Impact of the COVID-19 Lockdown on Primary Bariatric Surgery: A Comparative Study of Weight Loss Trajectories and Perioperative Outcomes

Abstract

Background:

The COVID-19 pandemic and subsequent lockdowns led to substantial changes in health-related behaviors and deferred elective surgeries. This study aimed to evaluate the impact of the COVID-19 lockdown on weight loss trajectories (WLT) and perioperative outcomes in primary bariatric surgery at a single tertiary referral center.

Methods:

Patients who underwent primary bariatric surgery immediately prior to the enforced lockdown (During-Lockdown Group, DLG) were compared with historical controls who underwent surgery before the lockdown (Before-Lockdown Group, BLG). WLT were estimated using a random intercept mixed-effects model with repeated measures. Cross-sectional group differences at 3- and 6-months post-surgery were examined. Perioperative outcomes were also analyzed.

Results:

A total of 273 patients were included (133 BLG, 140 DLG). The adjusted average weight decreased in both groups over time at different levels. The expected difference in weight between BLG and DLG at 3-months was 1.06 kg (P = .013) and at 6-months was .04 kg (P = .920), respectively. There were no statistically significant differences in postoperative complications, emergency department visits, re-admission rates, re-operation rates, or mortality.

Conclusions:

This study suggests that bariatric procedures performed before the COVID-19 lockdown were effective and safe. Despite a statistically significant difference in weight between groups at 3- and 6-months post-surgery, the difference was small and unlikely to be clinically significant. Short-term outcomes were comparable between the two groups. Further investigation is warranted to assess the impact of lifestyle changes during the lockdown on long-term bariatric surgery outcomes.

Introduction

On 11 March 2020, the World Health Organization (WHO) acknowledged COVID-19 as a global pandemic and on 22 March 2020, authorities in the United Arab Emirates (UAE) implemented stringent infection controls alongside a partial lockdown.^1,2 This partial lockdown led to substantial changes in health-related behaviours^1–3 and deferred millions of elective procedures, creating unprecedented surgical backlogs.^4,5 The reorganization of health care systems in the UAE resulted in the halting of elective surgeries.^4,5 Patients who had previously undergone bariatric surgery experienced limited access to the hospital for their follow-up visits due to the emergency state. The loss of this continuum of care, along with associated lifestyle changes, was expected to worsen bariatric surgery outcomes. Despite multiple studies reporting on bariatric surgery outcomes during the COVID-19 pandemic, there is still conflicting data concerning weight loss during the COVID-19 lockdown.^4,5 A main limitation of such studies is the discrepancy in the primary outcome measured. In contrast, weight loss trajectories (WLT) offer valuable insights by considering the weight changes over a continuous time interval in contrast to total weight loss (%TWL) and percent excess BMI loss (%EBMIL), which are single-point measures. As such, the aim of this study was to evaluate the impact of the COVID-19 lockdown on WLT and perioperative outcomes in primary bariatric surgery at a single tertiary referral center.

Methods

Study design and ethical approvals

This was a retrospective study conducted with the approval of the Institutional Review Board.

Population and data collection

Patients who underwent primary bariatric surgery immediately prior to the enforced lockdown from November 2019 to March 2020, During-Lockdown Group (DLG) were compared with historical controls of patients who underwent primary bariatric surgery before the lockdown from November 2018 to March 2019, Before-Lockdown Group (BLG). Demographic data, anthropometric data, and perioperative outcomes were collected and analyzed.

Selection criteria

Inclusion criteria were patients age ≥18 years, who underwent primary Roux-en-Y gastric bypass (RYGB) or sleeve gastrectomy (SG), and had a minimum 6-month post-surgery follow-up. The exclusion criteria were patients who underwent revisional bariatric surgery or had inadequate follow-up.

Data analysis

WLT were estimated using a random intercept mixed-effects model with repeated measures nested within patients. A third-degree polynomial was used to operationalize time. Baseline covariates were included in the model to control for preexisting group differences. Cross-sectional group differences 3- and 6-months post-surgery were examined by applying general linear hypothesis testing to mixed-effects results. Additionally, perioperative outcomes were analyzed.

