Overnight Thickness Variation in Diabetic Macular Edema

Abstract

purpose. To study the evening-to-morning variation in retinal thickness in patients with fovea-involving diabetic macular edema.

methods. Twelve eyes in 12 patients aged 39 to 78 years (mean, 57) with fovea-involving diabetic macular edema and 14 eyes in 7 healthy volunteers aged 30 to 70 years (mean, 57) were examined by optical coherence tomography, in the evening and in the morning after ≥6 hours of sleep in the recumbent position in darkness followed by 0.5 hour wakefulness in the same position in room light with both eyes open.

results. In patients with diabetic macular edema, macular thickness increased overnight, from 316 ± 72 μm in the evening to 336 ± 81 μm in the morning (P = 0.003). Visual acuity decreased from a mean of 41 ETDRS letters (Early Treatment of Diabetic Retinopathy Study; range, 4–61) in the evening, to a mean of 36 letters (range, 2–60) in the morning (P = 0.03). No overall change was found in mean arterial blood pressure (MABP; P = 0.48), blood glucose (P = 0.25), or corneal thickness (P = 0.26). The overnight change in macular thickness correlated directly with the change in MABP (r = 0.65, P = 0.03) but not with baseline MABP or blood glucose. The overnight increase in retinal thickness remained significant after statistical adjustment for the effect of arterial blood pressure (P = 0.002). Healthy subjects demonstrated no significant change in any parameter.

conclusions. In fovea-involving diabetic macular edema, a reduction in visual acuity accompanies overnight retinal thickening, the magnitude being related to the nocturnal change in blood pressure. The results indicate that deficient regulation of retinal capillary filling pressure promotes edema, but the bulk of the overnight increase in macular edema is caused by other mechanisms, of which postural variation in venous blood pressure and increased retinal metabolism in the dark merit further study.

Diabetic patients with macular edema occasionally report poorer reading vision early in the morning than later in the day, ¹ with reading improving spontaneously within 2 to 4 hours after rising. A variety of mechanisms may be postulated to explain this phenomenon. Our working hypothesis was that the change in visual acuity is caused by an increase in foveal edema during sleep in the recumbent position. To test the hypothesis we compared foveal thickness before and after sleep in patients with diabetic macular edema (DME) and in healthy subjects. 

Methods

The study adhered to the tenets of the Declaration of Helsinki and was approved by the medical ethics committee of Copenhagen County. Included were 14 eyes of 7 healthy volunteers (4 men, 3 women) aged 30 to 70 years (mean, 57) and 12 eyes in 12 patients aged 39 to 78 years (mean, 57; men, 3 women) with DME exceeding the ETDRS (Early Treatment Diabetic Retinopathy Study) clinically significant macular edema level and extending to within 500 μm of the center of the foveola. ² Inclusion required independent verification of edema by biomicroscopy, intravenous fluorescein angiography (IVFA), and optical coherence tomography (OCT) and ability to fixate the OCT fixation target. None of the participants had current or previous eye disease other than diabetic retinopathy or significant systemic disease other than diabetes and systemic complications of diabetes in the patient group. No participant had refractive error exceeding ±5 spherical diopters or ±3 cylindrical diopters. Applanation tonometry at study entry was below 21 mm Hg in all eyes. We enrolled the first 12 eligible volunteers with foveal-center–involving DME, as defined by the mean +2 SD of healthy subjects, according to Wang et al. ³  

