozaffarian, D., and Myers, S. S. 2016. Global Expanded Nutrient Supply (GENuS) model: a new method for estimating the global dietary supply of nutrients, PloS one, vol. 11, no. 1, e0146976
    
    [CrossRef](http://medrxiv.org/lookup/external-ref?access_num=10.1371/journal.pone.0146976&link_type=DOI) 
    
    [PubMed](http://medrxiv.org/lookup/external-ref?access_num=http://www.n&link_type=MED&atom=%2Fmedrxiv%2Fearly%2F2024%2F08%2F26%2F2024.08.26.24312574.atom) 

12. Stein, A. J., Meenakshi, J. V., Qaim, M., Nestel, P., Sachdev, H. P. S., and Bhutta, Z. A. 2005. Analyzing the health benefits of biofortified staple crops by means of the disability-adjusted life years approach: a handbook focusing on iron, zinc and vitamin A., Advance Access published 2005
    
    

13. Wang, X., Dou, Z., Feng, S., Zhang, Y., Ma, L., Zou, C., Bai, Z., Lakshmanan, P., Shi, X., and Liu, D. 2023. Global food nutrients analysis reveals alarming gaps and daunting challenges, Nature Food, vol. 4, no. 11, 1007–17
    
    

14. Wessells, K. R. and Brown, K. H. 2012. Estimating the global prevalence of zinc deficiency: results based on zinc availability in national food supplies and the prevalence of stunting, PloS one, vol. 7, no. 11, e50568
    
    [CrossRef](http://medrxiv.org/lookup/external-ref?access_num=10.1371/journal.pone.0050568&link_type=DOI) 
    
    [PubMed](http://medrxiv.org/lookup/external-ref?access_num=23209782&link_type=MED&atom=%2Fmedrxiv%2Fearly%2F2024%2F08%2F26%2F2024.08.26.24312574.atom) 
    
    [Web of Science](http://medrxiv.org/lookup/external-ref?access_num=23209782&link_type=ISI) 

15. Zimmermann, R. and Qaim, M. 2004. Potential health benefits of Golden Rice: a Philippine case study, Food Policy, vol. 29, no. 2, 147–68
    
    [CrossRef](http://medrxiv.org/lookup/external-ref?access_num=10.1016/j.foodpol.2004.03.001&link_type=DOI) 
    
    [Web of Science](http://medrxiv.org/lookup/external-ref?access_num=000221924100003&link_type=ISI)

 [1]: /embed/graphic-1.gif
 [2]: /embed/graphic-2.gif
 [3]: /embed/graphic-3.gif
 [4]: /embed/graphic-4.gif
 [5]: /embed/graphic-5.gif
 [6]: /embed/graphic-6.gif
 [7]: /embed/graphic-7.gif
 [8]: /embed/graphic-8.gif
 [9]: /embed/graphic-11.gif
 [10]: /embed/inline-graphic-1.gif