Results

A total of 273 patients were included in the study (133 BLG and 140 DLG). Of these, 62.7% were female with a similar mean age of (35.0 BLG versus 36.0 DLG years, P = .398) and BMI before surgery was (44.0 BLG versus 43.0 DLG kg/m², P = .360) between groups. Baseline characteristics were also similar. After surgery, the adjusted average weight decreased in both groups over the time at different levels. (Table 1, section P1). At 3- and 6-months, the adjusted average weights for BLG and DLG were statistically different. The expected difference in weight between BLG and DLG at 3-months was 1.06 kg (P = .013) and at 6-months was .04 kg (P = .920)—Figure 1. There were no statistically significant differences between BLG and DLG in post-operative major complications within 30-days (2.3% versus 6.4%, P = .092), ED visits (22.6% versus 15.7%, P = .149), re-admission rates (2.3% versus 3.6%, P = .522), re-operation rates (1.5% versus 3.6%, P = .280), and complications after 30-days (4.5% versus 2.9%, P = .465). No mortalities were reported.

FIG. 1.

Average Weight Loss Trajectories.

Table 1.

WLT: Mixed-Effects Result

									95% CI of estimate
Model section code	Model section name (group)		Trajectory parameter name	Estimate	Std. Error	Df	t	p	Min	Max
P1	Post-Surgery	Shared	Average wt at day 0 (intercept)	115.232	0.276	402.678	417.864	<.001	114.692	115.773
P1	Post-Surgery	BLG	Days Post-Op (linear)	−0.247	0.006	2580.561	−40.482	<.001	−0.259	−0.235
			Days Post-Op (quadratic)	0.001	4.04E-05	2524.544	16.714	<.001	0.001	0.001
			Days Post-Op (cubic)	−6.75E-07	7.05E-08	2490.405	−9.568	<.001	−8.13E-07	−5.36E-07
			DLG	Days Post-Op (linear) by DLG	−0.030	0.009	2563.314	−3.348	.001	−0.048	−0.013
				Days Post-Op (quadratic) by DLG	2.41E-04	6.18E-05	2599.207	3.906	<.001	1.20E-04	3.62E-04
				Days Post-Op (cubic) by DLG	−4.09E-07	1.11E-07	2562.491	−3.695	<.001	−6.26E-07	−1.92E-07
M	Trajectory Controls/Moderators		Intercept by Baseline Age	0.003	0.027	365.350	0.110	.913	−0.049	0.055
			Days Post-Op (linear) by Baseline Age	0.002	2.42E-04	2433.870	7.052	<.001	0.001	0.002
			Days Post-Op (quadratic) by Baseline Age	0.000	6.85E-07	2426.487	−5.101	<.001	−4.84E-06	−2.15E-06
			Intercept by Baseline Wt	0.965	0.014	365.824	71.017	<.001	0.939	0.992
			Days Post-Op (linear) by Baseline Wt	−0.001	1.22E-04	2436.832	−11.928	<.001	−0.002	−0.001
			Days Post-Op (quadratic) by Baseline Wt	1.50E-06	3.42E-07	2422.391	4.401	<.001	8.34E-07	2.17E-06
			Intercept by Male	−0.304	0.655	357.223	−0.465	.643	−1.588	0.979
			Days Post-Op (linear) by Male	−0.027	0.006	2414.375	−4.871	<.001	−0.038	−0.016
			Days Post-Op (quadratic) by Male	8.51E-05	1.56E-05	2405.284	5.451	<.001	5.45E-05	1.16E-04
			Intercept by T2D	0.371	0.698	354.476	0.532	.595	−0.996	1.739
			Days Post-Op (linear) by T2D	0.011	0.006	2421.641	1.882	.060	−4.58E-04	0.023
			Days Post-Op (quadratic) by T2D	−1.67E-05	1.66E-05	2409.683	−1.008	.314	−4.92E-05	1.58E-05
			Intercept by Smoker	0.517	0.711	369.590	0.727	.467	−0.876	1.911
			Days Post-Op (linear) by Smoker	−0.007	0.006	2427.749	−1.098	.272	−0.020	0.006
			Days Post-Op (quadratic) by Smoker	−9.38E-07	1.87E-05	2421.362	−0.050	.960	−3.75E-05	3.56E-05
			Intercept by OSA	0.135	0.635	363.366	0.212	.832	−1.110	1.380
			Days Post-Op (linear) by OSA	−0.018	0.005	2424.344	−3.406	.001	−0.029	−0.008
			Days Post-Op (quadratic) by OSA	4.26E-05	1.53E-05	2411.333	2.780	.005	1.26E-05	7.27E-05
			Intercept by CAD	1.718	1.626	332.195	1.056	.292	−1.470	4.905
			Days Post-Op (linear) by CAD	−0.026	0.011	2400.427	−2.322	.020	−0.048	−0.004
			Days Post-Op (quadratic) by CAD	7.56E-05	3.11E-05	2409.450	2.430	.015	1.46E-05	1.37E-04
			Intercept by HTN	−0.188	0.712	357.741	−0.264	.792	−1.584	1.208
			Days Post-Op (linear) by HTN	0.011	0.006	2415.066	1.898	.058	−3.73E-04	0.023
			Days Post-Op (quadratic) by HTN	−1.39E-05	1.65E-05	2404.704	−0.839	.402	−4.62E-05	1.85E-05
			Intercept by Hyperlipidemia	−0.070	0.556	367.108	−0.126	.900	−1.159	1.019
			Days Post-Op (linear) by Hyperlipidemia	−0.001	0.005	2457.931	−0.135	.893	−0.011	0.009
			Days Post-Op (quadratic) by Hyperlipidemia	3.08E-05	1.45E-05	2438.602	2.125	.034	2.40E-06	5.92E-05
			Intercept by Renal Insuf.	0.773	1.436	363.170	0.538	.591	−2.043	3.588
			Days Post-Op (linear) by Renal Insuf.	−0.004	0.013	2428.902	−0.340	.734	−0.029	0.020
			Days Post-Op (quadratic) by Renal Insuf.	2.68E-06	3.43E-05	2407.026	0.078	.938	−6.46E-05	7.00E-05
			Intercept by Sleeve	−0.931	0.784	355.389	−1.188	.235	−2.467	0.605
			Days Post-Op (linear) by Sleeve	0.031	0.007	2424.902	4.588	<.001	0.018	0.044
			Days Post-Op (quadratic) by Sleeve	−5.36E-05	1.89E-05	2414.664	−2.834	.005	−9.06E-05	−1.65E-05
			Intercept by Proc. Duration	−0.005	0.007	347.896	−0.659	.510	−0.019	0.009
			Days Post-Op (linear) by Proc. Duration	8.24E-05	5.89E-05	2439.360	1.398	.162	−3.31E-05	1.98E-04
			Days Post-Op (quadratic) by Proc. Duration	4.65E-08	1.61E-07	2434.445	0.289	.773	−2.69E-07	3.62E-07
			Intercept by LOS	0.180	0.176	326.078	1.022	.308	−0.165	0.526
			Days Post-Op (linear) by LOS	3.40E-04	0.001	2420.539	0.262	.793	−0.002	0.003
			Days Post-Op (quadratic) by LOS	−4.04E-06	3.47E-06	2408.552	−1.162	.245	−1.08E-05	2.77E-06