Study procedures included ETDRS refraction and determination of best-corrected visual acuity, ⁴ biomicroscopy with a Goldmann contact lens, fundus photography, IVFA, blood pressure manometry in the sitting position in the evening and in the recumbent position in the morning until immediately before rising to have OCT performed, and blood glucose measurement with a test strip. Retinal thickness was assessed with horizontal fovea-centered 4.52-mm wide OCT scans (OCT-1, software ver. CA A4.1, nominal axial resolution of 10 μm; Carl Zeiss Meditec, Dublin, CA). Macular thickness was defined as the mean of the entire scan, taken as seven evenly spaced positions along the scan, one in the center of the scan, presumably corresponding to the center of the foveola, and three on either side, extending to the end of the scan. Foveal thickness was defined as the value of the reading at the center of the scan. The assessment of the posterior and anterior surfaces of the neurosensory retina was guided by the automated tracing of a manufacturer-supplied tracking algorithm, which was evaluated visually, corrected as needed, and read with a ruler. All values were based on the mean of the two best-quality scans from the same session. All OCT procedures and evaluations were performed by the same observer. All patients were examined before photocoagulation treatment. 

Retinal thickness was measured in the evening between 8 and 10 PM, after the patient had been no less than 6 hours in the sitting or standing position and in the morning after the patient had been awake for 0.5 hour in normal room light while maintaining the recumbent position with both eyes opened. Visual acuity was expressed as the numbers of letters read on the ETDRS chart and/or the logMAR (logarithm of the minimum angle of resolution) value. Variations in observational parameters were analyzed by using Wilcoxon’s matched-pairs signed rank sum test and differences between groups using the Kruskal-Wallis nonparametric test. Correlations between parameters were analyzed with the Pearson test. All tests were made assuming a level of statistical significance of P < 0.05. 

Results

In healthy subjects, macular thickness over the full 4.52-mm width of the OCT scans was stable, at 239 ± 17 μm in the evening (mean ± SD) and 237 ± 15 μm in the morning (P = 0.50). Best-corrected visual acuity in the evening was 57 ± 4 letters (mean ± SD; logMAR, −0.04) and 57 ± 5 letters in the morning (logMAR −0.04; P = 0.45). Healthy subjects also demonstrated no significant overnight variation in MABP (P = 0.50), blood glucose (P = 0.20), or corneal thickness (P = 0.09). 

In patients with DME, macular thickness over the full 4.52-mm width of the OCT scan increased overnight by an average of 6.3%, from 316 ± 72 (mean ± SD) μm in the evening to 336 ± 81 μm in the morning (Figs. 1 2 ; P = 0.003; Table 1 ). The thickness of the center of the foveola increased overnight by an average of 11.9%, from 336 ± 103 μm in the evening to 376 ± 122 μm in the morning (P = 0.004). A concomitant decrease in best-corrected visual acuity was seen overnight, from 41 ± 14 letters (mean ± SD, range, 4–61) in the evening to 36 ± 11 letters in the morning (P = 0.03; Table 2 ). Of the 10 patients who experienced an increase in foveal center thickness of 5% or more relative to baseline, 7 had a decrease in acuity of five or more ETDRS letters overnight. The correlation between the changes in visual acuity and retinal thickness did not achieve statistical significance (r = 0.029; P = 0.20). 

The change in macular thickness in patients with DME correlated positively with the change in MABP (r = 0.65, P = 0.03; Fig. 3 ) but not with baseline blood pressure or blood glucose. After adjustment for the statistical effect of changes in MABP, the overnight increase in retinal thickness remained statistically significant (P = 0.002; parametric regression analysis). No overall change was found in MABP (P = 0.48), blood glucose (P = 0.25), corneal thickness (P = 0.26), or refraction (P = 0.63). 

Discussion

We have shown that the healthy retina is capable of maintaining a stable thickness from evening to morning and that in eyes with DME, it undergoes significant overnight thickening accompanied by a reduction in visual acuity. Apparently, the mechanisms that maintain the equilibrium between production and removal of extracellular fluid in the healthy retina are exhausted in DME, such that sleeping in the recumbent position induces a shift to a level of higher intraretinal fluid content. 

Our study independently confirms a previous OCT study that followed patients from 8 AM until 5 PM with a newer OCT instrument and a full macular-thickness mapping technique based on rapid sequences of radial transfoveal scans. ⁵ The study demonstrated that most of the overnight swelling subsides before noon. The levels of macular edema were comparable to those in the present study, in which we were able to document poorer visual acuity in the morning than in the evening. This is in agreement with the report of Sternberg et al. ¹ and the sporadic complaints of patients that prompted us to conduct this study. 

We have also shown that the overnight change in retinal thickness is proportional to variations in arterial blood pressure. This is of particular relevance in diabetes, because a large proportion of patients have impaired nocturnal blood pressure regulation. ⁶ This observation indicates that the retinal capillary filling pressure is important in producing DME. Nevertheless, the larger part of the overnight increase in macular thickness seems to occur independent of changes in arterial blood pressure. Other mechanisms involved in the maintenance of retinal edema include the osmotic effect of extravasated plasma macromolecules; the increased permeability of the retinal vessels; and the mechanical compliance of the swollen retina and its internal hydrostatic pressure gradients, both of which are entirely unknown; and the venous pressure in the retina. The orbital venous pressure increases when changing position from standing to lying down, from nearly zero to roughly the pressure in the right atrium. Changes of this magnitude do not occur in the retina, because venous pressure is never lower than the intraocular pressure. A change toward a more reclining position induces only a small increase in intraocular pressure in healthy subjects, but a considerably larger increase can be seen in subjects with autonomic failure, which is common among people with diabetes. ^{7 8} Increasing intraocular pressure will decrease the transmural pressure gradient in the retinal capillaries. Hence, such postural effects on macular edema should cause a conservative bias in the results of the study. 

We were unable to demonstrate any effect of overnight fluctuations in blood glucose. Our study population was unlikely to have been exposed to nocturnal hypoglycemia, because, of the 12 patients, only 4 received insulin treatment; and, of those, only 2 had type 1 diabetes. In addition, blood glucose measurements before and after sleep revealed no evidence of hypoglycemia. 

In summary, we have confirmed that an overnight increase in DME is the rule rather than the exception when the fovea is involved, and we have shown that macular thickening often is associated with a concomitant reduction in visual acuity. Variation in arterial blood pressure was found to influence the process of edema formation, but orthostatic variation in venous blood pressure ⁹ and increased retinal metabolism ¹⁰ in the dark are other mechanisms that should be considered. 

 Supported by the Danish Eye Health Society (Værn om Synet), the Danish Association of the Blind (Øjenfonden), and a Patient-Oriented Diabetes Research Career Award from the Juvenile Diabetes Research Foundation (Grant no. 8-2002-130).

 Submitted for publication July 28, 2004; revised December 13, 2004; accepted December 18, 2004.

 Disclosure: M. Larsen, None; M. Wang, None; B. Sander, None

 The publication costs of this article were defrayed in part by page charge payment. This article must therefore be marked “advertisement” in accordance with 18 U.S.C. §1734 solely to indicate this fact.

 

Figure 1.

 

Difference in macular thickness between evening and morning after an overnight sleep followed by a 0.5-hour rest with open eyes in room light in the recumbent position, in 14 eyes in 7 healthy control subjects and in 12 eyes in 12 patients with DME. Healthy subjects demonstrated no systematic variation, whereas a nocturnal–recumbent increase was measured in eyes with DME, from 316 ± 72 μm (mean ± SD) to 336 ± 81 μm (+6.3%; P = 0.003). In 8 of 12 eyes with DME, the increase exceeded the mean +2 SD of the variation found in healthy subjects.

Figure 2.

 

Transfoveal horizontal OCT from the right eye of a woman aged 53 years (patient 2) with DME. Two identically located scans were recorded, one in the evening (left) and one in the morning (right) after overnight sleep followed by a 0.5-hour rest with open eyes in the recumbent position. Best-corrected visual acuity was 20/32 in the evening and 20/80 in the morning. Maximum retinal thickness, which was achieved approximately 300 μm off the foveal center, was 420 μm in the evening and 496 μm in the morning, an 18% increase.

Table 1.