Discussion

Although this study demonstrated a significant difference in weight between groups at 3- and 6-months post-surgery, the difference was small and unlikely to be clinically significant. Additionally, there were no significant differences in postoperative outcomes between DLG and BLG. A retrospective study by Pereira et al. reported no difference in target weight loss at 1 year in a cohort that underwent bariatric surgery just before the pandemic in New York City, which was a major epicenter of the pandemic.³ A case-control study by Barranquero et al. observed reduced comorbidities but also found decreased excess weight loss at 1-year follow-up of RYGB patients intervened prior to the pandemic compared with the non-COVID group.⁶

Previous studies have reported that the lockdown, causing limited access to public places, led to unfavorable lifestyle changes such as increased caloric intake, more frequent snacking, reduced consumption of fresh fruits and vegetables and decreased physical activity.^2,3 However, an alternative perspective might suggest positive lifestyle changes as well. We hypothesize that patients reduced their outings with friends and decreased consumption of high caloric food from malls and restaurants. In UAE culture, social habits such as shared meals, families/friends eating together from one plate and waiting for the last person to finish the meal before stopping eating might have decreased due to the lockdown.⁷ We hypothesize that WLTs were not affected because patients exercised at home and telemedicine made follow-up visits more convenient and accessible. Telemedicine also allows physicians more time for follow-ups, as the halting of elective procedures created additional availability. However, these are hypotheses that warrant further investigation.