 

Characteristics and Retinal Thickness of Patients with DME

Pt. No.	Age (y) Gender	Eye	Diabetes Type	Insulin Treatment	Retinal Thickness (μm)
										Macula, Evening	Macula, Morning	Absolute Difference	Relative Difference (%)	Fovea, Evening	Fovea, Morning	Absolute Difference	Relative Difference (%)
1	53/M	OD	2	N	354	383	28	8	394	444	50	13
2	53/F	OD	2	N	299	346	47	16	325	434	109	34
3	55/F	OS	2	N	273	277	5	2	259	251	−8	−3
4	52/M	OD	2	N	321	328	7	2	367	410	43	12
5	48/M	OS	1	Y	262	279	16	6	272	342	69	25
6	59/M	OD	2	N	506	546	41	8	597	665	68	11
7	78/M	OD	2	N	270	292	21	8	340	360	20	6
8	51/M	OD	2	N	349	382	33	9	411	476	65	16
9	70/M	OD	2	N	308	307	−2	−1	251	254	3	1
10	68/M	OS	2	Y	353	374	21	6	332	348	16	5
11	39/F	OD	1	Y	261	285	24	9	282	319	37	13
12	53/M	OD	2	Y	230	236	6	3	204	214	10	5
Mean	56.6				316	336	21	6	336	376	40	11
SD	10.7				72	81	15	4	103	122	34	10

 Data were collected in the evening and in the morning after overnight sleep followed by a 0.5-hour rest with open eyes in room light in the recumbent position. Retinal thickness, macula represents the average of seven measurements of retinal thickness, evenly spaced along a 4.63-mm wide horizontal transfoveal optical coherence tomogram. Retinal thickness, fovea represents the center foveal thickness of the same tomogram.

Table 2.

 

Visual Acuity, MABP, and Plasma Glucose in Patients with DME

Pt. No.	Visual Acuity (ETDRS Letters)							MABP (mm Hg)							Plasma Glucose (mM)
		Evening	Morning	Absolute Difference	Relative Difference (%)	Evening	Morning	Absolute Difference	Relative Difference (%)	Evening	Morning	Absolute Difference	Relative Difference (%)
1	35	35	0	0	111	130	19	17	5.1	6.3	1.2	23.5
2	43	27	−16	−37	98	101	3	3	11.0	9.3	−1.7	−15.5
3	46	44	−2	−4	108	83	−25	−23	N/A	N/A	0.0
4	61	59	−2	−3	105	97	−8	−8	7.9	5.2	−2.7	−34.2
5	53	47	−6	−11	106	109	3	3	N/A	N/A	0.0
6	4	19	15	375	89	105	16	18	7.1	8.1	1.0	14.1
7	35	32	−3	−9	90	100	10	11	10.4	10.6	0.2	1.9
8	35	26	−9	−26	104	125	21	21	4.7	5.1	0.4	8.5
9	43	35	−8	−19	118	123	5	4	6.7	5.4	−1.3	−19.4
10	50	45	−5	−10	108	110	2	2	N/A	N/A	0.0
11	47	29	−18	−38	97	97	0	0	4.4	10.1	5.7	129.5
12	41	29	−12	−29	103	82	−21	−20	14.3	7.8	−6.5	−45.5
Mean	41	36	−6	16	103	105	2	2	8	8	0	7.0
SD	14	11	9	114	8	15	14	14	3	2	3	51.3

 Data were collected in the evening and in the morning after overnight sleep followed by a 0.5-hour rest with open eyes in room light in the recumbent position.

Figure 3.

 

Relative overnight change in macular thickness versus relative change in MABP in 12 patients with fovea-involving DME as assessed with horizontal transfoveal OCT. The overnight change in macular thickness increased in proportion to the change in MABP (r = 0.65, P = 0.03, Wilcoxon). An overnight increase was found in 11 of 12 eyes in 12 patients, and this finding was unaltered by statistical adjustment for the effect of MABP (P = 0.002; parametric regression analysis).