Even though perioperative COVID-19 infection has previously demonstrated negative impacts on patient safety, leading higher rates of surgical complications and mortality,⁸ our study did not find differences in postoperative outcomes including complications, readmissions, reoperations, and mortalities during lockdown. However, the study did not include information on whether patients had COVID-19 infection or its impact on perioperative outcomes.

Conclusion

This series, representing the impact of the COVID-19 lockdown on bariatric outcomes, suggests that procedures performed before the lockdown were effective and safe. Short-term outcomes are comparable as there was no clinically significant difference in WLT between BLG and DLG.

Footnotes

Acknowledgments

The results of this study were presented as a poster at the 25th IFSO World Congress in Miami, USA, from August 23 to 27, 2022. Website: https://scholar.google.com/citations?view_op=view_citation&hl=en&user=9FcJm20AAAAJ&cstart=20&pagesize=80&sortby=pubdate&citation_for_view=9FcJm20AAAAJ:SP6oXDckpogC.

Authors’ Contributions

J.S.B.: Study design, data analysis, data interpretation and article preparation. M.S.I.K.: Data collection, data interpretation, article preparation. G.R.V.: Study design, data interpretation, article preparation. J.T.D.: Study design, data interpretation, article preparation. G.D.G.: Data collection, data interpretation, article preparation. M.A.: Data analysis, data interpretation, article preparation. J.P.P.: Data analysis, data interpretation, article preparation. J.R.: Data analysis, data interpretation, article preparation. C.A.: Study design, data interpretation, article preparation. A.D.G.: Study design, data interpretation, article preparation. T.L.S.: Data collection, data interpretation, article preparation. R.C.: Data analysis, data interpretation, article preparation. M.K.: Data analysis, data interpretation, article preparation. J.H.R.: Study design, data analysis, data interpretation and article preparation.

Disclosure Statement

Authors have no conflict of interest to disclose.

Funding Information

No funding was received for this article.

References

Radwan

, Kitbi

, Hasan

, et al. Indirect health effects of covid-19: Unhealthy lifestyle behaviors during the lockdown in the United Arab Emirates. Int J Environ Res Public Health, 2021; 18(4); doi: 10.3390/ijerph18041964

Ismail

, Osaili

, Mohamad

, et al. Eating habits and lifestyle during covid-19 lockdown in the United Arab Emirates: A Cross-Sectional Study. Nutrients, 2020; 12(11); doi: 10.3390/nu12113314

Pereira

, Romero-Velez

, Skendelas

, et al. The COVID-19 pandemic did not affect target weight loss 1 year post bariatric surgery. Obes Surg, 2021; 31(11):4926–4932; doi: 10.1007/s11695-021-05672-8

Kooli

. COVID-19: Public health issues and ethical dilemmas. Ethics Med Public Health, 2021; 17:100635; doi: 10.1016/j.jemep.2021.100635

Alhareb

, Barajas-Gamboa

, Tat

, et al. Resuming bariatric surgery procedures during COVID-19 pandemic: Early surgical outcomes at a Tertiary referral center in the Middle East. Bariatr Surg Pract Patient Care, 2022; 17(2):115–120; doi: 10.1089/bari.2021.0039

Vitiello

, Berardi

, Musella

. Impact of COVID-19 lockdown on short-term weight loss in a Single Italian Institution: 1-year updated data. Obes Surg, 2022; 32(3):924–925; doi: 10.1007/s11695-021-05759-2

Ali

, Bernsen

, Baynouna

. Barriers to weight management among Emirati women T A qualitative investigation of health professionals’ perspectives. Int Q Community Health Educ, 2008; 29(2):143–159; doi: 10.2190/IQ.29.2.d

Crovari

, Inzunza

, Irarrázaval

, et al. Safety of bariatric surgery during the opening phase after the first wave of the COVID-19 Pandemic: Experience at an academic center. Obes Surg, 2021; 31(12):5376–5382; doi: 10.1007/s11695-021-05695-1